11~UIni~verrsnityrnFr~eerSnta~teI~~~~OO~
34300004855254
Universiteit Vrystaat
THE IMPACTS OF MULTILATERAL AND BILATERAL TRADE
AGREEMENTS ON AGRICULTURE TRADE IN SACU
by
MADIME REUBEN MOKOENA
Submitted in accordance with the requirement for the degree
PHILOSOPHIEA DOCTOR (PhD)
in the
FACULTY OF NATURAL AND AGRICULTURAL SCIENCES
DEPARTMENT OF AGRICULTURAL ECONOMICS
UNIVERSITY OF FREE STATE
BLOEMFONTEIN
SOUTH AFRICA
PROMOTER : PROF. A JOOSTE
CO-PROMOTER : PROF. ZG ALEMU
DECEMBER 2011
ACKNOWLEDGEMENTS
This thesis owes its existence to God. Through His Love and Powerful Spirit, He encouraged
my Study Promoters to provide good leadership from the initial to the final stage of this study.
There are many individuals who have provided inputs in various aspects of this study and I
give my thanks to all those who have been involved, several of whom I need to mention by
name. Firstly, I wish to thank my Promoter, Professor André Jooste, for his efforts and
encouragement to write this thesis as well as for his valuable comments he made from the
proposal stage until the final stage of this thesis.
Secondly, I wish to express my sincere appreciation to my Co-promoter, Professor Zerihun
Gudeta Alernu, for providing me with access to his econometric programme tool designed
specifically for Eviews software, which I have utilized during the analytical and/or modelling
phase of the study as well"as for his valuable comments he made after reading the first draft
chapters until the final stage of this thesis.
Thirdly, I would like to extend my special gratitude to the Anonymous Referees of the
Quarterly Journal of International Agriculture for reviewing my article entitled "Impacts
of the EU-SA TDCA's reciprocal preferential tariff quotas on market access for cheese
and wines"; which was finally accepted for publication. Their comments were also helpful in
improving the quality of this study.
My final thanks are due to my family for their patience, tolerance, encouragement and
support. I dedicate this study to my wife (Maledimo), my two daughters (Karnogelo and
Lebogang) and my two sons (Motheo and Thabang).
May God Bless Us All!'!
THE IMP ACTS OF MULTILATERAL AND BILATERAL TRADE
AGREEMENTS ON AGRICULTURE TRADE IN SACU
By
MADIME REUBEN MOKOENA
DEGREE PHD (AGRICULTURAL ECONOMICS)
DEPARTMENT AGRICULTURAL ECONOMICS
PROMOTER PROF. A. JOOSTE
CO-PROMOTER PROF. ZG ALEMU
ABSTRACT
International markets for agricultural products were characterised by,. amongst others,
quantitative restrictions, tariff-based protection, border protection, non-tariff barriers, ete
before 1995. Likewise, agricultural sector in South Africa (SA) was also faced by similar
trade distorting measures during the post-apartheid era. In response to globalisation
challenges, SA committed to move from protective to liberal trade regime in the agricultural
sector, as witnessed by its trade diplomacy engagements with the international community in
the context of multilateral, bilateral and/or regional approaches.
At the multilateral level, SA has successfully implemented its commitments as negotiated in
terms of the Agreement on Agriculture (AoA) during the Uruguay Round (UR) of General
Agreement on Tariffs and Trade (GATT) negotiations that gave birth to the World Trade
Organization (WTO). At the bilateral level SA 'has signed a Preferential Trade Agreement
(PTA) with the European Union (EU) called the Trade, Development and Co-operation
Agreement (TDCA) (better known as the EU-SA TDCA and includes a Free Trade
Agreement). At the regional level, the Southern African Customs Union (SACU) member
states including SA have signed a Protocol on Trade or a Regional Trade Agreement (RTA)
with the non-SACU countries of the Southern African Development Community (SADC).
The main objective of the study was to measure the impact of trade agreements on the
agricultural trade between SA and its trading partners. A gravity model using panel data was
Il
employed to analyze the ex-post impacts of the implementation of the trade treatments, i.e.
WTO AaA, EU-SA TDCA and SADC Trade Protocol on agricultural trade flows between SA
and its agricultural trading partners. Various statistical tests were undertaken to select the
suitable models for the datasets of total agricultural and selected agricultural products trade
flows between SA and its agricultural trading partners.
After the statistical tests were undertaken, 189 feasible models in total were selected, of which
. 161 were dynamic models and 28 were static models. Furthermore, 152 Fixed Effects (FE), 2
Random Effects (RE) and 7 pooled Ordinary Least Squares (OLS) estimators were found to
be efficient and suitable for the dynamic models; and 14 FE and 14 RE estimators were found
to be efficient and suitable for the static models. The highest number of selected dynamic
models suggested that passed trade is the predictor for current trade. The per capita ODPs of
SA and of its trading partners, the real effective exchange rates and distance have also played
a significant and expected role in influencing agricultural trade flows between SA and its
agricultural trading partners.
The results of the study have indicated that agricultural trade flows between SA and its
agricultural trade partners have responded positively to the implementation of WTO AaA.
The implementation of EU-SA TDCA and SADC Trade Protocol during the first five years
(for the period 2000 - 2004) have not delivered the expected results, as the majority of
agricultural trade flows between SA and EU countries as well as between SA and SADC
countries were not affected and some of the agricultural trade flows between SA and EU
countries as well as between SA and SADC countries were negatively affected. While the
majority of agricultural trade flows between SA and EU countries as well as between SA and
SADC countries were still not affected during the second five-year term (for the period 2005
- 2009), there were some improvements due to the significant positive effects of the EU-SA
TDCA implementation on three agricultural trade flows (i.e. total agricultural trade, total cut
flowers trade and total preserved fruits and nuts trade) as well as the significant positive
effects of the SADC Trade Protocol implementation on four agricultural trade flows (i.e. total
agricultural exports, total agricultural trade, total cut flowers trade and total fruits and
vegetable juices trade). However, the number of agricultural trade flows between SA and
ROW countries that have improved significantly for both periods were more than those of the
EU and SADC countries, even though ROW countries did not have a trade agreement with
SA.
iii
The implementation of the EU-SA TDCA and SADC Trade Protocol have created room for
potential increases of all the agricultural trade flows between SA and EU countries as well as
between SA and SADC countries for both periods. However, some of these potential
increases for the period 2000 - 2004 were diverted to the other markets. On average, during
the implementation of the EU-SA TDCA for the period 2000 - 2004, about 0.44% of
agricultural exports, 0.96% of cut flowers exports and 0.77% of wine exports from SA
destined for EU were diverted to other markets Furthermore, about 2.01% of SA's wine
imports that were supposed to have been soureed from the EU countries came from SA's
other wine trading partners; as well as the diversion of about 0.73% of total wine trade from
the SA and EU market to either SA and other wine trading partner market or EU and other
wine trading partner market. Similarly, the implementation of the SADC Trade Protocol led
to diversion of agricultural exports (about 0.43%), cut flowers exports (about 0.93%), total cut
flowers trade (about 0.92%), wine exports (about 0.73%), wine imports (about 1.45%) and \
total wine trade (about 0.35%) during the same period.
With regard to the implementation of the EU-SA TDCA and SADC Trade Protocol during the
period 2005 - 2009, there was no proof of trade diversion for all agricultural trade flows,
except that the was a trade creation for some of the agricultural trade flows between SA and
EU countries as well as between SA and SADC countries. In the case of the EU-SA TDCA,
there was trade creation on total agricultural exports, total agricultural trade, total preserved
fruits and nuts trade and total wine trade. In the case of the SADC Trade Protocol, there was
trade creation on total agricultural trade, cut flowers exports and preserved fruits and nuts
exports. In conclusion, these findings have clearly shown that tariff reductions alone are not
panacea to improve agricultural trade between SA and its major trading partners given the fact
that EU-SA TDCA and SADC Trade Protocol were mainly characterized by tariff phase
down schedules.
iv
TABLE OF CONTENTS
CONTENT PAGE
Acknowledgements i
Abstract ii
Table of Contents v
List of Tables xiv
List of Figures xvii
Acronyms xx
CHAPTER 1
INTRODUCTION
1.1 Background of the study 1
1.2 Problem Statement 4
1.3 Research objectives 7
1.4 Outline of the Study 8
v
CHAPTER2
TRADE AGREEMENTS AND THEAGIDCULTURALTRADE
LIBERALIZATION POLICY REFORM IN SOUTH AFIDCA
2.1 Introduction 9
2.2 A multilateral approach of South Africa's trade liberalization policy 10
2.3 A bilateral approach of South Africa's trade liberalization policy 12
2.4 A regional approach of South Africa's trade liberalization policy 15
2.5 Summary 18
CHAPTER3
LITERA TURE REVIEW OF THE IMPACTS OF TRADE AGREEMENTS ON THE
ECONOMIES OF DEVELOPED AND DEVELOPING COUNTIDES
3.1 Introduction 20
3.2 Implications of the World Trade Organization Agreement on Agriculture 20
3.3 Implications of the EU-SA Trade, Development and Cooperation Agreement 32
3.4 Implications of the SADC Trade Protocol on the Trade 35
3.5 Implications of the Selected Trade Agreements in the World 38
3.6 Summary 47
VI
CHAPTER4
METHODOLOGY OF THE STUDY
4.1 Introduction 49
4.2 Market equilibrium models 49
4.2.1 Partial models 49
4.2.1.1 AGLINK model 50
4.2.1.2 Country-Link System 50
4.2.1.3 European Simulation model.. 51
4.2.1.4 World Food Model. 52
4.2.1.5 FAPRI model 53
4.2.1.6 GAPsi model 53
4.2.1.7 SWOPSIM model 54
4.2.1.8 WATSIM model 54
4.2.1.9 ATPSM model 56
4.2.1.10 CAPRI model , 56
4.2.2 Economy-wide models 57
4.2.2.1 G-cubed model 58
4.2.2.2 GTAP model 59
4.2.2.3 GREEN model 60
4.2.2.4 INFORUM model 60
4.2.2.5 MEGABARE model 61
4.2.2.6 MICHIGAN BDS model 61
4.2.2.7 RUNS model 62
4.2.2.8 WTO housemodel 63
VII
4.3 Single equation econometric models 63
4.3.1 Import demand models 63
4.3.1.1 Almost Ideal Demand System (AIDS) model 65
4.3.1.2 Rotterdam Demand System (RDS) model.. 68
4.3.2 Gravity model 69
4.4 Model consideration and motivation 71
4.5 Theoretical framework and specification of gravity model.. 72
4.6 Data requirements and sources 82
4.7 Summary 83
viii
CHAPTER5
IMPACTS OF TRADE AGREEMENTS ON THE AGIDCULTURAL TRADE FLOWS
BETWEEN SOUTH AFIDCA AND ITS AGIDCULTURAL TRADING PARTNERS
5.1 Introduction 84
5.2 Statistical tests and selection of the suitable models 85
5.3 Effects of the control explanatory variables on agricultural trade flows 90
5.3.1 Gross Domestic Product per Capita (GDPPC) 90
5.3.2 Real Effective Exchange Rates (REER) 93
5.3.3 Distance (DIST) 95
5.4 The impacts of the WTO AoA on selected agricultural trade flows between
South Africa and its world agricultural trading partners 97
5.4.1 Aggregate agricultural trade flows between South Africa and the world 97
5.4.2 Cheese trade flows between South Africa and the world 98
5.4.3 Cut flowers trade flows between South Africa and the world 99
5.4.4 Frozen fruits and nuts trade flows between South Africa and the world l 00
5.4.5 Preserved fruits and nuts trade flows between South Africa and the world lOl
5.4.6 Fruits and vegetable juices trade flows between South Africa and the world .. 102
5.4.7 Wine trade flows between South Africa and the world 103
5.5 The impacts of the implementation of EU-SA TDCA on aggregate agricultural
and selected agricultural products trade flows between South Africa and the
EU countries 104
5.5.1 Agricultural trade flows between South Africa and the
EU countries 104
5.5.1.1 Agricultural exports from South Africa to the EU countries 105
5.5.1.2 Agricultural imports from the EU countries to South Africa 107
5.5.1.3 Agricultural trade (imports plus exports) between South Africa and the
EU countries 108
ix
5.5.2 Cheese trade flows between South Africa and the EU countries l1 0
5.5.2.1 Cheese exports from South Africa to the EU countries l11
5.5.2.2 Cheese imports from the EU countries to South Africa 113
5.5.2.3 Cheese trade (imports plus exports) between South Africa and the
EU countries 114
5.5.3 Cut flowers trade flows between South Africa and the
EU countries 115
5.5.3.1 Cut flowers exports from South Africa to the EU countries 116
5.5.3.2 Cut flowers imports from the EU countries to South Africa 118
5.5.3.3 Cut flowers trade (imports plus exports) between South Africa and the
EU countries 119
5.5.4 Frozen fruits and nuts trade flows between South Africa and the
EU countries ; 120
5.5.4.1 Frozen fruits and nuts exports from South Africa to the EU countries :.121
5.5.4.2 Frozen fruits and nuts imports from the EU countries to South Africa 123
5.5.4.3 Frozen fruits and nuts trade (imports plus exports) between
South Africa and the EU countries 124
5.5.5 Preserved fruits and nuts trade flows between South Africa and the
EU countries 125
5.5.5.1 Preserved fruits and nuts exports from South Africa to the
EU countries 126
5.5.5.2 Preserved fruits and nuts imports from the South Africa countries to
South Africa 127
5.5.5.3 Preserved fruits and nuts trade (imports plus exports) between
South Africa and the EU countries 129
5.5.6 Fruits and vegetable juices trade flows between South Africa and the
EU countries 130
x
5.5.6.1 Fruits and vegetable juices exports from South Africa to the
EU countries 132
5.5.6.2 Fruits and vegetable juices imports from the EU countries to
South Africa : 133
5.5.6.3 Fruits and vegetable juices trade (imports plus exports) between South Africa
and the EU countries 134
5.5.7 Wine trade flows between South Africa and the EU countries 136
5.5.7.1 Wine exports from South Africa to the EU countries 137
5.5.7.2 Wine imports from the South Africa countries to SA 138
5.5.7.3 Wine trade (imports plus exports) between South Africa and the
EU countries 140
5.6 The impacts of the implementation of SADe Trade Protocol (TP) on selected
agricultural trade flows between SA and the SADe countries ~ 141
5.6.1 Aggregate agricultural trade flows between South Africa and the
SADC countries 141
5.6.1.1 Agricultural exports from South Africa to the SADC countries 143
5.6.1.2 Agricultural imports from the SADC countries to South Africa 144
5.6.1.3 Agricultural trade (imports plus exports) between South Africa and the
SADC countries 145
5.6.2 Cheese exports from South Africa to the SADC countries 147
5.6.3 Cut flowers trade flows between South Africa and the SADC countries 148
5.6.3.1 Cut flowers exports from South Africa to the SADC countries 149
5.6.3.2 Cut flowers imports from the SADC countries to South Africa 151
5.6.3.3 Cut flowers trade (imports plus exports) between South Africa and the
SADC countries 152
5.6.4 Frozen fruits and nuts exports from South Africa to the SADC countries 154
5.6.5 Preserved fruits and nuts exports from South Africa to the SADC countries 155
Xl
5.6.6 Fruits and vegetable juices trade flows between South Africa and the
SADC countries 157
5.6.6.1 Fruits and vegetable juices exports from South Africa to the SADC
countries 159
5.6.6.2 Fruits and vegetable juices imports from the SADC countries to
South Africa · 160
5.6.6.3 Fruits and vegetable juices trade (imports plus exports) between
South Africa and the SADC countries 161
5.6.7 Wine trade flows between South Africa and the SADC countries 163
5.6.7.1 Wine exports from SA to the SADC countries 164
5.6.7.2 Wine imports from the SADC countries to South Africa 165
5.6.7.3 Wine trade (imports plus exports) between South Africa and the
SADC countries 166
5.7 The response of selected agricultural trade flows between South Africa and ROW
countries to the implementation EU-SA TDCA and SADC Trade Protocol.. ....... 168
5.7.1 Aggregate agricultural trade flows between South Africa and the ROW
countries 168
5.7.2 Cheese trade flows between South Africa and the ROW countries 169
5.7.3 Cut flowers trade flows between South Africa and the ROW countries 170
5.7.4 Frozen fruits and nuts trade flows between South Africa and the
ROW countries 171
5.7.5 Preserved fruits and nuts trade flows between South Africa and the
ROW countries 172
5.7.6 Fruits and vegetable juices trade flows between South Africa and the
ROW countries 173
5.7.7 Wine trade flows between South Africa and the ROW countries 174
5.8 Summary 175
xii
CHAPTER6
SUMMARY, CONCLUSIONS AND RECOMMENDATIONS
6.1 Introduction 178
6.2 Empirical results of the study 180
6.2.1 Selected suitable models for all the agricultural trade flows datasets 180
6.2.2 Effects of the control explanatory variables on agricultural trade flows 181
6.2.3 Impacts of trade agreements on agricultural trade flows 182
6.2.3.1 Impact of the implementation of the WTO AoA 183
6.2.3.2 Impact of the implementation of the EU-SA TDCA 183
6.2.3.3 Impact of the implementation of the SADC Trade Protocol 187
6.2.3.4 Agricultural trade response between SA and ROW countries during the
implementation of the EU-SA TDCA and SADC Trade Protocol 190
6.3 Conclusions of the study 191
6.4 Recommendations of the study 194
REFERENCES 196
APPENDICES 229
xiii
LIST OF TABLES
Table 2.1 Preferential tariff quotas of agricultural products under the
EU-SA DCA 15
Table 2.2 Non-SACU SADC countries' offer to SACU countries Tariff phase
down schedule for agricultural products under SADC Trade Protocol 18
Table5.1 Selection of the Estimator suitable for Agricultural Exports from SA
to the World 87
Table 5.2 Appendices for the statistical tests results towards the selections of the model
suitable for the datasets of agricultural trade flows between South Africa and
its agricultural trading partners 88
Table 5.3 Results for the selection of the model suitable for the datasets of agricultural
trade flows between South Africa and its agricultural trading partners 89
Table 5.4 Results for the income effects on agricultural trade flows between South Africa
and its agricultural trading partners 92
Table 5.5 Results for the exchange rates effects on agricultural trade flows between
South Africa and its agricultural trading partners 94
Table 5.6 Results for the distance effects on agricultural trade flows between South
Africa and its agricultural trading partners 96
Table 5.7 Results for agricultural trade flows between South Africa and the world 98
Table 5.8 Results for cheese trade flows between South Africa and the world 99
Table 5.9 Results for cut flowers trade flows between South Africa and the world 100
Table 5.10 Results for frozen fruits and nuts trade flows between South Africa and the
the world 101
Table 5.11 Results for preserved fruits and nuts trade flows between South Africa and
the world 101
Table 5.12 Results for fruits and vegetable juices trade flows between South Africa
the world 102
Table5.!3 Results for wine trade flows between SA and the world 103
Table 5.14 Utilisation of SA's export quotas under the EU-SA TDCA I04
xiv
Table 5.15 Results for total agricultural trade flows between South Africa and the
EU countries 105
Table 5.16 Results for cheese trade flows between South Africa and the
EU countries ·····.·111
Table 5.17 Results for cut flowers trade flows between South Africa and the
EU countries ····.····.···1·16
Table 5.18 Results for frozen fruits and nuts trade flows between South Africa and the
EU countries ·····121
Table 5.19 Results for preserved fruits and nuts trade flows between South Africa
and the EU countries 126
Table 5.20 Results for fruits and vegetable juices trade flows between South Africa
and the EU countries 131
Table 5.21 Results for wine trade flows between South Africa and the EU countries :136
Table 5.22 Results for total agricultural trade flows between South Africa and the
SADe countries 142
Table 5.23 Resul ts for cheese exports from South Africa to the SADe countries 147
Table 5.24 Results for cut flowers trade flows between South Africa and the
SADe countries 149
Table 5.25 Results for frozen fruits and nuts exports from South Africa to the
SADe countries 154
Table 5.26 Results for preserved fruits and nuts exports from South Africa to the
SADe countries 156
Table 5.27 Results for fruits and vegetable juices trade flows between South Africa and
the SADe countries 158
Table 5.28 Results for wine trade flows between South Africa and the
SADe countries 163
xv
Table 5.29 Results for agricultural trade flows between South Africa and the
Rest of World 169
Table 5.30 Results for cheese trade flows between South Africa and the
Rest of World 170
Table 5.31 Results for cut flowers trade flows between South Africa and the
Rest of World 171
Table 5.32 Results for frozen fruits and nuts trade flows between South Africa and the
Rest ofWorld 172
Table5.33 Results for preserved fruits and nuts trade flows between South Africa and the
Rest ofWorld 173
Table 5.34 Results for fruits and vegetable juices trade flows between South Africa
Rest of World 174
Table 5.35 Results for wine trade flows between South Africa and the Rest of World ... l 75
Table 6.1 Impact results for the implementation of WTO AoA on agricultural trade
flows between South Africa and its worldwide agricultural trading
partners 183
Table 6.2 Table 6.6: Impact results for the implementation of EU-SA TDCA
on agricultural trade flows between South Africa and EU countries 184
Table 6.3 Impact results for the implementation of SADC Trade Protocol on
agricultural trade flows between South Africa and SADC countries 187
Table 6.4 Responsiveness results of the agricultural trade flows between South Africa
and ROW countries during the of the implementation EU-SA TDCA and
SADC Trade Protocol 191
xvi
LIST OF FIGURES
Figure 5.1 Average actual and potential value of agricultural exports from South Africa
to the EU countries for the periods 2000 - 2004 and 2005 - 2009 106
Figure 5.2 Average actual and potential value of agricultural imports from the
EU countries to South Africa for the periods 2000 - 2004 and
2005 - 2009 108
Figure 5.3 Average actual and potential value of agricultural trade between South Africa
and the EU countries for the periods 2000 - 2004 and 2005 - 2009 119
Figure 5.4 Average actual and potential value of cheese exports from South Africa to EU
countries for the periods 2000 - 2004 and 2005 - 2009 112
Figure 5.5 Average actual and potential value of cheese imports from the EU countries to
South Africa for the periods 2000 - 2004 and 2005 - 2009 113
Figure 5.6 Average actual and potential value of cheese trade from the EU countries to
South Africa for the periods 2000 - 2004 and 2005 - 2009 115
Figure 5.7 Average actual and potential value of cut flowers exports from SA to
EU countries for the periods 2000 - 2004 and 2005 - 2009 117
Figure 5.8 Average actual and potential value of cut flowers imports from the EU
countries to South Africa for the periods 2000 - 2004 and
2005 - 2009 118
Figure 5.9 Average actual and potential value of cut flowers trade between South Africa
and the EU countries for the periods 2000 - 2004 and 2005 - 2009 120
Figure 5.10 Average actual and potential value of frozen fruits and nuts exports from South
Africa to EU countries for the periods 2000 - 2004 and 2005 - 2009 122
Figure 5.11 Average actual and potential value of frozen fruits and nuts imports from the
EU countries to South Africa for the periods 2000 - 2004 and
2005 - 2009 124
Figure 5.12 Average actual and potential value of frozen fruits and nuts trade between
South Africa and the EU countries for the periods 2000 - 2004 and
2005 - 2009 125
xvii
Figure 5.13 Average actual and potential value of preserved fruits and nuts exports
from South Africa to EU countries for the periods 2000 - 2004 and
2005 - 2009 127
Figure 5.14 Average actual and potential value of preserved fruits and nuts imports
from the EU countries to South Africa for the periods 2000 - 2004 and
2005 - 2009 128
Figure 5.15 Average actual and potential value of preserved fruits and nuts trade
between South Africa and the EU countries for the periods 2000 - 2004 and
2005 - 2009 130
Figure 5.16 Average actual and potential value of fruits and vegetable juices exports
from South Africa to EU countries for the period 2000 to 2004 132
Figure 5.17 Average actual and potential value of fruits and vegetable juices imports
from the EU countries to South Africa for the periods 2000 - 2004 and
2005 - 2009 , 134
Figure 5.18 Average actual and potential value of fruits and vegetable juices trade
between South Africa and the EU countries for the periods 2000 - 2004 and
2005 - 2009 135
Figure 5.19 Average actual and potential value of wine exports from South Africa to EU
countries for the periods 2000 - 2004 and 2005 - 2009 138
Figure 5.20 Average actual and potential value of wine imports from the EU countries to
South Africa for the periods 2000 - 2004 and 2005 - 2009 139
Figure 5.21 Average actual and potential value of wine trade between South Africa and
the EU countries for the periods 2000 - 2004 and 2005 - 2009 141
Figure 5.22 Average actual and potential value of agricultural exports from South Africa to
SADe countries for the periods 2000 - 2004 and 2005 - 2009 144
Figure 5.23 Average actual and potential value of agricultural imports from the SADe
countries to South Africa for the periods 2000 - 2004 and
2005 - 2009 145
Figure 5.24 Average actual and potential value of agricultural trade between South Africa
and the SADe countries for the periods 2000 - 2004 and 2005 - 2009 146
Figure 5.25 Average actual and potential value of cheese exports from South Africa to
XVIII
SADe countries for the periods 2000 - 2004 and 2005 - 2009 148
Figure 5.26 Average actual and potential value of cut flowers exports from South Africa to
SADe countries for the periods 2000 - 2004 and 2005 - 2009 150
Figure 5.27 Average actual and potential value of cut flowers imports from the
SADe countries to South Africa for the periods 2000 - 2004 and
2005 - 2009 152
Figure 5.28 Average actual and potential value of cut flowers trade between South Africa
and the SADe countries for the periods 2000 - 2004 and
2005 - 2009 153
Figure 5.29 Average actual and potential value of frozen fruits and nuts exports from
South Africa to SADe countries for the periods 2000 - 2004 and
2005 - 2009 155
Figure 5.30 Average actual and potential value of preserved fruits and nuts exports
from South Africa to SADe countries for the periods 2000 - 2004 and
200 - 2009 157
Figure 5.31 Average actual and potential value of fruits and vegetable juices exports
from South Africa to SADe countries for the periods 2000 - 2004 and
2005 - 2009 159
Figure 5.32 Average actual and potential value of fruits and vegetable juices imports
from the SADe countries to South Africa for the periods 2000 - 2004 and
2005 - 2009 161
Figure 5.33 Average actual and potential value of fruits and vegetable juices trade
between South Africa and the SADe countries for the periods 2000 - 2004 and
2005 - 2009 162
Figure 5.34 Average actual and potential value of wine exports from South Africa to
SADe countries for the periods 2000 - 2004 and 2005 - 2009 165
Figure 5.35 Average actual and potential value of wine imports from the SADe
countries to South Africa for the periods 2000 - 2004 and 2005 - 2009 166
Figure 5.36 Average actual and potential value of wine trade between South Africa and the
SADe countries for the periods 2000 - 2004 and 2005 - 2009 167
xix
ACRONYMS
AoA : Agreement of Agriculture
ABARE : Australian Bureau of Agricultural and Resource Economics
ACP : African, Caribbean and Pacific
AEC : African Economic Community
AFTA : ASEAN Free Trade Agreement
AGE : Applied General Equilibrium
AGOA : Africa Growth and Opportunity Act
AIC : Akaike Information Criteria
AIFTA : ASEAN-India Free Trade Agreement
ATPSM : Agricultural Trade policy Simulation Model
AU : African Union
BLNS : Botswana, Lesotho, Namibia and Swaziland
CAFTA : Central America Free Trade Agreement
CAP : Common Agricultural Policy
CDE : Constant Differences of Elasticities
CG : Cairns Group
CGE : Computable General Equilibrium
CLS : Country-Link System
COMESA : Common Market for Eastern and Southern Africa
CUSTA : Canada-U.S. Trade Agreement
DH : Doha Round
DIST ':'Distance
DTI : Department of Trade and Industry
EAC : East African Community
ECCAS : Economic Community of Central African States
ECO : Economic Cooperation Organization
ECOWAS : Economic Community of West African States
EFTA : European Free Trade Agreement
EU-SA TDCA : European Union-South Africa Trade, Development and
Co-operation Agreement
EU : European Union
EPAs : Economic Partnership Agreements
xx
ERS : Economic Research Service
ESIM : European Simulation Model
FAO : Food and Agriculture Organization
FAPRI : Food and Agricultural Policy Research Institute
FDI : Foreign Direct Investment
FE : Fixed Effects
FTA : Free Trade Agreement
FTAs : Free Trade Areas
GATT : General Agreement on Tariffs and Trade
GDPPC : Gross Domestic Product per Capita
G.E.F : Generalized Exponential Form
GTAP : Global Trade Analysis Project
GMOs : Genetically Modified Organisms
HSRC : Human Sciences Research Council
lCA : International Coffee Agreement
ICAC : International Cotton Advisory Committee
IGC : International Grains Council
IMF : International Monetary Fund
LA-AIDS : Linear Approximate Almost Ideal Demand System
MA : Market Analysis
MERCOSUR : Common Southern Market
MFN : Most Favoured Nation
MNCs : Multinational Corporations
MTS : Multi Trading System
NAMC : National Agriculture Marketing Council
NAFTA : North American Free Trade Agreement
NEPAD : New Partnership for Africa's Development
NTBs : Non-Tariff Barriers
OAU : Organization of African Unity
OIE : Organization International des Epizooties
OECD : Organisation for Economic Co-operation and Development
OLS : Ordinary Least Squares
PTA : Preferential Trade Agreement
QUAIDS : Quadratic AIDS
XXI
REER : Real Effective Exchange Rates
RTAs : Regional Trade Agreements
RSDAIDS : Restricted, Source-Differentiated Almost Ideal Demand System
RE : Random Effects
ROW : Rest of the World
SA : South Africa
SACU : Southern African Customs Union
SACU-US TIDCA : SACU-United States Trade, Investment, Development and
Co-operation Agreement
SADC : Southern Africa Development Community
SAM : Social Accounting Matrix
SC : Schwarz Criterion
SDAIDS : Source Differentiated, Almost Ideal Demand System
SMEs : Small and Medium Enterprises
SPS : Sanitary and Phytosanitary standards
SSA : Sub- Saharan African
SPs : Sensitive Products
TAs : Trade Agreements
TBT : Technical Barriers to Trade
TFP : Total Factor Productivity
TIPS : Trade and Investment Policy Strategies
TNCs : Trans National Companies
TP : Trade Protocol
TRQ : Tariff Rate Quota
UNCTAD : United Nations Conference on Trade and Development
URAoA : Uruguay Round Agreement on Agriculture
US : United States
US$ : United States Dollars
WFM : World Food Model
W&S : Wines and Spirits
WTO : World Trade Organization
WTOAoA : World Trade Organization Agreement on Agriculture
xxii
CHAPTER 1
Introduction
1.1 Background of the study
South Africa is one of the founder members of the General Agreement on Tariffs and Trade
(GATT). GATT was established in Geneva in 1947 to create a framework that would
regulate international trade through gradual reduction of trade barriers so as to stimulate
international commerce. The most important elements of the Agreement included those of
non-discrimination: the Most Favoured Nation (MFN) principle; reciprocity; transparency and
tariff reduction (Healy et al, 1998). The principal mechanisms for progress on trade
liberalisation through GATT have been the measures adopted at the periodic multilateral
negotiating rounds. In all, there have been nine rounds, starting with the 1947 Geneva Round
which established GATT, followed by the Annecy Round of 1949, the Torguay Round of
1950, the Geneva Round of 1956, the Dillon Round of 1960-61, the Kennedy Round of 1962-
67, the Tokyo Round of 1973-79, the Uruguay Round of 1986-93 and the Doha Development
Round that was launched in 2001.
The primary focus of the majority of rounds has been the promotion of multilateral tariff
reductions, and the extension of the agreed reductions to all members in accordance with the
MFN clause. Due to reasons of national food security and the fact that agriculture is often
considered a unique sector of the economy, agricultural sector trade was excluded from
GATT during the early rounds until it was placed on the GATT negotiating table during the
Uruguay Round that established the World Trade Organization (WTO), which is now the
main multilateral trade body. However, certain agricultural products have featured in other
negotiations as individual commodities. For example, the Dillon Round succeeded in cutting
tariffs on soya beans, cotton, vegetables and canned fruit to very low levels, and the
International Wheat Agreement and the International Dairy and Meat Agreement were
negotiated under the auspices of the Kennedy Round. In general, agricultural commodities
have remained off the negotiating table (Healy et al, 1998).
Despite GATT membership, South Africa has also engaged itself in the international
economy and participates effectively in the globalisation process through its involvement in
Introduction
various international organisations at regional, bilateral and multilateral levels. Apart from its
GATT/WTO membership, South Africa has also gained membership to the following
international organisations: International Grains Council (IGC), International Cotton
Advisory Committee (ICAC), International Coffee Agreement (ICA), Food and Agriculture
Organization (FAO), United Nations Conference on Trade and Development (UNCTAD),
World Organization for Animal Health - Organization International des Epizooties (OIE),
Codex Alimentarius, Cairns Group (CG), and others. These international organisations,
except the WTO, do not have trade packages but they do have an influence on trade
negotiations, especially the CG lobbying group.
On the other hand, South Africa has also joined hands with its African counterparts in
endeavouring to implement regional economic integration on a continental scale as initiated
by the Organization of African Unity (OAU), the predecessor of the African Union (AU). The
AU was launched in July 2002 in Durban, South Africa and aims at finalising the
establishment of the African Economic Community (AEC) by the year 2025, as agreed to by
34 African countries in Abuja, Nigeria in 1991 (Babarinde, 1996). The AU and NEPAD
(New Partnership for Africa's Development) have taken over from where the OAU ended and
will continue implementing the AEC. There are several approaches or arrangements that have
been developed or defined as ways of implementing the Abuja Treaty, and establishing the
AEC. South Africa is highly involved in most of them. These are the:
• Establishment of customs union level of integration in Africa,
• Establishment of free trade areas (FTAs) in Africa,
• Establishment of bilateral trade agreements between African countries, and
• Establishment of common market level of integration in Africa.
Regarding regional integration, South Africa is a member of the Southern African Customs
Union (SACU) and has played a leading role in the renegotiation of the SACU Agreement,
which was concluded in 2001. South Africa is also a member of the Southern Africa
Development Community (SADC), which established the Trade Protocol that was signed in
1999. Currently, South Africa does not hold membership in some of the established common
market integration levels in Africa, such as the Common Market for Eastern and Southern
Africa (COMESA), the on-off East African Community (EAC), the Economic Community of
2
Introduction
Central African States (ECCAS) and the Economic Community of West African States
(ECOW AS). These common market integration levels also have impacts on the South
African economy due to co-operation agreements which South Africa has with some of the
member countries who also happen to be SACU and SADC members. It should, however, be
noted that.sthere are ongoing initiatives that are structured around a possible tripartite FTA
between COMESA, EAC and SADC with the objective of advancing trade integration across
Africa and ensuring that African countries do not trade at a competitive disadvantage as
compared to other non-African trading partners (Sandrey, 2011).
Outside of Africa, South Africa has also concluded various trade agreements with other
countries or economic blocs and has also engaged in trade related negotiations with other
countries. For example, South Africa signed a Trade, Development and Co-operation
Agreement with the European Union (EU-SA TDCA) in 1999. In 1997, South Africa was
admitted as a qualified member of the Lomé Convention which was subsequently replaced by
the Cotonou Agreement, which was a co-operation agreement between the European Union
(EU) countries and the African, Caribbean and Pacific (ACP) countries. However, South
Africa did not qualify for all the benefits that ACP countries received under this agreement.
The Cotonou Agreement expired in 2007 and was replaced by Economic Partnership
Agreements (EPAs) between the EU and several developing and least developed country
groupings. However, the EPA negotiations between the EU and the SADC (including South
Africa) have not yet been concluded as some members of SADC have not yet signed, whereas
others have. On the other hand, SACU (including South Africa) concluded a Free Trade
Agreement (FTA) in 2006 with the non-EU countries which are affiliated to the European
Free Trade Agreement (EFTA), namely: Switzerland, Liechtenstein, Norway and Iceland
(NAMC, 2008).
Regarding America, South Africa was a beneficiary of the Africa Growth and Opportunity
Act (AGOA) of the United States (US), which provided preferential access for imports from
Sub-Saharan African countries into the US market. AGOA was signed in 2000 with the
expiry date of September 2008, but was extended to 2015. In addition, the US and SACU
negotiated the Trade, Investment, Development and Co-operation Agreement (SACU-US
TIDCA) that was concluded in 2008. TIDCA established a co-operative framework to
address non-tariff issues such as standards and customs procedures and also established
3
Introduction
commitments to enter into joint trade and investment promotion activities. In 2009 SACU
(including South Africa) signed a Preferential Trade Agreement (PT A) with the Mercosur
countries, namely: Brazil, Paraguay, Uruguay and Argentina. The SACU-MERCOSUR PTA
created a legal and institutional setting for resolving any trade friction that may arise in future
between the two regions, but its commercial value is limited (DTI, 20 10).
With regard to Asia, South Africa has worked closely with China to develop and implement
the Partnership for Growth and Development aimed at promoting value-added South African
exports to China and increasing inward investment by China in projects involving mineral
beneficiation. PTA negotiations with India are also currently underway. The relationship
with Japan is pursued through ongoing bi-national co-operative agreements.
This study will mainly focus on the multilateral, bilateral and regional trade agreements that
have been implemented in order to analyse their impacts on the agricultural sector trade
between South Africa and its trading partners. The specific focus will be on the impacts of
WTO, EU-SA TDCA and SADC FT A on South African agricultural trade.
1.2 Problem Statement
South Africa re-entered the international economy in the early 1990s at a time when the
process of globalization was beginning to gain momentum. To share in the benefits of
globalization, South Africa pursued a strategy of trade liberalization policy reform in the
context of multilateral, bilateral and/or regional approaches (Kusi, 2002). The distinguishing
characteristic of the reform policy was a willingness to expose the country to tariffs that were
often below the bound rates which were negotiated in the Uruguay Round of GA TT.
Whereas agricultural trade had been managed through quantitative controls, the Marrakech
Agreement called for the tariffication of all agricultural goods, and a phased reduction in the
tariffs. In this process, South Africa substantially liberalised its economy through reform of
the import regime and deregulation of the agricultural sector by reduction of domestic support
and export subsidies, as well as by harmonisation of Sanitary and Phytosanitary (SPS)
measures (Poonyth et ai, 2002; Jooste et ai, 2003). This led to increased trade openness in the
sector owing to the substantial elimination of trade restrictions, the rationalisation and
4
Introduction
simplification of the tariff regime as well as the reduction of tariff rates (NAMe, 1999; Loots,
2002).
Historically, South Africa had been trading agricultural products long before the adoption of
the trade liberalisation strategy. Given the policy reform in question, one would be interested
to ascertain the influence of the trade liberalisation policy on agricultural trade between South
Africa and its counterparts. To what extent do these trade agreements affect South Africa's
agricultural trade, as compared to other historical trade determinants? Are these trade
agreements significantly improving agricultural trade between South Africa and its trading
partners? If yes, how? Have they led to the attraction of South Africa's agricultural exports
to its trading partners or vice versa? Have the bilateral/regional trade agreements created or
diverted South African agricultural trade? Which trade agreement is more significant than the
others in terms of improving agricultural trade potential between South Africa and its
counterparts?
Generally, most countries, including South Africa, have adopted trade liberalisation policies
aimed at improving trade among themselves so as to improve economic growth, generate
employment, improve welfare gains, and the like. As agricultural trade liberalisation is part
of the trade liberalisation strategy of South Africa (DTI, 2010), it is necessary to analyse the
impacts of such trade agreements on agricultural trade between South Africa and its trading
partners and to compare them in order to ascertain which of these trade agreements is more
beneficial than the others. This analysis will also indicate how significant these trade
agreements are in terms of influencing agricultural trade between South Africa and its trading
partners.
Although much research had been conducted internationally into the effects of various trade
agreements on agricultural trade between different trading blocks, including developed 'and
developing countries, inadequate research has been conducted on the South African situation.
Several studies have attempted to answer the above questions, but with limited scope owing
to the fact that they were based on assumptions. Most of the South African case studies have
focused on the impacts of trade agreements on economic growth and welfare, but have
concentrated on a single agreement without comparing it to the others that also affect trade
between South Africa and its counterparts (see Davies, 1998; Penzorn and Kirsten, 1999;
5
Introduction
Andriamananjara and Hillberry, 2001). Other studies focused on only one aspect of trade,
mainly the export-side (see Kalaba, 2001; Cassim, 2001; Chauvin and Gaulier, 2002; Poonyth
et al, 2002; Nouve and Staatz, 2003). Few studies did compare the impacts of various trade
agreements on the above variables, but they were not specific to agricultural trade
liberalisation (see Lewis, 2001; Sandrey, 2006). Other studies focused on the impact of tariff
reductions on specific agricultural commodities (see Jooste, 1996, Jooste et al, 2001;
Oyewumi et ai, 2007). These studies have not addressed the questions of how, and to what
extent, these trade agreements influence agricultural trade between South Africa and its
counterparts, nor as to which one is more influential than the other.
It is not clear from the literature as to whether the trade agreements under review have led to
the attraction of South Africa's agricultural exports to trading partners, or the attraction of
South Africa's agricultural imports from such partners, or both. Therefore, it is difficult to
judge and generalise as to which trade agreement is more beneficial than the others insofar as
South African agricultural trade performance is concerned. Some of the agricultural products
under examination have been given preferential treatment, either reciprocally or non-
reciprocally, such as in-quota tariff rates and annual tariff phase-downs, effective from
implementation of such trade agreements. The question is: have the signatories of such
preferential trade agreements complied with what they had signed for? If they have fully
complied, it is expected that the volume of trade in agricultural products between South
Africa and its trading partners would have improved significantly during the implementation
of such trade agreements. This would in turn have positive effects on economic growth,
employment and welfare.
Therefore, there is a need to analyse the impacts of such preferential treatments on the imports
and exports of benefiting agricultural products between South Africa and its trading partners.
It is indeed necessary to ascertain how the trade agreements influence South Africa's
agricultural trade flows, i.e. whether they influenced South Africa to export more than
importing, or vice versa or both. In addressing the above questions, this study will analyse
the ex-post impacts of such preferential treatments on South Africa's agricultural trade at both
aggregate level (total agricultural imports and exports) and disaggregate or product level
(imports and exports of specific agricultural products).
6
Introduction
Moreover, this study will attempt to assess the impact of existing multilateral, bilateral and
regional trade agreements which South Africa has signed (WTO AaA, EU-SA TDCA and
SADC Trade Protocol (TP) on agricultural trade between South Africa and its trading
partners. Therefore, this study will generate new knowledge in terms of measuring the
compliance of the signatories to the trade agreements and will provide a useful contribution to
the understanding of the likely impacts of trade agreements on the volume of trade between
South Africa and its major trading partners. Furthermore, this study will also review the
literature on the impacts of various trade agreements on the agricultural and other economic
sectors of the developed and developing countries.
1.3 Research objectives
The overall objective of this study is to measure the impacts of the trade agreements under
review on agricultural trade between South Africa and its trading partners. The following are
the specific objectives:
1.3.1 To provide an overview of the trade agreements which have implications for
agricultural trade in South Africa;
1.3.2 To review the impacts of agricultural trade liberalisation policies in the context of the
trade agreements on the economic growth and welfare of South Africa and the
Southern Africa region, as well as of its trading partners;
1.3.3 To determine whether the trade agreements have a significant influence on agricultural
trade between South Africa and its trading counterparts;
1.3.4 To investigate whether the trade agreements have caused trade creation or trade
diversion;
1.3.5 To estimate trade potentials between South Africa and its trading partners owing to the
trade agreements.
7
Introduction
1.4 Outline of the Study
Chapter 2 provides an overview of various trade agreements and South Africa's agricultural
trade liberalisation policy, with a focus on the implemented trade agreements that South
Africa has signed. These are: a multilateral trade agreement with respect to the WTO AoA, a
bilateral trade agreement with respect to the EU-SA TDCA, and regional trade agreements
with respect to SACU and the SADC Trade Protocol. This chapter addresses the policy issues
around agricultural trade liberalisation with a view of unpacking what South Africa is offering
the international community, as well as what the international community is offering South
Africa in terms of agricultural trade provided by the trade agreements in question.
Chapter 3 outlines the impacts of trade liberalisation on developing countries including South
Africa. This encompasses a literature review of previous studies on the impacts of the trade
agreements in question on the economies of developed and developing countries, with a focus
on the. agricultural sector. Chapter 4 discusses the various models used in trade policy
analysis and summarises the theoretical framework of the model adopted in this study.
Furthermore, this chapter also discusses data requirements and provides the sources where
data was collected. Chapter 5 presents the empirical results of the study at both aggregate and
disaggregate levels, and Chapter 6 provides the summary, conclusions and recommendations
of this study.
8
CHAPTER2
TRADE AGREEMENTS AND THE AGRICULTURAL TRADE LIBERALISATION
Policy Reform in South Africa 1
2.1 Introduction
South Africa's agricultural sector, like those in most countries, was characterised by trade
distorting measures during the apartheid era, ranging from quantitative restrictions, price
controls, subsidies directly related to production quantities, and the like. These interventions
were aimed at supporting commercial farm incomes, promoting food self-sufficiency, and
stabilising prices (Van Schalkwyk, 1997; Jooste et al, 2003).
To reverse the years of recession and decades of "inward industrialisation strategies" trade
liberalisation became one of the central driving instruments for achieving accelerated
economic growth in South Africa. South Africa has also embarked on a process of trade
liberalisation policy reform in the context of multilateral, bilateral and regional approaches.
In the process, South Africa substantially liberalised the economy through reform of the
import regime and deregulation of the agricultural sector (Vink et ai, 2002).
This chapter addresses the policy issues around agricultural trade liberalisation with a view of
unpacking the trade benefits offered by the trade agreements under review. The following
sections of this chapter provide a detailed discussion on South Africa's reaction to the
globalisation policies through its engagements with the international community.
Furthermore, this chapter also describes the agricultural offers provided by the multilateral,
regional and bilateral trade agreements.
I NB: In this study, South Africa is referred to as including the Southern African Customs Union (SACU)
because the trade data of South Africa and the rest of SACU countries (i.e. Botswana, Lesotho, Namibia and
Swaziland) is combined due to the common external tariff. However, South African trade (exports and imports)
constitutes more than 90% of the total SACU trade (exports and imports).
9
Trade agreements and the agricultural trade liberalisation
policy reform In South Africa
2.2 A multilateral approach of South Africa's trade liberalisation policy
South Africa successfully participated in the negotiations of the Uruguay Round of GATT
and became a signatory of the Marrakech Agreement in 1994. Since then, South Africa's
trade regime has changed considerably as agriculture was brought under the multilateral trade
rules at the conclusion of the Uruguay Round and the establishment of the WTO in 1995
(Vink ef al., 2002). The Uruguay Round of GATT reinforced a rules-based system of trade: it
brought agriculture under the discipline of the trade rules of GATT and established a process
for reductions in support of agriculture. It also entrenched tariffs, through tariffication of non-
tariff barriers, as the currency of protection and it established the WTO, with the capability to
enforce the discipline which the various contracting countries agreed to (lngco and Townsend,
1998; Tsigas and Ingco, 2001).
In brief, the Uruguay Round Agreement on Agriculture (UR AoA) covers three main areas:
reductions in farm export subsidies, increases in import market access and cuts in domestic
producer subsidies. For example, on reductions in farm export subsidies, budget outlays of
industrialised countries were to be cut by 36% in value terms (24% for developing countries),
and the volume of subsidised exports for each commodity were to be reduced by 21% (14%
for developing nations) over the six years from 1995 to 2000 (10years to 2004 for developing
countries) from their 1986-90 base-period averages. Moreover, no export subsidies not in
place in the base year may be added. As far as cuts in domestic producer subsidies are
concerned, a common measure called the "Aggregate Measure of Support", which quantifies
the amount of domestic support to producers, was to be reduced by 20% (13.3% for
developing countries) over the implementation period from the 1986-88 level on average.
This information is obtainable from WTO website (www.wto.org).
The liberalisation of agricultural trade in South Africa started with the Marrakech Agreement
in 1994 and was given greater momentum after the first democratic government in South
Africa came into power in 1994, as part of the government's reorientation of the economy
from import substitution to an export-led growth strategy (Vink et al, 2002). South Africa's
agricultural offer to the WTO consisted of a five-year tariff reduction and rationalisation
programme, which entailed reducing to six the number oftariff categories that had previously
numbered over 100. Liberalisation of the agricultural sector first took the form of tariffication
10
Trade agreements and the agricultural trade liberalisation
policy reform In South Africa
of quantitative restrictions followed by the reduction in diversity of ad valorem tariffs
(Cassim et al, 2002).
As a result of these deregulation and trade liberalisation policies, South Africa committed
itself to various international obligations and implemented successfully all the Uruguay
Round rules on agriculture through the:
./' Introduction of the new Marketing of Agricultural Products Act in 1996 that resulted
in the elimination of all marketing boards, the removal of price regulation and single
channel markets by the end of 1997. This led to the reduction of domestic support
measures to WTO acceptable levels in 2000 .
./' Removal of export subsidies in July 1997 by the termination of the General Export
Incentive Scheme, except for sugar. For the latter, an industry arrangement exists for
local prices .
./' Replacement of import permits by import duties. This has already improved access to
the South African market.
South African agriculture is thus now generally free from trade distorting support measures.
Apart from the fiscal constraint that limits the extent to which it can support farmers, current
policy is predicated on the view that trade liberalisation will encourage efficient utilisation of
our scarce resources. Government strategy for growth and distribution is based on this trade
liberalisation approach. Improved market access is a key strategy for South Africa's
agricultural development. South Africa's interest has shifted to actively pursuing further
liberalisation of global markets and to the removal of trade distorting domestic support and
export subsidies by competitors. This is necessitated by South Africa's accession of
membership in the Cairns Group, which is a lobby group or informal association of
agricultural exporter members of the World Trade Organisation which share the common
objective of further liberalisation of global agricultural trade.
Many countries including South Africa have complied with the rules of the URAA of the
WTO, but high-income economies such as the EU and the US have exploited the loopholes of
the URAA, which has enabled them to provide more support while staying within their limits.
11
Trade agreements and the agricultural trade liberalisation
policy reform In South Africa
With the URAA nearly fully implemented, heterogeneous market interventions in the
economies in question still distort resource allocation and trade in agriculture (Fabiosa et al,
2003). As a result, many developing countries became disappointed with the limited
accomplishments achieved by the Marrakech Agreement. This disappointment led them to
voice their concerns as largely reflected in the Doha Declaration of the WTO (WTO, 2001).
Primarily, the lack of market access in high-income countries constrains trading opportunities
for developing economies because of tariff rate quotas (TRQs) and other trade barriers
(Martin and Winters, 1995; Anderson et al, 2001).
The new Doha Round of multilateral trade negotiations, as adopted in November 2001 in
Qatar, builds on the previous work of the Uruguay Round and, without prejudging the
outcome of the negotiations, members committed themselves to comprehensive negotiations
aimed at substantial improvements in market access; the reduction of, with a view to
eventually phasing out, all forms of export subsidies; and substantial reductions of trade-
distorting domestic support (WTO, 2001).
2.3 A bilateral approach of South Africa's trade liberalisation policy
South Africa has signed a Trade, Development and Co-operation Agreement with the EU
(better known as the EU-SA TDCA, which also includes a Free Trade Agreement). This
agreement was the culmination of five years of protracted negotiations and came into force in
January 2000. This reciprocal agreement entails the liberalisation of tariffs on 95% of EU
imports from South Africa over aIO-year period and on 86% of tariffs on South Africa's
imports from the EU over a 12-year period (Cassim et al, 2002). The main agricultural offers
of the EU-SA TDCA are as follows (EC Council, 1999):
o Agricultural tariff phase-down: According to the agreement, the EU will liberalise
approximately 61% of agricultural imports from South Africa over a ten-year
implementation period. South Africa will liberalise approximately 83% of
agricultural imports from the EU over 12 years. To achieve this, both sides have
placed products in tariff phase-down groups or lists based on the sensitivity of the
product or industry to tariff liberalisation. Certain sensitive products were placed on
'reserve lists'. Although tariff elimination is not envisaged for products on the
12
Trade agreements and the agricultural trade liberalisation
policy reform In South Africa
reserve list, the situation will be reviewed at a later stage. It is understood that
reviews will take place no later than five years after entry into force of the agreement.
The EU placed beef, certain dairy products, cut flowers, certain fresh deciduous
fruits, rice, maize, sugar, certain canned fruits and vegetables, certain fruit juices and
wine on the reserve list. South Africa placed beef, mutton, maize, wheat, certain
dairy products, and sugar on the reserve list.
o Agricultural tariff quotas: The EU has granted South Africa preferential tariff quotas for
cheese, cut flowers (including a separate quota for proteas), frozen strawberries, canned
fruit, fruit juices, sparkling wine and wine. In turn, South Africa also granted the EU
preferential tariff quotas for cheese, sparkling wine and wine. These quotas make up
approximately 13% of South Africa's agricultural trade with the EU. Table 1 provides
detailed information about these quotas.
o Agricultural safeguard clause (Article 16): The agricultural safeguard clause (Article
16) written into the agreement gives South Africa the right to challenge the EU if proof
can be found that increased imports of agricultural products are causing harm or
threatening to cause harm to the domestic industry. It calls for consultations to address
these problems, while it also allows for immediate action in cases where such action is
justified.
o Rules of origin: The rules of origin prohibit the deflection of trade within the free trade
agreement. They lay down specific criteria for imports enabling the importing country to
determine whether the imported product can be considered as originating in the exporting
country or not. According to the agreement, all South African exports to the EU subject
to preferential treatment under the agreement will have to be accompanied by a certificate
of origin certifying that the product in question meets the rules of origin. The South
African Revenue Service (Customs and Excise Division) will be responsible for issuing
rules of origin certificates (form EUR1).
o Co-operation in agriculture (Article 61): Article 61 of the agreement is aimed at the
promotion of sustainable rural development in South Africa through co-operation between
South Africa and the EU. Co-operation according to the article will take place through the
13
Trade agreements and the agricultural trade liberalisation
policy reform In South Africa
transfer of know-how, the establishment of joint ventures, and capacity building
programmes.
o Compromise agreement on port and sherry: A compromise agreement on port and
sherry paved the way for the conclusion of the overall agreement. It contains a number of
elements including a commitment from South Africa to phase out its use of the terms
'port' and 'sherry' on the international and SADC market (over five and eight years
respectively) and to review its use of the terms port and sherry on the domestic market
jointly with the EU no later than 10 years after the agreement is implemented.
o Wines and Spirits Agreements: The negotiations around the Wines and Spirits (W&S)
Agreements took a long time due to disagreements about the originality of the names
'port' and 'sherry'. A political compromise was reached in March 1999, under which
South Africa would phase out the names port and sherry within an agreed time period in
consideration for concessions of, in addition to preferential quotas, 15 million Eurosfor a
programme for restructuring the South African wines and spirits industry and for
marketing and distribution. The Wines and Spirits Agreements were signed in January
2002 and became effective immediately. Due to this delay, both parties agreed on a
formula to increase the wine quota to 42.02 million litres with effect from January 2002
which would then increase by 6.72 million litres per year until the end of the phase-down
period, as a compensation mechanism to take account of the fact that the quota had not
been opened in 2000 and 2001.
Historically, before the conclusion and implementation of the EU-SA TDCA, the EU had
been South Africa's main trading and investment partner, accounting for over 40% of its total
trade. Likewise, the EU's foreign investment in South Africa accounted for over 70% of its
total foreign direct investment (FDI), a figure that is likely to grow in the light of this
agreement. It is expected that the EU-SA TDCA will also strengthen and improve the access
of South Africa's agricultural products into the EU market and vice versa, as a resuIt of the
tariff cuts and quota allocations committed by both parties.
14
Trade agreements and the agricultural trade liberalisation
policy reform In South Africa
Table 2.1: Preferential tariff quotas of agricultural products under the EU-SA TDCA
HS Code Product Description Initial Quota Tariff Quota Duty AGF
European Union's offer to South Africa
0406 Cheese and curd 5000 tons Reduced by 100% of MFN 5%
0603 Cut flowers - roses, orchids & chrysanthemums 500 tons Reduced by 100% ofMFN 3%
Cut flowers - proteas 990 tons Reduced by 100% of MFN 5%
Other cut flowers I 100 tons Reduced by 75% of MFN 3%
0811 Frozen fruits and nuts 250 tons Reduced by 100% of MFN 3%
2008 Prepared or preserved fruits and nuts 60000 tons Reduced by 100% of MFN 3%
2009 Fruit and vegetable juices 5700 tons Reduced by 50% of MFN 3%
2204 Wine of fresh grapes - sparkling wine 450 000 litres Reduced by 100% of MFN 5%
Wine of fresh grapes - excluding sparkling wine 32 000 000 litres Reduced by 100% of MFN 3%
South Africa's offer to the European Union
0406 Cheese and curd 5000 tons Reduced by 50% of MFN 3%
2204 Wine of fresh grapes - sparkling wine 260 000 litres Reduced by 100% of MFN 5%
Wine of fresh grapes - excluding sparkling wine I 000 000 litres Reduced by 100% of MFN 5%
Source: EC Council, 1999 HS - Harmonised System AGF - Annual Growth Factor
--
2.4 A regional approach of South Africa's trade liberalisation policy
South Africa is a member of the Southern African Customs Union (SACU) and played a
leading role in the renegotiation of the 1969 SACU Agreement, which was concluded in
2001. The principal objective of the SACU Agreement as renegotiated is to maintain free
interchange of goods between member countries and to apply the same tariff and trade
regulations to imports from outside the common customs area on a basis that sustains. the
economic development of all the member countries. The SACU Agreement provides for a
common external tariff and a common excise tariff to this common customs area. All
customs and excise duties collected in the common customs area are paid into South Africa's
National Revenue Fund. The revenue is shared among members according to a revenue-
sharing formula as described in the agreement.
The advantage of the SACU Agreement is to ensure the easy flow of trade in the area and
provide an extended market for South African goods, including agricultural products, to the
BLNS (Botswana, Lesotho, Namibia and Swaziland) countries. The arrangement guarantees
15
Trade agreements and the agricultural trade liberalisation
policy reform In South Africa
a substantial income for the BLNS countries and saves them the costs of administering border
control and the import duties. In addition, the BLNS countries enjoy rebate facilities on the
importation of certain agricultural products, i.e. maize, wheat and dairy products, from third
countries. They only receive rebates on such products if their tariff rates into the customs
union are high.
In addition, SACU members, i.e. South Africa and the BLNS countries, are members of the
Southern Africa Development Community (SADC) and this community established the Trade
Protocol that was signed in 1996 and implemented in 2000. This SADC Trade Protocol is
part of a more comprehensive regional agreement for economic and political co-operation and
development. This co-operation will eventually extend to the rest of the continent through
NEPAD, for which South Africa has already taken the responsibility of leadership.
One of the objectives of the SADC Trade Protocol is to liberalise intra-regional trade in goods
and services on the basis of fair, mutually equitable and beneficial trade arrangements,
complemented by Protocols in other areas, with a view of establishing a Free Trade Area in
the SADC region (SADC Secretariat, 1999). The objective target of the SADC Trade
Protocol was to have 85% of all intra-SADe trade at zero tariffs by 2008 and to have the
remaining 15% liberalised by 2012. In order to achieve this objective target, the main
instrument of trade liberalisation is therefore the elimination of customs tariffs and non-tariff
measures on substantial intra-SADC trade (Hansohm et al, 2004).
According to this Protocol, the elimination of import duties (tariffs) would be carried out in
three categories. Category A (Elimination upon implementation), Category B (Phased
elimination within eight years), and Category C (Phased elimination not extending beyond
twelve years). By definition, Category A represents commodities that already attracted zero
or low tariff levels; Category B represents commodities with high tariff levels that constitute
significant sources of customs revenue; and Category C represents the sensitive products
whose imports are considered to be sensitive to domestic industrial and agricultural activities.
In this case, Categories A and B should have accounted for 85% of intra-SADC trade so that
by 2008 SADC would have been regarded as a free trade area in compliance with GATT
Article 24, whereas Category C is limited to a maximum of 15% of between 2008 and 2012.
16
Trade agreements and the agricultural trade liberalisation
policy reform In South Africa
The information on commodities covered under these categories is obtainable from the Tralae
website (www.tralac.org)
In acceding to the Protocol, all members (except Angola, the Democratic Republic of Congo
and Seychelles who were not then parties to the free trade process) tabled their
implementation instruments or plans in the form of tariff phase-down schedules according to
the categories in question. The country-specific offers were based on the principle of
reciprocity in such a way that tariff preferences would only be extended to member states who
were party to the process (Hansohm et al, 2004). Tariff liberalisation is based on an
asymmetric offer approach, taking into account member states' levels of development, as
follows:
» SACU's combined offer to non-SACU SADC member states entails five-step elimination
of tariffs on all SADC imports with the first cut upon implementation of the agreement.
These included a reduction of import tariffs by 60% on industrial and agricultural products
imported from SADC countries: The proposed SACU offer included the following:
../ The immediate liberalisation list contains all products with tariffs rates from l-
17% so that 19.2% of all SADC imports and 19.2% of tariff lines are covered .
../ A three year linear phase down that includes products between 18-25% tariffs and
covers 6.3% of SADC trade and 26.4% oftarifflines .
../ A five year linear phase down that contains 6.6% of SADC trade, 3.9% of tariffs
lines and includes all products with tariffs above 25% .
../ South Africa considers the following as sensitive products: dairy, wheat and
meslin, sugar and sugar confectionary, textiles, foot wear and vehicles.
» Different offers by non-SACU SADC countries to SACU countries have also been tabled.
All countries committed themselves, however, to completing the elimination of tariffs on
most products by the end of year eight of the implementation period and for sensitive
products by the end of year 12. Table 2.2 shows various offers for specific agricultural
products by non-SACU SADC countries to SACU countries.
17
Trade agreements and the agricultural trade liberalisation
policy reform In South Africa
Table 2.2: Non-SAeu SADe countries' offers to SAeU countries: Tariff phase-down
schedule for agricultural products under SADe Trade Protocol
Commodities Malawi Mauritius Mozambique Tanzania Zambia Zimbabwe
Dairy 8 yrs 5 yrs except 12 yrs 8 yrs 12 yrs 8 yrs
bird eggs & except for
honey 12 yrs milk-8 yrs
Deciduous fruits 8 yrs 5 yrs, 8 yrs, citrus- 8 yrs 8 yrs 8 yrs
strawberries- 12 12 yrs
years
Dried fruits 8 yrs 5 yrs 12 yrs 8 yrs 8 yrs o from year 1
Canned fruits 8 yrs 5 yrs 12 yrs 8 yrs 8 yrs 5 yrs
Fruit juices 8 yrs 5 yrs 12 yrs 12 yrs 8 yrs 8 yrs
,.
Wines and spirits 8 yrs 12 yrs 12 yrs 12 yrs 12 yrs 8 yrs
Wool and mohair o from year 1 o from year 1 o from year 1 7 yrs o from year 8 yrs
1
Wheat o from year 1 o from year 1 8 yrs 8 yrs 5 yrs 8 yrs
Wheat flour 8 yrs 12 yrs 8 yrs 12 yrs 12 yrs 8 yrs
Source: SADC Secretariat, 1999
In addition, core non-tariff barriers such as quantitative import and export restrictions should
have been eliminated immediately, However, non-tariff barriers not related to standards,
sanitary or phytosanitary requirements but solely for the purpose of managing trade had to be
removed by the end of year eight. This included single channel marketing regimes, restrictive
visa requirements, and others. Sanitary and phytosanitary standards (SPS) within the SADC
would be harmonised. The Protocol currently provides for "one-product-multiple-rules" rules
of origin, i.e. more than one rule could apply to the same product. Following the objection of
SACU, which felt that these rules would create the possibility of significant trade diversion in
the region, agreement was reached that product specific rules of origin should be developed
on a chapter-by-chapter basis, and that the Protocol would be amended accordingly.
2.5 Summary
This chapter has provided an overview of various trade agreements that have implications for
agricultural sector trade in South Africa. The overview has emphasised that South Africa has
18
Trade agreements and the agricultural trade liberalisation
policy reform In South Africa
responded positively to the challenges posed by the globalisation processes. This is indeed
necessitated by its engagements with the international community through its commitments to
multilateralisation, bilateralisation and regionalisation processes. Its commitments to
international obligations have led to a policy paradigm shift in the agricultural sector that has
resulted in trade liberalisation and deregulation policies.
In the light of these developments, expectations of more open trade regimes in the agricultural
sector are rising owing to eliminated and less strict trade restrictions, the rationalisation and
simplification of the tariff regime, as well as the reduction of tariff rates. The next chapter
provides a review of the literature on the impacts of the trade agreements on the agricultural
sector and other sectors of the economy in South Africa, as well as those in other developing
countries.
19
CHAPTER3
LITERATURE REVIEW OF THE IMPACTS OF TRADE AGREEMENTS ON THE
ECONOMIES OF DEVELOPED AND DEVELOPING COUNTRIES
3.1 Introduction
As mentioned in Chapter I, the core objective of this study is to measure the impacts of trade
agreements (as discussed in Chapter 2) on agricultural trade between South Africa and its trading
partners. This chapter reviews the literature on the studies that analysed the impacts of various
trade agreements on developed and developing countries. Indeed there are studies that have
analysed the impacts of various trade agreements, such as the multilateral trade agreements in the
context of the World Trade Organization Agreement on Agriculture (WTO AoA) as well as
Regional Trade Agreements (RTAs) in the context of customs unions, .preferential treatments,
free-trade arrangements, etc. These studies have analysed the impact of such trade agreements on
the economic growth, employment, trade and welfare of the developed and developing countries.
When reviewing the literature of above case studies, this chapter will initially focus on the
implications of the WTO AoA on the economies of the developed and developing countries with
reference to South Africa and the region. Furthermore, the impacts of implemented regional
trade agreements between South Africa and the European Union countries (in the context of EU-
SA TDeA) as well as between South Africa and SADC countries (in the context of SADC Trade
Protocol) will also be reviewed. In addition, the study will also attempt to review case studies on
the impacts of selected implemented and envisaged trade agreements on the economies of other
countries worldwide.
3.2 Implications of the World Trade Organization Agreement on Agriculture
As elaborated in Chapter 2, the mam goal of the World Trade Organization Agreement on
Agriculture (WTO AoA), commonly known as the Uruguay Round Agreement on Agriculture
(UR AoA), was to liberalise world trade through the creation of a framework that would regulate
intemational trade and stimulate international commerce. The WTO AoA has been described as
20
Literature review of the impacts of trade agreements on the
economies of developed and developing countries
one of the main accomplishments of the marathon seven-year Uruguay Round. It was signed
with the objective of bringing discipline into one of the most distorted sectors of international
trade, by controlling the unrestricted use of production and export subsidies and reducing tariff
and non-tariff barriers on imports of agricultural products.
There is a consensus view that, ceteris paribus, economies that are open to trade will grow faster
than countries that are closed. Wang and Winters (1998) argued that this consensus is especially
strong with respect to Africa, where decades of import substitution are thought to be partially
responsible for the continent's dismal economic performance. Sharma et al (1996) state that the
WTO AoA is expected to cause beneficial effects for aggregate world income, as inefficiencies in
production and trade will be removed gradually, but it is generally agreed that the impact on global
trade would be fairly small over the implementation period, reflecting the limited extent of the .
reforms achieved. However, there is also a common assertion made by the critics of globalisation
that trade liberalisation increases poverty. The proliferation of low-wage jobs and higher food
prices are some of the arguments brought forward in support ofthis argument (Hertel et al, 2003).
The question is who is right? Various impact studies on globalisation and trade liberalisation
. have attempted to answer this question and a sample of them are quoted below.
Studies by Harrison et al (1997) and Whalley (2000) on the effects of the global trade
liberalisation have found that openness is associated with more rapid growth. They estimated
annual increases in global GDP due to reductions in tariff and non-tariff barriers. Moreover,
most of the gains accrue to countries (including especially advanced countries) that offered the
most reductions in tariff and non-tariff barriers. Detailed studies of trade liberalisation suggest
that the benefits to the economy as a whole are far more than the costs (Edwards, 1989; Matusz
and Tarr, 1999). However, using a gravity model, Anderson and Van Wincoop (2004) found that
trade costs are large when broadly defined to include all costs involved in getting a good from
producer to final user. Both international trade costs and local distribution costs are very large
and together dominate the marginal cost of production. Trade costs also vary widely across
countries. On average, developing countries have significantly larger trade costs.
21
Literature review of the impacts of trade agreements on the
economies of developed and developing countries
As quoted by Jooste (2001), studies by Harrison et al (1995) and Hertel et al (1997) have shown
losses amounting to 0.24 per cent and 0.13 per cent of the SSA's base GDP respectively in the year
2005 as a result of the reforms under the Uruguay Round. Harrison et al (1995) concluded that there
exists a large potential for improvement, or even reversal, of the situation through domestic pal icy
reforms which are stated to be necessary for taking advantage of the new trading opportunities
opened up by the Uruguay Agreement. The OECD (1998a), while also recording welfare losses as a
result of the implementation of the AoA, supports the view of Harrison et al (1995). Dimaranan et al
(2003) also found consistent results, as they found that an across-the-board 50% cut in all
domestic support for OECD agriculture leads to welfare losses for most of the developing
regions, as well as for the combined total group of developing countries. The 50% cut in
domestic support also results in large declines in farm incomes in Europe and, to a lesser degree,
North America. They concluded that developing countries will be well advised to focus their
efforts on improved market access to the OECD economies, while permitting these wealthy
economies to continue - indeed even increase domestic support payments.
Goldin et al (1993), as quoted by Jooste (2001), have estimated worldwide benefits due to
liberalisation in the order of US$190 billion with tariff reductions in the order of 30 per cent.
About US$70 billion of this total would accrue to non-OECD countries. The total gains would
increase to US$430 billion with full agricultural reform, with the gain for non-OECD countries in
the order of US$180 billion. With the levels. of tariffication agreed in the Uruguay Round
agreement, the gains are much smaller, particularly for those agricultural exporters who do not
subsidise their agricultural activities.
Hathaway and Ingco (1995) share his optimism by stating that despite the substantial retreat by
the advocates of liberalisation, the Uruguay Round agreement on agriculture appears to hold
great promise. Cognisance is taken of the fact that some of the binding powers laid down during the
Uruguay Round may be weak, but the essence is that new rules have been laid down to which role
players must adhere in future. This view is also supported by Ingco and Townsend (1998) who
mention that several studies that have attempted to measure the impact of the Uruguay Round on
agriculture have indicated their concerns about the high cost of complying with the Uruguay Round
obligations and the limits these may put on developing strategies, whilst others raised concerns
22
Literature review of the impacts of trade agreements on the
economies of developed and developing countries
about the potential market losses due to the erosion in the value of preferential exports, as overall
cuts in tariffs will reduce the value of the preferences.
It is well known that OECD countries afford their agricultural sectors a high degree of protection.
For example, IFPRI (2003) indicated that total support to agriculture in OECD countries
amounted to US$311 billion in 2001, or about US$850 million per day, dwarfing the amount
those same countries give in development assistance. This protection costs developing countries
about US$24 billion annually in lost agricultural and agro-industrial income. Trade distorting
measures also displace more than US$40 billion of net agricultural exports per year from
developing countries. IFPRI (2003) also found that elimination of protectionism and subsidies of
the industrialized world's agriculture would triple developing countries' net agricultural trade.
This is indeed supported by Hertel et al (2000) who found that agricultural liberalisation in the
wake of the UR, i.e. 40% cuts in both market price support and domestic producer subsidies,
could still yield substantial benefits for the global economy in 2005 and the total gains amounted
to about US$70 billion. However, as Hertel and Martin (1999) have pointed out, the distribution
of these gains is quite different in the sense that while the rates of protection are higher in the
industrialized economies, they are the ones to capture the majority of the absolute gains from
liberalisation of food markets. However, when measured relative to initial income, developing
countries are also some of the biggest winners from cuts in agricultural protection.
Using a GTAP model, Diao et al (200 I) estimated the welfare effects that would occur if OECD
countries removed tariff, export subsidies, and domestic support. An interesting aspect of Diao et
aI's paper is that they calculated the contributions of each policy reform to the overall price
change. For instance, they find that the elimination of domestic support in OECD countries
would account for 30% of the rise in prices that would occur if all agricultural trade distortions
were removed. They also find that although the removal of OECD tariffs would create small
welfare gains for developing countries, the removal of OECD domestic support and export
subsidies would create a small loss for developing countries. These results are consistent with
the recent study by Rae and Strutt (2003), who apply four simulations of agricultural policy
reforms to an updated version of the GTAP model. For each of the three simulations that involve
the reduction of domestic support, they find that this reduction contributes negatively to the
23
'; '.' .;.
Literature review of the impacts of trade agreements on the
economies of developed and developing countries
overall welfare effect of that policy reform. In addition, Rae and Strutt (2003) predict that
developing countries' welfare would increase by US$2 billion if developed countries increase
Blue and Green Box domestic support. Taken as a group, the recent general equilibrium studies
suggest that the removal of OECD agricultural tariffs would benefit developing countries, but the
removal of OECD domestic support would not be beneficial.
Panagariya (2002) noted that 48 out of 63 low-income countries are net importers of food, and
that 31 of the world's 46 least developed countries are net importers of both food and agricultural
products. Further, of the 41 developing countries that are net exporters of agricultural products,
22 are net importers of food. If cereals prices increase, the welfare of net importers of food will
decline. Given this scenario, Panagariya (2002) argues that because agricultural price increases
benefit exporters but not importers, the benefits of OECD trade liberalisation would accrue to
middle-income developing countries in Latin America and Asia, who are actual or potential
exporters of currently-subsidized products. In contrast, least developed countries, which are
more likely to be net food importers, will. not as a group benefit from OECD agricultural
liberalisation. The predominance of low-income and net food importing countries in Africa
implies that the region would see a decrease in welfare if cereals prices increase with the
reduction ofOECD agricultural trade distortions.
The empirical work by Dimaranan et al (2003) supported Panagariya's argument, as they applied
the GTAP model to simulate various scenarios of OECD agricultural reform and predicted that
Sub-Saharan Africa would lose US$126.1 million, or .42%, if OECD countries halve domestic
support to agriculture. Furthermore, they suggested that this loss is due to many African
countries' status as net importers of subsidized agricultural products. Instead of building from
the GTAP model, Soledad Bos (2003) used a partial equilibrium approach and found that OECD
subsidy reduction would lead to welfare losses in African countries. Soledad Bos (2003) also
calculated changes in consumer and producer surplus that would occur in the maize markets of
five African countries if OECD countries were to reduce domestic support to agriculture by
100% or 50% and found negative net welfare changes in each of the countries she studies:
Uganda, Kenya, Zimbabwe, Botswana and Mozambique. Although this is an interesting result,
the scope of this study is rather narrow as it was limited to five countries and one food crop.
24
Literature review of the impacts of trade agreements on the
economies of developed and developing countries
Agriculture remains much more important in the economies of developing countries than it does
in the high-income countries. According to IFPRI (2003), an agricultural-led growth strategy
may produce greater multiplier effects for the rest of the economy than other alternatives in the
world's poorest countries. Furthermore, increased profits from agriculture encourage expanded
economic activity, causing dynamic effect in four areas, i.e. employment, land, capital and
technology. Developing countries remain small net exporters of agricultural commodities.
Further, consumers in developing countries spend over 30% of their incomes on food, which is
almost three times the share in industrial countries, making them much more vulnerable to
shocks. Agriculture's contribution to GDP in developing countries is also around three times as
high as its share in industrial countries (Hertel et al, 2000).
The major effects of international agricultural trade liberalisation will be higher prices and an
allocation shift in production .. A reduction in export subsidies will also raise the prices paid by
the importers (Bade, 1998). The developing countries have to open their domestic markets to
price signals in the world markets as part of their overall economic policy reforms, market
liberalisation, and market privatization. Therefore, they are more exposed than before to the
effects of price instabi Iity in the world market (Islam, 1996). Another effect of the agreement on
developing countries will be that they will be affected by a reduction in price support, which will
lead to a reduction in food surpluses and stocks in developed countries, and hence, a fall in food
aid availability (Karim and Kirschke, 2002).
Huber & Lehmann (2009) have analysed the consequences of world market prices for agricultural
production and the land-use patterns in the Swiss lowlands using a mathematical programming
model. Given a sufficient reduction in production costs, their results imply that income
maximizing farmers would focus on grassland based milk production. This would only lead to a
modest change in the existing landscape since their case study region was dominated by dairy
farms. If production costs remain high, agricultural production would shift to more extensive
production activities in order to maximize the sectoral income. However, if a certain level is
exceeded, farmers would merely cease production and cultivate their land in order to get direct
payments. This would change the land-use patterns considerably. The main driving forces
behind this development are the implementation of the direct payment system and the farmers'
25
Literature review of the impacts of trade agreements on the
economies of developed and developing countries
possibility to reduce their production costs, in particular, by means of structural change which
would result in more productive farms.
While the inclusion of agriculture into the multilateral trade rules during the Uruguay Round was
one of the biggest achievements of the GATT, Ingco and Townsend (1998) argued that its impact
on African countries was a subject of much controversy. Several studies raised concerns about
the high costs of complying with the obligations from the Uruguay Round and the limits these
may put on development strategies (UNCTAD, 1994; Weston, 1994; Konate, 1994; Greenaway,
1994). Other concerns relate to potential market losses for African countries from the erosion in
the value of its preferences in its export markets as overall cuts in tariffs will reduce the value of
the preferences (Davenport et al, 1994) and from terms-of-trade loses due to potentially higher
foods prices to net importers of food (GATT, 1994) as export subsidies are reduced.
This is also supported by Peacernaker-Arrand (2004) who found that the removal of wealthy
countries' subsidies would lead to welfare losses for most African countries, although the net,
effects are a small percentage of GDP in each country For example, the removal of developed
countries' subsidies for wheat and maize is more likely to have negative rather than positive
effects in most countries in sub-Saharan Africa - even for farmers. Although some countries
could benefit from the removal of subsidies in certain non-food crops, such as cotton, it is not in
the interest of most sub-Saharan African countries to pursue cuts in developed countries' cereals
subsidies as part of the next WTO agricultural agreement. Those who have proposed cuts in
developed countries' subsidies argue that because agriculture makes up a large share of
developing countries' economies, and because the majority of most developing countries'
populations are farmers, a decrease in wealthy countries' subsidies would benefit poor farmers
and developing countries. Peacemaker-Arrand (2004) concluded that the argument for subsidy
removal has two major flaws: it does not acknowledge the negative effect of agricultural price
increases in net-food-importing developing countries, and it assumes that poor farmers would
necessarily benefit from price increases in the goods they produce. In fact, because of the
substantial margins between producer and consumer prices, households that sell a greater
quantity of grain than they purchase may spend more on grain they purchase than they received
26
Literature review of the impacts of trade agreements on the
economies of developed and developing countries
in income from the grain they sold, because grain is priced higher when sold to consumers than
when purchased from farmers.
However, other studies (Sorsa, 1995) have shown that the Uruguay Round is unlikely to "burden"
Sub-Saharan African (SSA) countries with many new obligations and that most countries in the
region did not make meaningful liberalisation commitments in the Uruguay Round. Thus, by
resisting liberalisation and the opportunity to anchor domestic reform in an international
framework, Sub-Saharan African countries have forgone an opportunity to reap substantial gains
from the Uruguay Round (Ingco and Townsend, 1998). This is because, while the UR made
significant efforts to improve market access conditions, the general consensus after the UR was
that African countries did not go far enough in implementing their commitments, e.g. lowering
their bound duty rates (Ingco, 1995; Harrold, 1996). According to Hoekman (2002), this is due
to the fact that negotiators from the developing countries signed documents that most of them did
not fully understand. As a result, not much progress was made; in particular on market access for
textiles and also most of the developing countries did not anticipate the enormous burden of
implementing some of the WTO agreements. Most developing countries believe that the UR did
not produce fruitful results (Adhikari, 2000). This is supported by Ndirangu (2002), cited by
Makhura and Mokoena (2003), who argued that the Uruguay Round does not focus on addressing
the development needs and concerns of the majority of farmers in developing countries,
particularly in Africa.
Ingco and Townsend (1998) have also argued that while the Uruguay Round addressed the worst
distortions in world agriculture, it left many agricultural policy distortions especially in low-
income African countries outside its scope. Most of the African countries do not subsidize, but
tax agriculture either implicitly by giving higher protection to industry, or more explicitly by
taxing exports of many commodities or by maintaining government controlled domestic prices
below world prices (Schiff and Valdes, 1992). These distortions were not part of the Uruguay
Round agenda, and some of them are not even covered by the GA TT (examples include export
taxes or domestic pricing policies that "tax" agriculture). Actual liberalisation of industrial tariffs
in Africa within the Uruguay Round is also modest and did not reduce the existing bias against
agriculture. African trade policies have also suffered from frequent policy reversals and from the
27
Literature review of the impacts of trade agreements on the
economies of developed and developing countries
impact of exchange restrictions on trade flows. This is because, while the UR AaA sets rules on
international food trade and on domestic agriculture policy, these rules have accelerated the rapid
concentration of agribusinesses and undercut the ability of the poor countries to maintain food
self-sufficiency through subsistence agriculture. The UR AaA assumes that rather than being
self-sufficient in food, countries will buy their food in international markets using money earned
from exports. However, many less developed countries face low commodity prices for their
limited range of exports. During the first four years of the existence of the WTO, the prices of
agricultural commodities fell to record lows, while food prices remained high. This system hurts
both farmers and consumers, and paves the way for Trans National Companies (TNCs) to
dominate markets, especially in the poor countries. Therefore, the study d that rules are needed to
address the rapid concentration of TNCs in agribusiness because a small handful of companies
trade virtually all the world's. corn, wheat and soyabeans. This increased consolidation of
agribusiness in the hands of a few TNCs has led to near monopoly conditions in both the farm
supply industry and in the food processing and distribution systems.
. .
In support for above arguments, Ndirangu (2002) also d that the UR AaA tends to favour farmers
in developed countries rather than in Africa. It does not allow farmers in Africa to receive any
form of support other than that listed under the Green Box and the de minimis percentage of 10%
of agricultural production. Developed countries, which provided considerable support to their
farmers before the WTO was established, have continued to support them. These nations are
permitted to invoke the special safeguard clause, which allows imposition of additional duties in
case of increased imports without proof of injury in the domestic market, while African countries
do not have such recourse as they agreed to a ceiling binding during the Uruguay Round.
Agricultural trade has been more protected in the developed countries after the conclusion of the
Uruguay Round negotiations, while in Africa the farm sector became more exposed to external
forces and subsidized exports from developed nations. This has had adverse effects on
production and food security in the African continent. This is supported by Stevens et al
(undated) who indicated that the impact of UR AaA on food security will be negative due to fact
that it will alter world market conditions for agriculture with the likelihood of provoking changes
in both the level and the distribution of supply and demand. This will, in turn, alter the prices
that some countries receive for their exports and pay for their imports.
28
Literature review of the impacts of trade agreements on the
economies of developed and developing countries
Hertel etal (1998) evaluated the effects on Africa of tariff reductions in manufactures, textile and
clothing, and agriculture tariffs agreed under the Uruguay Round. Using the GTAP data they
found that the limited gains from the Uruguay Round in Africa are mainly because Africa does
not ease its trade restrictions as much as other countries, and so, as they put it, world trade
"bypasses the continent." Unlike textiles and clothing, which would suffer most as a result of the
Uruguay Round, production of cereals, non-grain crops, forestry and fish products would expand.
After simulating the domestic reforms in both trade and transportation sectors and in food grain
productivity, they discovered that, in both sectors, Africa lags significantly behind other low-
income countries, and institutional reforms could provide major gains at low cost. Tsigas and
Ingco (2001) also used the GTAP framework to assess the implication of improvements in
market access through quota expansion and lowering of in- and out-of-quota tariffs and found
that policy reform agreed in the UR Agreement on Agriculture and continuation of such reforms
would lead to significant gains for the world as a whole and for most regions.
This is supported by Pustovit and Schmitz (2003), who used a multi-commodity multi-country
comparative static trade model to analyse the impact of agriculture protection in OECD countries
on South African Agriculture. They found that developing (importing) countries could gain from
liberalisation of the OECD countries' agricultural policies if the disincentive effects of
production are taken into consideration and own policies are adjusted to more open markets (see
also Anderson et al, 200 I; Beghin et al, 2002; Diao et al, 2002; Hoekman and Anderson, 2000;
Martin and Winters, 1996). Furthermore, they indicated that liberalizing both OECD countries'
policies and South African agricultural policies could be the best way of contributing to
agricultural development and avoiding poverty and hunger. . hi addition, industrialized countries
would save money (welfare gain) which could be spend directly as development aid and
concluded that South Africa could benefit a lot by liberalizing trade and agricultural policies
world wide. These results are consistent with those of Chant et al (200 I), who used a computable
general equilibrium (CGE) model calibrated to a social accounting matrix (SAM) for South
Africa to assess the impact of agricultural trade liberalisation on agricultural productivity growth
and employment in South Africa. They found that agricultural trade liberalisation would have
positive impacts upon the economy, but that the extent of the benefits is dependent on whether
agriculture can remain competitive by improving productivity.
29
Literature review of the impacts of trade agreements on the
economies of developed and developing countries
However, Roberts (2000) found that liberalisation of trade in South Africa did not yield the
expected gains from incentives to export during the 1992-1997 period. Instead, while
manufacturing exports and imports increased, output growth faltered in most sub-sectors and
there were major reductions in employment. In many sub-sectors, improved trade performance
was associated with contractions in production and employment, while trade performance
deteriorated in sub-sectors with increasing employment. Furthermore, Kusi (2002) used a time
series regression model to analyse the impacts of trade liberalisation on South Africa's export
performance and found that there is a lack of a clear relationship between trade reforms and
improved export performance of the major sectors of the economy, such as finance and
insurance, agriculture, gold and uranium. In fact, external market conditions were the important
determinant of export performance across all sectors during the sample period.
The study by Jooste et al (2001) used a spatial partial equilibrium model to analyse the effect of
tariff reductions on the red meat industry in South Africa and found that the prices of livestock
and meat will drop substantially, whilst increased demand will be met largely by imports. The
consumers' welfare gains amounted to R2 829 million that translated to 0.49% increase in the
real gross national income. However, tariff reductions led to a substantial welfare loss by
producers of about R868 million that translated to about 2.71 % of real gross farm income and
10.72% of real net farm income. This is supported by Oyewumi et al (2007) who used a partial
equilibrium comparative static model to measure the welfare effects of further liberalisation in
the livestock industry of South Africa, particularly in meat products using four policy scenarios.
They found that a complete removal of tariffs on consumers will result in a welfare increase of
RI 880.8 million, which amounts to 0.33 per cent increase in real gross national income or 0.50
per cent increase in real disposable income. Whereas on the producers side, the welfare will drop
by R656.89 million, which represents a drop of 2.05 per cent in real gross farm income or 8.1 per
cent in net farm income. In this case, the tariff and Tariff Rate Quota (TRQ) liberalisation will
result in net welfare gains to society, but the impact on the agricultural sector would be much
more substantial in relative terms. They recommended it is worth considering the effects on
producers if further trade liberalisation is envisaged in the South African livestock industry.
30
Literature review of the impacts of trade agreements on the
economies of developed and developing countries
Apart from the UR AoA, there are also several studies that predicted the potential impacts of the
new Doha Round (OH) of the WTO on the economies of the developing countries and a sample
of them are provided here. The study by Fabiosa et al (2003) used a partial equilibrium model of
world agriculture to investigate the multilateral removal of all border taxes and farm programs
and their distortion of world agricultural markets. They found that net agricultural and food
exporters (i.e. developed countries) emerge with expanded exports, whereas net importing
countries (developing countries) with limited distortions before liberalisation are penalized by
higher world markets prices and reduced imports. On the other hand, Poonyth et al (2004) used
an agricultural trade policy simulation model (ATPSM) to assess the likely impact of the draft
Harbinson modalities/ for further commitment, along with the EU proposal ' and US proposal" on
the agricultural sector of the SADC countries. Their results showed that SADC as a group loses
in term of total welfare under all the three proposals and it is more pronounced under the
Harbinson proposal. In the case of the Harbinson and EU proposal, the loss is due to decrease in
consumer surplus and decrease in government revenue. Whereas in the case of the US proposal,
the loss in total welfare is due to a decrease in producer surplus and in government revenue.
The above findings by Fabiosa et al (2003) and Poonyth et al (2004) supports Short (2003) who
stated that "multilateral trade liberalisation is an indispensable part of development, but trade
alone is not an answer to poverty reduction, however, is one key driver of economic growth".
With these words, Short (2003) was trying to that without effective states with effective
institutions that pursue pro-poor policies, the poor will see little benefit from the trade
liberalisation of the OH and concluded that Doha Round is not only about helping the world's
poor developing countries, but it is also in developed countries' own self-interest and therefore
2 The Harbinson modalities proposed different reduction rates for developed and developing countries depending on the level of
the initial tariff. On market access, the Harbinson text proposes that countries be required to cut tariffs tariffs by a simple average,
except in-quota tariffs, for all agricultural products. Countries may reach this average in any way, subject to a minimum reduction
per tariff line, using bound tariffs as a base. Using a banded approach, tariff reductions shall be implemented in equal annual
instalments over a period of five years for developed countries and ten years for developing countries. In addition, the modalities
propose the elimination of export subsidies over a period of 9 years.
3 The key US proposal on tariff was the use ofa harmonisation formula that would reduce higher tariffs more deeply than lower
tariffs. For this, the Swiss formula was proposed meaning that all tariffs are reduced to below 25%. The other key proposal was to
apply the formula to applied tariffs.
4 The EU proposal was for the continuation of the UR approach, i.e. 36% average reduction of bound rates with a minimum 15%
cut for each tari ff Iine.
3 I
Literature review of the impacts of trade agreements on the
economies of developed and developing countries
both developed and developing countries stand to gain from a rules-based multilateral trading
system.
Surprisingly, Nyhodo et al (2009) used a static computable general equilibrium (CGE) model to
analyse the potential impact of Doha Development Agenda on the South African economy and
found that the South African economy would respond positively to world price changes, with
government and macro variables showing minimal but positive responses. Furthermore,
household consumption expenditures generally show positive changes, implying increased factor
incomes. They concluded that the overall effect is positive even though not all sectors will be
positively affected.
3.3 Implications of the EU-SA Trade, Development and Cooperation Agreement
. The EU-SA TDCA is a reciprocal agreement that entails the liberalisation of tariffs on 95% of
EU imports from South Africa over alO-year period and 86% of tariffs South Africa imports
from the EU over a 12-year period. There are several studies that have evaluated the impacts of
the EU-SA TDCA on trade, economic growth, employment, welfare, etc. For example, Davies
(1998) simulated a Free Trade Agreement (FT A) between the EU and South Africa using the
GTAP model and found a strong potential for trade diversion following an FTA. His study
showed that the FT A would cause a switch from cheaper sources to less-efficient EU products.
However, Laaksonen (2008) argued that there were no very strong signs of trade diversion to
trade between the EU and South Africa on the cost of trade with the rest of the world because the
significance of free trade between South Africa and the EU was difficult to pinpoint in the overall
trade picture during the early years of the new century, since there have been many other very
significant changes in the operating environment, including the economic rise of Asia and the
increase in raw material and energy prices. Laaksonen's argument lacked substance as it was not
supported by any substantial evidence. In fact, some studies found that both South Africa's
agricultural exports to the EU as well as EU's agricultural exports to South Africa for the period
2000 and 2009 have done well and concluded that the EU-SA TDCA has been a factor towards
this success (Sandrey, 20 10).
32
Literature review of the impacts of trade agreements on the
economies of developed and developing countries
Mokoena et al (2007) found consistent results with those of Davies (1998) as they also found that
the implementation of EU-SA TDCA resulted in the diversion of South Africa's agricultural
exports to other markets by about 0.51 % due to negative response of South Africa's agricultural
exports to the EU during the early stage of the EU-SA TDCA's implementation, but concluded
that the EU-SA TDCA has created a room for potential increase of the South African agricultural
exports in the EU market. This is in line with the findings of Sandrey (2006) who found that
while TDCA increased South Africa's imports from the EU by some R4.3 billion, about R2.7
million was trade diverted from other sources.
Furthermore, Mokoena et al (2008) used a gravity model to analyse the impacts of the EU-SA
TDCA's reciprocal preferential tariff quotas on cheese and wine trade flows and still found a
trade diversion on all wine trade flows, i.e. about 0.8% of exports, 2% of imports and 2.3% of the
total trade were diverted during the wine tariff quota implementation. However, there was no
proof of trade creation and diversion on both cheese imports and exports, except that there was a
diversion of about 4.1 % of the total cheese trade. In addition, their results showed that South
Africa's cheese exports to the EU and total cheese trade between the parties had declined during
the implementation of cheese tariff quotas. South Africa's wine exports to the EU and South
Africa's wine imports from the EU had also declined during the implementation of the wine tariff
quotas. However, the effects were insignificant on South Africa's cheese imports from the EU
and total wine trade between the two parties.
In contrast, Andriamananjara and Hillberry (2001) also applied the GTAP framework to analyse
the EU-South Africa FTA and found trade creation as the net effect as both South Africa's
exports and imports increased. In addition, their study incorporates dynamic effects of trade and
growth, adding estimates of the links between trade openness and total factor productivity (TFP)
shocks for South Africa. They found that the trade-induced growth is two percent of total growth
over the phase-in period. Also using the GTAP model, Penzhorn and Kirsten (1999) analysed the
impacts of the EU-SA TDCA on South African agriculture and found that both South Africa and
the EU will experience welfare gains as a result of the agreement. Furthermore, they determined
that the exports of dairy products to the EU would increase by another 35% while exports of
33
Literature review of the impacts of trade agreements on the
economies of developed and developing countries
vegetables and fruit, and other agricultural products will also increase by 25% and close to 30%,
respectivel y.
Similarly, Kalaba (2001) also analysed the effects of the EU-SA FTA on South Africa agriculture
with special reference to the competitiveness of fruits (i.e. grapes, pears and apples) in the EU
market using a source differentiated, almost ideal demand system (SOAIDS) model and found
that South African fruit exports to the EU are at least competitive among the selected suppliers,
i.e. US, Chile, Turkey and New Zealand. In addition, there was evidence of complementary
relationships between South African apples and those from the US, and that South Africa faces
strong competition in grapes from Chile and the US. However, South Africa's trade
liberalisation appears to have increased the exports of grapes to the EU. This was also supported
by Gay (2004) who used a trade simulation model to analyse the implications of the EU-SA
TO CA on the fruit trade and found that the EU-SA ToCA had a slight beneficial effect for South
African or~nge producers due to the small tariff cut for fresh oranges.
At the regional level, Mcfronald and Walrnsley (200 I) used the GTAP framework to analyse the
impacts of the EU-SA FTA on Botswana and they found that while the FTA may substantially
benefit the signatories, there are appreciable negative impacts for other states, especially South
Africa's immediate neighbours. Moreover, the reluctance of the EU to fully liberalise trade in
food and agriculture commodities results in a major reduction in the benefits for South Africa
without ameliorating substantively the adverse implications for other nations. Similarly, Tsolo et
al (2010) examined the impact of the EU-SA ToCA on trade patterns between the South Africa
and Botswana, Lesotho, Namibia and Swaziland (BLNS) using a random effects model and
found that the demand for imports by the BNLS countries is income elastic and price inelastic
implying that the imported goods from South Africa are necessary and consumers and producers
of the BLNS countries depend on them. Furthermore, they found that the EU-SA ToCA brought
about increased imports to the BLNS countries and that the volume of exports to South Africa
from the BLNS countries had increased after the agreement. These findings implied that imports
could have led to a crowding out of domestic production as a result of the EU-SA ToCA as well
as that the EU-SA ToCA could have benefited the BLNS countries by increasing their exports.
34
Literature review of the impacts of trade agreements on the
economies of developed and developing countries
3.4 Implications of the SADC Trade Protocol on Trade
The main objective of the SADe Protocol on Trade is to liberalise intra-regional trade in goods
and services on the basis of fair, mutually equitable and beneficial trade arrangements with a
view of establishing a Free Trade Area in the SADe region. There are several studies that have
evaluated the impacts of the SADe FTA on trade, economic growth, employment, welfare, etc.
Using a cross section econometric gravity model, Cassim (200 I) looked at the potential for trade
among SADe countries and found that specific areas where potential trade is less than actual
trade are mostly South African and Zimbabwean exports to the region. In case of South Africa,
in all instances, its potential exports are significantly lower than its actual exports.
Chauvin and Gaulier (2002) also used the gravity approach and found consistent results with that
of Cassim (200 I) in the sense that they found that South Africa's actual exports are all above
potential exports with other SADe countries. Nevertheless, even though other SADC countries'
combinations show some potential trade higher than actual trade, they seem nevertheless smaller
compared to Cassim's results. Surprisingly, findings by Poonyth et al (2002) are inconsistent
with the above. They also used the gravity model to evaluate the potential for trade integration in
the context of both the structural factors and growth behaviour of the region. They found that
South Africa's potential exports to selected SADC countries (i.e. Mauritius, Mozambique,
Malawi, Tanzania, Zambia and Zimbabwe) are significantly higher than its actual exports due to
the SADe FTA and concluded that there is room for improvement in the trade gap between
South Africa and its SADe partners.
A study by Diao and Robinson (2003) showed that the elimination of agricultural tariffs among
SADe countries would benefit real agricultural GDP in the region, national income and
agricultural output. Studies by Lewis (200 I) and by Lewis et al (2002) using computable general
equilibrium (CGE) modelling examined the impact of a FTA on SADC economies. They
concluded that the gains that can be achieved through trade expansion are limited given SADC's
small size relative to the global economy and the trade imbalances among its members. Nin-Pratt
et al 2008) added that the largest share of the gains would go to Zimbabwe, SACU, Malawi,
Mauritius and Tanzania, while Angola and the ORe would be negatively affected by the
35
Literature review of the impacts of trade agreements on the
economies of developed and developing countries
agreement. In addition they found that countries that benefit the most are those with a
comparative advantage for agriculture in the region (e.g. Zimbabwe), while still being inefficient
producers of regionally traded commodities. The inefficiency of the main regional exporters also
explains the negative welfare impacts of the agreement on countries with comparative
disadvantage in the region (net importers), like Angola and the ORe. This is because the
elimination of tariffs on regional imports in these countries would increase imports of wine, beer,
meal and flour of wheat; preparation of cereals, sugar and bakery products from inefficient
regional producers, with trade diversion dominating trade creation. It means inefficient
agricultural producers with a regional comparative advantage for agriculture would benefit from
trade creation with the rest of the world and as a result regional importers would be faced with
negative welfare effects because of increased imports from inefficient regional producers.
Evans (2001) assesses trade options for SA DC countries namely, an FTA, a Customs Union, or
open regionalism, by which SADC countries extend tariff reductions to all countries on a most
favoured nation (MFN) basis and concluded that trade creation dominates trade diversion in an
.FTA as intra-SADC trade increases by 9 percent while trade with the rest of the world (ROW)
.hardly changes. With free trade, trade creation was observed as SADC trade expanded by almost
7 percent, but with potential terms of trade costs. These results concur with the findings by Nin-
Pratt et al (2008) who estimated a total value of trade creation of$157 million or 0.92 per cent of
annual agricultural trade (from 2000 to 2005) of SADC countries, and a net effect between trade
creation and trade diversion of $129 million or 0.75 per cent of total agricultural trade.
In contrast, Mokoena et al (2007) analysed the ex-post impacts of the implementation of SADe
FTA on South Africa's total agricultural exports to SADe countries for the period 2000 to 2004
using a gravity model and found that the implementation of the SADC FTA resulted in the
diversion of South Africa's agricultural exports to other markets by about 0.5% Furthermore,
their results indicated that South Africa's agricultural exports responded negatively to the
implementation of the SADe FTA from 2000 to 2004, but concluded that the FTA has created a
room for potential increase of the South African agricultural exports in SA DC market. These
results are similar to the findings by Holden (1996) who indicated that regional trading blocs
36
Literature review of the impacts of trade agreements on the
economies of developed and developing countries
such as SADC encourage import substitution industrialisation and suggested that South Africa's
participation in a FTA would lead to trade diversion.
A study by Lewis et al (1999), cited in Poonyth et al (2002), concluded that the SADC FTA in
conjunction with the EU-SACU FTA improves welfare of all SADC countries. This is consistent
with the findings of Nin-Pratt et al (2008) who assessed the potential welfare impacts of a free
trade agreement (FT A) on the agricultural sector of Southern African countries and found that the
overall welfare effects of a FTA would be positive but small in most countries. At the country
level, they estimated that two-thirds of region-wide welfare gains from agricultural trade
liberalisation would go to low-income countries while almost one-third would go to SACU.
However, it should be noted that there are some complications brought about by the overlapping
nature of FTAs in the Southern African region, for example the EU-South Africa FTA has put a
strain on initiatives under way to form a free trade area within SADC and also puts into question
the continuing viability of SACU (Lewis, 2001). Added to this, Tsikata (1999) found that the
overlapping nature of membership of Southern African countries in many other schemes raises
questions over the consistency and feasibility of satisfying the conflicting obligations and
tensions among various institutions and their members.
In analysing the possible impact of various preferential trade agreements on South Africa and the
rest of SADC, Lewis (2001) concluded that while promotion of a SADC FTA will yield benefits
to all participants, SADC's small size relative to the global economy and the trade imbalances
among its members will likely limit the medium-term scope for trade expansion. However, he
observed that South Africa gains more from the FTA with the EU than it will from a SADC FTA,
and for the rest of SADC, the gains from greater access to the EU are proportionately even larger.
In a similar vein, Holden (1996) observed that South Africa has little incentive to seek
preferential treatment in the region, largely because of the economic divergence between it and
other countries in the region and because South Africa's share of regional exports remains small
relative to its exports to the rest of the world. Various studies using a gravity model have also
shown that the implementation of the FTA in SADC would have favourable effects on bilateral
trade (Longo and Sekkat 200 I; Subramanian and Tamirisa, 2001).
37
Literature review of the impacts of trade agreements on the
economies of developed and developing countries
In addition, Lewis (200 I) also used a World Bank database on tariff schedules for ten SADC
members, namely, SACU members, Malawi, Mauritius, Tanzania, Zambia and Zimbabwe, to
evaluate the possible impact of the proposed SADC FTA from the point of view of tariff
harmonisation and fiscal effects. The results indicated that fiscal considerations are likely to be
important in any form of SADC FTA and concluded with an observation that any regional trade
arrangement will involve differential benefits and losses among the individual countries,
suggesting that redistributive issues such as unemployment and income distribution should be
dealt with as first priority to increase chances of a viable arrangement.
Studies have also argued that the limited role that the SADC FTA could play in the region results
from the fact that tariffs are not the only obstacle to increased regional trade. To explain low
trade in Southern Africa several studies have stressed the importance of transport and transaction
costs, inadequate infrastructure, lack of diversification in sources of comparative advantage and
underdeveloped production structures (see for example, Cassim 2000; Chauvin and Gaulier 2002;
Davies 1996; Geda and Kibret 2002; Goldstein 2004; Holden 1996; Jenkins et al, 2000; Longo
and Sekkat 200 I; Nyirabu 2004; Radelet 1997). Furthermore, Nin-Pratt et al (2008) suggested
that the SADC region should be looking at regional policies and interventions beyond trade
arrangements, such as those targeting investment, agricultural productivity and diversification in
order to enhance benefits of regional trade liberalisation.
3.5 Implications of the Selected Trade Agreements in the World
Since early 1990s, Regional Trade Agreements (RTAs) have gained momentum and became a
very prominent feature of the Multilateral Trading System (MTS). As of 15 May 2011, about
489 RTAs were notified to the GATT/WTO of which 358 RTAs were notified under Article
XXIV of the GATT 1947 or GATT 1994; 36 under the Enabling Clause; and 95 under Article V
of the GATS. At the same date, 297 agreements were in force. Of these RTAs, Free Trade
Agreements (FT As) and partial scope agreements account for 90%, while customs unions
account for 10% (WTO website, 2011). This section reviews the literature on trade and welfare
impacts of selected RTAs.
38
Literature review of the impacts of trade agreements on the
economies of developed and developing countries
Fulponi et al (20 II) analysed the treatment of agriculture by RTAs and found that almost 60% of
the RTAs prohibit agricultural export subsidies and indicated that this should be seen as going
beyond the WTO-AoA commitments, thus making them "WTO-plus". Furthermore, they found
that countries which do not currently use export subsidies were a significant portion of those
prohibiting them in their RTAs. Overall, the RTAs examined do provide for increased trade
liberalisation compared to commitments under the WTO AoA, as evidenced by tariff elimination,
commitments on export subsidy elimination and sunset clauses on special agricultural safeguards.
However, few concrete commitments are found with respect to non-tariff measures such as
Sanitary and Phyto-sanitary (SPS) and Technical Barriers to Trade (TBT) measures.
On the other hand, Wainio et al (2011) examined the implications of selected trade agreements
(TAs) on U.S. agricultural trade with reference to recently concluded TAs between ASEAN
(Southeast Asia) countries and China and Australia/New Zealand, as well as pending TAs
between the United States and Korea, Colombia, and Panama. Their results suggest that TAs
between ASEAN countries and China and ASEAN countries and Australia! New Zealand would
result in moderate losses to U.S. agricultural exports of about $350 million to those countries, but
losses would be partially offset by gains in other markets. U .S. agricultural exports to Korea
would expand by an estimated $1.9 billion per year if the US-Korea TA were implemented. The
US-Colombia TA would result in an estimated $370 million in additional US exports per year.
US exports would realize smaller gains of about $50 million per year under the pact with
Panama. Empirical results confirm theoretical findings that trade created under TAs exceeds trade
diverted, but that results depend on the specific circumstances of each agreement. This study
finds that in the case of the recently implemented FTAs between the ASEAN countries and
China, Australia, and New Zealand, the potential for U.S. agricultural exporters to be affected as
a result of trade diversion is modest. This is because a large portion of U.S. exports to these
countries already faces duty-free or minimal tariffs. Where tariffs are significant, the United
States often faces only minimal competition from FTA members.
In contrast, the analysis of the pending U.S. trade agreement with Colombia reveals a different
story. Colombia has been active in negotiating additional FTAs with some key U.S. competitors.
One of these FTAs, with the four members of MERCOSUR, has already had significant effects
39
Literature review of the impacts of trade agreements on the
economies of developed and developing countries
on U.S. agricultural exports and market shares in the Colombian market. A majority of U.S.
agricultural exports to Colombia compete head-to-head with exports from MERCOSUR
countries. The U.S. market position in Colombia could soon be further eroded if the Canada-
Colombia FTA and the EU-Colombia FTA enter into force and competing exports from these
countries receive duty-free treatment. Duty-free access to the Colombian market would help
maintain and improve the competitive position of U.S. producers (Wainio et al, 2011).
Reeder, Torene, Jabara and Babula (2005) analysed the effects of the ANDEAN and
MERCOSUR pacts on the Venezuelan soybean trade and U.S. exports. They formulated a partial
equilibrium, deterministic, and Armington-type model of the Venezuelan market for soybeans
and meal by combining tariffs and the Andean price band variable levy into a single price wedge.
Their model results suggest that a combined MERCOSUR and Andean customs union under
either a high or a low world soybean product price scenario would noticeably benefit
MERCOSUR suppliers at the expense of the United States as well as adversely affect domestic
Venezuelan producers (soybean processors) and fellow Andean member Bolivia.
The study by Zahniser and Link (2002) analysed the impact of North American Free Trade
Agreement (NAFTA) and found that U.S. agricultural trade with Canada and Mexico has nearly
doubled since the implementation of NAFTA. While only a portion of this overall increase can
be attributed solely to the agreement, NAFTA has allowed competitive market forces to play a
more dominant role in determining agricultural trade flows among the three countries. By
dismantling numerous trade barriers, the agreement has contributed to an expansion in U.S.
agricultural exports and increased the domestic availability of various farm and food products. In
addition, NAFTA has established rules and institutions that mitigate potential trade frictions and
promote foreign direct investment. Conversely, many of the initial trepidations that were voiced
concerning declining agricultural employment and environmental degradation have not
materialized. They concluded that NAFTA should be judged not just in the context of the trade
gains associated with the agreement's agricultural provisions, but also in terms of the benefits
derived from "locking in" key trade, investment, and institutional reforms in an increasingly
integrated North American market.
40
Literature review of the impacts of trade agreements on the
economies of developed and developing countries
Yeboah et al(2009) analysed the trade effects of MERCOSUR and the Andean Community on
U.S. cotton exports to CBI countries using an import demand model on panel data for eight
cotton importing CBI countries from the US with annual observations from 1989-2007. Their
results indicated the elimination of tariffs by the eight CBI countries would increase U.S. cotton
exports by $2.3 million. About 88 percent of the increased U.S. cotton exports are due to trade
creation, and the remaining 12 percent is due to trade diversion. Trade creation effects are
substantially greater than trade diversion effects. The favorable trade creation effects indicate
that the U.S. - CBI agreement has been lucrative with respect to U.S. cotton exports to the region
for the period 1989 - 2007. The insignificant trade diversion effects on U.S. cotton exports to the
top eight CBI importers indicates that MERCOSUR and the ANDEAN Community have not
significantly interfered with U.S. cotton imports to the CBI. The insignificant trade diversion
effects on U.S. exports, indicates that MERCOSUR and the Andean Community pose an
insubstantial threat to U.S. exports to the top eight importing CBI countries. This study finds that
the trade creation effects of the U.S. - CBI agreement would be greater than the trade diversion
effects of MERCOSUR and the Andean Community. These results are congruent with the
empirical findings of other researchers; for example, Burfisher and Jones (1998) found that the
regional f!ee trade agreements have 'both trade creation and trade diversion effects in agriculture,
but trade creation dominates in most regional agreements.
Cafferata and Segura (2007) examined the possible economic impact of the Peru- United States
Trade Promotion Agreement (TPA) on Peruvian agriculture from the global and sectoral
perspectives, as well as from the point of view of products or agro-production chains. They
concluded that the elimination of tariffs on imports from the United States would have a limited
impact on the competitive position of most agricultural products on the domestic market,
basically due to the fact that most of the opening of trade in many agricultural subsectors
involves relatively low tariffs, and high protection is circumscribed, involving three lines of
products (rice, sugar and dairy products). IFPRI (2007) analysed the impacts of the US-Middle
East Free Trade Initiative on US trade with Jordan and Morocco and found that the effect of the
US-Jordan FTA will be small because Jordan's level of protection is already low and because
US-Jordan trade is small, while the effect of the US-Morocco FTA will be larger because
41
Literature review of the impacts of trade agreements on the
economies of developed and developing countries
Moroccan trade barriers are higher. Of particular importance, Morocco's wheat tariffs will be
phased out over ten years.
Using a Calculated General Equilibrium (CGE) model, Jansen et al (2007) analysed the impact of
the Central America Free Trade Agreement (CAFT A) on Agriculture and the Rural Sector in
Five Central American Countries focusing on the effects of tariff reductions and quotas under
CAFTA on macroeconomic indicators (economic growth, employment, imports, exports, etc.),
sector behaviour, income distribution and poverty. They found that the impact of CAFTA 's rules
regarding tariff reduction on the quotas for strategic (sensitive) products is small in the short or
medium term. Although small, the impact of CAFTA's trade liberalisation on economic growth
is positive. The simulation results also indicate that CAFTA does not increase poverty but rather
leads to a slight poverty reduction. The impact of tariff reductions under CAFTA on agricultural
sector growth is very small, but positive in Honduras and El Salvador, and negative in Nicaragua
and Costa Rica; while the impacts of increased quotas is .significant and positive only in the case
of Nicaragua.
Korinek and Melatos (2009) used a gravity model to analyse the trade impacts of selected
Regional Trade Agreements (RTAs) in Agriculture and these are the ASEAN Free Trade
Agreement (AFTA), the Common Market for Eastern and Southern Africa (COMESA) and the
Southern Cone Common Market (MERCOSUR). Their gravity estimates indicated that the
creation of AFT A, COMESA and MERCOSUR have increased trade in agricultural products
between their member countries. There was no robust indication of trade diversion with respect
to imports from outside the region. The agreements are therefore net trade creating. There was
no robust indication however that there has been strong trade creation with non-members in the
case of any of the RTAs under study. In some cases, lack of transport and communications
infrastructure, in addition to supply constraints, lessens the effect of the RTA on trade flows.
Trade costs such as transport and logistics seem to remain important factors in determining
agricultural trade flows. In some RTAs, countries have a comparative advantage in exporting
many of the same agricultural products, thereby decreasing the impact of the preferential market
access.
42
Literature review of the impacts of trade agreements on the
economies of developed and developing countries
Regarding the RTAs with the EU, the study by Henry et al (2006) analysed the impacts of the
EU-MERCOSUR trade agreement on agriculture competitiveness in Argentina and Brazil. They
found that the imports of cereals rise as a consequence of the partial liberalisation by nearly two
hundred and fifty thousand (250 000) tons or 0.8%. With producer prices dropping by -0.3%, EU
production decreases slightly. Exports increase slightly following the price reduction which
increases competitiveness of European production. Despite the price decrease total demand
decreases as well, which can mainly be attributed to less demand for feeding. Changes on
oilseeds markets can only be explained by cross effects from other markets, because in the
scenario no changes for oilseeds where specified. However, imports and producer pnces
decrease slightly, and net production and exports increase. Imports of meat increase due to the
expansion of TRQs, and the producer price falls by -0.4%. Production of meat as an aggregate is
unaffected, however, production of beef and poultry decrease by -0.2% and -0.1 % respectively.
That decrease is offset by higher production of pig meat. Demand is unaffected in percentage
whereas the EU can increase its exports slightly due to lower prices. On dairy markets, there are
hardly any changes in percentage terms, only exports of the EU and the producer price increase
slightly.
Using a computable general equilibrium (CGE) model (nicknamed MIRAGE), Boumellassa et al
(2006) analysed the economic impact of a potential free trade agreement (FT A) between the
European Union and ASEAN. They found that the gains accruing to ASEAN members are very
large, adding up to more than 2% of GDP in 2020. Accordingly, this potential agreement would
have an enormous impact on trade, production and welfare, as compared to other episodes of
trade liberalisation. The bulk of the gains (actually three quarter of the gains accruing to the
ASEAN) are associated with the liberalisation in services. All scenarios, including a
liberalisation in services, are associated with welfare gains shared by all countries taking part in
the agreement. The introduction of a list of sensitive products, as a result of political economy
constraints, will increase the overall expected welfare gains for the ASEAN and the EU.
Similarly, using a sample of 36 ACP countries, Morrissey and Zgovu (undated) estimated the
impact of Economic Partnership Agreements (EPAs) on ACP countries' agriculture trade (i.e.
imports), welfare and revenue effects assuming the elimination of tariffs on agricultural imports
43
Literature review of the impacts of trade agreements on the
economies of developed and developing countries
from the EU under EPAs. They found that over half of ACP countries are likely to experience
welfare gains even when assuming 'immediate' complete elimination of all tariffs on agriculture
imports from the EU and when excluding up to 20% of imports as sensitive products,. However,
although most LOCs gain (IOout of 13), most non-LOCs (about 60%) lose. The overall welfare
effect relative to GOP tends to be very small, whether positive or negative. While potential tariff
revenue losses were negligible, given that countries have at least ten years in which to implement
the tariff reductions, there is scope for tax substitution. They concluded that an important issue is
identifying the sensitive products (SPs) to be excluded because the exclusion of SPs reduced the
welfare gain (or increased the welfare loss) compared to estimates where no products were
excluded. Furthermore, Zgovu and Kweka (2006) applied a partial equilibrium model covering
all import products using 2003 trade data and examined the six-digit HS trade, tariff revenue and
net welfare effects of Malawi and Tanzania reciprocating EU's preferential tariff treatment under
the EU-ACP EPA. Their findings show that there will be welfare enhancing consumption and
trade creation effects but these will be swamped by strong welfare-lowering trade diversion and
tariff revenue losses leading to non-negligible net welfare losses.
With regard to RTAs in Asia and Oceania, Francis (2011) analysed in impacts of ASEAN-India
Free Trade Agreement (AlFTA) and established that ASEAN countries will gain significantly
increased market access in India in several semi-processed or processed agricultural products.
The reduced demand for local agricultural products as well as the increased imports of close
substitutes could lead to a fall in the prices of local crops and thus adversely affecting the
domestic agricultural sector. Further, Indian small and medium enterprises (SMEs) in
agriculture-related products and food products, as well as in some intermediate goods and light
manufacturing products are also likely to be negatively affected by the drastic tariff liberalisation
under the AlFTA, as average percentage tariff drops in Malaysia, Indonesia and Thailand's
Normal Track products are much lower than India's. However, import liberalisation in
intermediate goods will impel multinational corporations (MNCs) to undertake production
rationalisation across the region, particularly in the transport equipment and machinery sectors.
Similarly, the Centre for International Economics (2004) analysed the economic effects of the
Australia- Thailand Free Trade Agreement and found that the trade liberalisation undertaken as a
44
Literature review of the impacts of trade agreements on the
economies of developed and developing countries
result of the Agreement will deliver economic benefits to both Australia and Thailand. The gains
to Thailand are larger than for Australia due to Thailand having higher barriers to trade, and,
therefore, a less efficient economy, than Australia. This result also reflects the greater relative
importance of bilateral trade to Thailand than to Australia. Trade liberalisation improves
efficiency in the domestic sectors, and as a result both countries experience an increase in real
investment. In Austral ia, investment peaks at 0.1 per cent above the baseline in 2007 and stays at
0.02 per cent above the baseline after 2020. In Thailand, investment increases to a peak of 0.38
per cent higher above the baseline in 2013, and then reduces to 0.22 per cent above the baseline
in 2026. At the sectoral level, all sectors in both countries experience an increase in output. The
lowering of trade barriers is associated with more efficient domestic industries, while improving
access to markets of the bilateral trading partner. Domestic industries in both countries expand
their output as they move to meet increased consumption, export and investment demand.
Furthermore, using advanced econometric models (i.e. error-correction model), Victorio and
Rungswang (2008) analysed the effects of a free-trade agreement on Thailand's agricultural
imports from New Zealand and found that the FTA has increased the quantity of agricultural
products imported by Thailand from New Zealand. The empirical results of this study have
shown that higher relative prices for Thai agricultural products enticed more imports and the
theoretical influence of GDP was supported in sign, though not in terms of statistical
significance. Evidence was also found in support of the idea that any short-run disequilibrium is
returned to a long-run equilibrium and furthermore, that the FTA significantly influenced the
process of return. All of the findings corroborated economic predictions concerning the effects
upon trade of changes in commodity prices and of the dismantling of trade barriers.
In addition, a study by Toosi et al (2009) analysed the effect of regionalism on Iran's agricultural
trade with special reference to the Economic Cooperation Organization (ECO) region. They
estimated both standard and generalized gravity model to determine the effective factors on Iran
agricultural exports to ECO member countries and found that the ECO region could have a
positive effect on Iran agricultural trade because of very high similarity between Iran and the
other ECO members in religion, border, ethnic and language in relation to the other chosen trade
partners of Iran. By making a comparison between standard and generalized gravity model
45
Literature review of the impacts of trade agreements on the
economies of developed and developing countries
estimates, the results showed that a considerable share of the variability in the ECO agricultural
trade flows refers to uneconomic factors. This study also showed that Taj ikistan, Pakistan,
Kazakhstan and Azerbaijan in ECO region are more interested in importing agricultural products
from Iran and therefore concluded that these similarities amongst the ECO members might put
Iran in an advantageous position of expanding its agricultural exports by gradually reducing its
trade barriers in the ECO region.
Regarding RTAs in Africa, Sandrey and Jensen (2009) used the GTAP model to assess the
welfare and trade gains for the BLNS (Botswana, Lesotho, Namibia and Swaziland) from
envisaged FTAs between SACU and China as well as SACU and India. The results for a
SACU/China FTA show that there are comfortable welfare gains to South Africa, but negating
these are the labour market-related losses where employment falls by 0.13% and the real wage
declines by 0.19%. Scrutinising the production and trade results reveals that South Africa gains
modestly in the agricultural sector, but the big action is in the manufacturing sector. Both
Botswana and the rest of SACU (Lesotho, Namibia and Swaziland as one region) gain modestly
in terms of enhanced welfare of a little over one half of a percent of real GDP. Gains in the
production value of 'other agriculture', 'other meats', textiles and non-ferrous metals (NFM) are
recorded, while exports overall decline to South Africa but increase to both China and the rest of
the world. Overall imports into the rest of SACU increase by more than exports, with big
increases in textile imports from China leading the way.
For the Indian FTA, it was found that a simulation of comprehensive tariff reform in India is
dominated by the massive effects on South Africa's gold sector, and given the implausibility of
this they have opted for an alternative simulation that holds the Indian non-ferrous metal (gold)
tariffs at their initial value. Following declines in the exports of all manufacturing sectors except
non-ferrous metals, the relatively small changes show an overall reduction, while for Botswana's
import profile modest increases from India and the rest of the world more than displace South
African imports, with the latter leading to an overall decline in imports. Changes for trade in the
rest of SACU are even more modest, with slightly increased exports to India and a richer South
Africa just ahead of declines to the rest of the world. The direct effects of these FTA results are
46
Literature review of the impacts of trade agreements on the
economies of developed and developing countries
modest, with most of the changes coming about as the BLNS trade with South Africa changes at
the margin (Sandrey and Jensen, 2009).
Sandrey and Jensen (2009) also used the GT AP model to analyse the implication of the
SACU/MERCOSUR FTA on BLNS countries and found that there are comfortable welfare gains
to South Africa, while the rest of SACU (i.e. BLNS countries) had imperceptible welfare gains.
However, the production and trade results revealed that South Africa loses in agricultural
production due to increased agricultural imports from MERCOSUR countries that lead to a
marginal reduction in the prices of all agricultural products (and a decreased value of agricultural
output. They d that, while this is bad news for farmers, it translates into good news for
consumers as the reduced agricultural prices across the board are enough to marginally reduce the
consumer price index and therefore contributing positively to the overall welfare gains for South
Africa.
3.6 Summary
This chapter reviewed the literature 011 the impacts of various trade agreements on the economies
of the developed and developing countries including South Africa and the African region with
focus on the implications of the multilateral trade agreement in the context of the WTO AoA as
well as bilateral trade agreements in the context of the RTAs. Generally, there is a consensus
view that trade liberalisation benefits are far more than the costs, except that more benefits were
realised by the high income developed countries more especially in the context of the WTO AoA.
However, the majority of studies on the effects of the WTO AoA concluded that international
agricultural trade liberalisation resulted in high prices and in some instances this led to shifts in
production.
With regard to bilateral trade agreements, most of these studies indicated that RTAs resulted in
positive contributions to the economies of the developing countries, especially in improving their
welfare gains and expanding trade. Most studies have found that trade creation effects were
substantially greater than trade diversion effects due to the implementation of various RTAs
worldwide, and therefore concluded that the need for RTAs is thus greatest if the multilateral
47
Literature review of the impacts of trade agreements on the
economies of developed and developing countries
negotiations do not manage to facilitate trade on a broader scale. However, given the principal
objective of bilateral and regional free trade agreements to secure trade liberalisation and expand
market access for members, some studies have felt that the discriminatory nature of FTAs may
result in FTA members expanding their trade at the expense of non members who may become
less competitive purely on the basis of facing a higher tariff than the members.
Apart from reviewing trade and welfare of the trade agreements in question, this chapter has also
highlighted various models that were used to undertake such exercises. Gravity and CGE models
were the most commonly used models in the studies reviewed. Having learnt various methods
that were used in the impact studies of trade agreements, the next chapter provides a detailed
description of various models used in trade policy analysis with a view of selecting the suitable
model for the study.
48
CHAPTER4
METHODOLOGYOFTHESTUDY
4.1 Introduction
Investigation into the impact of trade policies and/or trade agreements on the economic sector in
a country or a country as a whole requires sophisticated modelling frameworks. Such models
include market equilibrium models (such as partial equilibrium and economy-wide models) and
single equation econometric models (such as import demand and gravity models). These models
have been used by many researchers to analyse the impacts of international trade policies. The
next sections of this chapter provide an overview of selected models of trade policy analysis
focusing on their uses, strengths and weaknesses with an ultimate objective of selecting the
suitable model for the study. Furthermore, the chapter provides the motivation why the model
has been considered as well as a detailed discussion on the theoretical framework and
specification of the model. Finally, the data requirements of the model and the sources of the
data are described.
4.2 Market equilibrium models
Many researchers have used market equilibrium models to address issues in international trade.
These models are used for the determination of equilibrium prices and quantities on sets of
markets in order to analyse the impacts of trade on various economic indicators such as economic
growth, welfare, employment, etc (see Tongeren and Van Meijl, 1999). They also contain the
response (behaviour) of economic agents to changes in prices; and prices adjust so as to clear
markets. There are two types of market equilibrium models, partial and economy-wide models,
which are discussed in detail below.
4.2.1 Partial models
These models treat international markets for a selected set of traded goods, e.g. agricultural
goods. In this case the agricultural system is considered as a closed system without linkages with
49
Methodology of the Study
the rest of the economy. The main area of application of partial equilibrium models is detailed
trade policy analysis to specific products. Partial models may be single- or multi-product (see
Francois and Reinert, 1997). The following list paraphrases the summary by Tongeren and Van
Meijl (1999) of global partial equilibrium models adapted to agricultural trade:
4.2.1.1 AGLINK model
The AGLlNK model is a recursive dynamic supply and demand model of world agriculture,
which uses (Nerlovian) partial adjustment relationships. AGLINK was developed by OECD in
co-operation with its member countries, and is presently used by government services of OECD
member countries. The model is used for analysis of the impacts of agricultural policies and for
forecasting the medium term development in supply, demand and prices for the principal
agricultural commodities produced, consumed and traded in member countries. One of the main
strengths of the AGLINK model is that the model structure closely represents the agricultural
situation in member countries. Hence, it has the ability to capture interactions between
commodities and between countries since it not only provides indications of directional
flows/impact, but also information on the magnitude of these impacts (Jooste, 2001).
Von Lampe (1999) states that a major shortcoming of the AGLINK model is its inflexibility and
inability to differentiate current regional aggregates embedded in the model further, namely the
rest of the OECD and the Rest of the World. He states that in aggregating important developing
countries such as China, India and the African Rim within a single region makes it difficult to
reflect the impact of the considerable changes in those regions on the world market. Another
shortcoming is the absence of important food crops in many southern hemisphere countries and
in Asia, since the substitution of these products in favour of higher-quality food cannot be
modelled.
4.2.1.2 Country-Link System
The Country-Link System (CLS) of the Economic Research Service (ERS) of the USDA is used
to conduct global supply, demand and trade projections in general, whilst different scenarios,
50
Methodology of the Study
such as the Asian crisis, could also be modelled. It also allows for individual country analyses.
It is a decentralised system that is linked to expertise based in different regions. Regional models
are then linked to each other to form a complete system capable of simultaneous multi-
commodity, multi-region solutions within the partial equilibrium framework over the medium
and long term. Another distinguishing feature of the CLS is that it has the capability to analyse
bilateral trade flows with the Armington facility (Landes, 1998).
Jooste (2001) paraphrased a number of major strengths of the CLS model. Firstly, the model has
broad coverage of the countries and commodities. Secondly, the model has established linkages
to regional and commodity expertise, supported by an appropriate software interface, since
analysts in different countries do not use the same software for model construction. Thirdly, the
model exhibits multi-commodity and multi-region consistency through a simultaneous solution
framework. Finally, the model can be adapted with relative speed as far as "non-model"
approaches are concerned.
Jooste (200 I) also identified a number of weaknesses associated with the model as follows:
Firstly, the non-standardised modelling format slows the process of linking models and
theoretical consistency cannot always be enforced in all models. Secondly, a lack of regional
expertise exists in some areas, whilst some models are also poorly maintained. Thirdly, some
key areas are not modelled endogenously. Finally, the model is not suitable for short-term
forecasting. According to Von Lampe (1999), the Country-Link system, in addition to including
several policy measures such as tariffs, quotas, etc., also considers a number of other exogenous
variables, i.e. changes in population, income and exchange rates, etc. He also regards the absence
of a number of products, such as pulses and various starchy products that are particularly
important for developing countries, as a minor disadvantage of the CLS.
4.2.1.3 European Simulation model
The European Simulation Model (ESIM) was initially developed co-operatively between the
USDA/ERS, Stanford University and Gëttingen University. ESIM is designed to forecast the
consequences of accession of the Central and Eastern European countries to the EU (see
51
Methodology of the Study
Tangerman and Josling, 1994). Besides EU enlargement ESIM is used to analyse the effects of
CAP (e.g. Agenda 2000) and WTO policies on agricultural markets and budgetary expenditure.
The major strengths of the ESIM model include a broad coverage of agricultural commodities
and the fact that it guarantees the theoretical conditions of homogeneity and symmetry.
However, the model also has some weaknesses which include limited coverage of region or
countries as well as the fact that it is a static model, and therefore cannot be used to model the
dynamics of trade.
4.2.1.4 World Food Model
The World Food Model (WFM) is a multi-product, dynamic partial equilibrium model and was
developed by the Food and Agriculture Organization (FAO). The model is designed to obtain
medium- and/or long-term projections (e.g. used in outlook of FAO on agricultural commodity
markets) and to simulate impacts of policy changes on prices, production, consumption and trade
of the most important agricultural products (FAO, 1993, 1994, 1998).
One of the major strengths of the WFM is that it is a dynamic model because it allows for the
outcome of one year or a sequence of years to influence the outcome of future years. According
to Von Lampe (1999), this enables the model to capture the adjustment paths of the market after
the introduction of certain shocks. It is a world model and, therefore, the regional coverage is
broad.
The main weakness of the model is that it does not satisfy all the laws of demand and supply (i.e.
homogeneity, additivity, symmetry and negativity). It is basically a determinist model and does
not contain stochastic elements. In principle, the WFM was not designed to simulate policy, but
rather concentrated on making projections of the world food situation. However, modifications
were made to the WFM to simulate the impact of trade liberalisation scenarios, more specifically
the impact of the Uruguay Round commitments. It covers only measurable Uruguay Round
commitments that encompass bound tariffs and their reductions, minimum access and limits on
subsidised exports. In essence, the model aims to examine the impact of production shocks on
52
Methodology of the Study
world price stability in order to verify if tariffication and reduction of tariffs have the expected
effect (FAO, 1998).
4.2.1.5 FAPRI model
The FAPRI model is a neoclassical, econometric partial and recursive dynamic model developed
by the Food and Agricultural Policy Research Institute (FAPRI) at Iowa State University. It is
basically an integrated set of models used to 'provide quantitative evaluations of national and
international agricultural policies and other exogenous factors that affect US and world
agriculture' (Devadoss et al, 1993). FAPRI has been used for several years in conducting US
policy evaluations. The set of models involves domestic livestock models, domestic crop
models, government cost and farm income models for the US linked to some world trade models.
The main strength of the model is that it has introduced the dynamics on both supply and demand
functions in a naïve adjustment model for most of the functions. It also includes projection
functions to generate projection of the exogenous variables for the next ten years. In addition, it
covers major agricultural and processed commodities. Its main weakness is that it does not report
information on sensitivity analyses.
4.2.1.6 GAPsi model
GAPsi, Gemeinsame AgrarPolitik - Simulation (Common Agricultural Policy Simulation), is a
partial multi-sector, multi-region, recursive dynamic equilibrium model developed and used at
the Institute of Market Analysis and Agricultural Trade Policy (MA) of the Federal Agricultural
Research Centre (see Salamon, 1998). This model is designed to evaluate EU agricultural
policies (e.g. CAP reform, Agenda 2000).
The model's strength is that it is used for providing both a baseline projection and the calculation
of alternative policy agreements. In addition, quantity instruments (such as quota and budget
restriction) are modelled explicitly. It is also dynamic in nature. However, the regional and
commodity coverage is limited.
53
Methodology of the Study
4.2.1.7 SWOPSIM model
The SWOPSIM (Static World Policy Simulation Model) is a standard multi-commodity, multi-
region partial equilibrium model originally developed by Roningen (1986) at the USDA to study
the impact of the GATT Uruguay Round. SWOPSIM models are designed to simulate the effects
of changes in producer and consumer support policies on production, consumption, and trade'
(Roningen, 1986). Generally, the framework has been employed to analyse the effect of policy
changes on agricultural activity and trade. Applications of the SWOPSIM modelling framework
have included: WTO trade liberalisation (e.g. the Uruguay Round); effects on agriculture from
EU enlargement and potential Eastern European EU membership; agricultural policy reform (e.g.
CAP); free trade hypotheses versus supply control; trade prospects and the opening up of Asian
markets; environmental change and global warming; the impacts of crop disease; trade
liberalisation impacts on production factor demand and the gains from trade (and comparative
advantage); effects of protection and exchange rate policies on agricultural trade; and welfare
analysis.
The advantage of the model is that it has been extended to capture trade flows usmg an
Armington-type specification (Dixit and Roningen, 1986), to include the permanent impact on
derived demand for factors following policy shifts (Liapsis, 1990) and to include medium and
long term projections (e.g. Roningen et al., 1990). In addition, the regional and commodity
coverage is broad and some of its applications provide the information on sensitivity analysis. Its
main weakness is that it does not include dynamics of trade.
4.2.1.8 WATSIM model
The WATSIM (World Agricultural Trade Simulation Model) is global-multi-region, multi-
commodity partial equilibrium model developed by the University of Bonn (Von Lampe, 1998).
WATSIM focuses on three target periods with different aims: Short-term shock analysis,
medium-term projections and policy analysis, and long-term projections and analysis of various
shift factors (e.g. income in Asia, productivity in transition countries). According to Von Lampe
(1999), the WATSIM includes a broad set of policy measures that influence domestic and world
54
Methodology of the Study
markets by altering price, production, demand and trade quantities. The model focuses mainly on
those key factors that will influence supply and demand prospects, for example, socio-economic
and natural variables that have a direct impact on supply and demand, urbanisation, changes in
real per capita income, etc. Von Lampe (1999), as quoted by Jooste (2001), summarised the
main strengths of the WATSIM model, which are paraphrased as follows:
• It is partial equilibrium in nature. In other words, the WATSIM does not account
endogenously for the linkages between other sectors and the agricultural sector, nor does it
account for the interrelationship with macro-economic conditions. Information and data on
the macro-economic environment are, however, introduced exogenously.
• It is multi-regional with multi-products. The multi-regional with multi-product approach
entails that the interaction between different regions and different products are captured
simultaneously if different scenarios are modelled. The model covers broadly the regions and
products.
• It is deterministic ID nature. In other words, uncertainty and risk associated with, for
example variability in weather conditions, are not accounted for. Average conditions are
assumed for particular target years. Endogenous changes in stock levels are furthermore only
accounted for when stock levels react to politically determined prices and when limited
export possibilities exist. Private stocks are assumed to be zero but could be included
exogenously.
• It is non-spatial. The WATSIM model does not account for trade flows or bi lateral
exchanges of products, whilst traded commodities are assumed perfect substitutes in that no
differentiation can be made between the imports and exports of a region's foreign trade
regime.
• It is synthetic. The behavioural parameters, i.e. income elasticities and price elasticities of
demand and supply are not estimated endogenously in the model, but are soureed from the
literature and other models.
The WATSIM model does, however, also have some weaknesses. Firstly, due do the lack of data
on agricultural policies in many developing countries, changes in policies of these countries.
cannot be simulated, and hence it is assumed that price incentives from the world market to
55
Methodology of the Study
domestic producers in such countries are transmitted fully. Secondly, issues such as market
access commitments and import tariffs applicable to net-exporting regions are not properly
represented in the model. Finally, it does not include dynamics of trade.
4.2.1.9 ATPSM model
The ATPSM (Agricultural Trade Policy Simulation Model) is a comparative-static, synthetic,
multi-commodity, multi-region partial equilibrium world trade model for agricultural products,
developed jointly by FAO and UNCTAD. ATPSM model is designed primarily for simulating
agricultural trade policies, notably in the context of the WTO Agreement on Agriculture (see
Poonyth and Sharma, 2003; Pustovit and Schmitz, 2003).
The main strengths of the model include the fact that it is synthetic and covers a broad spectrum
of regions and commodities. It has special features for modelling the Harbinson modalities along
with the EU and US proposals in the context of the WTO. Its main weakness is that it is static in
nature and cannot model the dynamics.
4.2.1.10 CAPRI model
The CAPRI (Common Agricultural Policy Regional Impact) model is a comparative static
equilibrium model, developed by the University of Bonn and funded by the European
Commission. The CAPRI model, commonly known as an EU-wide economic modelling system,
is designed to evaluate the regional and aggregate impacts of the Common Agricultural Policy
(CAP) and trade policies on production, income, markets, trade and the environment (see Britz
and Heckelei, 1997; Loehe and Britz, 1997; Heckelei et al, 1998)
The main strengths of the CAPRI model relate to the fact that it can simultaneously analyse the
effect of commodity market and policy developments in the individual regions of the EU as well
as the feedback from the regions to the EU and world markets. The special feature of the model
is that it is solved by iterating a supply module (which consists of individual programming
models for about 200 regions) and a market module (which follows the tradition of multi-
56
Methodology of tlie Study
commodity models). Based on aggregated supply quantities from regional models, the market
model returns market clearing prices. This means that an iterative process between the supply
and market components ultimately achieves a comparative static equilibrium. The main
weaknesses of model relate to the limited regional coverage as well as the fact that it is static in
nature and cannot model the dynamics.
4.2.2 Economy-wide models
These models capture implications of international trade for the economy as a whole, covering
the circular flow of income and expenditure and taking care of inter-industry relations. The
models have become a useful tool in analysing a number of varied trade policy issues, i.e. to
study the economic effects of trade policies such as tariffs and non-tariff barriers in a variety of
settings. Some are multi-country models that focus on analyzing the effects of global trade
policies or policy changes. Others focus on analyzing commercial policies of a single country,
where depending on whether the country is a developed or developing economy, the modelled
trade issues and policies can be quite diverse.
There are three broad classes of economy-wide models: macro-econometric models, input-output
models and applied general equilibrium (AGE) models. Macro-econometric models are
concerned with macro-economic phenomena such as inflation and exchange rates. Input-output
models provide a comprehensive description of inter-industry linkages and a full accounting of
primary incomes earned in production activities. AGE models do also usually contain full Input-
Output detail, but on top of that they contain equations that describe the behavioural response of
producers, consumers, importers and exporters and possibly other agents in the economy
(Francois and Reinert, 1997; Tongeren and van Meijl, 1999).
AGE models are specifically concerned with resource allocation issues, that is, where the
allocation of production factors over alternative uses is affected by certain policies or exogenous
developments. International trade is typically an area where such induced effects are important
consequences of policy choices. Needless to say, such induced effects are not visible in partial
models. In the face of changing international prices, resources will move between alternative
57
Methodology of the Study
uses within the domestic economy, or even between economies if production factors are
internationally mobile. Only if a complete description of the multi-sectoral nature of the
economy is provided, can such developmental issues be analysed. Tongeren and van Meijl
(1999) also paraphrased a list of various types of economy-wide models as follows:
4.2.2.1 G-cubed model
The G-cubed (Global Computable General Equilibrium Growth) model is a dynamic inter-
temporal general equilibrium and macroeconomic model initiated by McKibbin and Wilcoxen
(J 999). The G-cubed model aims at contributing to the ongoing policy debate on environmental
policy and international trade, with a focus on global warming policies. The model is a 'third
generation' model that combines insights from modern macroeconomics with typical multi-
sectorai resource allocation aspects. Key applications are economy-wide impacts of global
warming policies, and impacts of global macroeconomic shocks. It combines a conventional
AGE model representing the real sectors in a disaggregated way and a representation of financial
and capital assets and flows.
The main strength of the G-Cubed model is that it has sectoral detail and clear macroeconomic
structure, thus designed to provide a bridge between computable general equilibrium (CGE)
models that traditionally ignore the adjustment path between equilibria and macroeconomic
models that ignore individual behaviour and the sectoral composition of economies. The model
allows for analysis of the short-run dynamics and adjustment paths to a long run steady state.
The model employs full short run and long run macroeconomic closure with macro-dynamics at
an annual frequency around a long run Solow/Swan neo-classical growth model. The main
weakness of the model is that it does not have a specific agricultural focus because of its
concentration on macroeconomic phenomena (Tongeren et al, 200 I). In addition, while covering
the regions broadly, the commodity coverage is limited.
58
Methodology of the Study
4.2.2.2 GTAP model
The GTAP (Global Trade Analysis Project) model is a multi-region applied general equilibrium
model developed by Purdue University and IMPACT Project. The focus of the GTAP model is
directed towards the analysis of agricultural policy and trade (Francois et al, 1995; Hertel et al,
1995), although there have been GTAP related applications in non-agricultural trade-related
issues (McDougall and Tyers, 1994) as well as environmental policy analysis (Perroni and Wigle,
1997). European interest in GTAP has also grown, with a steady increase in the literature
examining the impacts of European enlargement to the East and CAP compatibility under the
Uruguay Round commitments (Hertel et al, 1997; Jensen et al, 1998), and modelling applications
based on the Agenda 2000 reform proposals (Slake et ai, 1999). More recently, database
development and modelling have also expanded in the direction of energy usage, climate change
and genetically modified organisms (GMOs).
The major strengths of the model include, among others, a broad coverage of regions and
commodities and the fact that the model has a global closure with respect to savings and
investments, which are treated in an analogous manner to all other goods and services. It has a
special feature of modelling consumption expenditures through a non-homothetic ·Constant
Differences of Elasticities of substitution (CDE) demand system (Hanoch, 1975, Surry 1989),
which allows budget shares to vary with income. The model has versions that allow recursive
and dynamic analysis and also allow sensitivity analysis depending on the modeller.
Furthermore, the model allows one region to be singled out for analysis by declaring the 'Rest of
World' as exogenous. GTAP is supported by a strong group of institutional stakeholders which
puts high requirements on the quality, timeliness and documentation of the data. However, the
weakness of the model is associated with the fact that it does not link individual country models
which are known to capture more regional economic and institutional details and, therefore, the
GTAP framework enforces uniform standards on regional and trade data.
59
Methodology of the Study
4.2.2.3 GREEN model
The GREEN (GeneRal Equilibrium ENvironmental) model is a relatively standard time-recursive
AGE model with global coverage. It was developed at the OECD Secretariat and used for the
assessment of policies that affect carbon emissions. The model has recently extensively been
used to assess implications of the Kyoto protocol on global climate change. The model
incorporates policy instruments such as ceilings (quotas) on emissions and tradable emission
permits.
The major strengths of the model include, amongst others, its broad coverage of regions as well
as its ability to model dynamics and conduct sensitivity analysis. However, the fact that the
model does not give special attention to the agricultural sector and specific polices related to
agriculture could be regarded as a major weakness. In addition the dataset of the model is not
publicly available and its commodity coverage is limited.
4.2.2.4 INFORUM model
/
The INFORUM (INterindustry FORecasting at the University of Maryland) model was founded
by Professor Clopper Almon in 1967. INFORUM models are internationally linked, dynamic
macroeconomic models with inter-industry linkages, and are used to produce annual forecasts for
a variety of industry indicators. The basic approach of INFORUM models is described by Almon
(1991). The TNFORUM system can be used to study the industrial and aggregate impacts of
macroeconomic developments such as changes in exchange rates, trade policy, and government
policy. Applications of INFORUM models to trade policy are relatively limited and tend to focus
on North America. The Canadian, Mexican and USA models were used by the Canadian
government (Department of External Affairs) in a study of the impacts of alternative free trade
agreements between the U.S. and Canada on the Canadian economy and later a similar study was
completed looking at the recently completed NAFTA accord (Almon et al, 1991). Richter (1994)
has examined the consequences of the full participation of Austria in the European Union.
Christou and Nyhus (1994) have examined broader aspects of European policy.
60
Methodology of the Study
The main strength of the model is that it treats a regional subset of economies because it has
features that link individual/single country models to a system. It covers a wide-range of
commodities that varies by country. The model can handle dynamics. One of the major
weaknesses of the model is that even though individual country models can capture more
regional economic and institutional detail, there are clear difficulties with this approach in terms
of consistency and maintenance. Indeed, the linked country models approach seems to be less
sustainable, and their contribution to global trade analysis has been rather limited. It is not a
global model, hence has limited regional coverage. Finally, the sensitivity analysis is not
systematically reported in the model.
4.2.2.5 MEGABARE model
The MEGABARE and its successor GTEM are recursive dynamic AGE models of the world
economy, which share their basic structure with the GTAP model, developed at the Australian
Bureau of Agricultural and Resource Economics (ABARE). These models build on the GTAP
model and database. The focus for the development of MEGABARE was to create a dynamic
general equilibrium model of the global economy suitable for analysis of international
greenhouse policy, but its scope includes broader issues relating to international trade policy,
especially agricultural trade reform.
The main strength of the model is that it is based on the GT AP model and therefore has a broad
coverage of regions and commodities. The model is well documented and publicly available. In
addition, it is theoretically consistent with the general equilibrium framework and can handle
dynamics. However, the sensitivity analysis is not reported in the model.
4.2.2.6 MICHIGAN BDS model
The MICHIGAN BOS (Brown-Deardorff-Stern) model, developed by Michigan State University,
is aptly described as a comparative static 'second generation' model, with monopolistic
competition in manufacturing sectors modelled in the Oixit-Stiglitz fashion. It evolved from
earlier work in the mid 1970s on the Tokyo Round of Multilateral Trade Liberalisation. The
61
Methodology of the Study
BOS model has been used to analyse the economic effects of the Canada-U.S. Trade Agreement
(CUSTA) and later to analyse NAFTA (Brown et al, 1992a, b, 1996), the extension of the
NAFTA to some major trading countries in South America, the formation of an East Asian
trading bloc, and the potential effects of integrating Czechoslovakia, Hungary, and Poland into
the EU (Brown et aI, 1996). Besides regional integration issues the model has been used to
analyse liberal isation of trade in services by Brown el al (1995) and by Brown et al (1996).
The mam strength of model is that it incorporates firm-level product differentiation and
economies of scale by default and therefore makes it possible to model imperfect competition. It
also covers regions and commodities very broadly. It is static in nature and therefore cannot
model dynamics. It does not report sensitivity analysis.
4.2.2.7 RUNS model
The RUNS (Rural Urban North South) model is a relatively standard time-recursive AGE model,
developed at the Free University of Brussels during the eighties by Burniaux (1987). RUNS2 has
subsequently been integrated into the OECD Development Centre's programme on Developing
Country Agriculture and International Economic Trends. The model is not currently in use at
OECD, but RUNS results are still likely to be referenced to date. The main goal of the model
was agricultural policy analysis, especially analysis of the impact of the common agricultural
policy (CAP) on developing countries and assessment of the Uruguay Round of multilateral trade
liberalisation.
The major strength of the model is its special feature of the Rural-Urban distinction, which is
represented by imperfect domestic factor mobility between rural and urban sectors. It is a
recursive dynamic model with broad coverage of regions and commodities. However, the
sensitivity analysis is not reported in the model. While the model is well documented, it is not
publicly available.
62
Methodology of the Study
4.2.2.8 WTO housemodel
The WTO housemodel is a standard AGE model, developed by Francois et al (1995). The model
was constructed to evaluate the results of the Uruguay Round of Multilateral trade liberalisation
and to support the WTO Secretariat in its preparations for the next round of negotiations. The
basic WTO model is a 'first generation' model, but various aspects of imperfect competition have
been added to it. The basic data as well as elasticity estimates are taken from the GT AP dataset.
The WTO housemodel exists in different versions. These are the basic version, which is the
standard perfect competition, constant returns, comparative static model with Armington
assumption for international trade; as well as an amended version, which assumes monopolistic
competition and scale economies internal to each firm.
The major advantage of this model is that it exists in different versions. The basic version is the
standard perfect competition, constant returns, comparative static model with Armington
assumption for international trade. The amended version assumes monopolistic competition and
scale economies internal to each firm, thus allowing modeling of imperfect competition. Quotas
(MFA and minimum market access) are modelled explicitly as inequality constraints. It is a
global model providing broad coverage of commodities and regions. It also reports sensitivity
analysis. The main weakness of model is that it cannot handle dynamics.
4.3 Single equation econometric models
The single equation econometric models, such as import demand and gravity models, are mainly
used to examine trade determinants, to predict trade potentials, to examine competitiveness and
responsiveness. They are commonly used in empirical studies of bilateral trade flows.
4.3.1 Import demand models
Estimation of demand functions consistent with economic theory has been a highly published
area in the last forty years. The majority of the papers follows the adoption of flexible functional
forms and relies heavily on duality theory. The Generalized Leontief (Diewert, 1971), the
63
Methodology of the Study
translog (Christensen et al, 1975), the Rotterdam Demand System (Theil, 1965, 1975 and Barten,
1964, 1968) and the Almost Ideal Demand System or AIDS (Deaton and Muellbauer, 1980) are
examples of popular demand models. Their functional forms are locally flexible, that is, they do
not put a priori restrictions on the possible elasticities. Instead, they possess enough parameters to
approximate any elasticity at a given point. These locally flexible functional forms often exhibit
small regular regions. Thus, a number of alternative flexible functional forms with larger regular
regions have been developed. Examples include the Quadratic AIDS model (QUAIDS) (Banks
ef al, 1997), the Laurent model (Barnett, 1983, 1985; Barnett and Lee, 1985; and Barnett et al,
1985) and the Generalized Exponential Form (GEF) (Cooper and McLaren, 1996).
The literature in applied economics shows that the AIDS and the Rotterdam models are
frequently used demand specifications (see Deaton and Muellbauer, 1980; Eales and Unnevehr,
1988; Lee, Seale and Jierwiriyapant, 1990; Alston ef al, 1990; Sparks ef al, 1990; Hayes et al,
1990; Green and Alston, 1990; Yang and Koo, 1994; Mixon and Henneberry, 1996; Kalaba,
2001). These models are product-specific, data-sensitive and static in nature and were used in the
above studies to estimate the responsiveness of consumers to certain imported goods as well as to
examine the price competitiveness of such goods from various suppliers (exporting countries) in
an importing country.
The success of the AIDS and Rotterdam models, according to Barnett and Seck (2008), is partly
due to the possibility of estimating some of their specifications without relying on procedure of
nonlinear estimation. In addition, theoretical restrictions can be imposed and tested with ease.
The AIDS model has a particularly attractive feature: the properties of the preference relations
that generate it are known. The AIDS is derived from a known cost function with the desired
properties. Studies that confront these two models have been rather rare, even though Deaton and
Muellbauer (1980) pointed out the striking similarity between these two models, after identifying
that the AIDS model with linear price (LA-AIDS) can be rewritten in difference form so that it
has the same dependent variables as the Rotterdam model in absolute price. Alston and Chalfand
(1993) developed a statistical test for the AIDS versus Rotterdam model using the approximation
expressing the AIDS in difference form and with approximately the same right hand side
variables.
64
Methodology of the Study
According to Kalaba (2001), the AIDS model and its rival Rotterdam model are similar in many
respects. Both have flexible functional forms, identical data requirements, are parsimonious with
respect to number of parameters, and are linear in parameters. Economic theory does not provide
a basis for choosing between the two models. Most researchers arbitrarily pick one model or the
other, but recent interest has focused on developing proper non-nested tests of the two demand
systems. Two prominent studies have presented techniques to select between the AIDS and the
Rotterdam demand systems (Alston and Chalfant, 1993; LaFrance, 1998). Alston and Chalfant
(1993) used a compound-model approach to select between the First Difference Almost Ideal
Demand System (FOAIDS) and the Rotterdam models, using U.S. meat demand data (beef, pork,
chicken, and fish). They found support for the Rotterdam model.
However, LaFrance (1998) pointed out that the least squares approach used by Alston and
Chalfant (1993) was biased and inconsistent because they had not considered endogeneityof
budget shares and their prices were not mean scaled in the Stone's index. Using the same data,
he conducted both a Lagrange multiplier test and a likelihood ratio test and failed to reject either
demand system. Compound model approaches typically have correct asymptotic size, but low
power (Pesaran, 1974). Thus, the failure to reject either null hypothesis may simply be the result
of using a test with low power. Most of the previous non-nested tests have been developed for
models that have the same dependent variables (see Pesaran, 1974). Coulibaly and Brorsen
(1999) show that a Cox's non-nested test based on the parametric bootstrap has high power, is
relatively easy to use, and is applicable to any model that can be simulated. The approach
appears promising as a method for selecting among functional forms in demand systems.
4.3.1.1 Almost Ideal Demand System (AIDS) model
Since its introduction by Deaton and Muellbauer (1980), the AIDS model has been widely used
in demand analysis. The majority of empirical applications follows Deaton and Muellbauer's
lead and replaces the translog price index with Stone's index to deflate income. This generates
the linear approximate almost ideal demand system (LA-AIDS), which is linear in the unknown
parameters and therefore simpler to estimate. Deaton and Muellbauer (1980) cautioned against
imposing symmetry on the LA-AIDS, and avoided doing so. They interpreted Stone's index as
65
Methodology of the Study
an approximation to the' 'true" translog index. Nevertheless, most applications of the LA-AIDS
test for and impose symmetry of the matrix of log-price coefficients (e.g. Anderson and BlundelJ,
1983; Moschini and Meilke, 1989). There really can be only one explanation for this practice;
the LA-AIDS is presumed to be the "true" model and symmetry of the matrix of log price
coefficients is presumed to be the correct way to obtain Slutsky symmetry and economic
rationality of the demand equations that are estimated.
The LA-AIDS has been criticized for reasons other than its failure to be consistent with
economically rational consumer choices. Eales and Unnevehr (1988) point out that budget shares
appear on both sides of the regression equations, producing simultaneity problems. Pashardes
(1993) and Buse (1998) criticize the errors in variables problem created by using of Stone's index
rather than the "true" translog price index on the right-hand-side of the regression equations.
Moschini (1995) argues that Stone's index is not a proper price index at all and that without some
mechanism to scale prices (e.g. at sample means), Stone's index leads to biased and inconsistent
parameter estimates.
It is also possible to use the AIDS model to analyse the import demand for products differentiated
by sources and this generates a restricted, source-differentiated almost ideal demand system
(RSDAIDS). According to Armington (1969), the problem of source differentiated AIDS
(SDAlDS) is the systematic simplifying of the product demand function to a point where it is
relevant to practical purposes of estimation. For example, the general Marshallian model runs
through a sequence of progressively restrictive assumptions, leading to a specification of product
demand function that preserves the relationship between demand, income, and prices. The
fundamental modification of the basic Marshallian model is the assumption of independence.
This assumption states that buyers' preferences for different products of any kind are independent
of their purchases of products of another kind. For example, an increase in purchases of Chilean
grapes does not change buyers' relative evaluation of New Zealand's apples (Kalaba, 2001).
Another assumption of the SOAIDS model, as paraphrased by Kalaba (2001), is that the
country's market share is unaffected by changes in the size of the market as long as relative
prices in that market are unchanged. The size of the market is a function of money income and
66
Methodology of the Study
prices of various goods. Therefore, demand for a product is a function of money income, the
price of each good and the price of product relative to prices of other products in the same
market. The growth in market share depends on the change in the product's price relative to
average change in prices in the market. Growth of the market depends mainly on changes in
. income and income elasticities of demand for the respective product. Although the AIDS model
has been criticized for its weakness, several studies preferred this model among others with
similar characteristics. The Armington model assumes that import demands are homothetic and
separable among import sources. Thus, within a market, trade patterns change only with relative
price changes, and elasticities of substitution between all pairs of products are identical and
constant. These are strong restrictions on demand and were rejected by several studies that have
tested these assumptions using alternative models (Winters, 1984; Alston et ai, 1990; Lee and
Brorsen, 1993). Winters suggested AIDS as an alternative to the Armington model. Alston et al
(1990) also presented the double log model and AIDS model as possible alternatives to the
Armington model.
Lee and Brorsen (1993) concluded that the Armington assumptions are inappropriate for
modelling agricultural import demands. The Armington restrictions had already been rejected by
Alston et al (1990) who used world cotton and wheat trade data. These restrictions also cause
specification errors by omitting relevant explanatory variables, like import prices from competing
sources within a group. The tests for non-nested models of AIDS and the double model log for
source differentiated U.S. beef import demands by Lee and Brorsen (1993) showed that both the
double-log import model and the AIDS model were appropriate for import demand. However,
the estimated elasticities using the AIDS model were more plausible than those from the double-
log model. In addition, the AIDS model permitted imposing the theoretical properties of
demand, while the double-log model only allowed homogeneity.
Empirical applications of the AIDS model to import demand have frequently assumed either
product aggregation or block separability (Yang and Koo, 1994). Under the product aggregation
assumption, products are not differentiated by sources and are perceived as the same (Hayes et ai,
1990). Moreover, the block separability assumption among goods allows estimation of share
equations for goods from different origins (Alston et al, 1990). For products that are similar and
67
Methodology of the Study
competing in the same market, the RSDAIDS is preferred. The RSDAIDS model is a more
general model and does not impose perfect substitutability assumptions.
4.3.1.2 Rotterdam Demand System (RDS) model
The Rotterdam model involves a nonlinear transformation of quantity on the left-hand side of the
demand equation (Kastens and Brester, 1996). Analysis by Barnett and Seck (2008) has shown
that the Rotterdam model is comparable to other popular flexible functional demand
specifications like the Almost Ideal Demand System. A Rotterdam specification was developed
to show how preference variables affect demand through their impacts on marginal utilities. A
change in a preference variable was viewed as resulting in changes in adjusted prices which were
decomposed into actual price changes minus preference-variable-induced changes in marginal
utilities. Restrictions on preference variables were considered through adjusted prices by
imposing restrictions on the marginal util ity elasticities with respect to the preference variables
(Brown and Lee, 2002).
The Rotterdam model was widely used to examine advertising and/or habit formation effects on
import demand. It is consistent with demand theory (Theil 1965; Barnett, 1979); it is as flexible
as any other local approximating form (Mountain, 1988); it lends itself to advertising applications
(e.g., Brown and Lee 1993; Duffy 1987, 1990); and prior testing indicated that the estimated
advertising effects from the Rotterdam model were similar to those obtained from its major rival,
the (linear approximate) Almost Ideal Demand System, and from a double-log specification
(Xiao, 1997).
Several approaches have been used to augment the Rotterdam specification to include advertising
effects. The most common approach, suggested by Theil (1980), is to view advertising as a "taste
shifter" that affects marginal utility. In this formulation, advertising enters the model as a price
deflator (e.g., Duffy 1987; Brown and Lee 1993). An alternative approach, advocated by Stigier
and Becker (1977), is to view advertising (or other information sources) as an input in the
household production function. In this formulation, advertising enters the (derived) demand
function for market goods as a separate shift variable along with prices and income (e.g.
68
Methodology of the Study
Kinnucan et ai, 1997). Testing the simple-shift specification against the taste-shift specification
using citrus data, Brown and Lee (1993) found them to be statistically equivalent.
Xiao et al (1998) also used both forms of the Rotterdam model to determine the sensitivity of
parameter estimates to model specification. The four-equation system consisted of demand
equations for fluid milk, fruit juices (chiefly orange and apple), soft drinks, and coffee and tea.
They treated the weak separability of the non-alcoholic drink group as a maintained hypothesis
and total group expenditure was used in place of income in the absolute-price form of the
Rotterdam model. In all four equations, advertising effects were statistically significant.
On the other hand, the Rotterdam model was used to model the various categories of apparel
demand by using habit formation models of the sort conceived by Manser (1976), Pollak and
Wales (1969), Blanciforti and Green (1983), Pollak and Wales (1992), and Holt and Goodwin
(1997), among others. In this case, the habit formation model was applied to a variant of the
differential demand system, otherwise known as the Rotterdam demand system, as introduced
originally by Barten (1964) and Theil (1965). The model is similar to that advocated by Theil
(1980) and employed by, among others, Brown and Lee (1997) for examining the stock effects of
advertising on consumption in a differential demand system context. Holt and Goodwin (1997)
were the first to incorporate a dynamic habit stock characterization into a system of differential
demand equations. The basic assumption is that habit stocks affect the marginal utility associated
with consuming each apparel item in the group and then showed how habit stock effects on utility
can be translated into effects on demand, and also showed that that habit formation may be
viewed as changing the perceived prices for all apparel items in the group.
4.3.2 Gravity model
The gravity model, developed in the 1960s, is a standard empirical framework for investigating
patterns of bilateral trade. It is derived as a reduced form of a broader class of structural models
(Anderson, 1979 and Bergstrand, 1986), as one of the popular tools in empirical studies
addressing issues in international trade (ITC, 2000; Bun and Klaassen, 2002; Nouve and Staatz,
2003). It has been used in pioneering works by Tinbergen (1962) and Pëynëhen (1963), who
69
Methodology of the Study
suggest the use of the Newtonian gravity concept to explain bilateral trade (attraction) by the
national incomes of the trading countries and the distance between them.
On this basis, a large number of studies were undertaken. Within this mushrooming literature,
gravity equations share a common design that can be customized for different purposes, which
are paraphrased by ITC (2000) as follows:
• Firstly, a gravity equation is bilateral. It explains a trade-related dependent variable, by
the combination of macroeconomic variables (size, income, exchange rates, prices, ete)
for both countries. Indicators of transportation costs between the two countries and more
generally market access variables are added.
• Secondly, a gravity equation may be used in order to estimate either determinants of the
volume or determinants of the nature of trade flows.
• Thirdly, theory definitively provides strong foundations to a modelling based on rough
indicators, which is quite useful when the purpose is to integrate a large number of
countries in the sample or when the statistical background for (developing) countries is
limited.
• Fourthly, there is inevitably a discrepancy between the theoretical model and the ideal
equation that would fit the data well. Border trade, seasonal trade, trade preferences or
regional integration may be controlled for with specific effects by pair of country; such a
solution however jeopardizes any attempt to use the model for forecasting purposes. This
justifies the introduction of cultural, historical or institutional determinants in equations
designed for an applied purpose.
• Lastly, given the type of variables under consideration, gravity-type econometric models
are estimated using rather aggregated data. Numerous studies have been running
equations on total exports.
A gravity model is a widely used method to explain trade patterns between countries using each
country's measures of "mass" and geographical distance between countries to assess changes in
trade flows (Otsuki et ai, 2001). Initially, gravity models were developed on a mostly empirical
basis, with researchers emphasizing that country size and transportation costs between countries
70
Methodology of the Study
were good predictors of trade volumes. And results were indeed positive, since such equations fit
the data quite well. However, the lack of theoretical foundations rapidly led scholars to
skepticism, and Anderson (1979), Helpman and Krugman (1985) and Bergstrand (1989) provided
the missing theoretical basis. While Bergstrand (1989) built a general equilibrium model of
world trade from which reduced equations may be derived, Helpman and Krugman (1985)
showed that the combination of comparative advantages and monopolistic competition provided
a coherent conceptual framework for empirical analysis.
Deardorff and Stern (1994), Engel and Rogers (1997), Frankel and Stein (1994) and Frankel et al.
(1995, 1996, 1997) and Wei and Parsley (1995) have found strong linkages between bilateral
trade and the proximity of its trading partners, where proximity is represented by distance,
adjacency and common language to reflect cultural similarities. It postulates that the volume of
trade between two countries is proportional to their economic sizes (capacity to supply exports
and to absorb imports) and inversely proportional to costs of trading. The distance between the
two trading units has traditionally served as a proxy for trading costs (Lairds and Yeats, 1990).
In a nutshell, the gravity model has the ability to predict patterns of bilateral trade with the
expectations that trade will increase with the economic mass of the countries but decrease with
distance that separates them (Poonyth et al, 2002).
4.4 Model consideration and motivation
Given the nature of this study and the types of research questions that need to be addressed, the
study will apply an econometric approach using the gravity model. ,The gravity trade
econometric model was considered in this study because of the following reasons: Firstly, the
gravity equation makes use of raw data without reliance on prior estimation of various
elasticities, etc. Secondly, the gravity equation can readily exploit panel data, and thereby
capture dynamic aspects of trade policy impacts. Lastly, the gravity equation singles out
distance between countries as a significant explanatory variable, which is desirable given South
Africa's location relative to its main trading partners.
71
Methodology of the Study
Gravity models have been used by many researchers to exarrune the impact of the factors
influencing trade performance, to examine whether a trade agreement led to trade creation or
trade conversion between trading partners, as well as to estimate trade potentials (see Tinbergen,
1962; Pëynëhen, 1963; Anderson, 1979; Bergstrand, 1986; Deardorff and Stern, 1994; Engel and
Rogers, 1997; Frankel and Stein, 1994; Frankel et al, 1995, 1996, 1997; Wei and Parsley, 1995;
Cassim, 2001; Poonyth et al, 2002; Chauvin and Gaulier, 2002; Bun and Klaassen, 2002; Nouve
and Staatz, 2003; Mokoena et al, 2008).
Gravity econometric equations are not sensitive to data, and hence could be estimated using
various types of data, i.e. cross-section, time-series and panel data, depending on the type of
research question to be addressed, and are applicable to both static or dynamic modelling (see
Bun and Klaassen, 2002). These equations can use various combinations of macro-economic
variables, such as gross domestic products and populations with geographic distance, ete; to
predict or forecast trade potentials. Hence, gravity equations have extensively been used in the
empirical literature on international trade (Havrylyshin and Pritchett, 1991; Frankel and Wei,
1993; Bayoumi and Eichengreen, 1995). The related econometric models can also be used to
predict trade patterns at the industry level (Bergstrand, 1989). In this case, the elasticities vary
across industries for a given macro-economic variable; and these elasticities are those which help
to predict future paths of specialization.
4.5 Theoretical framework and specification of gravity model
Gravity models have strong theoretical foundations both in traditional and in the new trade
theories (Wall, 1999; Cheng and Wall, 1999; Rose 2002; Evenett and KeIler, 2002). The lack of
rigorous theoretical underpinning has traditionally been the major criticism against gravity
models. However, Wall (1999) indicates that such criticism has been weakened since Deardorff
(1998) established a consistency between gravity models and variants of traditional trade
theories, such as the Ricardian and Heckschser-Ohlin models. Wall (1999) also points to "earlier
works by Anderson (1979) and Bergstrand (1986) who derived gravity equations from trade
models with product differentiation and increasing returns to scale" (Wall, 1999), suggesting that
gravity models may also be consistent with the new trade theory.
72
Methodology of the Study
Although early studies used cross-section analysis to estimate gravity models (Aitken, 1973;
Bergstrand, 1986), the analysis cannot answer a policy-related question of the impact of changes
in relative market size (or income) of countries on changes in the pattern of bilateral trade over
time (Kim et al, 2003). Temporal effects can be answered by using cross sectional time series
analysis, as discussed by Mátyás (1997); De Grauwe and Skudelny (2000); Wall (2000); Glick
and Rose (200 I). One reason is that the extra time series observations result in more accurate
estimates.
Using panel data models, Mátyás (1997) and Wall (2000) stressed the importance of including
country-pair specific effects, but ignored one potentially important aspect of trade, namely
dynamics. For countries that have traded a lot in the past, businesses have set up distribution and
service networks in the partner country. In addition, consumers have grown accustomed to the
partner country's products (habit formation). It is therefore very likely that current bilateral trade
between those countries is also high (Eichengreen and Irwin, 1997). Hence, passed trade affects
current trade. Ignoring this may lead to incorrect inference. Eiehengreen and Irwin (1997) and
De Grauwe and Skudelny (2000) therefore added lagged trade as a regressor to their gravity
model and showed that lagged trade is indeed important. This implies that the estimate for
-lagged trade represents not only dynamic effects, but also the impact of unobserved country-pair
specific time invariant factors, as these factors are present in both current and lagged trade.
Initially, gravity models were developed on a mostly empirical basis, with researchers
emphasising that country size and distance between countries were good predictors of trade
volumes. However, Anderson and van Wincoop (2003) argued that this commonly used
remoteness variable, which relies solely on distance, does not capture the entire range of factors
affecting bilateral trade flows and concluded that such gravity models suffer from an omitted
variable bias (see also Baldwin and Taglioni, 2006). To remedy this problem, Anderson and Van
Wincoop (2003) modified McCallum' s (1995) gravity equation (in which bilateral trade flows
between two regions depend on the output of regions, their bilateral distance and whether they
separated by a border) by adding multilateral resistance variables, which consist of country
specific price indices. Since the multilateral resistance variables as proposed by Anderson and
Van Wincoop (2003) are not observable, these authors propose, among others, the simultaneous
73
Methodology of the Study
use of both importer and exporter fixed effects to replace the resistance variables, yielding
coherent results (see Rose and van Wincoop, 2001; Eaton and Kortum, 2002). The use of both
exporter and importer fixed effects is supported by Helpman et al (2007) and Zwinkels and
Beugelsdijk (2010), as they have argued that the inclusion of exporter and importer fixed effects
allows for unbalanced bilateral trade flows even when all bilateral trade barriers are symmetric.
Furthermore, Si.ileyman (2010) proposed several extensions of the standard gravity model and
modified the traditional gravity equations by adding competitiveness that was composed of a
general and bilateral component and account for a flexible income response.
In this study, the gravity model is used to determine the impacts of bilateral and multilateral trade
agreements on trade flows of selected agricultural products as well as of aggregated agricultural
trade flows between South Africa and its trading partners. The estimated gravity equation in this
study is similar to that of Mátyás (1997) and Wall (2000), but was extended to incorporate the
work done by Eiehengreen and Irwin (1997) and De Grauwe and Skudelny (2000) so that it takes
into account the importance of both dynamics as well as controlling the country-pair specific
effects and/or the unobserved multilateral resistance variables. This gravity equation is then
expressed as follows:
In Y;j1 = ao + ai; + fJ"X" + ciit (1)
InYijt is the dependent variable, which the natural logarithms of real values of agricultural trade
flows between countries i and j (in all cases "i" denotes South Africa) and country j (in all cases
"j" denotes South Africa's trading partner) in year t. The values of all trade flows variables are
expressed in constant 2000 United States dollars (US$). Symbol J3n represents vector coefficients
associated with explanatory variables (Xn), as described in models below, whereas u's and Eijt
represent the intercepts and error term respectively. The model has two types of intercepts, i.e.
one common to all years and country pairs (uo) and one specific to the country pairs and common
to all years (Ui). It is assumed that the error term is normally distributed with zero mean and
constant variance for all observations and that the disturbances are pair-wise uncorrelated.
74
Methodology of the Study
For each trade flow, six models will be estimated with .each model being estimated two times,
firstly assuming the dynamic equation (i.e. includes the lagged dependent variable and secondly
assuming the static equation (i.e. excludes the lagged dependent variable), because there is no
economic justification for a priori selection criteria between the two models. Furthermore, the
gravity equation will be estimated using three models. The first and second models estimate the
period impacts (i.e. jointly from 2000 to 2004 and from 2005 to 2009 for EU-SA TDCA and
SADC Trade Protocol as well as jointly from 1995 to 1999 for WTO AaA) and the individual
yearly impacts (i.e. on annual basis from 2000 to 2004 and from 2005 to 2009 for EU-SA TDCA
and SA DC Trade Protocol as well as on annual basis from 1995 to 1999 for WTO AaA). The
third model estimates the trade direction impacts, i.e. whether the implementation of the EU-SA
TDCA and SADC Trade Protocol have created exports from South Africa to the EU and the
SADC countries respectively or diverted exports to other trading partners of South Africa. These
gravity equations are expressed as follows:
Models Explanatory Variables (Xn)
Dynamic Period Impact InYijt_p; InGOPPCit; InGOPPCjt; InREERt; 00004; InOISTij
InYijt_p;InGOPPCit; InGOPPCjt; InREERt; 00509; InOISTij
InYit_p; InGOPPCit; lnGOPPCt; InREERt; 09599; InDlSTi -r-,
Static Period Impact InGOPPCit; InGOPPCjt; InREERt; 00004; InDlSTij
InGOPPCit; InGOPPCjt; InREERt; 00509; InDlSTij
InGOPPCit; InGOPPCt; InREERt; 09599; InOISTi
Dynamic Yearly Impact InYijt_p;InGOPPCit; InGOPPCjt; InREERt; DOO; DOl; 002; 003; 004; i-msr,
InYijt_p;InGOPPCit; InGOPPCjt; InREERt; DOS; 006; 007; 008; 009; InOISTij
InYit_p; InGOPPCit; InGOPPCt; InREERt; 095; 096; 097; 098; 099; InDlSTi
Static Yearly Impact InGOPPCit; InGOPPCjt; InREERt; DOO; OOI; 002; 003; 004; InOISTij
InGOPPCit; InGOPPCjt; InREERt; DOS; 006; 007; 008; 009; InDlSTij
InGOPPCit; InGOPPCt; InREERt; 095; 096; 097; 098; 099; InOISTi
Dynamic Trade Direction Impact InYit_p; InGOPPCit; InGOPPCt; InREERt; PTAves; PTAno; InDlSTi
Static Trade Direction Impact InGOPPCit; InGOPPCt; InREERt; PTAyes; PTAno; InOISTij
Where:
The p-year lags of the dependent variables, The lag length of the dependent variable was
determined using the Akaike Information Criteria (AIC) and Schwarz Criterion (SC)
procedures, An ad hoc approach looking at the significance and sign of the new lags was also
followed in deciding the lag length in case where AIC and SC approaches recommends more
than a one-year lag, As a result, a one-year lag was determined because, in all models, lags
beyond the first one did not add to the predictive power of the model meaning that the second
and further lags were not statistically significant.
This variable has been included because, as mentioned in the introduction, historically before
the conclusion and implementation of the EU-SA TOCA, the EU has been South Africa's
main trading and investment partner accounting for over 40% of its total trade. For countries
75
Methodology of the Study
that have traded a lot in the past, businesses have set up distribution and service networks in
the partner country. In addition, consumers have grown accustomed to the partner country's
products (habit formation). It is therefore very likely that current bilateral trade between
those countries is also high (Eichengreen and Irwin, 1997). Hence, passed trade affects
current trade. Ignoring this may lead to incorrect inference. Eiehengreen and Irwin (1997)
and De Grauwe and Skudelny (2000) therefore added lagged trade as a regressor to their
gravity model and showed that lagged trade is indeed important. This implies that the
estimate for lagged trade represents not only dynamic effects, but also the impact of
unobserved country-pair specific time invariant factors, as these factors are present in both
current and lagged trade.
InGOPPCit and The logarithms of the real per capita gross domestic products for countries i and j in year t
InGOPPCjt respectively. The values of GDPPCs variables are also expressed in constant 2000 United
States dollars (US$). As the mass of two bodies determines the force of attraction between
them, as stated in the law of gravity, GDPPC of the trading countries represents both the
productive and consumption capacity that heavily determine the trade flow between them.
Many studies have included both the GDP and Population (POP) as explanatory variables for
bilateral trade flows with GDP serving as a proxy for output capacity of the exporting country
and for absorptive capacity of the importing country; whereas POP of the exporting country
serve as proxy for factor endowments (production capacity and POP of the importing country
as a proxy for market size (Dascal, Mattas and Tzouvelekas, 2002). In this study, the GDP
and POP were not included in the models separately due to the fact that they are highly
correlated. Furthermore, De Blasi, Seccia, Carlucci and Santeramo (undated) argued that
while the total GDP is appropriate for studies using aggregated data, in the case of a specific
agro-food product, this variable would overestimate the country's output capacity and
therefore emphasised the use ofGDP per capita.
Given the fact this study focuses on the aggregate agriculture and individual agro-food
products, which are regarded as subset of the total economy, the GDPPC is a stronger
variable explaining the income effect, as it serves as proxy for purchasing power of the
exporting and importing countries. GDP per capita has also been very commonly employed
(Sanso, Cuairan, and Sanz, 1993; Tamirisa, 1999). Gros and Gonciarz (1996) argued that the
per capita output is used to take into account the idea that as income increases, the share of
tradables in overall income might increase; i.e. for a given overall income a country with a
higher income per capita would trade more intensively (have more exports and imports) than
a poorer country. Therefore, it is expected that GDPPC would be positively related to both
exports and imports of agro-food products.
InREERit The logarithm of the real effective exchange rate of South African Rand to the base year
2000, which is an index measured as one rand (R 1.00) to a basket of 15 major currencies in
the world. Generally, it is expected that an appreciation of the importing country's currency
against trading partners' currencies would impact positively on imports because imports
become less expensive, whereas depreciation would affect imports negatively since they
become more expensive. On the other hand, a depreciation of the exporting country's
currency against trading partners' currencies means that its exports become cheaper in the
world markets, thus impacting positively on exports (Bergstrand, 1986; Koo, Karemera and
Taylor, 1994). Thus, the sign of the REER coefficient will depend on the depreciation or
appreciation of the South African Rand against trading partners' currencies.
00004/ 00509 The joint dummy variable for the periods 2000-2004 and 2005-2009, which firstly signifies
the implementation of the EU-SA TDCA and secondly the implementation of the SADC
Trade Protocol. This variable takes the value of 1 for the period of implementation of the
EU-SA TDCA between South Africa and the EU countries or 0 otherwise, as well as the
value of 1 for the period of implementation of the SADC Trade Protocol between South
Africa and the SADC countries or 0 otherwise.
76
Methodology of the Study
DOO,001, 002, The individual annual dummy variables for the implementation of EU-SA TDCA and SADC
003 and 004 Trade Protocol, which take the values of ones for the implementation or zeros otherwise.
005, 006, 007,
008 and 009
09599 The joint dummy variable for the period 1995-1999, which signifies the implementation of
the WTO URAA between South Africa and all its trading partners, which takes the value of 1
for the period 1995-1999 or 0 otherwise
095, 096, 097, The individual annual dummy variables for the implementation of WTO URAA, which take
098 and 099 the val ues of ones for the implementation or zeros otherwise.
. PTAycs and The dummy variables that are introduced when analysing trade creation and diversion .
PTA no Firstly, PTAycs represents the "both in" scenario, i.e. both South Africa and EU countries are
in the agreement, whereas PTAno represents the "in out" (otherwise) scenario, i.e. South
Africa in and EU countries out. Secondly, PTAycs represents the "both in" scenario, i.e. both
South Africa and SADC countries are in the agreement, whereas PTAno represents the "in
out" (otherwise) scenario, i.e. South Africa in and SADC countries out. In other words,
PTAyes is the same as D0004 and D0509, as it takes the value of 1 for the implementation of
the EU-SA TDCA and SADC Trade Protocol, i.e. from 2000 to 2004 and from 2005 to 2009,
and 0 otherwise (other trading partners). However, for EU-SA TDCA, PTAnotakes the value
of 1 for the periods 2000 to 2004 and 2005 to 2009 between South Africa and other trading
partners and 0 otherwise (EU countries). Whereas for SADC Trade Protocol, PTAnotakes the
value of 1 for the periods 2000 to 2004 and 2005 to 2009 between South Africa and other
trading partners and 0 otherwise (SADC countries). In analysing the trade direction, the a
priori criteria states that if the parameter estimate for PTAycs (<<bothin») is positive and
significant, there is trade creation due to regionalism. In contrast, if the parameter
estimate for PTAn. (edn out») is negative and significant, there is trade diversion
(International Trade Centre, 2000).
InDISTij The logarithm of a geographic distance between country i and country j in kilometres and
serves as a proxy for transportation costs. As a proxy for transportation costs, countries with
,short distance between each other are expected to trade more than those who are far apart due
to lower transaction cost. Distance can also be used as a proxy for the risks associated with
the quality of some of the perishable goods and the cost of personal contact between
managers and customers (ITC, 2000; Cheng and Wall, 2005; Gebrehiwet et ai, 2007).
The reason for the use of different dummy variables for trade agreements is that the
implementation of the EU-SA TDCA and SADC Trade Protocol started in 2000, while the
implementation of the WTO AoA, as concluded during the Uruguay Round negotiations, started
in 1995 and expired in 1999. It was expected that the new WTO AoA to be concluded during
Doha Round negotiations would be implemented with effect from 2000, but the negotiations
collapsed. However, the EU-SA TDCA and SADC Trade Protocol, dummy variables were
introduced for two periods of implementation, i.e. from 2000 to 2004 and from 2005 to 2009.
This is because both agreements were implemented effective from 2000 and normally the first
five years of implementation are characterised by too many administrative matters, i.e. putting
systems in place, and as a result the benefits of the trade agreements are not fully utilized. It is
expected that after five years of implementation, the businesses would have been well set-up as
77
Methodology of the Study
traders between the trading partners would have developed relationships. The other reason for
separating the two periods is that some of the models could not accommodate more than 5 year
dummies in addition to the control explanatory due to fact that the available data did not cover
longer period and as a result the degrees of freedom were insufficient.
The various forms of estimators were used when estimating these gravity equations, i.e. the
pooled ordinary least squares (OLS), fixed effects (FE) and random effects (RE). The pooled
OLS impose the restrictions that the unobserved factors are the same across country pairs,
meaning that aij = O. The fixed effects model, also known as the unobserved effects model or the
within transformation model, allows the unobserved country specific effects (e.g. language) to
differ according to the direction of trade, i.e. aij f. aji by assuming that the independent variables
are correlated with the unobserved country specific factors (ail Therefore, the fixed effects
model is a classical regression model that can be estimated using LSDV (least squares with a
dummy variable for each of the country pairs).
The random effects model assumes that the unobserved country specific factors (aij) are not
correlated with the explanatory variables and aij is subsumed into the error term to form a
composite error term as Vijt = aij + ë!il. The ideal assumptions of random effects model include all
of the fixed effects assumptions plus the additional requirement that aij is independent of all
explanatory variables in all time periods. However, due to the fact that aij is in the composite
error in each time period, the Vijl are expected to be serially correlated across time. The feasible
generalised least squares (FGLS) transformation will therefore be used to eliminate the serial
correlation in the errors (Wooldridge, 2002).
After the various models had been estimated, tests for autocorrelation were conducted using the
Durbin-H (in the case of dynamic models) and Durbin Watson (in the case of static models)
statistics. Where necessary, the Cochrane-Orcutt procedure was used to correct the
autocorrelation problem as follows: Firstly, equation 1 was estimated in order to obtain Eijt.
Secondly Et was then regressed on Eijt-I in order to estimate the rho (p) as follows:
Ciil = pCiil-1+ /liil····························································································· (2)
78
Methodology of the Study
Thirdly, the original variables in equation I were transformed using the estimated rho to create a
new estimating equation to be used for the creation of pseudo-GLS estimator. In this case
equation 1 was multiplied by p to obtain:
PY;.il = pao + pa, + paii....+ P(JiilXiil + peiil'··········· .. ···· .. ········ .... ··· .. ··························· (3)
One time period lag was introduced to equation 3 as follows:
pY. = pa + pa + pa ..+ pi] X + pe.. . ···(4)
yl-lOl !I yl yl-I yl-I'
Equation 4 was then subtracted from equation 1 to obtain:
A serially correlated error term is eliminated by substituting Eijt by Eijt-I - !lijt (see equation 2
above) in equation 5 as follows:
(yyl-I -pY y'-lO')=(I-p)a +(I-p)a +(I-p)a U..!l'+ (J.. (X ~II -ptX' yl-I )+1r/yl'... ·····················(6)
Equation 6 is then simplified by redefining each transformed variable as follows:
yy~l =a 0k' +a I k' +a !.I .k' + (Jy.l.X'.yl + r1/y.f.' 1.=1 . ....• .T , ················································(7)
In this case, Y; =Y;il-I - PY;il-l; X~, =(Xiil - pXi/I-I; and k' =1- p. Therefore, in equation 7,
the first-order autocorrelation problem is corrected. All forms of gravity equations, as described
above, were estimated in the form of equation 7.
Once all the models are estimated, a number of statistical tests were undertaken to test the
efficiency of all the estimators. In a nut-shell these tests were used to select the best estimator
suitable for the data. The following steps were followed: In step one, the poolability tests using
79
Methodology of the Study
the Wald statistic were conducted to test the significance of country-specific effects jointly on
trade flows. The Wald test is distributed like the F-statistic as follows:
= (eu'ell - e'e)/(N -1) :::::F (N-ll,(NT-N-K) ......................................•...... (8)
e'e/(NT - N - K)
In this case, ell'ell stands for the restricted residual sum of squares in a least squares regression,
i.e. constrained by Ho. Whereas, e' e stands for the unrestricted residual sum of squares in a least
squares regression, i.e. unconstrained by Ho (see Baltagi, 1995; Green, 2000; Wooldridge, 2002).
The null and alternative hypotheses of the F-test are as follows:
Ho: aij = 0, i.e. country specific effects are jointly insignificant
HI: aij *- aji *- 0, country specific effects are jointly significant
If the joint null that the coefficients are insignificant is rejected, a choice between the fixed effect
and random effects models was made in step two by applying the Hausman's test (Greene, 2000;
Wooldridge,2003). The Hausman test is distributed like Chi-square statistic as follows:
X2[K] = [/In! - Pili;] 'i:-I [Pn! - {JIIIJ··········· ····················································(9)
In this case, l[K] stands for the chi-square with K degrees of freedom whereas {JF!! and /3I1E are
the estimated coefficients of the fixed effects and random effects models respectively. The
symbol L: represents the estimated covariance matrices of the slope estimator in the fixed effects
model minus estimated covariance matrices of the slope estimator in the random effects model,
excluding the constant term. In a nut-shell i: is the covariance matrix of the difference vector
, ,
(/3,;£ - /3I1E)· The null and alternative hypotheses of the Chi-square test are as follows:
Ho: unobserved effects are uncorrelated with other regressors in the model
HI: unobserved effects are correlated with other regressors in the model
The choice between the dynamic and static models was made by looking whether the lagged
variable is significant or not. The suitable model will then be used to calculate the agricultural
80
Methodology of the Study
trade flow potentials with countries between which trade is supposed to have reached its potential
during the periods 2000 to 2004 and 2005 to 2009, i.e. agricultural imports, agricultural exports
and agricultural trade (imports plus exports). In this case, this is between South Africa and EU
countries as well as South Africa and SADC countries due to the implementation of the EU-SA
TOCA and SADC Trade Protocol respectively.
The calculation is done in the following three steps, as outlined by the International Trade Centre
(2000). Firstly, the model is simulated to obtain the unadjusted simulated values of agricultural
trade flows. Secondly, the unadjusted simulated values of agricultural trade flows are corrected
for overall deviation to obtain the adjusted simulated values of agricultural trade flows. Thirdly,
the agricultural trade flow potentials are then obtained by taking the normal average of the
unadjusted simulated and adjusted simulated values of agricultural trade flows. In summary,
after estimating the unadjusted simulated values of agricultural trade flows using equation 1, they
are corrected in equation 10 as follows:
LYik
Yij' = Yij. k£.i, (10)
r,
k .k",.i
Where:
Y;; Adjusted simulated values of agricultural trade flows between countries iandj.
Y;i Unadjusted simulated values of agricultural trade flows between countries i and j.
IY;k Correction factor with the numerator and denominator representing the sum of observed ork,k"'j actual and the sum of the unadjusted simulated bilateral agricultural. trade flows respectively to
a group countries of which the agricultural trade flows should have reached a potential due to
kI,fh;jk regional integration, excluding the residual associated with the destination market.
The agricultural trade flow potentials are then calculated by taking the average on the adjusted
and unadjusted simulated bilateral values of agricultural trade flows as follows:
..................................................................................... ·.. ·.. ··(11)
81
Methodology of the Study
4.6 Data requirements and sources
In order to estimate the above gravity trade equations, secondary data were required. These data
included both time-series and cross-sectional dimensions, thus permitting a special econometric
technique adapted to panel data modelling. South Africa's agricultural trading partners, i.e.
countries, were the cross-sectional units whereas the time series dimension were the years. The
datasets of total agricultural trade flows and the selected agricultural products trade flows
between South Africa and its agricultural trading partners for WTO models covered the period
1994 to 2004, whereas for EU-SA TDCA and SADC Trade Protocol models covered the period
1994 to 2009. The reasons for the different periods of datasets for WTO models compared to
EU-SA TDCA and SADC Trade Protocol models have already been stated above when
substantiating the use of different dummy variables for trade agreements. The WTO datasets
could have covered the period 1994 to 1999, but the datasets for some of the agricultural trade
flows could not allow the model estimation due to insufficient degrees of freedom. Therefore,
the datasets were extended with five more years to 2004 in order to improve the degrees of
freedom.
8:.
The gravity model required data on imports, exports, countries' per capita incomes (GDPPCs),
real effective exchange rates (REER) and distances. The imports and exports data on selected
agricultural products were obtained from the trade databases of the Trade and Investment Policy
Strategies (TIPS) and Eurostat of the European Commission. GDPPCs were obtained from the
World Development Indicators database of the World Bank, International Financial Statistics
Database of the International Monetary Fund (IMF) as well as from United Nations Statistical
Database. REER was obtained from the South African Reserve Bank. Geographical distances
between South Africa and its trading partners (i.e. city to city distance in kilometres) were
obtained from GIS Centre of the Human Sciences Research Council (HSRC) and also from the
Bali and Indonesian website (www.indo.com/distances).
82
Methodology of the Study
4.7 Summary
This chapter provided a detailed review of models of trade policy analysis. These models
included among others market equilibrium models, i.e. partial equilibrium and economy-wide
models, as well as single equation econometric models, i.e. import demand and gravity models.
While providing an overview of these models, emphasis was given on describing the strengths
and the weaknesses of the models with an objective of identifying the suitable model for this
study.
Having identified the gravity model as the suitable one for the study, this chapter, furthermore,
provided a detai led theoretical framework and the specifications of the gravity model. The next
chapter provides the empirical results of the study on the impacts of trade agreements on South
Africa's agricultural trade flows at both aggregate level (sector level imports and exports) and
disaggregate level (product level imports and exports), obtained from estimating the model In
question.
83
, -'t., .:.••,
CHAPTERS ..:~
IMPACTS OF TRADE AGREEMENTS ON THE AGRICULTURAL TRADE FLOWS BETWEEN
SOUTH AFRICA AND ITS AGRICULTURAL TRADING PARTNERS
5.1 Introduction
This chapter presents the empirical results of the impacts of the trade agreements, which were
discussed in Chapter 2, on the agricultural trade flows between South Africa and its
agricultural trading partners at both aggregate and disaggregate levels. The aggregate level of
analysis focuses on the total agricultural trade flows, i.e. total agricultural exports from South
Africa to its trading partners; total agricultural imports from South Africa's trading partners to
South Africa; as well as total. agricultural trade (imports plus ex.ports) between South Africa.. '
and its trading partners. The disaggregate level of analysis focuses on the trade flows of
selected agricultural products, ,which are cheese and curds (HS0406), cut flowers (HS0603),
frozen fruits and nuts -(HS0811), presyrved fruits and nuts (HS2008), fruits and vegetable
juices (HS2009) and wines (HS2204)6. These products were selected because they were
"
given preferential treatment in the trade agreements under analysis.
The following sections of this chapter provide detailed discussion of the results of the impacts
of trade agreements on South Africa's agricultural trade flows at the aggregate and
disaggregate levels. The trade agreements covered are the WTO AoA, EU-SA TDCA and
SADC Trade Protocol (TP) respectively. The impact results focus on the responsiveness of
the aggregate and disaggregate agricultural trade flows between South Africa and its
agricultural trading partners due to the implementation of the trade agreements in question.
That is, whether the implementation of the trade agreements has 'had significant effects on
agricultural trade flows between South Africa and its agricultural trading partners or not. The
analysis goes further in estimating the trade potentials arising from the implementation of the
trade agreements. Further analysis investigates whether the implementation of the trade
agreements has led to trade creation or trade diversion, with emphasis on EU-SA TDCA and
SADC Trade Protocol. That is whether the implementation of the EU-SA TDCA and SADC
Trade Protocol created agricultural trade (imports, exports and total trade) between South
5 In the rest of the study, this product is referred to as "cheese".
6 In the rest of the study, this product is referred to as "wine".
84
-',
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
Africa and the EU and SADC countries or diverted agricultural trade to other South African
trading partners?
The results are set out as follows:
• section 5.2 presents the statistical test results for the selection of the suitable models
for the dataset of each trade flow;
• section 5.3 discusses the effects of the control explanatory variables (i.e. Income,
exchange rates and distance) on agricultural trade flows.
The impact results based on the selected suitable models, which cover the period and yearly
impacts as well as the trade direction and trade potential results in the case of EU-SA TDCA
and SADC Trade Protocol, are presented as follows:
• section 5.4 presents WTO's AoA impact results for the period 1995-1999;
• section 5.5 presents EU-SA TDCA impact results for the periods 2000-2004 and
2005-2009;
• section 5.6 presents SADC Trade Protocol impact results for the periods 2000-2004
and 2005-2009; and
• section 5.7 presents the responsrveness of agricultural trade flows between South
Africa and rest of the world (non-EU and non-SADC countries) during the
implementation of the EU-SA TDCA and SADC Trade Protocol, for the periods
2000-2004 and 2005-2009.
5.2 Statistical tests and selection of the suitable models
This section discusses the procedure followed when selecting the suitable models for all the
trade flows covered in this study. As discussed in Chapter 4, tests for autocorrelation were
conducted on the estimated various models using the Durbin-H (in the case of dynamic
85
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
models) and Durbin- Watson (in the case of static models) statistics. Where necessary, the
Cochrane-Orcutt procedure was used to correct autocorrelation problems. A number of
statistical tests were undertaken to test the efficiency of the estimators for all the models in
order to select the best estimator suitable for the data of each trade flow.
The following steps were followed: In step one, the poolability tests, using the Wald statistic,
were conducted to test the significance of country-specific effects jointly on trade flows.
Where the joint null hypothesis that the coefficients are insignificant was rejected, a choice
between the dynamic FE model or dynamic RE model and a choice between the static FE
model or static RE model were made in step two by applying Hausman's tests (Greene,
2000). In step three, a choice between the dynamic model or the static model was made by
looking at the significance level of the lagged dependent variable, i.e. if significant, then the
dynamic model was preferred or otherwise the static model was preferred, if insignificant.
Using the dataset of South Africa's total agricultural exports to all South Africa's world
agricultural trading partners from 1994 to 2004 as an example, the results of the above
statistical tests for the selection of the suitable model are presented in Table 5.1 below (which
is also set out in Appendix SA). The results show that the poolability tests, which are based
on the Wald statistic, have proved that country-specific effects are important in determining
the total agricultural export from South Africa to all South Africa's agricultural trading
partners in the world. This means that the hypothesis that country effects are jointly
insignificant is to be rejected and has, therefore, led to the conclusion that the pooled OLS
estimator was inefficient.
The autocorrelation tests on the dynamic models for South Africa's agricultural exports to all
world agricultural trading partners have proved that there were no first-order autocorrelation
problems in both FE and RE estimators in all models (i.e. period impact and yearly impact).
However, the autocorrelation tests on the static models have detected first-order
autocorrelation problems in both FE and RE estimators for period impact and yearly impact
models. After correction, using the Cochrane-Orcutt procedure, the autocorrelation problem
disappeared only in the FE estimators, while the problem was still present in the RE
estimators.
86
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
,
Table 5.1: Selection of the Estimator suitable for Agricultural Exports from South
Africa to the World
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Wald Period Impact 1190 5 3.69* OLS No 1309 4 22.44* OLS No
Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 1190 9 3.67* OLS No 1309 8 22.76* OLS No
FE or RE Yes FE or RE Yes
Durbin Period Impact 1190 5 -0.62 FE-no auto Yes 1309 4 1.46 FE-no auto No
Watson - - - FE-auto - 1190 4 1.89** FE-auto Yes
Statistic 1190 6 -1.31 RE-no auto Yes 1309 6 1.21 RE-no auto No
(jar - - - RE-auto - 1190 6 1.81 RE-auto No
static) or Yearly Impact 1190 9 -0.56 FE-no auto Yes 1309 8 1.46 FE-no auto No
Durbin-H - - - FE-auto - 1190 8 1.92** FE-auto Yes
Statistic 1190 10 -1.34 RE-no auto Yes 1309 9 1.22 RE-no auto No
(for - - - RE-auto 1190 9 1.84 RE-auto No
dynamic) -
Hausman Period Impact N/A 5 471.79* FE Yes - - - FE -
Test RE No RE -
Statistic Yearly Impact N/A 9 464.07* FE Yes - - - FE -
RE No RE -
~B: " •• & ••• denote significance at the 1.,5 and 10 percent levels respectively. N & K denote the sample size and the number of regressors respectively. OLS,
FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. "no auto & -auto denote estimation assuming no
autocorrelation problem and estimation with the correction of I" order autocorrelation problem respectively.
Source: Author's calculations
The tests for unobserved effects, based on the Hausman statistics, in all the dynamic models
have proved that the hypothesis, that unobserved country specific effects are uncorrelated
with explanatory variables, cannot be accepted at the 99% level, and therefore indicate that
the FE estimators were efficient. However, due to the insignificant autocorrelation tests on
the static RE estimators for both periods, the Hausman statistics were not calculated and in
this case only the FE estimators for static period and yearly impact models were selected.
The statistical tests results which established the selection of the suitable models, as well as
the selected suitable dynamic and static models for all agricultural trade flows (exports,
imports and trade) between South Africa and its agricultural trading partners, are presented in
their respective Appendices as shown in Table 5.2 below.
87
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
Table 5.2: Appendices for the statistical tests results towards the selection of the model
suitable for datasets of agricultural trade flows between South Africa and its
agricultural trading partners
Trade Flows WTOAoA EU-SA TDCA SADC TP ROW
Appendices Appendices Appendices Appendices
Agric X 5A and 5B 5AQ and 5AR 50B and 50C 5EU and 5EY
Agric M 5C and 50 5AT and 5AU 50E and 50F 5EW and 5EX
Agric T 5E and 5F 5AW and 5AX 50H and 501 5EY and 5EY
HS0406 X 5G and 5H 5Al and 5BA 50K and 50L 5FA and 5FB
HS0406 M 51 and 5J 5BC and 5BO - 5FA and 5FB
HS0406 T 5K and 5L 5BF and 5BG - 5FC and 5FO
HS0603 X 5M and 5N 5BI and 5BJ 50N and 500 5FE and 5FF
HS0603 M 50 and 5P 5BL and 5BM 50Q and 50R 5FG and 5FH
HS0603 T 5Q and 5R 5BO and 5BP 50T and 50U 5Fl and 5FJ
HS0811 X 5S and 5T 5BR and 5BS 50W and 50X 5FK and 5FL
HS0811 M 5U and 5Y 5BU and 5BY - 5FM and 5FN
HS0811 T 5W and 5X 5BX and 5BY - -
HS2008 X 5Y and sz 5CA and 5CB 50l and 5EA 5FO and 5FP
HS2008 M 5AA and 5AB 5CO and 5CE - 5FQ and 5FR
HS2008 T 5AC and 5AO 5CG and 5CH - 5FS and 5FT
HS2009 X 5AE and 5AF 5CJ and 5CK 5EC and 5EO 5FU and 5FY
HS2009 M 5AG and 5AH 5CM and 5CN 5EF and 5EG 5FW and 5FX
HS2009 T 5AI and 5AJ 5CP and 5CQ 5EI and 5EJ 5FY and 5Fl
HS2204 X 5AK and 5AL 5CS and 5CT 5EL and 5EM 5GA and 5GB
HS2204 M 5AM and 5AN 5CY and 5CW 5EO and 5EP 5GC and 5GO
HS2204 T 5AO and 5AP 5CY and 5Cl 5ER and 5ES 5GE and 5GF
NB: X, M and T stand for exports, imports and trade. AGRIC - Total agriculture, I-IS0406 - Cheese & curd, I-IS0603 - Cut flowers,
I-IS0811 - Frozen fruits & nuts, HS2008 - Preserved fruits & nuts, HS2009 - Fruits & vegetable juices, & HS2204 - Wines.
Sou rce: Author's calculations
The results of the selected suitable models are presented in Table 5.3 below. These results
have shown that 189 models, in total, were found to be efficient and suitable for the datasets
of the selected agricultural trade flows, of which 161 were dynamic models and 28 were static
models. With regard to the dynamic models; 152 FE, 2 RE and 7 pooled OLS estimators
were found to be efficient and suitable; whereas 14 FE and 14 RE estimators were found to be
efficient and suitable for the static models.
Furthermore, 42 estimators were found to be suitable for the selected WTO AoA models,
which consisted of 38 dynamic models (with 38 FE estimators selected as efficient and
suitable) and 4 static models (with 2 FE and 2 RE estimators selected as efficient and
suitable). Regarding the selected ROW models, 38 dynamic FE estimators were found to be
efficient and suitable. With regard to the selected EU-SA TDCA models, 63 estimators were
found to be suitable and consisted of 53 dynamic models (with 51 FE and 2 OLS estimators
selected as efficient and suitable) and 10 static models (with 7 FE and 3 RE estimators
selected as efficient and suitable). With regard to the selected models for the SADC Trade
88
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
Protocol, 46 estimators were found to be suitable, which consisted of 32 dynamic models
(with 25 FE, 2 RE and 5 OLS estimators selected as efficient and suitable) and 14 static
models (with 5 FE and 9 RE estimators selected as efficient and suitable).
Table 5.3: Results for the selection of the model suitable for datasets of agricultural
trade flows between South Africa and its agricultural trading partners
Trade Flows Period Impact Yearly Impact Direction Impact
Model Model Model
WTO EU-SA SADC ROW WTO EU-SA SADC ROW EU-SA SADC
AoA TDCA TP AoA TDCA TP TDCA TP
Detailed results
Agric X DFE DFE SFE DFE DFE DFE DFE DFE DFE DFE
Agric M DFE DFE DFE DFE DFE DFE SRE DFE DFE DFE
Agric T DFE DFE DFE DFE DFE DFE DRE DFE DFE DFE
HS0406 X DFE DFE DOLS DFE DFE DFE DOLS DFE DFE DFE
HS0406 M DFE DFE - DFE DFE DFE - DFE DFE -
HS0406 T DFE DFE - - DFE DFE - - DFE -
HS0603 X DFE DFE SFE DFE DFE DFE SFE DFE DFE DFE
HS0603 M DFE SRE DFE DFE DFE SRE DFE DFE DFE DFE
HS0603 T DFE DOLS SRE DFE DFE DOLS SRE DFE DFE DFE
HS08II X DFE DFE DFE DFE DFE DFE DFE DFE DFE DFE
HS0811 M DFE DFE - DFE DFE DFE - DFE DFE -
HS0811 T DFE SFE - - DFE SFE - - DFE -
HS2008 X DFE DFE DFE DFE DFE DFE DFE DFE DFE DFE
HS2008 M DFE SFE - DFE DFE SFE - DFE DFE -
HS2008 T SRE DFE - DFE SRE DFE - DFE SRE -
HS2009 X DFE DFE DOLS DFE DFE DFE DOLS DFE DFE DFE
HS2009 M DFE DFE SRE DFE DFE DFE SRE DFE DFE DFE
HS2009 T SFE DFE DOLS DFE SFE DFE SFE DFE SFE SFE
HS2204 X DFE DFE DFE DFE DFE DFE DRE DFE DFE DFE
HS2204 M DFE SFE SRE DFE DFE SFE SRE DFE DFE DFE
HS2204 T DFE DFE SRE DFE DFE DFE SRE DFE DFE DFE
Summarised results
DATASETS FOR: DFE ORE DOLS SFE SRE TOTAL
WTO AoA Trade Flows 38 0 0 2 2 42
ROW Trade Flows 38 0 0 0 0 38
EU-SA TDCA Trade Flows 51 0 2 7 3 63
SADe TP Trade Flows 25 2 5 5 9 46
TOTAL 152 2 7 14 14 189
161 28
NB: X, M and T stand for exports, imports and trade. AGRIC - Total agriculture, I-IS0406 - Cheese & curd, I-IS0603 - Cut flowers, I-IS0811 -
Frozen fruits & nuts, HS2008 - Preserved fruits & nuts, I-IS2009 - Fruits & vegetable juices, & I-IS2204 - Wines. OFE, ORE, OOLS stand for
dynamic fixed effects, dynamic random effects and dynamic ordinary least squares; whereas SFE and SRE stand for static fixed effects
and static random effects.
Source: Author's calculations
The dominance of the dynamic models has justified the importance of dynamics in trade and
supports many economic arguments which have suggested that past trade is the predictor for
current trade. This means that the historical relationships between South Africa and its
89
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
trading partners have played a significant role in determining the majority of current
agricultural trade flows between South Africa and its agricultural trading partners.
Furthermore, the poolability tests have indicated that the joint country-pair specific effects
were not important in the majority of selected models. These findings conform to the
expectations of De Grauwe and Skudelny (2000) and Eiehengreen and Irwin (1997), as they
have found that both the dynamics (or past trade) and the unobserved country-pair specific
effects were indeed important in determining the majority of current agricultural trade flows
between South Africa and its agricultural trading partners.
For example, using the results for agricultural trade flows between South Africa and the world
as presented in Table 5.7, these results showed that the previous year's total agricultural
exports (i.e. InYijt-1) had significantly improved the current total agricultural exports from
South Africa to the world by 0.34%. Similarly, the current total agricultural imports from the
world to South Africa as well as the current total agricultural trade (exports plus imports)
between South Africa and the world have also improved significantly by 0.17% and 0.33%
respectively owing to the importance of dynamics in trade. The results of dynamic effects on
all agricultural trade flows between South Africa and its trading partners should be interpreted
in a similar manner as above.
5.3 Effects of the control explanatory variables on agricultural trade flows
The control explanatory variables that were included when estimating models are: per capita
Gross Domestic Products of South Africa and of its trading partners (GDPPCs); real effective
exchange rates (REER); and geographic distances between South Africa and its trading
partners (DIST). Their effects on agricultural trade flows between South Africa and its
agricultural trading partners are discussed in the following subsections.
5.3.1 Gross Domestic Product per Capita (GDPPC)
The results for the income effects on the agricultural trade flows between South Africa and its
agricultural trading partners are presented in Table 5.4 below. In total, 56 agricultural trade
flows between South Africa and its agricultural trading partners were positively affected by
the income and 5 of them were negatively affected by income. However, 68 agricultural trade
flows between South Africa and its agricultural trading partners were not affected by the
90
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
Income. The per capita GDP of South Africa (GDPPC1) had positively affected 12
agricultural export flows and 12 agricultural import flows between South Africa and its
agricultural trading partners, but had negatively affected 2 agricultural export flows. On the
other hand, the per capita GDP of South Africa's trading agricultural partners (GDPPCJ) had
positively affected 9 agricultural export flows and 20 agricultural imports flows between
South Africa and its agricultural trading partners. Similarly, the sum of per capita GDPs of
South Africa and its agricultural trading partners (GDPPCIJ) had also positively affected 12
total agricultural trade (exports plus imports) flows between South Africa and its agricultural
trading partners and negatively affected 2 total agricultural trade flows.
The results of the total agricultural trade flows between South Africa and the world as
presented in Table 5.7 indicate that the per capita GDPs have played a significant positive role
in determining agricultural trade flows between South Africa and the world. On average, a
1% increase in South Africa's GDP per capita led to an increase of between 11.83% and
12.66% in South Africa's agricultural exports to the world; as well as an increase of between
]6.39% and 16.47% in South Africa's agricultural imports from the world. The results have
also shown that a 1% increase in per capita GDPs of South Africa's agricultural trading
partners in the world led to a 0.4% increase in South Africa's agricultural exports to the
world. In cases where the coefficient of the GDPPC is significantly negative, it shows that a
]% increase in South Africa's GDP per capita led to a decrease of such trade flow between
South Africa and its trading partners. For example, an examination of the results of the cut
flowers trade flows between South Africa and the EU countries as presented in Table 5.17
shows that the percentage increase in the sum of per capita GDPs of South Africa and EU
countries had negatively affected total cut flowers trade between South Africa and the EU
countries. On average, a 1% increase in the sum of per capita GDPs of South Africa and EU
countries led to decrease of between 5.99% and 6.05% in total cut flowers trade between
South Africa and the EU countries. The coefficients of the GDPPCs for other agricultural
trade flows should be interpreted similarly as above, taking into account that insignificant
coefficient means percentage increase in GDPPC had no effects on such agricultural trade
flow between South Africa and its agricultural trading partners.
91
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
Table 5.4: Results for the income effects on agricultural trade flows between South
Africa and its agricultural trading partners
Trade GDPPC, GDPPCJ GDPPCIJ
Flows (South Africa) (Trading Partners) (South Africa + Trading
Partners)
World EU SADe ROW World EU SADe ROW World EU SADe ROW
Detailed results
Agric X + + 0 + + + 0 0 na na Na Na
Agric M + 0 + + + + + + na na Na Na
Agric T na na na Na na na Na na + + 0 +
HS0406 X 0 0 0 - 0 0 0 + na na Na Na
HS0406 M 0 + na + 0 0 Na 0 na na Na Na
HS0406 T na na na Na na na Na na 0 0 Na Na
HS0603 X 0 0 + - 0 0 0 0 na na Na Na
HS0603 M 0 0 + 0 0 + 0 + na na Na Na
HS0603 T na na na na na na Na na + - + +
HS0811 X 0 0 0 0 0 + 0 0 na na Na Na
HS0811 M 0 0 na 0 + 0 Na + na na Na Na
HS081] T na na na na na na Na na 0 0 Na Na
HS2008 X 0 + 0 0 0 0 0 + na na Na Na
HS2008 M 0 0 na + + 0 Na 0 na na Na Na
HS2008 T na na na na + na Na na 0 - na 0
HS2009 X + + + + + 0 + + na na na na
HS2009 M + 0 + + 0 0 + 0 na na na na
HS2009 T na na na na na na Na na + 0 0 +
HS2204 X 0 + + + ·0 + + 0 na na na na
HS2204 M 0 0 + + na + 0 0 na na na na
HS2204 T na na na na na na Na na + + + +
Summarised results
EFFECTS GDPPC,- X GDPPC,- M GDPPCJ - X GDPPC,,- M GDPPC",-T TOTAL
Positive effects 12 12 9 11 12 56
Negative effects 2 0 1 0 2 6
No effects 14 13 18 14 9 68
NB: X, M and T stand for exports, imports and trade respectively. AGRIC - Total agriculture, I-IS0406 - Cheese & curd, I-IS0603 - Cut flowers,
I-IS0811 - Frozen fruit, & nuts, I-IS2008 - Preserved fruits & nuts, I-IS2009 - Fruits & vegetable juices, & I-IS2204 - Wines. + means positive effects,
and - means negative effects, and (I means no effects. na means not applicable
Source: Author's calculations
The results have clearly shown that agricultural exports, imports and total trade flows between
South Africa and its agricultural trading partners are income-elastic. Therefore, these results
support the a priori expectation that countries with a higher income per capita would trade
more intensively (have more exports and imports) than a poorer country because the per
capita incomes of the trading countries reflect in both the productive and consumption
capacity which heavily determine the trade flow between them. In some cases where the
income effects are negative, the reason may be that the majority of countries included were
poorer and therefore might have spent much of their increased extra income on internal
developmental programs such as infrastructural development projects rather than increasing
92
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
their imports from South Africa and other countries. Thus, South Africa's exports to these
poorer countries had declined even though their income had increased. Alternatively, these
poorer countries might have decided to purchase agricultural products locally in order to
boost their local economies.
5.3.2 Real Effective Exchange Rates (REER)
The results for the real effective exchange rates effects on the agricultural trade flows between
South Africa and its agricultural trading partners are presented in Table 5.5 below. In total,
13 agricultural trade flows between South Africa and its agricultural trading partners were
positively affected by the exchange rates and 11 of them were negatively affected. On the
other hand, 52 agricultural trade flows between South Africa and its agricultural trading
partners were not affected by the exchange rates. The results have indicated that the real
effective exchange rates had positively affected 4 agricultural export flows, 7 agricultural
import flows and 2 total agricultural trade (exports plus imports) flows between South Africa
and its agricultural trading partners. However, they had negatively affected 7 agricultural
export flows and 4 total agricultural trade (exports plus imports) flows between South Africa
and its agricultural 'trading partners.
It should be noted that during the period under review the exchange value of the South Africa
currency (Rand) had been appreciating against a basket of currencies in real terms (Motsumi
et al, 2008). Using the results of the total agricultural trade flows between South Africa and
the world as presented in Table 5.7, South Africa's total agricultural exports to the world had
declined by about 0.6% owing to the appreciation in the real effective exchange rate. On the
other hand, the total agricultural imports and aggregate agricultural trade between South
Africa and its agricultural trading partners had not been affected by the appreciation of the
real effective exchange rate. An examination of the results of the cut flowers trade flows
between South Africa and the EU countries as presented in Table 5.17, shows that the
appreciation of the real effective exchange rate had increased South Africa's cut flowers
imports from the EU countries by between 6.39% and 7.7%. In this case, exports suffered
while imports gained owing to the appreciation of the South Africa Rand against a basket of
currencies. The effects of exchange rates on the total agricultural trade flows (imports plus
exports) might take any direction depending on which trade flow dominates this variable, i.e.
93
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
whether the import value is higher than the export value or vice versa. For example, the
results presented in Table 5.17 show that South Africa's total cut flowers trade with the EU
countries had declined by between 0.69% and 0.72% owing to the appreciation of the real
effective exchange rate. In contrast, the results presented in Table 5.33 show that South
Africa's total cut flowers trade with the ROW countries had increased by 1.23% owing to the
appreciation of the real effective exchange rate.
Table 5.5: Results for the exchange rates effects on agricultural trade flows between
South Africa and its agricultural trading partners
Trade Flows World EU SADC ROW
Detailed results
Agric X - + 0 -
Agric M 0 0 0 0
Agric T 0 0 0 -
HS0406 X - 0 0 -
HS0406 M + 0 Na +
HS0406 T 0 0 Na Na
HS0603 X 0 0 0 -
HS0603 M 0 + + 0
HS0603 T 0 - - +
HS0811 X - 0 0 0
HS0811 M 0 + na 0
HS0811 T 0 0 na Na
HS2008 X 0 + + 0
HS2008 M 0 0 na 0
HS2008 T 0 0 na 0
HS2009 X 0 0 0 0
HS2009 M 0 0 0 0
HS2009 T 0 0 0 0
HS2204 X - 0 + -
HS2204 M + 0 0 +
HS2204 T + 0 0 0
Summarised results
EFFECTS Exports (X) Imports (M) Trade (T) TOTAL
Positive effects 4 7 2 13
Negative effects 7 0 4 11
No effects 16 18 18 52
NB: X, M and T stand for exports, imports and trade respectively. + means positive effects, and - means negative effects,
and 0 means no effects. na means not applicable
Source: Author's calculations
Certain research findings by Bergstrand (1985) and Koo et al (1994) have shown that an
appreciation ofthe importing country's currency against its trading partners' currencies would
impact positively on imports because imports become less expensive, whereas depreciation
would affect imports negatively since they become more expensive. On the other hand, a
depreciation of the exporting country's currency against its trading partners' currencies means
94
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
that its exports become cheaper in the world markets, thus impacting positively on exports.
Given these findings, these results have clearly shown that the effects of the real effective
exchange rates on the majority of these agricultural trade flows have met the a priori
expectations. However, there are few cases were the effects of the real effective exchange
rate on exports flows did not met the expectations, for example, total agricultural exports and
preserved fruits and nuts exports to the EU countries as wel! as preserved fruits and nuts
exports and wine exports to the SADe countries. The coefficients of the REER for other
agricultural trade flows should be interpreted similarly as above.
5.3.3 Distance (DIST)
The results for the distance effects on the agricultural trade flows between South Africa and
its agricultural trading partners are presented in Table 5.6 below. Owing to the fact that FE
estimators were suitable and efficient for the majority of selected models as shown in Table
5.3 above, the distance factor was eliminated in these models. As a result, only 11 selected
models had a distance included as one of the explanatory variables. The results have shown
that, in total, the distance had negatively affected 8 agricultural trade flows between South
Africa and its agricultural trading partners, but only 1 agricultural trade flow was positively
affected. However, 2 agricultural trade flows were not affected by distance. For example, the
results of cheese exports from South Africa to the SADe countries as presented in Table 5.25,
show that the distance between South Africa and SADe countries contributed to the
significant decrease of cheese exports from South Africa to the SADe countries by 0.41%.
Furthermore, an examination of the results of the cut flowers trade flows between South
Africa and the SADe countries as presented in Table 5.26, shows that the distance between
South Africa and SADe contributed to a significant decrease of total cut flowers trade
between South Africa and the SADe countries of between 3.03% and 3.15%. The
coefficients of the DIST for other agricultural trade flows should be interpreted in this
manner.
The overall effects of distance on agricultural trade flows between South Africa and its
agricultural trading partners were in line with the a priori expectations because distance
between trading partners is regarded as a proxy for transportation costs, meaning that
countries with short distances between each other are expected to trade more than those which
95
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
are far apart owing to lower transaction cost. Furthermore, distance can also be used as a
proxy for the risks associated with the quality of some perishable goods and for the cost of
personal contact between managers and customers (ITC, 2000; Cheng and Wall, 2005;
Gebrehiwet et al, 2007). Given the geographic location of South Africa in relation to its
major trading partners, as well as the fact that most of the agricultural products traded are
perishable, it was expected that distance would have negative effects on South Africa's
exports and imports of agricultural products.
Table 5.6: Results for the distance effects on agricultural trade flows between South
Africa and its agricultural trading partners
Trade Flows World EU SADC ROW
Detailed results
Agric X na na Na na
Agric M na na Na na
Agric T na na - na
HS0406 X na na - na
HS0406 M na na na na
HS0406 T na na na na
HS0603 X na na na na
HS0603 M na - na na
HS0603 T na + - na
HS0811 X na na na 0
HS0811 M na na na na
HS0811 T na na na na
HS2008 X na na na na
HS2008 M na na na na
HS2008 T na na na na
HS2009 X na na - na
HS2009 M na na - na
HS2009 T na na - na
HS2204 X na na - na
HS2204 M na na 0 na
HS2204 T na na na na
Summarised results
EFFECTS Exports (X) Imports (M) Trade (T) TOTAL
Positive effects 0 0 I 1
Negative effects 3 2 3 8
No effects I I 0 2
NB: X, M and T stand for exports, imports and trade respectively. AGRIC - Total agriculture, HS0406 - Cheese & curd, HS0603 - Cut
flowers, I-IS081I - Frozen fruits & nuts. HS2008 - Preserved fruits & nuts, I-IS2009 - Fruits & vegetable juices, & I-IS2204 - Wines. +
means positive effects, and - means negative effects, and 0 means no effects. na means not applicable
Source: Author's calculations
96
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
5.4 The impacts of the WTO AoA on selected agricultural trade flows between South
Africa and its world agricultural trading partners
This section discusses the impacts of the multilateral trade agreement, the Agreement of
Agriculture (AoA) of the World Trade Organization (WTO), on the total agricultural trade
flows and the selected agricultural products trade flows between South Africa and the world
(i.e. all South Africa's agricultural trading partners including the EU and SADe member
states). In this case, the datasets cover all the trade flows (i.e. exports, imports and total trade)
of the total agriculture and all the selected agricultural products.
5.4.1 Aggregate agricultural trade flows between South Africa and the world
The results for total agricultural trade flows (exports, imports and trade) between South
Africa and the world are presented in Table 5.7 below. As shown in this table, the cross-
sections information refers to the number of countries that traded agricultural products with
South Africa during the period under review (1994 to 2004), which are referred to as South
Africa's agricultural trading partners in the rest of this study. For example, this table shows
that South Africa exported agricultural products to 119 countries in the world and imported
agricultural products from 127 countries during the period under analysis. Furthermore,
South Africa traded agricultural products with 109 countries in the world, i.e. during this
period South Africa exported agricultural products to these countries and also imported
agricultural products from them during the same period. All the adjusted R2 values confirm
that the predictive power of the selected models is good, i.e. the selected variables "best fit"
the model. The cross-sections information as well as the adjusted R2 estimates for all the
other selected agricultural trade flows, as presented in their respective Tables below, should
be interpreted similarly as above.
The results from the period impact model indicate that during the implementation of the
WTO's AoA from 1995 to 1999, South Africa's agricultural exports to the world had
significantly increased by 0.44%. On annual basis, the most significant increase in South
Africa's agricultural exports to the world occurred in 1999 by 0.42%. On the other hand, total
agricultural trade (imports plus exports) between South Africa and the world had significantly
7 Note: In interpreting all the results, exports refer to the value of exports, imports refer to the value of imports
and trade refers to the value of trade.
97
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
declined by 0.23% and on annual basis, significant decreases occurred in 1998 and 1999 by
0.27% and 0.22% respectively. However, the results showed that South Africa's agricultural
imports from the world did not respond significantly to the implementation of the WTO's
AoA.
Table 5.7: Results for agricultural trade flows between South Africa and the world,
1994-2004
Variables Period Impact Model Yearly Impact Model
Ex~orts Im~orts Trade Ex~orts Im~orts Trade
Constant
InYiit-1 0.34* 0.17* 0.33* 0.34* 0.17* 0.33*
InGOPPCit 12.66* 16.39* 11.83* 16.47*
InGOPPCit 0.40*** 0.05 0.38 0.03
InGOPPCiit -0.02 -0.04
REERt -0.60** 0.46 -0.15 -0.14 0.72 -0.22
09599 0.44* -0.25 -0.23**
095 -0.08 -0.34 -0.23
096 0.36 -0.53 -0.21
097 0.16 -0.31 -0.06
098 0.18 -0.42 -0.27**
D99 0.42** -0.18 -0.22**
InDISTï
Adjusted R2 0.82 0.75 0.88 0.82 0.75 0.88
Observations 1190 1270 1090 1190 1270 1090
Cross-Sections 119 127 109 119 127 109
*, ** & *** denote significance at the 1,5 and 10 l2ercent levels resl2ectivel~.
Source: Author's calculations
5.4.2 Cheese trade flows between South Africa and the world
The results for all cheese trade flows (exports, imports and trade) between South Africa and
the world are presented in Table 5.8 below. The results from the period impact model
indicate that during the implementation of the WTO's AoA for the period 1995 to 1999,
South Africa's cheese exports to the world had significantly increased by 1.23% and, on
annual basis, the most significant increase in South Africa's cheese exports to the world
occurred in 1997 by 2.08%. Surprisingly, while the joint period effect of the implementation
of the WTO's AoA on total cheese trade (imports plus exports) was insignificant, the
individual yearly significant positive effects were observed in 1996 and 1997, as total cheese
trade between South Africa and its cheese trading partners in the world had improved by
1.11% and 0.93% respectively during these years. However, the results showed that South
Africa's cheese imports from the world did not respond significantly to the implementation of
the WTO'sAoA.
98
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
Table 5.8: Results for cheese trade flows between South Africa and the world, 1994-2004
Variables Period Impact Model Yearly Impact Model
Ex~orts Im~orts Trade Ex~orts Im~orts Trade
Constant -15.15
InYiit-1 0.30* 0.38* 0.31 * 0.38*
InGOPPCit -1.41 31.18** -8.10 16.61
InGOPPCit -0.43 1.00 -0.49 1.03
InGOPPCiit 8.74* 4.61
REERt -3.67* 3.82* 1.50 -4.12** 3.58*** 0.80
09599 1.23*** -0.10 0.45
095 1.31 -1.37 -0.02
096 0.57 0.08 1.11**
097 2.08** 0.03 0.93***
098 0.81 -0.21 0.06
099 1.21 -1.18 0.10
InOlSTï -7.43**
Adjusted R2 0.67 0.76 0.47 0.67 0.76 0.83
Observations 320 240 100 320 240 110
Cross-Sections 32 24 10 32 24 10
*, ** & ***denote significance at the 1,5 and 10~ercent levels res~ectivel~.
Source: Author's calculations
5.4.3 Cut flowers trade flows between South Africa and the world
The results for all cut flowers trade flows (exports, imports and trade) between South Africa
and the world are presented in Table 5.9 below. The results from the period impact model
indicate that during the implementation of the WTO's AaA over the period 1995 to 1999,
South Africa's cut flowers exports to the world had significantly increased by 0.92% and, on
annual basis, the significant increase in South Africa's cut flowers exports to the world
occurred only in 1999 by 1.15%.
Surprisingly, while the joint period effect of the implementation of the WTO's AaA on cut
flowers imports was insignificant, the individual yearly significant negative effects were
observed in 1996 and 1997, when South Africa's cut flowers imports had declined by 2.78%
and 1.89% respectively. However, the results showed that total cut flowers trade between
South Africa and its worldwide cut flowers trading partners did not respond significantly to
the implementation of the WTO's AaA.
99
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
Table 5.9: Results for cut flowers trade flows between South Africa and the world, 1994-
2004
Variables Period Impact Model Yearly Impact Model
Ex~orts Im~orts Trade Ex~orts Im~orts Trade
Constant
InYiit-1 0.34* 0.18* 0.24* 0.33* 0.17* 0.20*
InGOPPCit 7.81 5.78 8.89 12.59
InGOPPCit 1.46** 0.98 1.41*** 0.64
InGOPPCiit 4.32 2.45
REERt -1.69* 1.74 -0.23 -1.10 4.81 * 0.66
09599 0.92* -0.55 0.36
095 0.52 -2.15 -0.82
096 0.73 -2.78** -0.21
097 0.61 -1.89*** 0.07
098 0.45 -1.08 -0.14
099 1.15* -0.22 0.24
InOISTj'
Adjusted R2 0.77 0.57 0.80 0.77 0.58 0.80
Observations 560 260 220 560 260 220
Cross-Sections 56 26 22 56 26 22
*, ** & *** denote significance at the 1,5 and 10 Eercent levels resQectivel~.
Source: Author's calculations
5.4.4 Frozen fruits and nuts trade flows between South Africa and the world
The results for all frozen fruits and nuts trade flows (exports, imports and trade) between
South Africa and the world are presented in Table 5.10 below. The results from the period
impact model indicate that the implementation of the WTO's AoA over the period 1995 to
1999 had no effects on all frozen fruits and nuts trade flows (exports, imports and trade)
between South Africa and the world. Similar results were observed from the yearly impact
model with the exception oftotal frozen fruits and nuts trade (imports plus exports) where the
individual yearly significant positive effects were observed in 1997 when total frozen fruits
and nuts trade between South Africa and its frozen fruits and nuts trading partners in the
world improved by 3.78%.
Surprisingly, while the joint period effect of the implementation of the WTO's AoA on cut
flowers imports was insignificant, the individual yearly significant negative effects were
observed in 1996 and 1997, when South Africa's cut flowers imports had declined by 2.78%
and 1.89% respectively. However, the results showed that total cut flowers trade between
South Africa and its worldwide cut flowers trading partners did not respond significantly to
the implementation of the WTO's AoA.
100
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
Table 5.10: Results for frozen fruits and nuts trade flows between South Africa and the
world, 1994-2004
Variables Period Impact Model Yearly Impact Model
Ex~orts Im~orts Trade Ex~orts Im~orts Trade
Constant 53.36 81.07
InYiit-1 0.15* 0.48* 0.29* 0.18* 0.49* 0.30*
InGOPPCit 7.16 -7.48 -1.81 -10.19
InGOPPCit -0.32 -0.44 -0.26 -0.42***
InGOPPCiit 1.19 3.93
REERt -1.83 3.52 0.35 -4.20*** 2.05 -2.46
09599 0.11 -1.07 0.52
095 1.93 -0.34 2.60
096 0.00 -1.34 1.43
097 2.24 0.03 3.78**
098 -0.20 1.98 1.51
099 0.38 -2.52 0.34
InOlSTï -0.18 -0.15
Adjusted R2 0.41 0.27 0.41 0.42 0.29 0.42
Observations 210 230 110 210 230 110
Cross-Sections 21 23 11 21 23 11
*, ** & *** denote significance at the 1,5 and 10 Ecrcent levels resEectively.
Source: Author's calculations
5.4.5 Preserved fruits and nuts trade flows between South Africa and the world
The results for all preserved fruits and nuts trade flows (exports, imports and trade) between
South Africa and the world are presented in Table 5.11 below.
Table 5.11: Results for preserved fruits and nuts trade flows between South Africa and
the world, 1994-2004
Variables Period Impact Model Yearly Impact Model
Ex~orts Im~orts Trade Ex~orts lm~orts Trade
Constant 0.96
InYiit_1 0.27* 0.30* 0.27* 0.28*
InGOPPCit 15.53** 25.96** 19.10** 26.57***
InGOPPCit 3.36* 1.98 3.49* 1.31
InGOPPCiit 3.32 3.28*
REERt -0.29 0.33 -0.60 -0.80 5.23* -0.92
09599 0.38 0.15 0.18
095 1.15 -3.53* 0.46
096 0.75 -3.30* 0.23
097 0.53 -2.53** 0.02
098 0.72 -0.58 -0.17
099 0.55 -0.69 -0.20
InOlSTï -1.62*
Adjusted R2 0.67 0.55 0.73 0.67 0.57 0.74
Observations 850 380 352 850 380 320
Cross-Sections 85 38 32 85 38 32
*, ** & *** denote significance at the 1,5 and 10 Ecrcent levels resEectivel~.
Source: Author's calculations
101
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
The results from the period impact model indicate that the implementation of the WTO's AoA
during the period 1995 to 1999 had no effects on all preserved fruits and nuts trade flows
(exports, imports and trade) between South Africa and the world. Similar results were
observed from the yearly impact model with the exception of frozen fruits and nuts imports
where individual yearly significant negative effects were observed in 1995, 1996 and 1997, as
South Africa's preserved fruits and nuts imports from its frozen fruits and nuts trading
partners in the world had declined by 3.53%, 3.30% and 2.53% in 1995, 1996 and 1997
respectively.
5.4.6 Fruits and vegetable juices trade flows between South Africa and the world
The results for all fruits and vegetable juices trade flows (exports, imports and trade) between
South Africa and the world are presented in Table 5.12 below. The results from the period
and yearly impact models indicate that the implementation of the WTO's AoA over the period
1995 to 1999 had no effects on fruits and vegetable juices exports and imports between South
Africa and the world.
Table 5.12: Results for fruits and vegetable juices trade flows between South Africa and
the world, 1994-2004
Variables Period Impact Model Yearly Impact Model
Ex~orts Im~orts Trade Ex~orts Im~orts Trade
Constant
InYiit_1 0.21 * 0.21 * 0.09*** 0.21 * 0.20* 0.09***
InGOPPCit 13.82*** 28.64** 8.37 43.48*
InGOPPCjt 2.00** 2.20 2.43* 1.97
InGOPPCijt 3.83** 4.27**
REERt 0.35 -0.17 0.71 -1.92*** 1.34 -0.33
09599 -0.31 0.44 -0.54***
095 1.12 0.12 0.19
096 0.91 -0.27 -0.08
097 0.86 -0.45 0.35
098 0.27 1.38 -0.11
099 -0.61 0.87 -0.65**
InDlSTj'
Adjusted R2 0.66 0.56 0.65 0.67 0.56 0.66
Observations 860 410 400 860 410 400
Cross-Sections 86 41 40 86 41 40
*, **& ***denote significance at the 1,5 and 10percent levels resEectively.
Source: Author's calculations
However, the total fruits and vegetable juices trade (imports plus exports) had significantly
declined by 0.54% during the implementation period and the individual yearly significant
negative effects were observed in 1999, when total fruits and vegetable juices trade between
102
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
South Africa and its fruits and vegetable juices trading partners in the world declined by
0.65%.
5.4.7 Wine trade flows between South Africa and the world
The final results for all wine trade flows (exports, imports and trade) between South Africa
and the world are presented in Table 5.8 below. The results from the period impact model
indicate that the implementation of the WTO's AaA from 1995 to 1999 had significantly
improved all the wine trade flows between South Africa and the World. For example, South
Africa's wine exports to the world significantly increased by 0.75% during the
implementation period and on annual basis, the significant increase in South Africa's wine
exports tothe world occurred in 1999, at 0.71%.
Table 5.13: Results for wine trade flows between South Africa and the world, 1994-2004
Variables Period Impact Model Yearly Impact Model
Ex~orts Im~orts Trade Ex~orts Im~orts Trade
Constant
InYiit.! 0.28* 0.12* 0.36* 0.27* 0.13* 0.37*
InCOPPCit 25.83* 25.94* 23.77* 28.25*
InCOPPCit 0.73 0.74 0.64 0.70
InCOPPCiit 6.88* 7.39*
REERt -0.95*** 1.28 0.13 0.18 -0.18 -0.90
09599 0.75** 1.39* 0.43***
095 -0.35 2.37* 1.09**
096 -0.09 2.25* 1.12*
097 0.37 1.95* 1.27*
098 0.23 2.12* 0.82**
099 0.71 *** 1.33** 0.31
InDlSTï
Adjusted R2 0.73 0.76 0.84 0.73 0.76 0.84
Observations 1020 430 420 1020 430 420
Cross-Sections 102 43 42 102 43 42
-, ** & *** denote significance at the 1,5 and 10 Eercent levels resEectively.
Source: Author's calculations
Furthermore, South Africa's wine imports from the world significantly increased by 1.39%
and on annual basis, significant increases occurred in all years by 2.37%, 2.25%, 1.95%,
2.12% and 1.33% in 1995, 1996, 1997, 1998 and 1999 respectively. Similarly, total wine
trade (imports plus exports) between South Africa and the world significantly increased by
0.43% and on annual basis significant increases occurred in 1995, 1996, 1997 and 1998 by
1.09%, 1.12%, 1.27% and 0.82% respectively.
103
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
5.5 The impacts of the implementation of EU-SA TDCA on aggregate agricultural
and selected agricultural products trade flows between South Africa and the EU
countries
This section discusses the impacts of the bilateral trade agreement between South Africa and
the EU countries (i.e. EU-SA TDCA) on the total agricultural trade flows and the selected
agricultural products trade flows (which had received preferential tariff quotas as noted in
Chapter 2) between South Africa and the EU countries for the period 1994-2009. These are
cheese and curd (HS0406), cut flowers (HS0603), frozen fruits and nuts (HS0811), preserved
fruits and nuts (HS2008), fruits and vegetable juices (HS2009) and wines (HS2204). Table
5.14 below shows the utilisation of quotas for these products by South Africa exporters to the
EU market on annual basis from 2000 to 2009 as well as the average quota utilisation for the
periods 2000-2004 and 2005-2009. Neither the official information on the utilisation of
cheese export quotas, nor the import quotas of cheese and wines, have been published at the
time of writing.
Tablê 5.14: Utilisation of South Africa's export quotas under the EU-SA TDCA
Year HS0603 HS0811 HS2008 HS2009 HS2204
2000 29.25% 25.00% 46.33% 70.35% 1.55%
2001 54.87% 32.37% 50.48% 100.00% 6.39%
2002 62.65% 38.22% 58.97% 75.20% 30.55%
2003 61.21% 62.82% 61.79% 54.97% 60.36%
2004 68.88% 54.75% 59.14% 41.41% 68.53%
2005 65.25% 3.00% 53.00% 22.50% 48.00%
2006 55.25% 0.00% 53.67% 58.50% 65.50%
2007 53.00% 0.00% 57.00% 21.50% 60.00%
2008 53.25% 0.00% 45.87% 58.20% 56.49%
2009 47.28% 0.00% 46.33% 37.00% 74.00%
2000 -2004 55.37% 42.63% 55.34% 68.39% 33.48%
2005 -2009 54.81% 0.60% 51.17% 39.54% 60.80%
Source: EC Taric Website (http://ec.europa.eu/taxation customs)
5.5.1 Aggregate agricultural trade flows between South Africa and the EU countries
This subsection provides the results of the impacts of the implementation of the EU-SA
TDCA on South Africa's agricultural exports to the EU countries; South Africa's agricultural
imports from the EU countries; as well as total agricultural trade (import plus exports)
between South Africa and EU countries for the periods 2000-2004 and 2005-2009. The
results are presented in Table 5.15 below.
104
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
Table 5.15: Results for total agricultural trade flows between South Africa and the EU
countries
Models Variables Exports Imports Trade
2000 2005 2000 2005 2000 2005
2004 2009 2004 2009 2004 2009
Constant
InYiit-1 0.46* 0.46* 0.28* 0.28* 0.43* 0.43*
... InGDPPCit 1.15 1.15 1.57 1.57
Col
ell InGDPPCit 2.12* 2.12* 2.40*** 2.40***
e,
-Qj InGDPPCiit 1.50* 1.50*
.5"0
"0 0 REERt 0.15 0.15 0.30 0.30 0.03 0.03
.o;:~ 00004 / 00509 -0.24** 0.24 -0.01 0.49 0.66* 0.66*
~'" InDlSTii
Adjusted R2 0.96 0.96 0.86 0.86 0.96 0.96
Observations 225 225 225 225 225 225
Cross-Sections 15 15 15 15 15 15
Constant
InYiit_1 0.47* 0.47* 0.29* 0.29* 0.45* 0.45*
InGDPPCit 5.84* 5.84* 4.74 4.74
InGDPPCit 2.01 * 2.01 * 2.31*** 2.31***
InGDPPCiit 1.65* 1.65*
ti
ell REERt 0.43** 0.43** 0.47 0.47 0.09 0.09
e,
ë 'ii DOO/ 005 -0.20*** -0.21 -0.13 0.18 -0.07 0.74*
_"0
;;.-.0 001/006 -0.18 -0.94** -0.38 -0.35 -0.13 0.32**
;:~ 002/007 -0.05 -0.89*** -0.82 -0.18 -0.15 0.61 *
ell
>'"- 003/008 0.28 -0.95*** -0.65 -0.39 0.32** 0.61 *
004/009 0.18 -0.76 -0.64 -0.17 0.30** 0.71 *
InDlSTii
Adjusted R2 0.97 0.97 0.86 0.86 0.96 0.96
Observations 225 225 225 225 225 225
Cross-Sections 15 15 15 15 15 15
Constant
InYiit_1 0.38* 0.38* 0.23* 0.23* 0.37* 0.37*
c
0 InGDPPCit 2.06** 2.06** 5.39** 5.39**.-:::
,C_ol InGDPPCit 0.03 0.03 0.80* 0.80*'" InGDPPCiit 0.46* 0.46*
QQj
::"g REERt -0.54* -0.54* -0.32 -0.32 -0.48* -0.48*
.£~ PTAvcs -0.46*** 0.52*** 0.01 0.03 0.16 0.93*~ PTA no -0.44* 0.60** 0.33 -0.22 0.23** 0.94*
"'0,_"ell InDISTii
Eo-< Adjusted R
2 0.86 0.86 0.76 0.76 0.89 0.89
Observations 1785 1785 1905 1905 1635 1635
Cross-Sections 119 119 127 127 109 109
*, **& *** denote significance at the 1,5 and 10Eercent levels res12ectively
Source: Author's calculations
5.5.1.1 Agricultural exports from South Africa to the EU countries
The results show that agricultural exports from South Africa to the EU countries had
significantly declined by 0.24% during the implementation of the EU-SA TDCA over the
period 2000-2004 and on annual basis, the significant decrease occurred in 2000 by 0.2%.
While the joint period effect of the implementation of the EU-SA TDCA on South Africa's
105
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
agricultural exports to the EU countries was insignificant for the period 2005-2009, the
individual yearly significant negative effects were observed in 2006, 2007 and 2008, when
agricultural exports from South Africa to the EU countries significantly declined by 0.94%,
0.89% and 0.95% respectively during these years. On average, 0.44% of South Africa's
agricultural exports destined for the EU market were diverted to other agricultural trading
partners of South Africa over the period 2000-2004. In contrast, for the period 2005-2009,
the implementation of the EU-SA TDCA led to the creation of 0.52% of South Africa's
agricultural exports market in the EU countries. The results for the average actual and
potential agricultural exports from South Africa to the EU countries for the periods 2000-
2004 and 2005-2009 are presented in Figure 5.1 in log values, and the dollar values are
presented in Appendix SAS.
22.00
20.00 - ~~ • jl!:; ~ ~-'-!:iI!:
18.00 H ~H ~ U ~~ ~~~
~~
16.00 -=--l: ~.~ '-=~ ~ ~ ~ ~ ~ ~~~~~~r!:~~~
_ 14.00 «r=~~~~rt~~rHtfYfrr~t~dr~it ~I ~~-~iI!;-iI~~~ II~I~!;U-jI!:;-~-~jI! ~~~~~~~ 12.00 ~r-r-r-~r-~~~
:;-iI!~:-iI!:U- ~
.s
'; 10.00 ~~~~~~~~~~~~~~ Fjl!:; ~~~~~~~~~~~il!: ~cu:::J ~~~~~~~~V~~~~~~,~J~U~f~i~i~ F~~~~~~~~~~~~~rt~ 8.00
6.00 ~~~~~~f'~~s ~f'rt;;.:;.:;;.:;.:I;.:;.I:<£:<£I:<£:<;£:<£:<~£:<£:~4.00 F~~~~~~~~~~ifI~~I ~~~~
~~~~rt~
2.00
0.00 r=~~~~~~rr~rIrr~r~~t~~~~~~~~~~~~~~~~~~~~
<to) <to) <to) <to) <to) <t O)<t O)<t O)<t O)<t 0) <to) <to) <to) <to) <to)
0000 0000 00000000000000 0000 0000 0000 0000 00 0000 00 0000 0000 0000 00
NN N .Nn N NN NN NN N N NN NN NN N N N N N NN6J, 0 0 J,6 J.,60 0 0 0 0 00 00 0J.60 0.n 00 J,6 J,600 00 0th0o 0.n 00 0.n 00 0.n 0aJ0..
N N N
0 .n cJ,N
0 0 0 0 0 00 00 0 0 0 0 00 00 00 0 0 0 0 0 00 00 00 0000
N N N N N N NN N N N N N NN NN NN N N N N N NN N N NN
AUT BEL DNK FIN FRA DEU GRC IRL ITA LUX NLD PRT ESP SWE GBR
Countries and Periods
~ Actual == Potential
Figure 5.1: Average actual and potential value of agricultural exports from South Africa
to the EU countries for the periods 2000-2004 and 2005-2009
The results show that South African agricultural exporters had outperformed and exhausted
the estimated potential capacity for the absorption of South Africa's agricultural exports in
Denmark, Finland, France, Greece, the Netherlands, Spain and Sweden; but had under-scored
or underachieved in Austria, Belgium, Germany, Italy, Ireland, Luxembourg, Portugal and the
106
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
United Kingdom for the period 2000-2004. For the period 2005-2009, South African
agricultural exporters still outperformed and exhausted the estimated potential capacity for the
absorption of South Africa's agricultural exports in Denmark, Finland, Italy, Ireland,
Luxembourg, the Netherlands, and Sweden; whereas they under-scored in Austria, Belgium,
France, Germany, Greece, Portugal, Spain and the United Kingdom.
5.5.1.2 Agricultural imports from the EU countries to South Africa
The results indicate that the implementation of the EU-SA TDCA for both periods had no
effects on the agricultural imports from the EU countries to South Africa. The results also
show no proof of creation or diversion of South Africa's agricultural exports from either the
EU countries or South Africa's other agricultural trading partners.
The results for the average actual and potential agricultural imports from the EU countries to
South Africa for the periods 2000-2004 and 2005-2009 are presented in Figure 5.2 in ·log
values, and the dollar values are presented in Appendix 5AV. The results show that South
African agricultural importers had outperformed and exhausted the estimated potential
capacity for the absorption of South Africa's agricultural imports from Austria, Italy,
Luxembourg and Spain; but had under-scored or underachieved the estimated potential
capacity for the absorption of South Africa's agricultural imports from Belgium, Denmark,
France, Finland, Germany, Greece, Ireland, the Netherlands, Portugal, Sweden and the United
Kingdom for the period 2000-2004.
During the period 2005-2009, South African agricultural importers outperformed and
exhausted the estimated potential capacity for the absorption of South Africa's agricultural
imports from Austria, Belgium, Germany, Italy, France, Luxembourg, the Netherlands,
Portugal and Spain, whereas they under-scored or underachieved the estimated potential
capacity for the absorption of South Africa's agricultural imports from Denmark, Finland,
Greece, Ireland, Sweden, and the United Kingdom.
107
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
22.00
20.00
18.00
16.00
~ 14.00
III
glil) 12.00
III
.c.u
10.00
:
nl
> 8.00
6.00
4.00
2.00
0.00
~m~m~m~m~m~m~m~m~m~m~m~m~m~m~m
000000000000000000000000000000
N0N0N0N0N0N0NN0N0N0N0N0NN0N0N0N0N0N0NN0N0N0N0N0NN0N0N0N00000
oI~oI ~oI ~oI ~Io~Io~I o~I oI~oI~oI~oI ~oI ~Io~Io~Io~I I I I I I I I I I I I I I
000000000000000000000000000000
0NN0N0N0N0N0NN0N0N0N0N0NN0N0N0N0N0N0NN0N0N0N0N0NN0N0N0N00000
AUT BEL DNK FIN FRA DEU GRC IRL ITA LUX NLD PRT ESP SWE GBR
Countries and Periods
~ Actual = Potential
Figure 5.2: Average actual and potential value of agricultural imports from the EU
countries to South Africa for the periods 2000-2004 and 2005-2009
5.5.1.3 Agricultural trade (imports plus exports) between South Africa and the EU
countries
The results show that total agricultural trade between South Africa and the EU countries had
significantly improved by 0.66% during the implementation of the EU-SA TDCA over the
period 2005-2009. On annual basis, significant increases occurred in all years as follows:
0.74% in 2005, 0.32% in 2006, 0.61% in both 2007 and 2008, and 0.71% in 2009.
Surprisingly, while the joint period effect of the implementation of the EU-SA TDCA on total
agricultural trade between South Africa and the EU countries was insignificant for the period
2000-2004, individual yearly significant positive effects were observed in 2008 and 2009,
when total agricultural trade between South Africa and the EU countries significantly
increased by 0.32% and 0.3% respectively. On average, the implementation of the EU-SA
TDCA led to the creation of 0.93% of total agricultural trade between South Africa and the
EU countries for the period 2005-2009. However, there was no proof of creation or diversion
of total agricultural trade between South Africa and the EU countries to other agricultural
trading partners of South Africa or EU countries. The results for the average actual and
108
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
potential total agricultural trade between South Africa and the EU countries for the periods
2000-2004 and 2005-2009 are presented in Figure 5.3 in log values, and the dollar values are
presented in Appendix SAY.
22.00
20.00 ~ ~ ~..._ ~ ~~~ ~ ~~f
18.00 f-r;;-i-i-i or;_jj;~ ~-F-~ ~l---,'~~~~4
~~~~~~ P~~~=~~~~rt~-~'~~~~~~~rl~F~~n ~~~~~~~~~16.00 ~~t~-~~~ ~14.00 ~~~~~~~~~~~~~~~ ~ei ~~~~~~~~~~ ~~~~~ ~~~~~r:~.;~~~~~i'~.::~'~~~~~~~ r. ~~~~~~VIstil) 12.00 ~P~~~~~~rl ~ ~~~~ ~ P~~~:~~~
VI ~~~~~~~~~~ E~i~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
???&
~~~~~~~~~III 10.00 ~~~
:::J
i>i 8.00 ~-i-i-i-ifl3:J:J::J::~~~~~~~-.~c::~C~~~~~~~-~l~~~~~~~~~~~-t;~~~~~-~-
g~~~_~~~~~~~~?~~?~~?~~?~~~?~~?~~?~???????r~l~~~~~~6.00
4.00 ~~~~~~~~~~~~~f"~~V~'?~V????~
~~.~A~~~.~~~~.
2.00 P~~~~~~~rl ~~~~~~~~~~~~~~~
~~~~~~
~~~~ H ~P~~?~~~??d
0.00 ~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~
g 0en 0<:t e0n<:0t 0en 0<:t0en 0<:ten <:ten <:t <:ten en<:t en<:t en <:ten <:ten <:ten <:ten0 0 0 00 0 00 000 0000 000
en
0 00 0
<:t 00 0
N N N N N N NN NN N NN NN00N N0 N0N0 N00
0 00 00 00N N N0N0 N00 0
000 000N NN NN0
0I Jo. 6 tho Ili) cV. 6J.. 6 tho J.6 Jo. 6 tho I IlI)O Jo. I I I I I I I IOll) 0 li) Oll) Oll)
00 00 00 0000 00 0000 00 0 00 00 0 0 00 00 0 00
0 0 0 0 00
N N N NN N N NN00NN0 0NN0 N0N0 N0 N0 N0N0 N0N0 N0 N0N0 N0 N0 N0 N0 N0N0
AUT BEL DNK FIN FRA DEU GRC IRL ITA LUX NLD PRT ESP SWE GBR
Countries and Periods
f..iActual ;; Potential
Figure 5.3: Average actual and potential value of agricultural trade between South
Africa and the EU countries for the periods 2000-2004 and 2005-2009
The results show that South Africa's agricultural traders (importers and exporters) operating
between South Africa and EU countries had outperformed and exhausted the estimated
potential capacity for the absorption of total agricultural trade in the South Africa-Austria,
South Africa-Denmark, South Africa-Finland, South Africa-Greece, South Africa-
Netherlands, South Africa-Spain and South Africa-Sweden markets. On the other hand, they
under-scored or underachieved in the South Africa-Belgium, South Africa-France, South
Africa-Germany, South Africa-Ireland, South Africa-Luxembourg, South Africa-Portugal and
South Africa-United Kingdom markets over the period 2000-2004. For the period 2005-
2009, South Africa's agricultural traders (importers and exporters) outperformed and
exhausted the estimated potential capacity for the absorption of total agricultural trade in the
109
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
South Africa-Austria, South Africa-Denmark, South Africa-Germany, South Africa-Italy,
South Africa-Luxembourg, South Africa-Netherlands, and South Africa-Sweden markets;
whereas they under-scored or underachieved in the South Africa-Belgium, South Africa-
Finland, South Africa-France, South Africa-Greece, South Africa-Ireland, South Africa-
Portugal, South Africa-Spain and South Africa-United Kingdom markets.
5.5.2 Cheese trade flows between South Africa and the EU countries
This subsection provides the results of the impacts of the implementation of the EU-SA
TDCA's reciprocal cheese in-quota tariff preferences on South Africa's cheese exports to the
EU countries; South Africa's cheese imports from the EU countries; as well as total cheese
trade (import plus exports) between South Africa and EU countries for the periods 2000-2004
and 2005-2009. The results are presented in Table 5.16.
110
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
Table 5.16: Results for cheese trade flows between South Africa and the EU countries
Models Variables Exports Imports Trade
2000 2005 2000 2005 2000 2005
2004 2009 2004 2009 2004 2009
Constant
InYiit_] 0.28** 0.28** 0.30* 0.30* 0.26** 0.26**
.... InGDPPCit 6.00 6.00 -2.10 -2.10
oto:sl InGDPPCit 7.90 7.90 3.72 3.72
Q..
51i InGDPPCiit 0.85 0.85_"0
"0 0 REERt 2.49 2.49 0.84 0.84 0.39 0.39
.o;:~ 00004 /00509 -3.53*** -2.89 0.59 -0.08 -0.88 0.20
<lJ
Cl. InDlSTii
Adjusted R2 0.40 0.40 0.67 0.67 0.77 0.77
Observations 90 90 195 195 90 90
Cross-Sections 6 6 13 13 6 6
Constant
InYiit_] 0.30** 0.30** 0.35* 0.35* 0.28** 0.28**
InGDPPCit -10.86 -10.86 33.60** 33.60**
InGDPPCit 7.05 7.05 -3.02 -3.02
InGDPPCiit -0.64 -0.64
to:is REERt 1.07 1.07 1.28 1.28 0.05 0.05
Q..
51i DOO/ DOS -3.90 -1.55 -0.13 -3.63*** 0.29 0.61
_"0
>,0 001/006 -6.23*** 2.06 -0.38 -4.82*** 1.99 0.72
ï:~
o:s 002/007 -5.56 2.53 -0.82 -5.84*** 6.41* 0.91
..<.lJ. 003/008 -7.27 0.51 -0.65 -6.71 *** 2.24 1.20004/009 -10.57*** 0.55 -0.64 -12.16* 1.79 -1.24
InDlSTii
Adjusted R2 0.39 0.39 0.76 0.76 0.78 0.78
Observations 90 90 195 195 90 90
Cross-Sections 6 6 13 13 6 6
Constant
InYiit-1 0.40* 0.40* 0.32* 0.32* 0.23*** 0.23***
.c2: InGDPPCit -7.72 -7.72 6.85 6.85
t..i. InGDPPCit 1.09 1.09 0.89 0.89<lJ InGDPPCiit 2.49 2.49
Q1i
.. "0 REERt -1.88** - 1.88** 1.91*** 1.91*** 0.87 0.87
....s.~. 0 PTA.e, -2.3 7* 1.19 0.91 -1.55 -0.56 -0.06
1.;0..
<lJ PTA no -0.96 2.01 *** -0.74 -1.66 -0.16 -2.09**
"o.0.:s. mmsr,
E- Adjusted R2 0.70 0.70 0.64 0.64 0.68 0.68
Observations 480 480 360 360 150 150
Cross-Sections 32 32 24 24 10 10
-, ** & *** denote significance at the 1,5 and 10 Eercent levels resEectively
Source: Author's calculations
5.5.2.1 Cheese exports from South Africa to the EU countries
The results indicate that cheese exports from South Africa to the EU countries had
significantly declined by 3.53% during the implementation of the EU-SA TDCA for the
period 2000-2004 and on annual basis, significant decreases occurred in 2001 and 2004 by
6.23% and 10.57% respectively. However, the effects of the implementation of the EU-SA
111
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
TDCA on South Africa's cheese exports to the EU countries for the period 2005-2009 were
insignificant. There was no proof of trade creation or diversion of South Africa's cheese
exports to the EU during the implementation of the EU-SA TDCA over both periods.
12.00
10.00 -
8.00 -
~
1/1
g110 6.00 ~ -
1/1
III ~
::::I
>IV 4.00 - - ~H~ ~I ~~
- - ~ F~~ ~2.00 - r-II -~~~
0.00 ~ ~ f1 ~ ~~ ~ ~~~. ~ I~~
"<t al "<t al "<t al "<t al "<t al "<t al
0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N
6 J, 6 J, 6 J, 6 J, 6 J, 6 ILJ")
0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N
FRA DEU GRC NlD ESP GBR
Countries and Periods
~Actual = Potential
Figure 5.4: Average actual and potential value of cheese exports from South Africa to
EU countries for the periods 2000-2004 and 2005-2009
The results for the average actual and potential cheese exports from South Africa to the EU
countries for the periods 2000-2004 and 2005-2009 are presented in Figure 5.4 in log values,
and the dollar values are presented in Appendix SBB. The results show that South Africa's
cheese exporters had outperformed and exhausted the estimated potential capacity for the
absorption of South Africa's cheese exports in the Netherlands and Spain, but had under-
scored or underachieved in France, Germany, Greece and the United Kingdom over the period
2000-2004. For the period 2005-2009, South African cheese exporters outperformed and
exhausted the estimated potential capacity for the absorption of South Africa's cheese exports
in France, Germany and Greece; whereas they under-scored or underachieved in Spain and
the United Kingdom.
112
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
5.5.2.2 Cheese imports from the EU countries to South Africa
Tbe results show that the implementation of the EU-SA TDCA over both periods had no joint
period effects on South Africa's cheese imports from the EU countries, but on annual basis,
there were significant decreases in South Africa's cheese imports from the EU countries
which occurred in all years over the period 2005-2009, as follows: 3.63% in 2005, 4.82% in
2006,5.84% in both 2007,6.71% in 2008 and 12.16% in 2009. However, there was no proof
of creation or diversion of South Africa's cheese imports from the EU countries. The results
of the average actual and potential cheese imports from the EU countries to South Africa for
the periods 2000-2004 and 2005-2009 are presented in Figure 5.5 in log values, and the
dollar values are presented in Appendix 5BE.
16.00
14.00
12.00
~ 10.00
1/1
-IlO..Q 8.001/1
QI
::J
111 6.00>
4.00
2.00
0.00
o::t 0'1 o::t 0'1 o::t 0'1 o::t 0'1 o::t 0'1 o::t 0'1 o::t 0'1 o::t 0'1 o::t 0'1 o::t 0'1 o::t 0'1 o::t 0'1 o::t 0'1
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N N N N N N N N N N N N N N N
6 J, 6 J, I I I I0 LIl 0 LIl 6 J, 6 ILIl 6 J, 6 ILIl 6 J, 6 J, 6 J, 6 ILIl 6 J,
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N N N N N N N N N N N N N N N
AUT BEL DNK FRA DEU GRC IRL ITA NLD PRT ESP SWE GBR
Countries and Periods
~ Actual = Potential
Figure 5.5: Average actual and potential value of cheese imports from the EU countries
to South Africa for the periods 2000-2004 and 2005-2009
The results show that South Africa's cheese importers had outperformed and exhausted the
estimated potential capacity for the absorption of Soutb Africa's cheese imports from
Belgium, Italy, Ireland and Portugal; but had under-scored or underachieved the estimated
potential capacity for the absorption of South Africa's cheese imports from Austria, Denmark,
113
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
France, Germany, Greece, the Netherlands, Spain, Sweden and the United Kingdom over the
period 2000-2004.
During the period 2005-2009, South Africa's cheese importers had outperformed and
exhausted the estimated potential capacity for the absorption of South Africa's cheese imports
from Denmark, France, Italy, Portugal, Spain and Sweden; but had under-scored or
underachieved the estimated potential capacity for the absorption of South Africa's cheese
imports from Austria, Belgium, Germany, Greece, Ireland, the Netherlands and the United
Kingdom.
5.5.2.3 Cheese trade (imports plus exports) between South Africa and the EU countries
The results show that the implementation of the EU-SA TDCA over both periods had no joint
period effects on total cheese trade between South Africa and the EU countries, but on annual
basis there was a significant increase in total cheese trade between South Africa and the EU
countries which occurred in 2002 by 6.41%. On average, for the period 2005-2009, the
implementation of the EU-SA TDCA led to the diversion of total cheese trade between South
Africa and the EU countries to other cheese trading partners of South Africa or EU countries
by 2.09%. However, there was no proof of creation or diversion of total cheese trade between
South Africa and the EU countries over the period 2000-2004.
The results for the average actual and potential total cheese trade between South Africa and
the EU countries for the periods 2000-2004 and 2005-2009 are presented in Figure 5.6 in log
values, and the dollar values are presented in Appendix SBH. The results show that South
Africa's cheese traders (importers and exporters) operating between South Africa and EU
countries had outperformed and exhausted the estimated potential capacity for the absorption
of total cheese trade in the South Africa-France market, South Africa-Germany market, South
Africa-Greece market, South Africa-Netherlands market and South Africa-United Kingdom,
but had under-scored or underachieved in the South Africa-Spain market over the period
2000-2004.
In contrast, for the period 2005-2009, South Africa's cheese traders (importers and exporters)
operating between South Africa and EU countries had outperformed and exhausted the
estimated potential capacity for the absorption of total cheese trade in the South Africa-Spain
114
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
market, but had under-scored or underachieved in the South Africa-France market, South
Africa-Germany market, South Africa-Greece market and South Africa-United Kingdom
market.
16.00
14.00
12.00
~ 10.00
1/1
gbO 8.00
1/1
Cl
::::I
iV 6.00
>
4.00
2.00
0.00
'<t 0\ '<t 0\ '<t 0\ '<t 0\ '<t 0\ '<t 0\
0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N
2> J. 2> J. 2> J. 2> J. 2> J. 2> J.
0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N
FRA DEU GRC NLD ESP GBR
Countries and Periods
~ Actual :: Potential
Figure 5.6: Average actual and potential value of cheese trade from the EU countries to
South Africa for the periods 2000-2004 and 2005-2009
5.5.3 Cut flowers trade flows between South Africa and the EU countries
This subsection provides the results of the impacts of the implementation of the EU-SA
TDCA's non-reciprocal cut flowers in-quota tariff preferences on South Africa's cut flowers
exports to the EU countries; South Africa's cut flowers imports from the EU countries; as
well as total cut flowers trade (import plus exports) between South Africa and EU countries
for the periods 2000-2004 and 2005-2009. The results are presented in Table 5.7 below.
115
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
Table 5.17: Results for cut flowers trade flows between South Africa and the EU
countries
Models Variables Exports Imports Trade
2000 2005 2000 2005 2000 2005
2004 2009 2004 2009 2004 2009
Constant 1276.5** 1276.5** -253.34* -253.34*
InYiit-J 0.51 * 0.51 * 0.75* 0.75*
.... InGDPPCit -1.56 -1.56 -13.69 -13.69
~u InGDPPCQ._ it
-2.00 -2.00 40.63* 40.63*
E Q,j InGDPPCiit -6.05* -6.05*
_'0
'0 0 REERt -0.40 -0.40 7.70** 7.70** -0.69*** -0.69***
.52 :;; D0004 / 00509 -0.51 0.41 2.24*** -2.83 -0.42*** 1.05*
:..
Q,j
Q., InDISTii -176.47* -176.47* 35.35* 35.35*
Adjusted R2 0.83 0.83 0.31 0.31 0.96 0.96
Observations 210 210 105 105 JOS 105
Cross-Sections 14 14 7 7 7 7
Constant 1460.7** 1460.7** -246.26* -246.26*
InYiit-J 0.52* 0.52* 0.76* 0.76*
InGDPPCit 5.99 5.99 -38.35 -38.35
InGDPPCil -2.95 -2.95 41.00* 41.00*
InGDPPCiit -5.99* -5.99*
REERt -0.27 -0.27 6.39** 6.39** -0.72*** -0.72***
DOO/005 0.73 -0.58 0.88 0.06 -0.49** 1.22*
001/006 1.09 -0.76 -0.60 3.04 -0.47 0.99*
002/007 3.37*** -0.61 -3.20 2.76 -0.01 1.12*
003/008 2.94** -1.87 -1.03 4.82 0.42 1.00*
D04/D09 3.05 -1.35 -2.12 2.95 . -0.24 0.89*
InDlSTii -174.80** -174.80** 34.51 * 34.51 *
Adjusted R2 0.83 0.83 0.34 0.34 0.96 0.96
Observations 210 210 105 105 105 105
Cross-Sections 14 14 7 7 7 7
Constant
InYiit-J 0.49* 0.49* 0.26* 0.26* 0.33* 0.33*
InGDPPCit -4.03 -4.03 4.11 4.11
InGDPPCit 0.56 0.56 0.71 0.71
InGDPPCiit 3.44*** 3.44***
REERt -0.60 -0.60 0.58 0.58 0.44 0.44
PTA ves -1.62* 0.51 -0.36 -1.82 -0.70 -0.25
PTA no -0.59*** 1.38** 1.00 -0.61 -0.25 0.32
InDlSTii
Adjusted R2 0.78 0.78 0.62 0.62 0.81 0.81
Observations 840 840 390 390 330 330
Cross-Sections 56 56 26 26 22 22
-, ** & *** denote significance at the 1,5 and 10 percent levels respectively
Source: Author's calculations
5.5.3.1 Cut flowers exports from South Africa to the EU countries
The results show that the implementation of the EU-SA TDCA over both periods had no joint
period effects on South Africa's cut flowers exports to the EU countries, but on annual basis
there were significant increases in South Africa's cut flowers exports to the EU countries
which occurred in 2002 and 2003, at 3.37% and 2.94% respectively. On average, for the
116
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
period 2000-2004, the implementation of the EU-SA TDCA led to the diversion of South
Africa's cut flowers exports destined for the EU countries to other South African cut flowers
trading partners by 0.59%. However, the results show no proof of creation or diversion of
South Africa's cut flowers exports to the EU countries over the period 2005-2009. The
results for the average actual and potential cut flowers exports from South Africa to the EU
countries over the periods 2000-2004 and 2005-2009 are presented in Figure 5.7 in log
values, and the dollar values are presented in Appendix 5BK.
20.00
16.00
~ 12.00
VI
0.0
.2
VI
QI
~ 8.00
>
4.00
0.00
oq~q~q~q~q~q~q~q~q~q~q~q~q~q~0000000000000000000000000000NN0 000 06~N6N~N6N~N6NN~NN0 06~N6N~N0 0 06NN~N6N~N06N~N06N~N0N06NN0NN0 0 0 000 0 0 0 0 oIo~6~6~6~
o0000010000000000000010001000000000100000000000 0 0 0 0 010101010 0 0 0 0NNNNNNNNNNNNNNNNNNNNNNNNNNNN
AUT BEL DNK FIN FRA DEU GRC IRL ITA NLD PRT ESP SWE GBR
Countries and Periods
~ Actual == Potential
Figure 5.7: Average actual and potential value of cut flowers exports from South Africa
to EU countries for the periods 2000-2004 and 2005-2009
The results show that South Africa's cut flowers exporters had outperformed and exhausted
the estimated potential capacity for the absorption of South Africa's cut flowers exports in
Austria, Belgium, Denmark, Finland, France, Germany, Italy, Ireland, the Netherlands,
Portugal, Spain, Sweden and the United Kingdom; but had under-scored or underachieved in
Greece over the period 2000-2004. During the period 2005-2009, South Africa's cut flowers
exporters outperformed and exhausted the estimated potential capacity for the absorption of
South Africa's cut flowers exports in Austria, Belgium, Germany, Greece, Italy, Ireland, the
117
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
Netherlands, Portugal, and the United Kingdom; whereas they under-scored or underachieved
in Denmark and Finland.
5.5.3.2 Cut flowers imports from the EU countries to South Africa
The results show that the implementation of the EU-SA TDCA over both periods had no
effects on the cut flowers imports from the EU countries to South Africa and that there was no
proof of creation or diversion of South Africa's cut flowers imports from either the EU
countries or South Africa's other cut flowers trading partners. The results for the average
actual and potential cut flowers imports from the EU countries to South Africa for the periods
2000-2004 and 2005-2009 are presented in Figure 5.8 in log values, and the dollar values are
presented in Appendix 5BN.
12.00
,:.;
10.00 ~
I ~
8.00 W
1:;):
III
110 ~ ~ ~-~ 6.00 2 =III 2~ 2
CU
:J H
n ~~>i 4.00 - - " -~~ ~lE L~ ~ iIE- - ~ ~~ r ==2.00 ~~ ~ ~
~ ~ ~ ~~~ 2~ ~
0.00 ~ ~ ~ z~ ~~~ .
~,~.~E ~ ~
oo::t en oo::t en "oo::t en oo::t en oo::t en oo::t en oo::t en
0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N N N
6 J, 6 J, 6 J, 6 J, 6 J, 6 I I1.11 6 1.11
0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N N N
DEU ESP FRA ITA NLD PRT GBR
Countries and Periods
~Actual = Potential
Figure 5.8: Average actual and potential value of cut flowers imports from the EU
countries to South Africa for the periods 2000-2004 and 2005-2009
The results show that South Africa's cut flowers importers had outperformed and exhausted
the estimated potential capacity for the absorption of South Africa's cut flowers imports from
France, Portugal, the Netherlands and Spain; but had under-scored or underachieved the
118
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
estimated potential capacity for the absorption of South Africa's cut flowers imports from
Germany, Italy and the United Kingdom over the period 2000-2004. In contrast, over the
period 2005-2009, South Africa's cut flowers importers outperformed and exhausted the
estimated potential capacity for the absorption of South Africa's cut flowers imports from
France and the Netherlands; but under-scored or underachieved the estimated potential
capacity for the absorption of South Africa's cut flowers imports from Germany and Spain.
5.5.3.3 Cut flowers trade (imports plus exports) between South Africa and the EU
countries
The results show that total cut flowers trade between South Africa and the EU countries had
significantly declined by 0.42% during the implementation of the EU-SA TDCA over the
period 2000-2004 but on annual basis, a significant increase occurred in 2000 by 0.49%. In
contrast, for the period 2005-2009, the implementation of the EU-SA TDCA had significantly
increased the total cut flowers trade between South Africa and the EU countries by 1.05%.
(
On annual basis, significant increases occurred in all years as follows: 1.22% in 2005, 0.99%
in 2006,1.12% in 2007,1% in 2008 and 0.89% in 2009. The results show no proofoftrade
creation or diversion of cut flowers trade between South Africa and the EU countries to the
other markets.
The results for the average actual and potential total cut flowers trade between South Africa
and the EU countries for the periods 2000-2004 and 2005-2009 are presented in Figure 5.9
in log values, and the dollar values are presented in Appendix 5BQ. The results show that
South Africa's cut flowers traders (importers and exporters) operating between South Africa
and EU countries had outperformed and exhausted the estimated potential capacity for the
absorption of total cut flowers trade in the South Africa-Italy, South Africa-Netherlands,
South Africa-Portugal and South Africa-United Kingdom markets; but had under-scored or
underachieved in the South Africa-France, South Africa-Germany and South Africa-Spain
markets over the period 2000-2004.
Over the period 2005-2009, South Africa's cut flowers traders (importers and exporters)
operating between South Africa and EU countries had outperformed and exhausted the
estimated potential capacity for the absorption of total cut flowers trade in the South Africa-
119
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
Germany, South Africa-Netherlands and South Africa-Spain markets, but had under-scored or
underachieved in the South Africa-France market.
18.00
16.00
14.00
12.00
~
III
l1li 10.00g
III
III 8.00
:::I
i>V 6.00
4.00
2.00
0.00
'<t en '<t en '<t en g en '<t en '<t en '<t en0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N N N
0 J, 0 ILIl 0 J, 0 J, 0 J, 0 J, I0 J,
0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N N N
FRA DEU ITA NLD PRT ESP GBR
Countries and Periods
:1. Actual = Potential
Figure 5.9: Average actual and potential value of cut flowers trade between South Africa
and the EU countries for the periods 2000-2004 and 2005-2009
5.5.4 Frozen fruits and nuts trade flows between South Africa and the EU countries
This sub-section provides the results of the impacts of the implementation of the EU-SA
TDCA's non-reciprocal frozen fruits and nuts in-quota tariff preferences on South Africa's
frozen fruits and nuts exports to the EU countries; South Africa's frozen fruits and nuts
imports from the EU countries; as well as total frozen fruits and nuts trade (import plus
exports) between South Africa and EU countries over the periods 2000-2004 and 2005-2009.
The results are presented inTable 5.18.
120
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
Table 5.18: Results for frozen fruits and nuts trade flows between South Africa and the
EU countries
Models Variables Exports Imports Trade
2000 2005 2000 2005 2000 2005
2004 2009 2004 2009 2004 2009
Constant
- InYiit-t 0.28* 0.28* 0.25** 0.25**InGOPPC;t -14.70 -14.70 16.68 16.68("0:"1 InGOPPCit 20.32 20.32 10.82 10.82
E0._Q.l InGOPPCiit 1.46 1.46
....."0
"0 0 REERt 1.64 1.64 5.12*** 5.12*** 1.06 1.06
o~
'c 00004 / DOS09 -3.56*** 0.24 1.47 -4.97 -1.07 0.32
Q.l
0. iemsr,
Adjusted R2 0.55 0.55 0.25 0.25 0.40 0.40
Observations 105 105 105 105 90 90
Cross-Sections 7 7 7 7 6 6
Constant
InYiit-t 0.29* 0.29* 0.27* 0.27*
InGDPPCit -21.98 -21.98 58.31 58.31
InGOPPCit 17.16 17.16 7.09 7.09
InGDPPCiit 6.72 6.72
REERt 0.71 0.71 6.70** 6.70** 0.96 0.96
DOO/ DOS -10.55* 2.09 -0.07 -9.12 -0.93 0.70
DOl / D06 -17.98* 1.68 2.45 -13.79*** 0.23 -1.75
D02 / D07 -26.85* 2.57 -0.05 -12.12 5.06 0.29
D03 / D08 -18.31 * 4.01 1.21 -15.59*** 1.96 -0.39
D04/ D09 -25.85* 0.58 -1.62 -15.20*** 0.30 0.35
InDlSTii
Adjusted R2 0.54 0.54 0.26 0.26 0.28 0.28
Observations 105 105 105 105 90 90
Cross-Sections 7 7 7 7 6 6
Constant
InYiit-t 0.40* 0.40* 0.25* 0.25* 0.42* 0.42*
InGDPPCit 3.00 3.00 -2.79 -2.79
InGDPPCit 0.32 0.32 8.17* 8.17*
InGOPPCijt -8.84 -8.84
REERt -1.86*** -1.86*** 2.49 2.49 -2.28 -2.28
PTA,.s -0.77 -0.60 1.39 -0.41 -0.57 1.70
PTA no 0.22 0.77 0.41 -1.74 -0.48 0.29
mnrsr,
Adjusted R2 0.57 0.57 0.35 0.35 0.47 0.47
Observations 315 315 345 345 165 165
Cross-Sections 21 21 23 23 II II
*, ** & *** denote significance at the 1,5 and 10 percent levels respectively
Source: Author's calculations
5.5.4.1 Frozen fruits and nuts exports from South Africa to the EU countries
The results show that frozen fruits and nuts exports from South Africa to the EU countries had
significantly declined by 3.26% during the implementation of the EU-SA TDCA over the
period 2000-2004. On annual basis, significant decreases occurred in all years as follows:
10.55% in 2000,17.98% in 2001,26.85% in 2002,18.31% in 2003 and 25.85% in 2004.
121
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
However, the implementation of the EU-SA TDCA over the period 2005-2009 had no effects
on South Africa's frozen fruits and nuts exports to the EU countries. Furthermore, there was
no proof of creation in South Africa's frozen fruits and nuts exports market in the EU
countries or diversion of South Africa's frozen fruits and nuts exports destined for the EU
market to other South African frozen fruits and nuts trading partners for both periods The
results for the average actual and potential frozen fruits and nuts exports from South Africa to
the EU countries for the periods 2000-2004 and 2005-2009 are presented in Figure 5.10 in
log values, and the dollar values are presented in Appendix SBT.
16.00
14.00
12.00
1h 10.00
III
toO
-.2 8.00III
.1a11
nl 6.00
:>
4.00
2.00
0.00
'<t 0\ '<t 0\ '<t 0\ '<t 0\ '<t 0\ '<t 0\ '<t 0\
0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N N N
6 J, 6 J, 6 J, 6 I ILI'! 6 J, 6 LI'! 6 J,
0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N N N
AUT BEL FRA DEU NLD SWE GBR
Countries and Periods
~ Actual :: Potential
Figure 5.10: Average actual and potential value of frozen fruits and nuts exports from
South Africa to EU countries for the periods 2000-2004 and 2005-2009
The results show that South Africa's frozen fruits and nuts exporters had outperformed and
exhausted the estimated potential capacity for the absorption of South Africa's frozen fruits
and nuts exports in Belgium, France, the Netherlands and the United Kingdom; but had
under-scored or underachieved in Austria, Germany and Sweden for the period 2000-2004.
For the period 2005-2009, South Africa's frozen fruits and nuts exporters continued to
outperform and exhaust the estimated potential capacity for the absorption of South Africa's
122
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
frozen fruits and nuts exports in Belgium; but they under-scored or underachieved in France,
Germany, the Netherlands and the United Kingdom.
5.5.4.2 Frozen fruits and nuts imports from the EU countries to South Africa
The results show that the implementation of the EU-SA TDCA for both periods had no joint
period effects on the frozen fruits and nuts imports from the EU countries to South Africa and
also show no proof of creation or diversion of South Africa's frozen fruits and nuts imports
from either the EU countries or South Africa's other frozen fruits and nuts trading partners.
However, on an annual basis, negative effects of the implementation of the EU-SA TDCA on
South Africa's frozen fruits and nuts imports from the EU countries were detected in 2006
(13.79%),2008 (15.59%) and 2009 (15.2%).
The results for the average actual and potential frozen fruits and nuts imports from the EU
countries to South Africa for the periods 2000-2004 and 2005-2009 are presented in Figure
5.11 in log values, and the dollar values are presented in Appendix 5BW. The results show
that South Africa's frozen fruits and nuts importers had outperformed and exhausted the
estimated potential capacity for the absorption of South Africa's frozen fruits and nuts
imports from Belgium, Denmark, Greece and the Netherlands; but had under-scored or
underachieved the estimated potential capacity for the absorption of South Africa's frozen
fruits and nuts imports from France, Germany and the United Kingdom over the period 2000-
2004.
In contrast, for the period 2005-2009, South Africa's frozen fruits and nuts importers
outperformed and exhausted the estimated potential capacity for the absorption of South
Africa's frozen fruits and nuts imports from Belgium and the Netherlands, but under-scored
or underachieved the estimated potential capacity for the absorption of South Africa's frozen
fruits and nuts imports from Denmark, Germany, Greece and the United Kingdom.
123
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
16.00
14.00
12.00
ih 10.00
III
l1li
.-.Q 8.00III
GI
:::J
iii
> 6.00
4.00
2.00
0.00
q- 0\ q- 0\ q- 0\ q- 0\ q- 0\ q- 0\ q- 0\
0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0, 0, 0 0 0 0 0 0
0 0 0 0
N N N N N N N N N N N N N N
6 Lh 0 LIl 6 Lh ,0 Lh 6 Lh 6 Lh , ,0 LIl
0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N N N
BEL DNK FRA DEU GRC NLD GBR
Countries and Periods
~ Actual = Potential
Figure 5.11: Average actual and potential value of frozen fruits and nuts imports from
the EU countries to South Africa for the periods 2000-2004 and 2005-2009
5.5.4.3 Frozen fruits and nuts trade (imports plus exports) between South Africa and the
EU countries
The results show that the implementation of the EU-SA TDCA had no effects on total frozen
fruits and nuts trade between South Africa and the EU countries and also show no proof of
trade creation or diversion during both periods. The results for the average actual and
potential total frozen fruits and nuts trade between South Africa and the EU countries for the
periods 2000-2004 and 2005-2009 are presented in Figure 5.12 in log values, and the dollar
values are presented in Appendix 5BZ.
The results show that South Africa's frozen fruits and nuts traders (importers and exporters)
operating between South Africa and EU countries had outperformed and exhausted the
estimated potential capacity for the absorption of total frozen fruits and nuts trade in the South
Africa-France, South Africa-Netherlands and South Africa-United Kingdom markets; but had
124
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
under-scored or underachieved in the South Africa-Austria, South Africa-Belgium and South
Africa-Germany markets over the period 2000-2004. Over the period 2005-2009, South
Africa's frozen fruits and nuts traders (importers and exporters) outperformed and exhausted
the estimated potential capacity for the absorption of total frozen fruits and nuts trade in the
South Africa-Belgium and South Africa-Netherlands markets; whereas they under-scored or
underachieved in the South Africa-Germany and South Africa-United Kingdom markets.
16.00 .,..-------------------------
o'<t eon '<t en '<t en o'<t en o'<t en 'o<t eno oo o o o oo o o o o o o o o o oo
N N N N N N N N N N N N
oI J, oI Ilil 6 J, 6 I I Io o o o o loil oo loil o6
J,
o o o o o oo o o o o o o oo
N N N N N N N N N N N N
AUT BEL FRA DEU NLD GBR
Countries and Periods
"'Actual = Potential
Figure 5.12: Average actual and potential value of frozen fruits and nuts trade between
South Africa and the EU countries for the periods 2000-2004 and 2005-2009
5.5.5 Preserved fruits and nuts trade flows between South Africa and the EU countries
This subsection provides the results of the impacts of the implementation of the EU-SA
TDCA's non-reciprocal preserved fruits and nuts in-quota tariff preferences on South Africa's
preserved fruits and nuts exports to the EU countries; South Africa's preserved fruits and nuts
imports from the EU countries; as well as total preserved fruits and nuts trade (import plus
exports) between South Africa and EU countries for the periods 2000-2004 and 2005-2009.
The results are presented in Table 5.19 below.
125
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
Table 5.19: Results for preserved fruits and nuts trade flows between South Africa and
the EU countries
Models Variables Exports Imports Trade
2000 2005 2000 2005 2000 2005
2004 2009 2004 2009 2004 2009
Constant
- InYiil-1 0.18** 0.18** 0.15** 0.15** 0.39* 0.39*InGOPPCil 0.76 0.76 4.17 4.17uC': InGOPPCjt -0.54 -0.54 0.57 0.57
Q.._
_E "I0II InGOPPCiit -1.74*** -1.74***
"0 0 REERt 1.09** 1.09** 0.06 0.06 -0.40 -0.40
.:..2.. ;;; 00004 / 00509 -0.04 0.34 -0.91 0.97 -0.17 0.72*
III
Q.. InDlSTii
Adjusted R2 0.68 0.68 0.61 0.61 0.88 0.88
Observations 225 225 135 135 135 135
Cross-Sections 15 15 9 9 9 9
Constant
InYiit-1 0.17** 0.17** 0.15** 0.15** 0.42* 0.42*
InGOPPCit 10.92*** 10.92*** 4.32 4.32
- InGOPPCit -1.37 -1.37 0.16 0.16InGOPPCiit -1.76*** -1.76***uC': REERt 1.49** 1.49** -0.05 -0.05 -0.43 -0.43
Q..
_Ea:";0 DOO / DOS -0.50 -0.81 -0.47 0.93 -0.33** 0.91 *
>.0 001/006 -0.56 -1.48 -2.51 *** 1.09 -0.10 0.55**
-.::;E 002/007 -1.53 -1.79 -5.26** 1.32 0.24 0.75*
C':
III
;;.- 003/008 0.26 -2.08 -2.49 0.81 0.44** 0.71 *
004/009 -0.68 -2.18 -2.55 0.85 0.32 0.59*
InOISTii
Adjusted R2 0.68 0.68 0.60 0.60 0.88 0.88
Observations 225 225 135 135 135 135
Cross-Sections 15 15 9 9 9 9
Constant
InYiit-c 1
0.46* 0.46* 0.35* 0.35* 0.18* 0.18*
0 InGOPPCit -4.66 -4.66 7.47 7.47:;:
u InGOPPCit 0.83 0.83 3.07*** 3.07***
.I.I.I.
Qa:; InGOPPCiil 1.10 1.10
:::'g REERt -0.63 -0.63 -0.45 -0.45 -0.08 -0.08
~:;E PTAvcs -1.21** 1.11 -0.37 -0.51 -0.23 0.43***r.. PTA no -0.21 0.82 0.69 -0.20 0.59** 0.85*III
"0
.C..':. mmsr,Adjusted R2Eo-< 0.68 0.68 0.57 0.57 0.81 0.81
Observations 1275 1275 570 570 480 480
Cross-Sections 85 85 38 38 32 32
", ** & *** denote siBnificance at the 1,5 and 10 ~ercent levels res~ectively
Source: Author's calculations
5.5.5.1 Preserved fruits and nuts exports from South Africa to the EU countries
The results show that the implementation of the EU-SA TDeA had no effects on preserved
fruits and nuts exports from South Africa to the EU countries and also show no proof of trade
creation or diversion. The results for the average actual and potential preserved fruits and
nuts exports from South Africa to the EU countries for the periods 2000-2004 and 2005-2009
126
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
are presented in Figure 5.13 in log values, and the dollar values are presented in Appendix
5CC. The results show that South Africa's preserved fruits and nuts exporters had
outperformed and exhausted the estimated potential capacity for the absorption of South
Africa's preserved fruits and nuts exports in Austria, Belgium, Denmark, Finland, France,
Germany, Greece, Italy, Luxembourg, the Netherlands, Portugal and the United Kingdom; but
had under-scored or underachieved in Ireland, Spain and Sweden over the period 2000-2004.
18.00
~
16.00 ~~ ~ ~~
~
~~~~ ~ ~ jil ~ ~ - - ~~
14.00 V~rt - :~ll - :r- l~~:-~ ~~~:
~~~~r-: ~~~~ ~~~t~ _~: ,~;: ~I~;:~~~~12.00 ~~~~~U~~V~~
~ V~~~~~~~~ ~~~?~~?~~,~;r~~,~;~~ 10.00 r~,;r,;r,;~rri ~~~~~~~~~~~~~~.e~:~ól~:~g ~~~~~~~~~~
'" 8.00
~~~~~~~~~~~~~~~~~ ,-: ~~~~~~~H~
QI
:::l F~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~jil ~I~~ï>a 6.00 ~~~~~~~jIl~~~~~~~~~ r~~-~
~~~~~~~~~r'~jt;~~~~
r~-~~~~~~~q4.00 j!: ~ j!: rjl.:; ~ ~ V~ri~~~~~~~~~~~~~~~~ ~r-r-~~j!;f
2.00 P????~~???????ri ~~~~~~~~~~
0.00 ~~~~~~~ ~~~~~~~~~~
~~~~~~~~~~i~i~!=~j~It~:~~~
0~m0~000
m0~0m~0m0~0m0~m0~0m0~0m~0m0~0m~m~m~m~m~m
NNN0N0N0NN0N0N0N0NN0N0N0N0N0NN0N0N0
00000
N0NN0N0N0
000000000
N0N0NN0N000000
oI~oI ~oI ~Io~Io~Io~I o~I oI~oI~oI ~oI ~Io~Io~Io~I o~I I I I I I I I I I I I I I00000000000000000000000000000000000000000000000000000000000NNNNNNNNNNNNNNNNNNNNNNNNNNNNNN 0
AUT BEL DNK FIN FRA DEU GRC IRL ITA LUX NLD PRT ESP SWE GBR
Countries and Periods
~ Actual = Potential
Figure 5.13: Average actual and potential value of preserved fruits and nuts exports
from South Africa to EU countries for the periods 2000-2004 and 2005-2009
For the period 2005-2009, South Africa's preserved fruits and nuts exporters still
outperformed and exhausted the estimated potential capacity for the absorption of South
Africa's preserved fruits and nuts exports to Austria, Belgium, France, Germany, the
Netherlands, Sweden and the United Kingdom; whereas they under-scored or underachieved
in Denmark, Finland, Greece, Italy, Ireland, Portugal and Spain.
5.5.5.2 Preserved fruits and nuts imports from the EU countries to South Africa
The results show that the implementation of the EU-SA TDCA over both periods had no joint
effects on the preserved fruits and nuts imports from the EU countries to South Africa.
127
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
However, on an annual basis, there were significant declines in South Africa's preserved
fruits and nuts imports from the EU countries by 2.51% and 5.26% in 2001 and 2002
respectively. The results have also shown no proof of creation or diversion of South Africa's
preserved fruits and nuts imports from either the EU countries or South Africa's other
preserved fruits and nuts trading partners during the implementation of the EU-SA TDCA for
both periods. The results of the average actual and potential preserved fruits and nuts imports
from the EU countries to South Africa for the periods 2000-2004 and 2005-2009 are
presented in Figure 5.14 in log values, and the dollar values are presented in Appendix 5CF.
16.00
14.00 - ,.
12.00
~ 10.00
1/1
s110 8.00
1/1
.11.:1
>nl
6.00
4.00
2.00
0.00
'<t 0'1 '<t 0'1 '<t 0'1 '<t 0'1 '<t 0'1 '<t 0'1 '<t 0'1 '<t 0'1 '<t 0'1
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N N N N N N N
I I I I
0 LIl 0 LIl 6 ..n 6 ..n 6 ..n 6 ..n I I I I0 LIl 6 LIl 6 LIl
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N N N N N N N
AUT BEL FRA DEU GRC ITA NLD ESP GBR
Countries and Periods
~ Actual = Potential
Figure 5.14: Average actual and potential value of preserved fruits and nuts imports
from the EU countries to South Africa for the periods 2000-2004 and 2005-2009
The results show that South Africa's preserved fruits and nuts importers had outperformed
and exhausted the estimated potential capacity for the absorption of South Africa's preserved
fruits and nuts imports from Austria, Belgium, France, Germany, Greece, Italy, the
Netherlands and Spain; but had under-scored or underachieved the estimated potential
capacity for the absorption of South Africa's preserved fruits and nuts imports from the
United Kingdom for the period 2000-2004. In contrast, over the period 2005-2009, South
]28
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
Africa's preserved fruits and nuts importers outperformed and exhausted the estimated
potential capacity for the absorption of South Africa's preserved fruits and nuts imports from
Austria, Greece, the Netherlands and the United Kingdom; but under-scored or underachieved
the estimated potential capacity for the absorption of South Africa's preserved fruits and nuts
imports from Belgium, France, Germany, Italy and Spain.
5.5.5.3 Preserved fruits and nuts trade (imports plus exports) between South Africa and
the EU countries
The results show that the total preserved fruits and nuts trade between South Africa and the
EU countries had significantly improved by 0.72% during the implementation of the EU-SA
TDCA over the period 2005-2009. On an annual basis, significant increases occurred in all
years as follows: 0.91% in 2005,0.55% in 2006,0.75% in 2007,0.71 % in 2008 and 0.59% in
2009. Surprisingly, while the joint period effect of the implementation of the EU-SA TDeA
on total preserved fruits and nuts trade between South Africa and the EU countries was
insignificant for the period 2000-2004, the individual yearly significant negative and positive
effects were observed in 2000 and 2003 respectively. In 2000, the total preserved fruits and
nuts trade between South Africa and the EU countries significantly declined by 0.33%, but
increased by 0.44% in 2003. Furthermore, on average, the implementation of the EU-SA
TDCA led to the creation of 0.43% in the total preserved fruits and nuts trade between South
Africa and the EU countries over the period 2005-2009. However, there was no proof of
creation or diversion of total preserved fruits and nuts trade between South Africa and the EU
countries to other preserved fruits and nuts trading partners of South Africa or EU countries
during the period 2000-2004.
The results for the average actual and potential total preserved fruits and nuts trade between
South Africa and the EU countries for the periods 2000-2004 and 2005-2009 are presented in
Figure 5.15 in log values, and the dollar values are presented in Appendix SCI. The results
show that South Africa's preserved fruits and nuts traders (importers and exporters) operating
between South Africa and EU countries had outperformed and exhausted the estimated
potential capacity for the absorption of total preserved fruits and nuts trade in the South
Africa-Belgium, South Africa-France and South Africa-Greece markets; but had under-scored
or underachieved in the South Africa-Austria, South Africa-Germany, South Africa-Italy,
129
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
South Africa-Netherlands, South Africa-Spain and South Africa-United Kingdom markets
over the period 2000-2004. During the period 2005-2009, South Africa's preserved fruits
and nuts traders (importers and exporters) outperformed and exhausted the estimated potential
capacity for the absorption of total preserved fruits and nuts trade in the South Africa-France,
South Africa-Germany, South Africa-Netherlands and South Africa-United Kingdom
markets; whereas they under-scored or underachieved in the South Africa-Austria, South
Africa-Belgium, South Africa-Greece, South Africa-Italy and South Africa-Spain markets.
However, the differences between the actual and the potential values were small.
18.00
16.00
14.00
12.00
~
."
gl1li 10.00
."
III 8.00
:::J
>ra 6.00
4.00
2.00
0.00
'<t en '<t en '<t en '<t en '<t en '<t en '<t en '<t en '<t en
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N N N N N N N
I
0 JJ 6 JJ 6 JJ 6 JJ 6 JJ 6 JJ 6 ILIl 6 JJ 6 JJ
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N N N N N N N
AUT BEL FRA DEU GRC ITA NLD ESP GBR
Countries and Periods
~Actual :: Potential
Figure 5.15: Average actual and potential value of preserved fruits and nuts trade
between South Africa and the EU countries for the periods 2000-2004 and 2005-2009
5.5.6 Fruits and vegetable juices trade flows between South Africa and the EU
countries
This subsection provides the results of the impacts of the implementation of the EU-SA
TDCA's non-reciprocal fruits and vegetable juices in-quota tariff preferences on South
Africa's fruits and vegetable juices exports to the EU countries; South Africa's fruits and
vegetable juices imports from the EU countries; as well as total fruits and vegetable juices
130
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
trade (import plus exports) between South Africa and EU countries for the periods 2000-2004
and 2005-2009. The results are presented in Table 5.20 below.
Table 5.20: Results for fruits and vegetable juices trade flows between South Africa and
the EU countries
Models Variables Exports Imports Trade
2000 2005 2000 2005 2000 2005
2004 2009 2004 2009 2004 2009
Constant
InYiit-1 0.11 *** 0.11 *** 0.36* 0.36* 0.17** 0.17**
- InGDPPCit 16.92** 16.92** 9.11 9.11<.I
«:l InGDPPCit 1.53 1.53 5.74 5.74
cE._<lj InGDPPCiit 4.55 4.55
....,"0
"0 0 REERt 1.10 1.10 1.74 1.74 0.45 0.45
.o;:~ 00004 / 00509 -0.97 -2.55 -1.25 -1.81 -0.29 -0.10
<lj
Q.. InDISTii
Adjusted 2R 0.70 0.70 0.63 0.63 0.56 0.56
Observations 210 210 195 195 195 195
Cross-Sections 14 14 13 13 13 13
Constant
InYiil-1 0.11 *** 0.11 *** 0.36* 0.36* 0.16** 0.16**
InGDPPCit 16.94 16.94 -0.59 -0.59
InGDPPCit 0.96 0.96 5.88 5.88
- InGDPPCiit 4.05 4.05<.I
«:l REERt 0.93 0.93 1.22 1.22 0.61 0.61
c.
Ea:; DOO/ DOS 0.07 -2.41 -1.42 -0.48 -0.31 -1.37***
_"0
-2~0 001/006 0.48
-2.55 -0.28 0.06 -0.09 0.25
002/007 1.66 -2.01 -0.55 0.52 0.17 0.10
«:l
<lj
;;.... 003/008 2.41 -2.52 -1.30 1.05 0.25 0.90
004/009 1.60 -2.84 -1.05 -0.05 -0.22 0.08
InOlSTii
Adjusted 2R 0.69 0.69 0.63 0.63 0.58 0.58
Observations 210 210 195 195 195 195
Cross-Sections 14 14 13 13 13 13
Constant
InYiit-1 0.27* 0.27* 0.24* 0.24* 0.20* 0.20*
.c-s: InGDPPCit 8.29** 8.29** 13.86** 13.86**<...I. InGDPPCit 2.17* 2.17* 1.32 1.32<lj
Qa:; InGDPPCiit 4.29* 4.29*
... "0 REER, -0.33 -0.33 -0.39 -0.39 -0.09 -0.09
...:::.;;0.. PTAve, -0.44 -1.04 -1.60*** -1.90 -0.23 -0.08
I;i;.. 0.45 -0.44 -0.39 -0.49 0.67** 0.75*
<lj PTA no
".0«.:l. InOlSTii
E-< Adjusted 2R 0.65 0.65 0.62 0.62 0.68 0.68
Observations 1290 1290 645 645 630 630
Cross-Sections 86 86 43 43 42 42
*, ** & *** denote significance at the 1,5 and 10 ~ercent levels res~ectivel~
Source: Author's calculations
131
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
5.5.6.1 Fruits and vegetable juices exports from South Africa to the EU countries
The results show that the implementation of the EU-SA TDCA for both periods had no effects
on South Africa's fruits and vegetable juices exports to the EU countries. Furthermore, the
results show no proof of creation in South Africa's fruits and vegetable juices exports market
in the EU countries or diversion of South Africa's fruits and vegetable juices exports destined
for the EU market to other South African fruits and vegetable juices trading partners for both
periods. The results for the average actual and potential fruits and vegetable juices exports
from South Africa to the EU countries for the periods 2000-2004 and 2005-2009 are
presented in Figure 5.16 in log values, and the dollar values are presented in Appendix SeL.
18.00
16.00
14.00
12.00
~
III
gbO
10.00
III
III 8.00
:::I
ii
> 6.00
4.00
2.00
0.00
<t Ol <t Ol <t Ol <t Ol <t Ol <t Ol <t Ol <t Ol <t Ol <t Ol <t Ol <t Ol <t Ol <t Ol
0000000000000000000000000000
0000000000000000000000000000
NNNNNNNNNNNNNNNNNNNNNNNNNNNN
oI I I I I I I I I I I I I I I I I I I I I I I I I I I Ili) 0 li) 0 li) 0 li) 0 li) 0 li) 0 li) 0 li) 0 li) 0 li) 0 li) 0 li) 0 li) 0 li)
0000000000000000000000000000
0000000000000000000000000000
NNNNNNNNNNNNNNNNNNNNNNNNNNNN
AUT BEL DNK FIN FRA DEU GRC IRL ITA NLD PRT ESP SWE GBR
Countries and Periods
II!Actual = Potential
Figure 5.16: Average actual and potential value of fruits and vegetable juices exports
from South Africa to EU countries for the period 2000 to 2004
The results show that South African fruits and vegetable juices exporters had outperformed
and exhausted the estimated potential capacity for the absorption of South Africa's fruits and
vegetable juices exports in Denmark, Finland, Greece, Italy, the Netherlands and Sweden; but
had under-scored or underachieved in Austria, Belgium, France, Germany, Ireland, Portugal,
132
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
Spain and the United Kingdom over the period 2000-2004. For the period 2005-2009, South
Africa's fruits and vegetable juices exporters continued to outperform and exhaust the
estimated potential capacity for the absorption of South Africa's fruits and vegetable juices
exports in Belgium, Denmark, France, Germany, Italy, Ireland, the Netherlands, Portugal,
Spain and the United Kingdom; whereas they under-scored or underachieved in Austria,
Finland and Sweden.
5.5.6.2 Fruits and vegetable juices imports from the EU countries to South Africa
The results also show that the implementation of the EU-SA TDCA over both periods had no
effects on the fruits and vegetable juices imports from the EU countries to South Africa.
Furthermore, the results show no proof of creation in South Africa's fruits and vegetable
juices import market from the EU countries or diversion of South Africa's fruits and
vegetable juices import market from the EU countries to other South African fruits and
vegetable juices trading partners.
The results for the average actual and potential fruits and vegetable juices imports from the
EU countries to South Africa for the periods 2000-2004 and 2005-2009 are presented in
Figure 5.17 in log values, and the dollar values are presented in Appendix seo. The results
show that South Africa's fruits and vegetable juices importers had outperformed and
exhausted the estimated potential capacity for the absorption of South Africa's fruits and
vegetable juices imports from Austria, Belgium, Germany, Greece, Ireland, Italy, the
Netherlands, Portugal, Sweden and the United Kingdom; but had under-scored or
underachieved the estimated potential capacity for the absorption of South Africa's fruits and
vegetable juices imports from Denmark, France and Spain over the period 2000-2004.
For the period 2005-2009, South Africa's fruits and vegetable juices importers had
outperformed and exhausted the estimated potential capacity for the absorption of South
Africa's fruits and vegetable juices imports from Belgium, France, Ireland, Italy, Portugal,
Sweden and Spain; but under-scored or underachieved the estimated potential capacity for the
absorption of South Africa's fruits and vegetable juices imports from Austria, Denmark,
Germany, Greece, the Netherlands and the United Kingdom.
133
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
16.00
14.00
12.00
~ 10.00
1/1
l1li
.-.2 8.00
c1/u1
::I
i>V
6.00
4.00
2.00
0.00
'<t (TI '<t (TI '<t (TI '<t (TI '<t (TI '<t (TI '<t (TI '<t (TI '<t (TI '<t (TI '<t (TI '<t (TI '<t (TI
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
N N.n N N N .Nn N .N N N N N N N N N N N N N N N N N6 6 ILIl 6 6 n 6 .n 6 .n 6 .n 6 .n 6 .n 6 .n 6 .n 6 .n N N6 ILIl
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N N N N N N N N N N N N N N N
AUT BEL DNK FRA DEU GRC IRL ITA NLD PRT ESP SWE GBR
Countries and Periods
~ Actual = Potential
Figure 5.17: Average actual and potential value of fruits and vegetable juices imports
from the EU countries to South Africa for the periods 2000-2004 and 2005-2009
5.5.6.3 Fruits and vegetable juices trade (imports plus exports) between South Africa
and the EU countries
The results show that the implementation of the EU-SA TDCA over both periods had no joint
effects on the total fruits and vegetable juices trade between South Africa and the EU
countries, but there was a significant decline of 1.37% which occurred in 2000. However,
there was no proof of creation or diversion of total fruits and vegetable juices trade between
South Africa and the EU countries to other fruits and vegetable juices trading partners of
South Africa or EU countries.
The results for the average actual and potential total fruits and vegetable juices trade between
South Africa and the EU countries for the periods 2000-2004 and 2005-2009 are presented in
Figure 5.18 in log values, and the dollar values are presented in Appendix SeR. The results
show that South Africa's fruits and vegetable juices traders (importers and exporters)
operating between South Africa and EU countries had outperformed and exhausted the
estimated potential capacity for the absorption of total fruits and vegetable juices trade in the
134
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
South Africa-Denmark, South Africa-Greece, South Africa-Italy and South Africa-
Netherlands markets; but had under-scored or underachieved in the South Africa-Austria,
South Africa-Belgium, South Africa-France, South Africa-Germany, South Africa-Ireland,
South Africa-Portugal, South Africa-Spain, South Africa-Sweden and South Africa-United
Kingdom markets over the period 2000-2004.
18.00
16.00
14.00
12.00
~
~ 10.00
.2
~II> 8.00
.2
>nl 6.00
4.00
2.00
0.00 ..,. 0'\ ..,. 0'\ <:t 0'\ <:t 0'1 <:t 0'\ <:t 0'\ <:t 0'\ ..,. 0'\ <:t CTI <:t CTI ..,. 0'\ <:t 0'\ <:t CTI
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N N N N N N N N N N
6 ,;., 6 '"';J '"';J '"';J '"';JLI') 6 ,;., 6 LI') 6 LI') 6 LI') 6 ,;., 6 ,;., 6 ,;., 6 6 ,;., 6 ,;., 6
0 0 0 0
0 0 0 0 818 0 0 0 0 0 0 0 0 0 0 0 0 0 o';"10 0 0 0 0 0 0 0 0 0 0 0 0 o 0 0 0 0 00 0 0 0 0
N N N N N N N N N N N N N N N N N N N N N N N N N ~I
AUT BEL DNK FRA DEU GRC IRL ITA NLD PRT ESP SWE GBR
Countries and Periods
~ Actual == Potential
Figure 5.18: Average actual and potential value of fruits and vegetable juices trade
between South Africa and the EU countries for the periods 2000-2004 and 2005-2009
However, over the period 2005-2009, South Africa's fruits and vegetable juices traders
(importers and exporters) outperformed and exhausted the estimated potential capacity for the
absorption of total fruits and vegetable juices trade in the South Africa-Belgium, South
Africa-France, South Africa-Greece, South Africa-Ireland, South Africa-Italy, South Africa-
Netherlands, South Africa-Portugal and South Africa-Spain markets; whereas they under-
scored or underachieved in the South Africa-Austria, South Africa-Denmark, South Africa-
Germany, South Africa-Sweden and South Africa-United Kingdom markets.
135
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
5.5.7 Wine trade flows between South Africa and the EU countries
This subsection provides the results of the impacts of the implementation of the EU-SA
TDCA's reciprocal wine in-quota tariff preferences on South Africa's wine exports to the EU
countries; South Africa's wine imports from the EU countries; as well as total wine trade
(import plus exports) between South Africa and EU countries for the periods 2000-2004 and
2005-2009. The results are presented in Table 5.21.
Table 5.21: Results for wine between South Africa and the EU countries
Models Variables Exports Imports Trade
2000 2005 2000 2005 ·2000 2005
2004 2009 2004 2009 2004 2009
Constant
InYiit-1 0.37* 0.37* 0.62* 0.62*
InGDPPCit 3.72*** 3.72*** -2.04 -2.04
InGDPPCit 3.80* 3.80* 6.17* 6.17*
InGDPPCiit 3.13* 3.13*
REERt 0.01 0.01 1.11 I.Il 0.31 0.31
D0004 / D0509 -0.44** 0.02 -1.69* 0.84 0.07 0.24
InDISTii
Adjusted R2 0.95 0.95 0.81 0.81 0.96 0.96
Observations 225 225 240 240 225 225
Cross-Sections 15 15 15 15 15 15
Constant
1nYiit_1 0.34* 0.34* 0.64* 0.64*
InGDPPCit 12.96* 12.96* 8.90 8.90
InGDPPCit 3.85* 3.85* 5.27** 5.27**
InGDPPCiit 3.41 * 3.41 *
REERt 0.54 0.54 1.56 1.56 0.38 0.38
DOO/ DOS -0.50 -1.05*** -1.80** -0.19 -0.06 0.46**
DOl / D06 -0.58 -1.99** -3.30* -1.54 -0.09 -0.12
D02 / D07 -0.78 -2.22** -3.92** -1.06 -0.02 0.11
D03 / D08 -0.16 -2.43** -2.04 -1.85 0.28 0.06
D04/ D09 -0.45 -1.73*** -1.96 -1.97 0.31 0.32
InDISTii
Adjusted R2 0.95 0.95 0.81 0.81 0.96 0.96
Observations 225 225 240 240 225 225
Cross-Sections 15 15 15 15 15 15
Constant
InYiit_1 0.30* 0.30* 0.13* 0.13* 0.50* 0.50*
InGDPPCit 3.58 3.58 4.40 4.40
InGDPPCit 0.32 0.32 1.62** 1.62**
InGDPPCiit 3.95* 3.95*
REERt -1.19* -1.19* 1.95* 1.95* 0.39*** 0.39***
PTAve, -0.67 0.79 -0.69 0.06 -0.24 0.33***
PTA no -0.77** 1.04*** -2.01 * -0.81 -0.51 *** 0.34***
InDISTii
Adjusted R2 0.74 0.74 0.75 0.75 0.87 0.87
Observations 1530 1530 675 675 660 660
Cross-Sections 102 102 45 45 44 44
*, ** & *** denote significance at the 1,5 and 10 percent levels respectively
Source: Author's calculations
136
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
5.5.7.1 Wine exports from South Africa to the EU countries
The results show that wine exports from South Africa to the EU countries had significantly
declined by 0.44% during the implementation of the EU-SA TOCA for the period 2000-2004.
Surprisingly, the joint period effect of the implementation of the EU-SA TDCA on South
Africa's wine exports to the EU countries was insignificant for the period 2005-2009, but the
individual yearly significant negative effects were observed in all years as wine exports from
South Africa to the EU countries had significantly declined by 1.05%, 1.99%, 2.22%, 2.43%
and 1.73% respectively during these years. On average, 0.77% of the South Africa's wine
exports destined for the EU market were diverted to other wine trading partners of South
Africa over the period 2000-2004. However, there was no proof of creation of South Africa's
wine exports market in the EU countries or diversion of South Africa's wine exports destined
for the EU countries to other of South Africa's wine trading partners.
The results for the average actual and potential wine exports from South Africa to the EU
countries for the periods 2000-2004 and 2005-2009 are presented in Figure 5.19 in log
values, and the dollar values are presented in Appendix 5eu. The results show that South
Africa's wine exporters had outperformed and exhausted the estimated potential capacity for
the absorption of South Africa's wine exports in Austria, Denmark, Finland, Germany,
Luxembourg, the Netherlands and Portugal; but had underscored or underachieved in
Belgium, France, Greece, Italy, Ireland, Spain and Sweden and the United Kingdom for the
period 2000-2004. For the period 2005-2009, South Africa's wine exporters still
outperformed and exhausted the estimated potential capacity for the absorption of South
Africa's wine exports in Denmark, Germany, Greece, Italy, Luxembourg, Spain and Sweden;
whereas they under-scored or underachieved in Austria, Belgium, Finland, France, Ireland,
the Netherlands, Portugal and the United Kingdom.
137
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
22.00
20.00 ~
18.00
~. ~ ~ .. ,,_ ~. p-,: ~ ~~Af16.00 ~ ~~
14.00 l~~~~~~~~~~~~~~~~~~~~~~ ~~ .- ~~
F~rt
~ -,:~ -,:
III ~~~I!E~ .H~f
bO
- 12.00
~~
.2
~~~~~ ~~~ ~,:: r
III 10.00 .~ .. ~.~~~~~~~~~~
III
:::I ~~~~~~~~~~~~~~~~~~~~~ ~ § rrrrrtti>V 8.00
6.00 ~~~~~~~~~~~~~~~~~~~~~ ~ uu
4.00 ~~~~~~~~~~~~~~~~~~~~~ ~ ~ rrrt
2.00 ~~~~~~~~~~~~~~~~~~~~~
~ ~ ~ rrrt
Z~Z~~~~~0.00 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~ r=r=rt
~0m0~0m0~0m~0m0~0m0~m0~0m0~0m~0m0~0m0~m0~0m0~0m~0m0~0m000000
N0N0N0N0N0N0NN0N0N0N0N0N0N0NN0N0N0N0N0N0NN0N0N0N0N0N0NN00000
oI~oI ~oI ~o• ~Io~Io~Io~I oI~oI~oI~oI ~oI I I I I I I I I I I I I I I I I I00000000000000000000000000000000000000000
~0o0~0o~0o0~0000
NNNNNNNNNNNNNNNNNNNNNNNN0N0N0N0N0NN00000
AUT BEL DNK FIN FRA DEU GRC IRL ITA LUX NLD PRT ESP SWE GBR
Countries and Periods
~Actual = Potential
Figure 5.19: Average actual and potential value of wine exports from South Africa to
EU countries for the periods 2000-2004 and 2005-2009
5.5.7.2 Wine imports from the EU countries to South Africa
The final results for South Africa's wine imports from the EU countries are presented in
Table 5.29 below. The results show that South Africa's wine imports from the EU countries
had significantly declined by 1.69% during the implementation of the EU-SA TDeA over
period 2000-2004. On an annual basis, these decreases occurred in 2000, 2001 and 2002 by
1.8%, 3.3% and 3.92% respectively. Furthermore, the implementation of the EU-SA IDeA
over the period 2000-2004 led to a diversion of South Africa's wine imports, which were
ordinarily soureed from the EU countries, by 2.01%. However, the implementation of the
EU-SA IDeA over period 2005-2009 had no effects on South Africa's wine imports from
the EU countries and there' was no proof of creation in South Africa's wine import market
from the EU countries or diversion of South Africa's wine import market from the EU
countries to other South African wine trading partners.
138
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
The results for the average actual and potential wine imports from the EU countries to South
Africa over the periods 2000-2004 and 2005-2009 are presented in Figure S.20 in log values,
and the dollar values are presented in Appendix sex.
18.00
16.00 ~~-
14.00 i!;:~ s, ,.,:
~H~ ,~,~~ ~p~~::~ ~-12.00
~ ~. ~;;;: iIr~ ~~
III ~~ ~ ~;;;: ::~~~~ ~~
11.0 10.00
.2 • r: V~~~r;i;~;:~i~. ~H~ ~V~~~~~~~~~~H-II ~~~cuI 8.00 ~~
::::J
'iii jfj:~~~
:
> ~??~ ~~~;;;:~~
~~~
,.;-~?~r.~?;;
~;:~d ~~~~~~~~
~~ ~~~~~~
6.00
usu p~~=~~~~ ~~ :~~~;V;;: ~?~?~~~~~~~~~,~,~~~?rt4.00 ::~~;;;:~~~~~ ??d~~~~~~~
2.00 u~~~r~ n ~~~:~~V~~~~~~~?~~?jI:~~~1E ~~~~ ~~?~~~~~~;~~~;;~;;V;:;~:;;:~I~~~~~~~~~?~d~~~~~~~~0.00 .~.~~~~.~ ~~~~ ~~~~~
0'<10"0\ 0'<1"00\ 0'<I" 00\'<01" 00\'<01" 00\'<01" 00\'<01" 0\'<1" 00\'<01" 0\'<1" 0\'<1" 0\'<1" 0\'<1" 0\'<1" 0\'<1" 0\
N0N0 N0 N0N0 N0N0 00 00 00 0
000 00 000 00 00 00 00 00 0NN NN NN NN NN NN0 N0N0 N0N0 N0N0 0NN0 N0N0 N0
cJ, 86 J,6 J,6 J,6 J,6
' ,
00 00 00 0LA0c 00 00 0LA0c L0/'IO J,600 00 00 00 00 00 00 000 00NN N NN NN NN NN NN NN NN N 8
J,6 J,6 J,6 J,6
00 00 00LAc00 00.n
NN8 N0N0 N0N0 N0N0 N0N0 N0N0
AUT BEL DNK FIN FRA DEU GRC IRL ITA LUX NLD PRT ESP SWE GBR
Countries and Periods
~Actual = Potential
Figure 5.20: Average actual and potential value of wine imports from the EU countries
to South Africa for the periods 2000-2004 and 2005-2009
The results show that South Africa's wine importers had outperformed and exhausted the
estimated potential capacity for the absorption of South Africa's wine imports from Austria,
Belgium, France, Greece, Italy and Sweden; but had under-scored or underachieved the
estimated potential capacity for the absorption of South Africa's wine imports from Denmark,
Finland, Germany, Ireland, the Netherlands, Portugal and Spain over the period 2000-2004.
However, the actual and potential of South Africa's wine imports from Luxembourg and the
United Kingdom were close to equal.
In contrast, during the period 2005-2009, South Africa's wine importers outperformed and
exhausted the estimated potential capacity for the absorption of South Africa's wine imports
from Austria, France, Germany, Italy, Ireland, Luxembourg and Sweden; but under-scored or
underachieved the estimated potential capacity for the absorption of South Africa's wine
139
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
imports from Belgium, Denmark, Finland, Greece, the Netherlands, Portugal, Spain and the
United Kingdom.
5.5.7.3 Wine trade (imports plus exports) between South Africa and the EU countries
The results show that the implementation of the EU-SA TDCA for both periods had no joint
effects on the total wine trade between South Africa and the EU countries, but there was a
significant decline of 0.46% that occurred in 2005. The implementation of the EU-SA TDCA
during the period 2000-2004 led to a diversion of 0.51% of the total wine trade between
South Africa and the EU countries to either South Africa and its other wine trading partners or
to EU countries and their other wine trading partners. However, the implementation of the
EU-SA TDCA led to the creation of 0.93% of total wine trade between South Africa and the
EU countries over the period 2005-2009.The results for the average actual and potential total
wine trade between South Africa and the EU countries for the periods 2000-2004 and 2005-
2009 are presented in Figure 5.21 in log values, and the dollar values are presented in
Appendix 5DA.
The results show that South Africa's wme traders (importers and exporters) operating
between South Africa and the EU countries had outperformed and exhausted the estimated
potential capacity for the absorption of total wine trade in the South Africa-Austria, South
Africa-Belgium, South Africa-Denmark, South Africa-Finland, South Africa-Germany, South
Africa-Luxembourg, South Africa-Netherlands and South Africa-United Kingdom markets;
but had under-scored or underachieved in the South Africa-France, South Africa-Greece,
South Africa-Ireland, South Africa-Italy, South Africa-Portugal and South Africa-Spain
markets over the period 2000-2004. However, the actual and potential total wine trade
between South Africa and Sweden was closer to equal. Over the period 2005-2009, South
Africa's wine traders (importers and exporters) outperformed and exhausted the estimated
potential capacity for the absorption of total wine trade in the South Africa-Austria, South
Africa-Denmark, South Africa-France, South Africa-Germany, South Africa-Italy, South
Africa-Luxembourg and South Africa-Sweden markets; whereas they under-scored or
underachieved in the South Africa-Belgium, South Africa-Finland, South Africa-Greece,
South Africa-Ireland, South Africa-Netherlands, South Africa-Portugal, South Africa-Spain
and South Africa-United Kingdom markets.
140
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
18.00
16.00
14.00
~ 12.00
aII.Io
g 10.00
III
Ol
~ 8.00
>
0~m~0000
m0~0m~m~m~m~m~m000000000000000000000
~m0~0m0~0m0~m0~0m0~0m~m00000000000000000000000
06
NN~N6N~NN6N~N6NN~N6N~NN6N~N6NN~N6NNNNNNNNNNN
00000000000000000000000000000000
~060~060~060~060~0600~060~6~
NNNNNNNNNNNNNNNNNNNN0N0N0NN0N0N0N0N0NN000000
AUT BEL DNK FIN FRA DEU GRC IRL ITA LUX NLD PRT ESP SWE GBR
Countries and Periods
II! Actual = Potential
Figure 5.21: Average actual and potential value of wine trade between South Africa and
the EU countries for the periods 2000-2004 and 2005-2009
5.6 The impacts of the implementation of SADC Trade Protocol (TP) on selected
agricultural trade flows between South Africa and the SADC countries
This section discusses the impacts of the SADe Trade Protocol regional trade agreement on
the total agricultural trade flows and the selected agricultural products trade flows between
South Africa and the SADe countries for the periods 2000-2004 and 2005-2009. The
selected agricultural products include cheese and curd (HS0406), cut flowers (HS0603),
frozen fruits and nuts (HS0811), preserved fruits and nuts (HS2008), fruits and vegetable
juices (HS2009) and wines (HS2204).
5.6.1 Aggregate agricultural trade flows between South Africa and the SADe
countries
This subsection provides the results of the impacts of the implementation of the SADe Trade
Protocol on South Africa's agricultural exports to the SADe countries; South Africa's
agricultural imports from the SADe countries; as well as total agricultural trade (import plus
141
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
exports) between South Africa and SADe countries for the periods 2000-2004 and 2005-
2009. The results are presented in Table 5.22.
Table 5.22: Results for agricultural trade flows between South Africa and the SADe
countries
Models Variables Exports Imports Trade
2000 2005 2000 2005 2000 2005
2004 2009 2004 2009 2004 2009
Constant
InViit_) 0.26* 0.26* 0.48* 0.48* 0.30* 0.30*
InGOPPCit 1.02 1.02 3.19 3.19
InGOPPCit -0.17 -0.17 0.40*** 0.40***
InGOPPCiit 0.65 0.65
REERt -0.25 -0.25 -0.26 -0.26 -0.21 -0.21
D0004 /00509 -0.56* 0.84** 0.01 0.23 -0.20** 0.87*
InOJSTii
Adjusted R2 0.87 0.87 0.93 0.93 0.92 0.92
Observatións 90 90 90 90 90 90
Cross-Sections 6 6 6 6 6 6
Constant 15.40* 15.40*
InViit-) 0.34* 0.34* 0.49* 0.49* 0.38* 0.38*
InGOPPCit -0.98 -0.98 9.17*: 9.17**
InGOPPCit -0.19 -0.19 0.47** 0.47**
InGOPPCiit 0.20 0.20
REERt -0.27 -0.27 0.06 0.06 -0.17 -0.17
DOO /005 -0.53 * 1.25* 0.08 -0.70 -0.20** 0.98*
001/006 -0.89* 0.93 0.27 -0.74 -0.28** 0.68*
D02/007 -1.00* 0.78 0.91 ** -0.97 -0.03 0.63*
003/008 -0.69* 1.64** 1.30* - 1.64 0.10 1.15*
004/009 -0.89* 1.30*** 0.78*** - 1.33 0.06 0.91 *
mmsr, -0.72* -0.72*
Adjusted R2 0.89 0.89 0.93 0.93 0.87 0.87
Observations 90 90 90 90 90 90
Cross-Sections 6 6 6 6 6 6
Constant
InViit-) 0.38* 0.38* 0.23* 0.23* 0.37* 0.37*
InGOPPCit 2.06** 2.06** 5.39** 5.39**
InGOPPCit 0.03 0.03 0.80* 0.80*
InGOPPCiit 0.46* 0.46*
REERt -0.54* -0.54* -0.32 -0.32 -0.48* -0.48*
PTAv.s -0.69** 0.45 0.01 0.04 -0.01 0.83*
PTA no -0.43 * 0.59** 0.31 -0.20 0.24** 0.95*
mnisr,
Adjusted R2 0.86 0.86 0.76 0.76 0.89 0.89
Observations 1785 1785 1905 1905 1635 1635
Cross-Sections 119 119 127 127 109 109
*, ** & ** * denote significance at the I, 5 and 10 percent levels respectively
Source: Author's calculations
142
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
5.6.1.1 Agricultural exports from South Africa to the SADe countries
The results show that agricultural exports from South Africa to the SADe countries had
significantly declined by 0.56% during the implementation of the SADe Trade Protocol over
the period 2000-2004. On an annual basis, these significant decreases occurred in all years as
follows: 0.53% in 2000, 0.89% in 2001, 1% in 2002, 0.69% in 2003, and 0.89% in 2004. On
the other hand, during the implementation of the SADe Trade Protocol over the period 2005-
2009, the agricultural exports from South Africa to the SADe countries had significantly
improved by 0.84% and on annual basis, the significant increases occurred in 2005 by 1.25%,
in 2008 by 1.64%, and in 2009 by 1.3%. On average, 0.69% of South Africa's agricultural
exports destined for the SADe market were diverted to other agricultural trading partners of
South Africa for the period 2000-2004. However, for the period 2005-2009, the
implementation of the SADe Trade Protocol led to the creation of about 0.59% of South
Africa's agricultural exports market in the SADe countries.
The results for the average actual and potential agricultural exports from South Africa to the
SADe countries for the periods 2000-2004 and 2005-2009 are presented in Figure 5.22 in
log values, and the dollar values are presented in Appendix 5DD. The results show that
South Africa's agricultural exporters had outperformed and exhausted the estimated potential
capacity for the absorption of South Africa's agricultural exports in Malawi, Mauritius,
Mozambique and Zambia; but had under-scored or underachieved in Tanzania and Zimbabwe
for the period 2000-2004. In contrast, for the period 2005-2009, South Africa's agricultural
exporters outperformed and exhausted the estimated potential capacity for the absorption of
South Africa's agricultural exports in Mauritius, Zambia and Zimbabwe; whereas they under-
scored or underachieved in Malawi, Mozambique and Tanzania. However, the differences
between the actual and the potential values were small.
143
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
20.00
18.00
16.00
14.00
~ 12.00
III
gbO 10.00
III
III
.:! 8.00
ru
> 6.00
4.00
2.00
0.00
<:t 0"\ <:t 0"\ <:t 0"\ <:t 0"\ <:t 0"\ <:t 0"\
0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N
I I 6 I I I0 LIl LIl 0 JJ 6 LIl 6 JJ 6 JJ
0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N
MWI MUS MOl TlA 1MB lWE
Countries and Periods
~Actual = Potential
Figure 5.22: Average actual and potential value of agricultural exports from South
Africa to SADe countries for the periods 2000-2004 and 2005-2009
5.6.1.2 Agricultural imports from the SADe countries to South Africa
The results show that the implementation of the SADe Trade Protocol had no joint period
effects on the agricultural imports from the SADe countries to South Africa for both periods
and show no proof of creation or diversion of South Africa's agricultural imports from either
the SADe countries or South Africa's other agricultural trading partners. However, on an
annual basis, the individual yearly positive effects were observed in 2002, 2003 and 2004,
when the agricultural imports from the SADe countries to South Africa increased by 0.91%,
1.3% and 0.78% respectively during those years.
The results for the average actual and potential agricultural imports from the SADe countries
to South Africa for the periods 2000-2004 and 2005-2009 are presented in Figure 5.23 in log
values, and the dollar values are presented in Appendix 5DG. The results show that South
Africa's agricultural importers had outperformed and exhausted the estimated potential
capacity for the absorption of South Africa's agricultural imports from Malawi, Tanzania,
144
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
Zambia and Zimbabwe; but had under-scored or underachieved in Mauritius and
Mozambique over the period 2000-2004. In contrast, for the period 2005-2009 South
Africa's agricultural importers outperformed and exhausted the estimated potential capacity
for the absorption of South Africa's agricultural imports from Mauritius, Tanzania and
Zambia, but had under-scored or underachieved in Malawi, Mozambique and Zimbabwe.
20.00
18.00
16.00
14.00
~ 12.00
III
glID 10.00
III
QI
-!= 8.00>II 6.00
4.00
2.00
0.00
'<t Ol '<t Ol '<t Ol '<t Ol '<t Ol '<t Ol
0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N
6 I I ILIl 6 LIl 6 J, 6 J, 6 J, 6 LIl
0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N
MWI MUS Mal TlA IMB IWE
Countries and Periods
~Actual ;; Potential
Figure 5.23: Average actual and potential value of agricultural imports from the SADe
countries to South Africa for the periods 2000-2004 and 2005--2009
5.6.1.3 Agricultural trade (imports plus exports) between South Africa and the SADe
countries
The results show that total agricultural trade between South Africa and the SADe countries
had significantly declined by 0.2% during the implementation of the SADe Trade Protocol
over the period 2000-2004 and on an annual basis, significant decreases occurred in 2000 and
2001 by 0.2% and 0.28% respectively. However, the implementation of the SADe Trade
Protocol for the period 2005-2009 contributed to a 0.87% increase in total agricultural trade
between South Africa and the SADe countries and on an annual basis, significant increases
occurred in all years as follows: 0.98% in 2005, 0.68% in 2006, 0.63% in 2007, 1.15% in
145
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
2008, and 0.91% in 2009. On average, the implementation of the SADe Trade Protocol led
to the creation of 0.83% in total agricultural trade between South Africa and the SADe
countries over the period 2005-2009. However, there was no proof of creation or diversion
of total agricultural trade between South Africa and the SADe countries to other agricultural
trading partners of South Africa or SADe countries during the period 2000-2004. The results
for the average actual and potential total agricultural trade between South Africa and the
SADe countries for the periods 2000-2004 and 2005-2009 are presented in Figure 5.24 in
log values, and the dollar values are presented in Appendix 5DJ.
22.00
20.00
18.00
16.00
14.00
~
III
gaD 12.00
III 10.00
al
-
>r
=a 8.00
6.00
4.00
2.00
0.00
'<t 0'1 '<t 0'1 '<t 0'1 '<t 0'1 '<t 0'1 '<t 0'1
0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N
I I I I I I
0 LI'l 0 LI'l 6 J, 6 LI'l 6 J, 6 LI'l
0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N
MWI MUS Mal TZA 1MB lWE
Countries and Periods
iii'.: Actual = Potential
Figure 5.24: Average actual and potential value of agricultural trade between South
Africa and the SADe countries for the periods 2000-2004 and 2005-2009
The results show that South Africa's agricultural traders (importers and exporters) operating
between South Africa and the SADe countries had outperformed and exhausted the estimated
potential capacity for the absorption of total agricultural trade in the South Africa-Malawi,
South Africa-Zambia and South Africa-Zimbabwe markets; but had under-scored or
underachieved in the South Africa-Mauritius, South Africa-Mozambique and South Africa-
146
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
Tanzania markets for the period 2000-2004. Over the period 2005-2009 South Africa's
agricultural traders (importers and exporters) operating between South Africa and the SADe
countries had outperformed and exhausted the estimated potential capacity for the absorption
of total agricultural trade in the South Africa-Mauritius, South Africa-Zambia and South
Africa-Zimbabwe markets; whereas they had under-scored or underachieved in the South
Africa-Malawi, South Africa-Mozambique and South Africa-Tanzania markets.
5.6.2 Cheese exports from South Africa to the SADC countries
This subsection provides the results of the impacts of the implementation of the SADe Trade
Protocol on South Africa's cheese exports to the SADe countries for the periods 2000-2004
and 2005-2009. South Africa did not import cheese from the SADe countries during these
periods. The final results for cheese exports from South Africa to the SADe countries are
presented in Table 5.23 below. The results show that the implementation of the SADe Trade
Protocol had no effects on South Africa's cheese exports to the SADe countries and show no
proof of creation of South Africa's cheese exports market in the SADe countries or diversion
of South Africa's cheese exports destined for the SADe market to other South African cheese
trading partners for either period.
Table 5.23: Results for cheese exports from South Africa to the SADC countries
Variables Period Impact Yearly Impact Export Direction
Model Model Model
2000 2005 2000 2005 2000 2005
2004 2009 2004 2009 2004 2009
Constant 21.36 21.36 -6.95 -6.95
InYiit-1 0.70* 0.70* 0.71 * 0.71 * 0.40* 0.40*
InGDPPCit -1.90 -1.90 1.51 1.51 -7.72 -7.72
InGDPPCit -0.01 -0.01 -0.01 -0.01 1.09 1.09
REERt 0.12 0.12 0.33 0.33 -1.88** -1.88**
00004 / 00509 -0.10 0.65
DOO/ DOS -0.12 0.37
001/006 -0.24 -0.24
002/007 -0.15 -0.07
003/ D08 0.51 -0.11
004/ D09 0.10 -0.31 -0.34 2.02
PTAves 0.33 1.84
PTA no'
InDISTï -0.41 ** -0.41 ** -0.41 ** -0.41 **
Adjusted R2 0.78 0.78 0.77 0.77 0.70 0.70
Observations 90 90 90 90 480 480
Cross-Sections 6 6 6 6 32 32
*, ** & *** denote significance at the 1,5and 10 rercent levels resrectively.
Source: Author's calculations
147
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
The results for the average actual and potential cheese exports from South Africa to the
SADe countries for the periods 2000-2004 and 2005-2009 are presented in Figure 5.25 in
log values, and the dollar values are presented in Appendix 5DM. The results show that
South Africa's cheese exporters had outperformed and exhausted the estimated potential
capacity for the absorption of South Africa's cheese exports in Zambia and Zimbabwe; but
had under-scored or underachieved in Malawi, Mauritius, Mozambique and Tanzania over the
period 2000-2004. In contrast, during the period 2005-2009 South Africa's cheese exporters
had outperformed and exhausted the estimated potential capacity for the absorption of South
Africa's cheese exports in Malawi, Mozambique and Tanzania; whereas they had under-
scored or underachieved inMauritius, Zambia and Zimbabwe
16.00
14.00
12.00
10.00
~
OIl
gbil 8.00
OIl
QI
:s
i>i
6.00
4.00
2.00
0.00
<t C'l <t C'l <t C'l <t C'l <t C'l <t C'l
0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N.n N N N N N NI I 6 I I I I0 LJ') LJ') 6 6 LJ') 0 LJ') 6 .n
0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N
MWI MUS MOl TlA 1MB lWE
Countries and Periods
~ Actual ;; Potential
Figure 5.25: Average actual and potential value of cheese exports from South Africa to
SADC countries for the periods 2000-2004 and 2005-2009
.5.6.3 Cut flowers trade flows between South Africa and the SADC countries
This subsection provides the results of the impacts of the implementation of the SADe Trade
Protocol on South Africa's cut flowers exports to the SADe countries; South Africa's cut
flowers imports from the SADe countries; as well as total cut flowers trade (import plus
148
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
exports) between South Africa and the SADC countries for the periods 2000-2004 and 2005-
2009. The results are presented in Table 5.24.
Table 5.24: Results for cut flowers between South Africa and the SADC countries
Models Variables Exports Imports Trade
2000 2005 2000 2005 2000 2005
2004 2009 2004 2009 2004 2009
Constant -0.53 -0.53
InY
- iit
•1 0.33* 0.33*
InGDPPCit 14.65*** 14.65*** 18.97 18.97
<.ol
e<I InGDPPCit 1.11 1.11 0.50 0.50
Q.._
E Q.l InGDPPCiit 4.64* 4.64*
_"0
"0 0 REERt -1.40 -1.40 0.87 0.87 -1.24*** -1.24***
.o;:~ D0004 / DOS09 -0.43 -1.88 0.09 -2.60 0.70 1.41 *
Q.l
Q., InDlSTii -3.03* -3.03*
Adjusted R2 0.64 0.64 0.72 0.72 0.44 0.44
Observations 96 96 90 90 96 96
Cross-Sections 6 6 6 6 6 6
Constant -2.36 -2.36
InYiit_1 0.30* 0.30*
InGDPPCit 11.16 11.16 69.18* 69.18*
InGDPPCit 1.03 1.03 1.04 1.04
InGDPPCiit 4.69* 4.69*
REERt -1.63 -1.63 3.86** 3.86** -0.80 -0.80
DOO / DOS 1.47** -1.35 -4.69*** -9.18** 0.51 1.15**
DOl / D06 -0.47 -1.56 -10.78* -12.11 * -1.50*** 1.67*
D02 / D07 -0.04 -1.12 -15.89* -14.39* -1.28 1.80*
D03 / D08 -0.62 -1.22 -8.64*** -16.06* 0.33 1.56**
D04/ D09 -2.42 -0.40 -18.39* -14.28* 0.21 1.76*
InDlSTij -3.15* -3.15*
Adjusted R2 0.63 0.63 0.73 0.73 0.63 0.63
Observations 96 96 90 90 90 90
Cross-Sections 6 6 6 6 6 6
Constant
InYiit-1 0.49* 0.49* 0.26* 0.26* 0.33* 0.33*
InGDPPCit -4.03 -4.03 4.11 4.11
InGDPPCit 0.56 0.56 0.71 0.71
InGDPPCiit 3.44*** 3.44***
REERt -0.60 -0.60 0.58 0.58 0.44 0.44
PTAv.s -0.85 1.38** 1.73*** 0.86 0.20 0.59
PTA no -0.93 * 1.14** 0.30 -!.59 -0.65*** -0.44
InDISTij
Adjusted R2 0.78 0.78 0.62 0.62 0.8! 0.8!
Observations 840 840 390 390 330 330
Cross-Sections 56 56 26 26 22 22
*. ** & ** * denote significance at the 1, 5 and 10 percent levels respectively
Source: Author's calculations
5.6.3.1 Cut flowers exports from South Africa to the SADC countries
The results show that the implementation of the SADe Trade Protocol had no joint effects on
South Africa's cut flowers exports to the SADC countries over both periods; but on an annual
149
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
basis, the average South African cut flowers exports to the SADe countries had significantly
increased by 1.47% in 2000. On average, 0.93% of South Africa's cut flowers exports
destined for the SADe market were diverted to other cut flowers trading partners of South
Africa for the period 2000-2004. Over the period 2005-2009, the implementation of the
SADe Trade Protocol led to the creation of 1.38% in South Africa's cut flowers exports
market in the SADe countries. The results for the average actual and potential cut flowers
exports from South Africa to the SADe countries for the periods 2000-2004 and 2005-2009
are presented in Figure 5.26 in log values, and the dollar values are presented in Appendix
SDP.
14.00
12.00
10.00
~
III 8.00
gl1li
III
41 6.00
:::I
;;
>
4.00
2.00
0.00
'<t O'l '<t O'l '<t O'l '<t O'l '<t O'l '<t O'l
0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N
I I
0 LIl 6 tA 6 tA 6 tA 6 tA 6 tA
0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N
MWI MUS MOl TZA 1MB lWE
Countries and Periods
~ Actual ;; Potential
Figure 5.26: Average actual and potential value of cut flowers exports from South Africa
to SADe countries for the periods 2000-2004 and 2005-2009
The results show that South Africa's cut flowers exporters had outperformed and exhausted
the estimated potential capacity for the absorption of South Africa's cut flowers exports in
Malawi, Mauritius and Mozambique; but had under-scored or underachieved in Tanzania,
Zambia and Zimbabwe over the period 2000-2004. Similarly, for the period 2005-2009,
South Africa's cut flowers exporters had also outperformed and exhausted the estimated
potential capacity for the absorption of South Africa's cut flowers exports in Malawi,
150
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
Mauritius and Mozambique; but had under-scored or underachieved in Tanzania, Zambia and
Zimbabwe.
5.6.3.2 Cut flowers imports from the SADe countries to South Africa
The results show that the implementation of the SADe Trade Protocol had no joint period
effects on the cut flowers imports from the SADC countries to South Africa for both periods.
Surprisingly, on an annual basis, the individual yearly negative effects were observed in all
years as follows: 4.69% in 2000, 10.78% in 2001, 15.89% in 2002, 8.64% in 2003, and
18.39% in 2004; and then 9.18% in 2005,12.11% in 2006,14.39% in 2007,16.06% in 2008,
and 14.28% in 2009.
The implementation of the SADe Trade Protocol during the period 2000-2004 led to the
creation of 1.73% in South Africa's cut flowers import market from the SADe countries.
However, for the period 2005-2009, there is no proof of creation in South Africa's cut
flowers import market from the SADe countries or diversion of South Africa's cut flowers
import market from the SADe countries to other South African cut flowers trading partners
owing to the implementation of the SADe Trade Protocol.
The results for the average actual and potential cut flowers imports from the SADC countries
to South Africa for the periods 2000-2004 and 2005-2009 are presented in Figure 5.27 in log
values, and the dollar values are presented in Appendix SDS. The results show that South
Africa's cut flowers importers had outperformed and exhausted the estimated potential
capacity for the absorption of South Africa's cut flowers imports from Malawi, Tanzania,
Zambia and Zimbabwe; but had under-scored or underachieved in Mauritius and
Mozambique over the period 2000-2004. For the period 2005-2009, South Africa's cut
flowers importers outperformed and exhausted the estimated potential capacity for the
absorption of South Africa's cut flowers imports from Malawi, Tanzania and Zambia;
whereas they under-scored or underachieved in Mauritius, Mozambique and Zimbabwe.
151
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
16.00
14.00 -
12.00
10.00
~
ga'".o 8.00
O'"l
:::J
'jij 6.00
>
4.00
2.00
0.00
<:t 0"1 <:t 0"1 <:t 0"1 <:t 0"1 <:t 0"1 <:t 0"1
0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0, 0 0N N, N N N N N N N N N, N,
0 VI 6 J.. 6 J.. 6 J.. 6 J.. 0 VI
0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N
MWI MUS MOl TZA 1MB lWE
Countries and Periods
fil: Actual = Potential
Figure 5.27: Average actual and potential value of cut flowers imports from the SADC
countries to South Africa for the periods 2000-2004 and 2005-2009
5.6.3.3 Cut flowers trade (imports plus exports) between South Africa and the SADC
countries
The results show that total cut flowers trade between South Africa and the SADe countries
had significantly increased by 1.41% during the implementation of the SADe Trade Protocol
over the period 2005-2009. On an annual basis, these significant increases occurred in all
years as follows: 1.15% in 2005, 1.67% in 2006, 1.8% in 2007, 1.56% in 2008, and 1.76% in
2009. However, there were no joint effects during the implementation of the SADe Trade
Protocol over the period 2000-2004; but there was a significant decrease of total cut flowers
trade between South Africa and the SADe countries that occurred in 2001 by 1.5%. On
average, the implementation of the SADe Trade Protocol led to the diversion of 0.65% of
total cut flowers trade between South Africa and the SADe countries to either South Africa's
other cut flowers trading partners or to the SADe countries' other cut flowers trading partners
for the period 2000-2004. However, there was no proof of creation or diversion of total cut
flowers trade between South Africa and the SADe countries to other cut flowers trading
partners of South Africa or of the SADe countries during the implementation of the SADe
152
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
Trade Protocol over the period 2005-2009.The results for the average actual and potential
total cut flowers trade between South Africa and the SADe countries for the periods 2000-
2004 and 2005-2009 are presented in Figure 5.28 in log values, and the dollar values are
presented in Appendix 5DV.
18.00
16.00
14.00
12.00
~
III 10.00
gIll)
III 8.00
QI
:::J
iV
> 6.00
4.00
2.00
0.00
.". C'1 .". (TI .". (TI .". (TI .". (TI .". (TI
0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N
6 J, I J, I I I I 6 I I I0 0 lil 0 lil lil 0 lil
0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N
MWI MUS Mal TlA 1MB lWE
Countries and Periods
~ Actual = Potential
Figure 5.28: Average actual and potential value of cut flowers trade between South
Africa and the SADe countries for the periods 2000-2004 and 2005-2009
The results show that South Africa's cut flowers traders (importers and exporters) operating
between South Africa and the SADe countries had outperformed and exhausted the estimated
potential capacity for the absorption of total cut flowers trade in the South Africa-Malawi,
South Africa-Mauritius, South Africa-Tanzania, South Africa-Zambia and South Africa-
Zimbabwe markets; but had under-scored or underachieved only in the South Africa-
Mozambique market for the period 2000-2004. For the period 2005-2009, they had
outperformed and exhausted the estimated potential capacity for the absorption of total cut
flowers trade in the South Africa-Tanzania, South Africa-Zambia and South Africa-
Zimbabwe markets; whereas they had under-scored or underachieved in the South Africa-
Mauritius and South Africa-Mozambique markets. However, South Africa's cut flowers
153
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
traders (importers and exporters) operating between South Africa and the SADe countries
had nearly matched the potential in the South Africa-Malawi market.
5.6.4 Frozen fruits and nuts exports from South Africa to the SADC countries
This subsection provides the results of the impacts of the implementation of the SADe Trade
Protocol on South Africa's frozen fruits and nuts exports to the SADe countries for the
periods 2000-2004 and 2005-2009. South Africa did not import frozen fruits and nuts from
SADe countries in either period. The results for frozen fruits and nuts exports from South
Africa to the SADe countries are presented in Table 5.25 below. The results show that the
implementation of the SADe Trade Protocol had no effects on South Africa's frozen fruits
and nuts exports to the SADe countries in either period. Furthermore, the results show no
proof of creation of South Africa's frozen fruits and nuts exports market in the SADe
countries or diversion of South Africa's frozen fruits and nuts exports destined for the SADe
market to other South African frozen fruits and nuts trading partners in either period.
Table 5.25: Results for frozen fruits and nuts exports from South Africa to the SADC
countries
Variables Period Im pact Yearly Impact Export Direction
Model Model Model
2000 2005 2000 2005 2000 2005
2004 2009 2004 2009 2004 2009
Constant
InYijt_1 0.38* 0.38* 0.37* 0.37* 0.40* 0.40*
InGDPPCit 1.68 1.68 10.60 10.60 3.00 3.00
InGDPPCjt -0.80 -0.80 -0.71 -0.71 0.32 0.32
REERt -1.28 -1.28 1.03 '-0.76 -1.86*** -1.86***
00004 / 00509 -1.89 1.62
DOO/ DOS 1.05 0.57
001/006 0.08 -0.37
002/007 4.57 -0.36
003/008 1.93 -0.80
004/009 3.06 -0.43
PTAyes 0.44 1.09
PTA no 0.70 0.10
InDISTï
Adjusted R2 0.70 0.70 0.69 0.69 0.57 0.57
Observations 90 90 90 90 315 315
Cross-Sections 6 6 6 6 21 21
*, ** & *** denote significance at the I, 5 and 10percent levels respectively. t-values are in parentheses
Source: Author's calculations
154
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
The results for the average actual and potential frozen fruits and nuts exports from South
Africa to the SADe countries for the periods 2000-2004 and 2005-2009 are presented in
Figure 5.29 in log values, and the dollar values are presented in Appendix 5DY. The results
show that South Africa's frozen fruits and nuts exporters had outperformed and exhausted the
estimated potential capacity for the absorption of South Africa's frozen fruits and nuts exports
in Mauritius, Zambia and Zimbabwe but had under-scored or underachieved in Malawi and
Mozambique and over the period 2000-2004. However, over the period 2005-2009 South
Africa's frozen fruits and nuts exporters had outperformed and exhausted the estimated
potential capacity for the absorption of South Africa's frozen fruits and nuts exports in
Malawi, Mozambique, Tanzania and Zambia whereas they had under-scored or
underachieved in Mauritius and Zimbabwe.
12.00
MWI MUS Mal TlA 1MB lWE
Countries and Periods
~ Actual == Potential
Figure 5.29: Average actual and potential value of frozen fruits and nuts exports from
South Africa to SADe countries for the periods 2000-2004 and 2005-2009
5.6.5 Preserved fruits and nuts exports from South Africa to the SADe countries
This subsection provides the results of the impacts of the implementation of the SADe Trade
Protocol on South Africa's preserved fruits and nuts exports to the SADe countries for the
periods 2000-2004 and 2005-2009. South Africa did not import preserved fruits and nuts
155
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
from the SADe countries in either period. The results for preserved fruits and nuts exports
from South Africa to the SADe countries are presented in Table 5.26 below. The results
show that the implementation of the SADe Trade Protocol had no joint effects on South
Africa's preserved fruits and nuts exports to the SADe countries in either period, but there
were significant positive increases of South Africa's preserved fruits and nuts exports to the
SADe countries which occurred on annual an basis as follows: 1.07% in 2001, 1.76% in
2002, 1.86% in 2003, and 1.91% in 2004.
Table 5.26: Results for preserved fruits and nuts exports from South Africa to the
SADe countries
Variables Period Impact Yearly Impact Export Direction
Model Model Model
2000 2005 2000 2005 2000 2005
2004 2009 2004 2009 2004 2009
Constant
InYiit-1 0.39* 0.39* 0.37* 0.37* 0.46* 0.46*
InGDPPCit 2.37 2.37 4.83 4.83 -4.66 -4.66
InGDPPCit -0.04 -0.04 -0.04 -0.04 0.83 0.83
REERt -0.94** -0.94** -0.79 -0.79 -0.63 -0.63
00004 / 00509 0.28 0.58
000·/ DOS 0.36 0.27
001/006 1.07*** 0.13
002/007 1.76** --0.23
003/008 1.86* 0.03
004/009 1.91** 0.22
PTAyes -0.03 1.81**
PTA no -0.41 0.79
InDISTj'
Adjusted 2R 0.86 0.86 0.85 0.85 0.68 0.68
Observations 90 90 90 90 1275 1275
Cross-Sections 6 6 6 6 85 85
*, ** & *** denote significance at the 1,5and 10 Eercent levels resEectively.
Source: Author's calculations
On average, the implementation of the SADe Trade Protocol during the period 2005-2009
led to the creation of 1.81% in South Africa's preserved fruits and nuts exports market in the
SADe countries. However, the results show no proof of creation of South Africa's preserved
fruits and nuts exports market in the SADe countries or diversion of South Africa's preserved
fruits and nuts exports destined for the SADe market to other South African preserved fruits
and nuts trading partners over the period 2000-2004.
156
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
The results for the average actual and potential preserved fruits and nuts exports from South
Africa to the SADe countries for the periods 2000-2004 and 2005-2009 are presented in
Figure 5.30 in log values, and the dollar values are presented in Appendix 5EB. The results
show that South Africa's preserved fruits and nuts exporters had outperformed and exhausted
the estimated potential capacity for the absorption of South Africa's preserved fruits and nuts
exports in Malawi and Zambia but had under-scored or underachieved in Mauritius,
Mozambique, Tanzania and Zimbabwe over the period 2000-2004. During the period 2005-
2009, South Africa's preserved fruits and nuts exporters outperformed and exhausted the
estimated potential capacity for the absorption of South Africa's preserved fruits and nuts
exports in Malawi, Tanzania, Zambia and Zimbabwe whereas they under-scored or
underachieved inMauritius and Mozambique.
16.00
14.00
12.00
~ 10.00
III
bG
-.E! 8.00III
III
:::I
i>ii
6.00
4.00
2.00
0.00
<:t 0"1 <:t 0"1 <:t 0"1 <:t 0"1 <:t 0"1 <:t 0"1
0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N
0 I I I I ILf1 0 J, 0 Lf1 0 Lf1 0 J, 0 Lf1
0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N
MWI MUS MOZ TZA 2MB ZWE
Countries and Periods
~Actual = Potential
Figure 5.30: Average actual and potential value of preserved fruits and nuts exports
from South Africa to SADC countries for the periods 2000-2004 and 2005-2009
5.6.6 Fruits and vegetable juices trade flows between South Africa and the SADC
countries
This subsection provides the results of the impacts of the implementation of the SADe Trade
Protocol on South Africa's fruits and vegetable juices exports to the SADe countries; South
157
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
Africa's fruits and vegetable juices imports from the SADC countries; as well as total fruits
and vegetable juices trade (import plus exports) between South Africa and the SADC
countries for the periods 2000-2004 and 2005-2009. The results are presented in Table 5.27.
Table 5.27: Results for fruits and vegetable juices trade flows between South Africa and
the SADe countries
Models Variables Exports Imports Trade
2000 2005 2000 2005 2000 2005
2004 2009 2004 2009 2004 2009
Constant -16.60 -16.60 -190.8*** -190.8*** 2.08 2.08
InYiit-1 0.72* 0.72* 0.75* 0.75*
InGDPPCit 2.69 2.69 24.79*** 24.79***
InGDPPCit 0.13*** 0.13*** 1.40* 1.40*
InGDPPCiit 0.43 0.43
REERt -0.01 -0.01 0.09 0.09 -0.13 -0.13
D0004 / D0509 0.15 0.15 -3.37 -3.37 0.00 0.61 *
InDISTii -0.23 ** -0.23** -2.0 1** -2.01** -0.21 *** -0.21 ***
Adjusted R2 0.91 0.91 0.15 0.15 0.91 0.91
Observations 90 90 90 90 90 90
Cross-Sections 6 6 6 6 6 6
Constant -68.93** -68.93** -434.3** -434.3** 2.07 2.07
InYiit-1 0.72* 0.72* 0.75* 0.75*
InGDPPCit 9.01 * 9.01 * 54.99** 54.99**
InGDPPCit 0.13*** 0.13*** 1.56* 1.56*
InGDPPCiit 0.42 0.42
REERt 0.38 0.38 1.14 1.14 -0.12 -0.12
DOO/ DOS -0.15 -0.58 -1.31 -5.98 -0.03 0.76*
DO] / D06 -0.64*** -1.28*** -4.46 -8.70*** -0.02 0.35***
D02/ D07 -0.53 -1.35*** -1.02 -10.17*** 0.79*** 0.66*
D03/ D08 -0.59 -1.57*** -4.19 -9.72 0.43 0.59*
D04/ D09 -0.95 -1.18 -1.03 -13.23** 0.43 0.70*
InDlSTii -0.23** -0.23 ** -2.50*** -2.50*** -0.20 -0.20
Adjusted R2 0.92 0.92 0:19 0.19 0.91 0.91
Observations 90 90 90 90 90 90
Cross-Sections 6 6 6 6 6 6
Constant
InYiit_1 0.27* 0.27* 0.24* 0.24* 0.20* 0.20*
InGDPPCit 8.29** 8.29** 13.86** 13.86**
InGDPPCit 2.17* 2.17* 1.32 1.32
InGDPPCiit 4.29* 4.29*
REERt -0.33 -0.33 -0.39 -0.39 -0.09 -0.09
PTAves 0.05 -0.64 0.22 -1.87 0.33 0.62
PTA no 0.33 -0.54 -0.87 -0.83 0.57*** 0.39
InDISTii
Adjusted R2 0.65 0.65 0.62 0.62 0.68 0.68
Observations 1290 1290 645 645 630 630
Cross-Sections 86 86 43 43 42 42
*, ** & *** denote significance at the 1,5 and 10 percent levels respectively
Source: Author's calculations
158
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
5.6.6.1 Fruits and vegetable juices exports from South Africa to the SADC countries
The results show that the implementation of the SADe Trade Protocol had no joint effects on
South Africa's fruits and vegetable juices exports to the SADe countries in either period, but
there were significant decreases in South Africa's fruits and vegetable juices exports to the
SADe countries that occurred on an annual basis as follows: 0.64% in 2001, 1.28% in 2006,
1.35% in 2007, and 1.57% in 2008. The results show that there was no proof of creation of
South Africa's fruits and vegetable juices exports market in the SADe countries or diversion
of South Africa's fruits and vegetable juices exports destined for the SADe market to other
South African fruits and vegetable juices trading partners in either period. The results for the
average actual and potential fruits and vegetable juices exports from South Africa to the
SADe countries for the periods 2000-2004 and 2005-2009 are presented in Figure 5.31 in
log values, and the dollar values are presented in Appendix SEE.
18.00
16.00 ~ ~ ~ ~
14.00 ~~
12.00 ~V~~J~~
~~d~~~~~~~d ~
~
III 10.00
g110 -I~~ ~~dHV~~~d~~E~-
III 8.00
III
:;,
;;
> 6.00
4.00 V~~ ~~~~~~~~~d
2.00 ~~~~~~~~~~~
~~~~~~~~~~~~
0.00
'<t 0\ '<t 0\ '<t 0\ '<t 0\ '<t 0\ '<t 0\
0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N
6 IIII 6 tA 6 I I I IIII 0 III 0 tA 6 tA
0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N
MWI MUS MOl TlA 1MB lWE
Countries and Periods
~Actual :: Potential
Figure 5.31: Average actual and potential value of fruits and vegetable juices exports
from South Africa to SADC countries for the periods 2000-2004 and 2005-2009
The results show that South Africa's fruits and vegetable juices exporters had outperformed
and exhausted the estimated potential capacity for the absorption of South Africa's fruits and
vegetable juices exports in Mauritius and Zambia but had under-scored or underachieved in
159
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
Malawi, Mozambique and Zimbabwe over the period 2000-2004. During the period 2005-
2009 South Africa's fruits and vegetable juices exporters outperformed and exhausted the
estimated potential capacity for the absorption of South Africa's fruits and vegetable juices
exports in Tanzania and Zimbabwe whereas they under-scored or underachieved in Malawi,
Mauritius, Mozambique and Zambia.
5.6.6.2 Fruits and vegetable juices imports from the SADe countries to South Africa
The results show that the implementation of the SADe Trade Protocol had no joint period
effects on the fruits and vegetable juices imports from the SADe countries to South Africa in
either period. However, on an annual basis, the individual annual negative effects were
observed in 2006, 2007 and 2009, when the fruits and vegetable juices imports from the
SADe countries to South Africa decreased by 8.7%, 10.17% and 13.23% respectively during
those years. The results also indicate there was no proof of creation of South Africa's fruits
and vegetable juices import market from the SADe countries or diversion of South Africa's
fruits and vegetable juices import market from the SADe countries to other South African
fruits and vegetable juices trading partners owing to the implementation of the SADe Trade
Protocol.
The results for the average actual and potential fruits and vegetable juices imports from the
. -
SADe countries to South Africa for the periods 2000-2004 and 2005-2009 are presented in
Figure 5.32 above in log values, and the dollar values are presented in Appendix 5EH. The
results show that South Africa's fruits and vegetable juices importers had outperformed and
exhausted the estimated potential capacity for the absorption of South Africa's fruits and
vegetable juices imports from Mauritius and Zambia; but had under-scored or underachieved
in Mozambique and Zimbabwe over the period 2000-2004. During the period 2005-2009,
South Africa's fruits and vegetable juices importers outperformed and exhausted the
estimated potential capacity for the absorption of South Africa's fruits and vegetable juices
imports from Mauritius and Zimbabwe; whereas they under-scored or underachieved in
Malawi, Mozambique and Tanzania.
160
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
14.00
12.00
10.00
~- 8.00
(II)
j.2 I6.00 1 ---------
4.00
2.00 r- =
I ~
0.00 T gl~ <:T 0'1 <:T 0'1 <:T 0'1
o 0 oo oo 0 oo oo o0 o
N N N N N N N
6o J0o, 6 Jo, 6o o J
<;J
LI)
o 0 o o 0 o
.. 6
0 o oo oo
N N N N N N N N
MWI MUS Mal TlA 1MB lWE
Countrie and Periods
~ Actual ~ Potential
Figure 5.32: Average actual and potential value of fruits and vegetable juices imports
from the SADC countries to South Africa for the periods 2000-2004 and 2005-2009
5.6.6.3 Fruits and vegetable juices trade (imports plus exports) between South Africa
and the SADC countries
The results show that the total fruits and vegetable juices trade between South Africa and the
SADe countries had significantly increased by 0.61% during the implementation of the
SADe Trade Protocol over the period 2005-2009 and on annual basis this significant increase
occurred in all years as follows: 0.76% in 2005, 0.35% in 2006, 0.66% in 2007, 0.59% in
2008, and 0.7% in 2009. However, the implementation of the SADe Trade Protocol had no
joint effects on total fruits and vegetable juices trade between South Africa and the SADe
countries over the period 2004-2004, whereas a significant increase of 0.79% occurred in
2002. However, the results show no proof of creation or diversion of total fruits and
vegetable juices trade between South Africa and the SADe countries to other fruits and
vegetable juices trading partners of South Africa or SADe countries during the period 2005-
2009.
161
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
The results for the average actual and potential total fruits and vegetable juices trade between
South Africa and the SADe countries for the periods 2000-2004 and 2005-2009 are
presented in Figure 5.33 in log values, and the dollar values are presented in Appendix 5EK.
The results show that South Africa's fruits and vegetable juices traders (importers and
exporters) operating between South Africa and the SADe countries had under-scored or
underachieved the estimated potential capacity for the absorption of total fruits and vegetable
juices trade in all markets; i.e. the South Africa-Mauritius, South Africa-Mozambique, South
Africa-Zambia and South Africa-Zimbabwe markets, meaning that no market was exhausted
over the period 2000-2004. Similarly, no market was exhausted during the period 2005-
2009, as South Africa's fruits and vegetable juices traders (importers and exporters) operating
between South Africa and the SADe countries had also under-scored or underachieved in all
markets, i.e. the South Africa-Malawi, South Africa-Mauritius, South Africa-Mozambique,
South Africa-Tanzania and South Africa-Zimbabwe markets.
18.00
16.00
14.00
12.00
~
t'i"l) 10.00g
Q'"I 8.00
:::l
iii
> 6.00
4.00
2.00
0.00
'<t Ol '<t Ol '<t Ol '<t Ol '<t Ol '<t Ol
0 0 0 0 0 0 0 0 0 0 o o
0 0 0 0 0 0 0 0 0 0 o o
N N N N N N N N N N N N
0 J, 0 J, I0 J, 0 J, 0 J, o II/)
0 0 0 0 0 0 0 0 0 0 o o
0 0 0 0 0 0 0 0 0 0 o o
N N N N N N N N N N N N
MWI MUS MOl TZA 1MB lWE
Countries and Periods
il'! Actual = Potential
Figure 5.33: Average actual and potential value of fruits and vegetable juices trade
between South Africa and the SADe countries for the periods 2000-2004 and 2005-2009
162
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
5.6.7 Wine trade flows between South Africa and the SADC countries
This subsection provides the results of the impacts of the implementation of the SADe Trade
Protocol on South Africa's wine exports to the SADe countries; South Africa's wine imports
from the SADe countries; as well as total wine trade between South Africa and SADe
countries for the periods 2000-2004 and 2005-2009. The results are presented in Table 5.28.
Table 5.28: Results for wine trade flows between South Africa and the SADC countries
Models Variables Exports Imports Trade
2000 2005 2000 2005 2000 2005
2004 2009 2004 2009 2004 2009
Constant -185.70*** -185.70***
InYiit-1 0.38* 0.38* 0.36* 0.36*
- InGDPPCit 1.38 1.38 21.58*** 21.58***~<J InGDPPCit 0.32*** 0.32*** -0.26 -0.26Q._
E a. InGDPPCiit 2.78** 2.78**
_"0
"0 0 REERt -0.04 -0.04 3.50 3.50 0.09 0.09
o:;E
'c 00004 / 00509 -0.16 0.58*** -0.41 -4.06 0.13 0.37a.
Q., InDlSTii 0.40 0.40
Adjusted R2 0.53 0.88 0.58 0.58 0.88 0.88
Observations 60 90 90 90 90 90
Cross-Sections 6 6 6 6 6 6
Constant -72.56* -72.56* -114.92 -114.92
InYijt_1 0.55* 0.55* 0.47* 0.47*
InGDPPCit 9.72* 9.72* 13.87 13.87
InGDPPCit 0.25** 0.25** 0.03 0.03
InGDPPCiit 3.85* 3.85*
REERt 0.50** 0.50** 3.04 3.04 0.15 0.15
DOO / DOS -0.15 -0.44 -0.83 -3.74 0.20 0.60**
001/006 -0.36*** -1.49* 1.83 -0.76 -0.13 -0.22
002/007 -0.26 -1.62** 1.56 -2.31 0.07 -0.07
003/008 -0.27 -1.86** 1.88 -3.27 0.08 -0.17
004/009 -0.60 -1.44** -0.49 -2.24 0.16 -0.03
InDISTij -0.37** -0.37** -0.77 -0.77
Adjusted R2 0.57 0.84 0.58 0.58 0.90 0.90
Observations 60 90 90 90 90 90
Cross-Sections 6 6 6 6 6 6
Constant
InYiit-1 0.30* 0.30* 0.13* 0.13* 0.50* 0.50*
InGOPPCit 3.58 3.58 4.40 4.40
InGOPPCit 0.32 0.32 1.62** 1.62**
InGOPPCiit 3.95* 3.95*
REERt -1.19* -1.19* 1.95* 1.95* 0.39*** 0.39***
PTAv.s -1.20*** 0.20 -2.31 * 0.53 -0.15 -0.04
PTA no -0.73** 1.05** -1.45* 0.93** -0.47*** 0.39**
InDISTij
Adjusted R2 0.74 0.74 0.75 0.75 0.87 0.87
Observations 1530 1530 675 675 660 660
Cross-Sections 102 102 45 45 44 44
*, ** & *** denote significance at the 1,5 and 10percent levels respectively
Source: Author's calculations
163
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
5.6.7.1 Wine exports from South Africa to the SADe countries
The results show that wine exports from South Africa to the SADe countries had significantly
increased by 0.58% during the implementation of the SADe Trade Protocol over the period
2005-2009. Surprising, on an annual basis South Africa's wine exports to the SADe
countries had significantly declined by 1.49% in 2006, 1.62% in 2007, 1.86% in 2008, and
1.44% in 2009. However, the implementation of the SADe Trade Protocol had no joint
effects on South Africa's wine exports to the SADe countries over the period 2004-2004, but
there was a significant decrease ofO.36% that occurred in 2001. On average, 0.73% of South
Africa's wine exports destined for the SADe market were diverted to other wine trading
partners of South Africa over the period 2000-2004. The results show no proof of creation in
South Africa's wine exports market in the SADe countries or diversion of South Africa's
wine exports destined for the SADe market to other South African wine trading partners
during the period 2005-2009.
The results for the average actual and potential wine exports from South Africa to the SADe
countries for the periods 2000-2004 and 2005-2009 are presented in Figure 5.34 in log
values, and the dollar values are presented in Appendix SEN. The results show that South
Africa's wine exporters had outperformed and exhausted the estimated potential capacity for
the absorption of South Africa's wine exports in Mauritius, Mozambique and Tanzania; but
had under-scored or underachieved in Malawi, Zambia and Zimbabwe over the period 2000-
2004. Similarly, over the period 2005-2009 South Africa's wine exporters outperformed and
exhausted the estimated potential capacity for the absorption of South Africa's wine exports
in Mauritius, Tanzania, and Zambia, whereas they under-scored or underachieved in Malawi,
Mozambique and Zimbabwe.
164
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
18.00
16.00
14.00
12.00
~
III 10.00
g110
III 8.00
~QI
ru
> 6.00
4.00
2.00
0.00
'<t 0'1 '<t 0'1 '<t 0'1 '<t 0'1 '<t 0'1 '<t 0'1
0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N
0 ILIl 6 LA I I 6 I I0 LIl LIl 6 LA 6 LIl
0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N
MWI MUS MOl TZA 1MB lWE
Countries and Periods
~Actual = Potential
Figure 5.34: Average actual and potential value of wine exports from South Africa to
SADC countries for the periods 2000-2004 and 2005-2009
5.6.7.2 Wine imports from the SADC countries to South Africa
The results show that the implementation of the SADe Trade Protocol had no effects on the
wine imports from the SADe countries to South Africa in either period. The implementation
of the SADe Trade Protocol during the period 2000-2004 had led to the diversion 1.45% of
South Africa's wine import market from the SADe countries to other South African wine
trading partners. However, for the period 2005-2009 the results show no proof of creation of
South Africa's wine import market from the SADe countries or diversion of South Africa's
wine import market from the SADe countries to other South African wine trading partners
owing to the implementation of the SADe Trade Protocol.
The results for the average actual and potential wine imports from the SADe countries to
South Africa for the periods 2000-2004 and 2005-2009 are presented in Figure 5.35 above in
log values, and the dollar values are presented in Appendix SEQ. The results show that
South Africa's wine importers had outperformed and exhausted the estimated potential
165
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
capacity for the absorption of South Africa's wine imports from Mauritius and Zimbabwe; but
had under-scored or underachieved South Africa's wine imports from Malawi and Zambia
over the period 2000-2004. For the period 2005-2009, South Africa's wine importers
outperformed and exhausted the estimated potential capacity for the absorption of South
Africa's wine imports from Malawi and Mauritius but they under-scored or underachieved
South Africa's wine imports from Mozambique, Tanzania and Zimbabwe.
7.00 -,--------------------------
~ 3.00
::l
~
2.00
1.00
'<t 0'1 '<t 0'1 '<t 0'1 '<t 0'1 0'1 '<t 0'1
0 0 0 0 0 0 0 0 o o o
0 0 0 0 0 0 0 0 o o o
N N N N N N N N N N N
6 I I I IIJ') 6 IJ') 6 JJ 6 IJ') JJ 6 IJ')
0 0 0 0 0 0 0 0 o o o
0 0 0 0 0 0 0 0 o o o
N N N N N N N N N N N
MWI MUS Mal TZA 1MB lWE
Countries and Periods
~Actual = Potential
Figure 5.35: Average actual and potential value of wine imports from the SADC
countries to South Africa for the periods 2000-2004 and 2005-2009
5.6.7.3 Wine trade (imports plus exports) between South Africa and the SADC countries
The results show that the implementation of the SADe Trade Protocol had no joint period
effects on total wine trade between South Africa and the SADe countries in either period, but
a significant increase of total wine trade between South Africa and the SADe countries was
observed in 2005 by 0.6%. On average, the implementation of the SADe Trade Protocol led
to the diversion of 0.65% oftotal wine trade between South Africa and the SADe countries to
either South Africa's other wine trading partners or to the SADe countries' other wine trading
partners over the period 2000-2004. The results for the average actual and potential total
166
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
wine trade between South Africa and the SADe countries for the periods 2000-2004 and
2005-2009 are presented in Figure 5.36 in log values, and the dollar values are presented in
Appendix SET.
18.00
16.00
14.00
12.00
~
III
till 10.00.-sIII 8.00
CII
:::J
iV
> 6.00
4.00
2.00
0.00
<t 0\ <t 0\ <t 0\ <t 0\ <t 0\ <t 0\
0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N
I
0 JJ 6 I I I ILIl 0 LIl 6 J, 6 LIl 6 JJ
0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0
N N N N N N N N N N N N
MWI MUS MOl TZA 1MB lWE
Countries and Periods
~Actual == Potential
Figure 5.36: Average actual and potential value of wine trade between South Africa and
the SADe countries for the periods 2000-2004 and 2005-2009
The results show that South Africa's wine traders (importers and exporters) operating
between South Africa and the SADe countries had outperformed and exhausted the estimated
potential capacity for the absorption of total wine trade in the South Africa-Malawi market;
but under-scored or underachieved in the South Africa-Mauritius, South Africa-Zambia and
South Africa-Zimbabwe markets over the period 2000-2004. For the period 2005-2009,
South Africa's wine traders (importers and exporters) operating between South Africa and the
SADe countries had outperformed and exhausted the estimated potential capacity for the
absorption of total wine trade in the South Africa-Mauritius and South Africa-Tanzania
markets; whereas they had under-scored or underachieved in the South Africa-Malawi, South
Africa-Mozambique and South Africa-Zimbabwe markets.
167
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
5.7 The response of selected agricultural trade flows between South Africa and ROW
countries to the implementation EU-SA TDCA and SADC Trade Protocol
This section discusses various responses of total agricultural trade flows and the selected
agricultural products trade flows between South Africa and the ROW countries (non-EU and
non-SADC countries) for the periods 2000-2004 and 2005-2009. During these periods, the
EU-SA TDCA and the SADC Trade Protocol were implemented and therefore this section
reports on how South Africa's agricultural trade flows with non-EU and non-SADC countries
performed, given that during these periods South Africa's agricultural trade flows were given
preferential access in the EU and SADC markets through tariff phase-down schedules and in-
quota tariff preferences. Therefore, this section will examine the datasets of all the trade
flows (i.e. exports, imports and total trade) of total agriculture and all the selected agricultural
products between South Africa and the ROW countries.
5.7.1 Aggregate agricultural trade flows between South Africa and the ROW countries
The results for all agricultural trade flows (exports, imports and trade) between South Africa
and the ROW countries are presented in Table 5.29 below. The results show that, during the
period 2000-2004, South Africa's total agricultural exports to the ROW countries had
significantly declined by 0.48%, and on an annual basis significant decreases occurred in all
years as follows: 0.8% in 2000, 0.85% in 2001, 1.49% in 2002,0.81 % in 2003, and 1.28% in
2004. However, during the period 2005-2009, South Africa's total agricultural exports to the
ROW countries had significantly increased by 0.59%.
Surprisingly, on an annual basis the results provide a contrasting picture as there were
significant declines in South Africa's total agricultural exports to the ROW countries that
occurred in 2006, 2007, 2008 and 2009 by 1.07%, 1.3%, 1.32% and 1.18% respectively.
There were no joint period effects on South Africa's total agricultural imports from the ROW
countries in either period, but there were significant individual yearly effects for the period
2005-2009 that were observed in all years as follows: 2.02% in 2005, 3.16% in 2006,3.62%
in 2007, 4.04% in 2008, and 3.27% in 2009.
168
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
Table 5.29: Results for agricultural trade flows between South Africa and the ROW
countries
Variables Period Impact Model Yearly Impact Model
Exports Imports Trade Exports Imports Trade
2000 2005 2000 2005 2000 2005 2000 2005 2000 2005 2000 2005
2004 2009 2004 2009 2004 2009 2004 2009 2004 2009 2004 2009
Constant
InYiit.1 0.38* 0.38* 0.23* 0.23* 0.36* 0.36* 0.38* 0.38* 0.22* 0.22* 0.37* 0.37*
InCOPPCit 2.20 2.20 5.73** 5.73** 9.85* 9.85* 19.54* 19.54*
InCOPPCit 0.02 0.02 0.81 * 0.81 * 0.02 0.02 0.81 * 0.81 *
InCOPPCiit 0.45* 0.45* 0.46* 0.46*
REERt -0.59* -0.59* -0.37 -0.37 -0.54* -0.54* -0.12 -0.12 0.45 0.45 -0.52* -0.52*
00004 /00509 -0.48** 0.59** 0.32 -0.31 0.25** 0.95*
DOO / DOS -0.80* -0.29 0.29 -2.02** 0.09 1.10*
001/006 -0.85* -1.07** 0.24 -3.16* 0.13 0.78*
002/007 -1.49* -1.30** 0.23 -3.62* 0.30 0.82*
003/008 - -0.81 ** -1.32** 0.62 -4.04* 0.47** 1.04*
004/009 -1.28* -1.18** 0.42 -3.27* 0.55* 1.00*
InOISTï
Adjusted R2 0.82 0.82 0.72 0.72 0.86 0.86 0.82 0.82 0.72 0.72 0.86 0.86
Observations 1470 1470 1590 1590 1320 1320 1470 1470 1590 1590 1320 1320
Cross-Sections 98 98 106 106 88 88 98 98 106 106 88 88
*, ** & * ** denote significance at the I, 5 and IOpercent levels respectively.
Source: Author's calculations
The results also show that during the period 2000-2004 the total agricultural trade between
South Africa and the ROW countries had significantly improved by 0.25% and on an annual
basis significant increases occurred in 2003 and 2004 by 0.47% and 0.55% respectively.
Similarly, for the period 2005-2009, the total agricultural trade between South Africa and the
ROW countries had significantly improved by 0.95%, and on an annual basis significant
increases occurred in all years as follows: 1.1% in 2005, 0.78% in 2006, 0.82% in 2007,
1.04% in 2008, and 1% in 2009.
5.7.2 Cheese trade flows between South Africa and the ROW countries
The results for all cheese trade flows (exports and imports) between South Africa and the
ROW are presented in Table 5.30 below. The results showed that .during the period 2005-
2009 South Africa's cheese exports to the ROW countries had significantly increased by
3.41%. There were no joint period effects on South Africa's cheese imports from the ROW
countries in either period, but there were significant negative individual annual effects over
these periods. For example, cheese imports from the ROW countries had declined by 6.72%
169
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
and 8.64% in 2002 and 2004 respectively as well as by 8.1% and 12.97% in 2005 and 2009
respecti vely.
Table 5.30: Results for cheese trade flows between South Africa and the ROW countries
;
ariables Period Impact Model Yearly Impact Model
Exports Imports Trade Exports Imports Trade
2000 2005 2000 2005 2000 2005 2000 2005 2000 2005 2000 2005
- - - - - - - - - - - -
2004 2009 2004 2009 2004 2009 2004 2009 2004 2009 2004 2009
onstant - - - - - - - - - - - -
Yiit-I 0.41 * 0.41 * 0.31* 0.31 * - - 0.40* 0.40* 0.33* 0.33* - -
GDPPCit -15.63** -15.63** 20.75 20.75 - - -12.38 -12.38 48.33*** 48.33*** - -
GDPPCit 2.63 2.63 -1.32 -1.32 - - 2.75*** 2.75*** -1.79 -1.79 - -
GDPPCiit - - - - - - - - - - - -
lEERt -3.27* -3.27* 4.08** 4.08** - - -3.10* -3.10* 5.49** 5.49** - -
P004 / D0509 -1.12 3.41 ** -1.01 -4.24 - - - - - - - -
(.lo / D05 - - - - - - -1.34 2.81 -2.84 -8.10*** - -
pt / D06 - . - - - - - -0.84 3.27 -4.17 -8.35 - -
02/007 - - - - - - -1.64 2.64 -6.72*** -9.26 - -
03/008 - - - - - - -1.52 2.67 -2.00 -11.36 - -
04/ D09 - - - - - - -1.92 2.15 -8.64** -12.97** - -
DISTï - - - - - - - - - - - -
~justed R2 0.59 0.59 0.60 0.60 - - 0.59 0.59 0.61 0.61 - -
bservations 300 300 135 135 - - 300 300 135 135 - -
oss-Sections 20 20 9 9 - - . 20 20 9 9 - -
** & * ** denote significance at the 1, 5 and 10percent levels respectively.
Source: Author's calculations
5.7.3 Cut flowers trade flows between South Africa and the ROW countries
The results for all cut flowers trade flows (exports, imports and trade) between South Africa
and the ROW are presented in Table 5.31 below. The results show that during the period
2000-2004 South Africa's total cut flower exports to the ROW countries had significantly
declined by 1.14%, and on an annual basis significant decreases occurred in all years as
follows: 1.36% in 2000,2.385% in 2001, 2.58% in 2002, 1.97% in 2003, and 2.34% in 2004.
However, during the period 2005-2009 South Africa's cut flowers exports to the ROW
countries had significantly increased by 1.96% but individual yearly effects were
insignificant. There were no joint period effects on cut flowers imports and total cut flowers
trade between South Africa and the ROW countries in either period, but there were significant
negative individual annual effects observed in certain years in both periods. For example,
South Africa's cut flowers imports from the ROW countries declined by 8.38% in 2007,
170
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
whereas total cut flowers trade between South Africa and the ROW countries declined by
1.41%, 1.91% and 2.67% in 2000,2001 and 2002 respectively.
Table S.31: Results for cut flowers trade flows between South Africa and the ROW
countries
Variables Period Impact Model Yearly Impact Model
Exports Imports Trade Exports Imports Trade
2000 2005 2000 2005 2000 2005 2000 2005 2000 2005 2000 2005
- - - - - - - - - - - -
2004 2009 2004 2009 2004 2009 2004 2009 2004 2009 2004 2009
I--0nstant - - - - - - - - - - - -
nYjito• 0.51• 0.51 • 0.19" 0.19" 0.45' 0.45' 0.50' 0.50' 0.22* 0.22' 0.45' 0.45'
nGOPPCjt -7.18*' -7.18" -11.68 -11.68 - - -2.01 -2.01 21.24 21.24 - -
nGOPPCit 0.72 0.72 4.04*** 4.04'" - - 0.78 0.78 3.42 3.42 - -
nGOPPCjit - - - - 5.16*** 5.16**' - - - - 7.90· 7.90*
EERt -1.09** -1.09** 0.14 0.14 0.99 0.99 -0.83 -0.83 2.11 2.11 1.23*" 1.23*"
)0004 / D0509 -1.14' 1.96* -0.14 0.80 -1.05 -0.75 - - - - - -
)00/005 - - - - - - -1.36* 1.24 -0.28 -4.03 -1.41'" -0.35
~01 /006 - - - - - - -2.38' 1.14 0.52 -4.03 -1.91*** -1.06
p02 / 007 - - - - - - -2.58** 1.10 -0.72 -8.38**' -2.67*** -2.06*"
P03/008 - - - - - - -1.97*" 0.55 -0.21 -7.19 -1.38 -1.43
P04/009 - - - - - - -2.34" 0.30 2.10 -6.29 -0.98 -2.34'
nDlSTjj - - - - - - - - - - - -
djusted R2 0.75 0.75 0.52 0.52 0.73 0.73 0.75 0.75 0.53 0.53 0.74 0.74
)bservations 540 540 195 195 135 135 540 540 195 195 135 135
ross-Sections 36 36 13 13 9 9 36 36 13 13 9 9
, ** & *** denote significance at the 1,5 and 10percent levels respectively.
Source: Author's calculations
S.7.4 Frozen fruits and nuts trade flows between South Africa and the ROW countries
The results for all frozen fruits and nuts trade flows (exports and imports) between South
Africa and the ROW are presented in Table 5.32 below. The results show that there were no
joint period effects or individual yearly effects on all frozen fruits and nuts trade flows
between South Africa and the ROW countries in either period.
171
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
Table 5.32: Results for frozen fruits and nuts trade flows between South Africa and the
ROW countries
Variables Period Impact Model Yearly Impact Model
Exports Imports Trade Exports Imports Trade
2000 2005 2000 2005 2000 2005 2000 2005 2000 2005 2000 2005
2004 2009 2004 2009 2004 2009 2004 2009 2004 2009 2004 2009
Constant
InYiit_1 0.43* 0.43* 0.24* 0.24* 0.43* 0.43* 0.25* 0.25*
InGOPPCit 7.16 7.16 -13.89 -13.89 2.57 2.57 -39.46 -39.46
InGOPPCit 2.75 2.75 8.45* 8.45* 2.73 2.73 8.66* 8.66*
InGOPPCiit
REERt -1.58 -1.58 1.35 1.35 -1.93 -1.93 0.07 0.07
D0004 /00509 1.09 -0.70 0.35 0.54
DOO /DOS 0.98 -0.53 0.66 5.59
DOl /006 3.37 0.49 -1.22 2.24
D02 /007 5.02 0.48 -0.16 7.31
D03 /008 2.10 -0.47 -3.60 8.33
D04 /009 -1.77 1.38 -1.73 6.01
InDlSTï
Adjusted R2 0.53 0.53 0.36 0.36 0.52 0.52 0.38 0.38
Observations 135 135 210 210 135 135 210 210
Cross-Sections 9 9 14 14 9 9 14 14
*, ** & * ** denote significance at the 1, 5 and 10percent levels respectively.
Source: Author's calculations
5.7.5 Preserved fruits and nuts trade flows between South Africa and the ROW
countries
The results for all preserved fruits and nuts trade flows (exports, imports and trade) between
South Africa and the ROW are presented in Table 5.33 below. The results show that the total
preserved fruits and nuts trade between South Africa and the ROW countries had significantly
increased by 0.69% during the period 2000-2004 and by 0.75% during the period 2005-2009.
Furthermore, on an annual basis the significant increases occurred in 2004 by 1.33%, in 2005
by 0.93%, in 2006 by 0.74%, in 2007 by 0.67% and in 2008 by 0.89%.
However, there were no joint period effects on either preserved fruits and nuts exports or
imports between South Africa and the ROW countries for both periods, but there were
significant positive and negative individual yearly effects that occurred in both periods. For
example, in 2007 and 2009 South Africa's preserved fruits and nuts exports to the ROW
countries increased by 3.73% and 3.5% respectively. On the other hand, South Africa's
preserved fruits and nuts imports from the ROW countries declined by 3.59% and 5.8% in
2001 and 2002 respectively.
172
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
Table 5.33: Results for preserved fruits and nuts trade flows between South Africa and
the ROW countries
Variables Period Impact Model Yearly Impact Model
Exports Imports Trade Exports Imports Trade
2000 2005 2000 2005 2000 2005 2000 2005 2000 2005 2000 2005
2004 2009 2004 2009 2004 2009 2004 2009 2004 2009 2004 2009
onstant
nYiit·1 0.46* 0.46* 0.20* 0.20* 0.17* 0.17* 0.45* 0.45* 0.20* 0.20* 0.17* 0.17*
nGOPPCit -7.24 -7.24 15.86*** 15.86*** 11.87 11.87 35.25** 35.25**
nGOPPCjt 1.80*** 1.80*** 0.34 0.34 1.73 1.73 0.15 0.15
nGOPPCijt 1.76 1.76 1.89 1.89
~EERt -0.94 -0.94 -1.20 -1.20 -0.16 -0.16 0.21 0.21 -0.09 -0.09 -0.15 -0.15
>0004 /00509 -0.56 1.03 -0.28 -0.22 0.69*** 0.75*
)00/ DOS - -0.49 -1.28 -1.18 -2.72 0.47 0.93*
>01 /D06 0.61 -2.73 -3.59** -3.83 0.04 0.74***
~02 /D07
1.03 3.73*** -5.80** -4.80 0.54 0.67***
~03 /D08 1.13 -3.68 -2.03 -5.30 1.12 0.89**
i>04/D09
0.97 3.50*** -3.39 -4.63 1.33*** 0.35
hOlSTï
djusted R2 0.63 0.63 0.53 0.53 0.78 0.78 0.64 0.64 0.53 0.53 0.78 0.78
)bservations 960 960 390 390 "/.: .. 300 300 960 960 390 390 300 300
ross-Sections 64 64 26 26 20 20 64 64 26 26 20 20
, ** & >I< ** denote significance at the 1, 5 and 10percent levels respectively.
Source: Author's calculations
5.7.6 Fruits and vegetable juices trade flows between South Africa and the ROW
countries
The final results for all fruits and vegetable juices trade flows (exports, imports and trade)
between South Africa and the ROW are presented in Table 5.34 below. The results show that
the total fruits and vegetable juices trade between South Africa and the ROW countries had
significantly increased by 0.81% during the period 2000-2004 and, on an annual basis the
significant increases occurred in 2000, 2001, 2003 and 2004 by 0.95%, 2.22%, 2.23% and
1.43% respectively.
Similarly, the results also show that the total fruits and vegetable juices trade between South
Africa and the ROW countries had significantly increased by 0.66% during the period 2005-
2009 and, on an annual basis significant increases occurred in 2005 by 0.63% and by 0.93%
in 2008 and 2009. However, there were no joint period effects on both fruits and vegetable
juices exports and imports between South Africa and the ROW countries in either period.
However, on an annual basis there were significant increases of South Africa's fruits and
173
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
vegetable juices exports to the ROW countries in 2000, 2001, 2002 and 2004 by 1.06%,
1.34%,3.22% and 2.51% respectively.
Table 5.34: Results for fruits and vegetable juices trade flows between South Africa and
the ROW countries
Variables Period Impact Model Yearly Impact Model
Exports Imports Trade Exports Imports Trade
2000 2005 2000 2005 2000 2005 2000 2005 2000 2005 2000 2005
2004 2009 2004 2009 2004 2009 2004 2009 2004 2009 2004 2009
Constant
InYiit•1 0.30* 0.30* 0.20* 0.20* 0.21 * 0.21 * 0.29* 0.29* 0.20* 0.20* 0.21 * 0.21 *
InGDPPCit 5.43 5.43 15.86*** 15.86*** 16.03*** 16.03*** 35.25** 35.25**
InGDPPCit 3.16* 3.16* 0.34 0.34 3.10* 3.10* 0.15 0.15
InGDPPCiit 4.74* 4.74* 4.52* 4.52*
REERt -0.62 -0.62 -1.20 -1.20 -0.25 -0.25 0.01 0.01 -0.09 -0.09 -0.21 -0.21
00004 / 00509 0.62 -0.09 -0.40 -0.22 0.81*** 0.66**
DOO/'D05 1.06** * -1.44 -0.63 -2.72 0.95** 0.63***
001/006 1.34*** -2.11 -1.20 -3.83 0.71 0.52
002/007 3.22* -2.63 -2.02 -4.80 2.22* 0.56
003/008 2.51 * -2.80 1.22 -5.30 2.23* 0.93**
004/009 0.97 -2.61 1.45 -4.63 1.43** 0.93**
InDISTï
Adjusted R2 0.61 0.61 0.53 0.53 0.73 0.73 0.61 0.61 0.53 0.53 0.73 0.73
Observations 990 990 360 360 345 345 990 990 360 360 345 345
Cross-Sections 66 66 24 24 23 23 66 66 24 24 23 23
*, ** & *** denote significance at the I, 5 and IOpercent levels respectively.
Source: Author's calculations
5.7.7 Wine trade flows between South Africa and the ROW countries
The results for all wine trade flows (exports, imports and trade) between South Africa and the
ROW are presented in Table 5.35 below. The results show that during the period 2000-2004
South Africa's wine exports to the ROW countries had significantly declined by 0.92%. On
an annual basis, these significant decreases occurred in all years as follows: 1.22% in 2000,
2.4% in 2001, 2.98% in 2002, 1.75% in 2003, and 2.77% in 2004. However, during the
period 2005-2009 South Africa's wine exports to the ROW countries had significantly
increased by 1.15%. Surprisingly, on an annual basis the results provide a contrasting picture
as there were significant declines in South Africa's wine exports to the ROW countries that
occurred in 2008 and 2009 by 3.12% and 3.04% respectively.
Furthermore, results also show that South Africa's wine imports from the ROW countries had
significantly declined by 1.57% over the period 2000-2004 and on an annual basis, this
significant decrease occurred only in 2000 by 1.34%. Similarly, during the period 2000-
174
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
2004, the total wine trade between South Africa and the ROW countries had significantly
declined by l.57% and on an annual basis this significant decrease occurred only in 2000 by
0.93%. However, there were significant joint period effects and individual yearly effects on
both South Africa's wine imports from the ROW countries and the total wine trade between
South Africa and the ROW countries during the period 2005-2009.
Table 5.35: Results for wine trade flows between South Africa and the Rest of World
Variables Period Impact Model Yearly Impact Model
Exports Imports Trade Exports Imports Trade
2000 2005 2000 2005 2000 2005 2000 2005 2000 2005 2000 2005
... - - - - - - - - - - - -2004 2009 2004 2009 2004 2009 2004 2009 2004 2009 2004 2009
!constant - - - - - - - - - - - -
InYiit_1 0.30* 0.30* 0.21 * 0.21 * 0.48* 0.48* 0.30* 0.30* 0.21 * 0.21 * 0.47* 0.47*
nCOPPCit 3.52 3.52 4.26 4.26 - - 19.54* 19.54* 13.40 13.40 - -
nCOPPCit 0.26 0.26 0.79 0.79 - - 0.28 0.28 0.68 0.68 - -
nCOPPCijt - - - - 4.29* 4.29* - - - - 4.13* 4.13*
REERt -1.35* -1.35* 2.45* 2.45* 0.54 0.54 -0.42 -0.42 2.98* 2.98* 0.56 0.56
p0004 /00509 -0.92** 1.15*** -1.57** -0.76 -0.86*** 0.40 - - - - - -
DOO / DOS - - - - - - -1.22* -0.92 -1.34**- -2.03 -0.93*** 0.34
001/006 - - - - - - -2.40* -1.86 -0.91 -2.41 -0.66 0.37
P02 /007 - - - - - - -2.98* -2.26 0.34 -3.06 0.89 0.52
p03 /008 - - - - - - -1.75** -3.12** 1.00 -3.33 0.18 0.57
P04/009 - - - - - - -2.77* -3.04** -0.48 -2.51 0.63 0.41
nOISTï - - - - - - - - - - - -
<\djusted R2 0.67 0.67 0.69 0.69 0.80 0.80 0.67 0.67 0.69 0.69 0.80 0.80
Dbservatlons 1215 1215 360 360 345 345 1215 1215 360 360 345 345
:ross-Sections 81 81 24 24 23 23 81 81 24 24 23 23
'" ** & *** denote significance at the I, 5 and IOpercent levels respectively.
Source: Author's calculations
5.8 Summary
This chapter has presented empirical results of the ex-post impacts of the implementation of
the trade agreements under analysis, namely WTO AoA, EU-SA TDCA and the SADC Trade
Protocol, on agricultural trade flows between South Africa and its agricultural trading
partners. Various statistical tests were undertaken to select the suitable models for the
datasets of total agricultural and selected agricultural products trade flows between South
Africa and its agricultural trading partners. In total, 189 models were selected, of which 161
were dynamic models (which comprised 152 FE, 2 RE and 7 OLS estimators) and 28 were
static models (which comprised 14 FE and 14 RE estimators). The higher number of selected
dynamic models with FE estimators justified the importance of dynamics as well as the
importance of the unobserved country-pair specific effects in trade analysis. The per capita
175
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
ODPs of South Africa and of its trading partners, the real effective exchange rates and
distance have also played a significant and expected role in influencing agricultural trade
flows between South Africa and its agricultural trading partners.
The agricultural trade flows between South Africa and its agricultural trade partners have
responded positively to the implementation of the WTO AoA. However, the implementation
of the EU-SA TDCA and SADC Trade Protocol during the first five years (for the period
2000-2004) have not delivered the expected results, as some of the agricultural trade flows
between South Africa and EU countries as well as between South Africa and SADC countries
were negatively affected. The majority of agricultural trade flows between South Africa and
EU countries as well as between South Africa and SADC countries were not affected.
Similarly, the majority of agricultural trade flows between South Africa and EU countries as
well as between South Africa and SADC countries were not affected during the second five-
year term (for the period 2005-2009), but there were some improvements owing to the
significant positive effects of the EU-SA TDCA implementation on three agricultural trade
flows (i.e. total agricultural trade, total cut flowers trade and total preserved fruits and nuts
trade) as well as the significant positive effects of the SADC Trade Protocol implementation
on four agricultural trade flows (i.e. total agricultural exports, total agricultural trade, total cut
flowers trade and total fruits and vegetable juices trade). Surprisingly, the number of
agricultural trade flows between South Africa and ROW countries that have improved
significantly for both periods were greater than those of the EU and SADC countries, even
though ROW countries did not have trade agreements with South Africa.
The implementation of the EU-SA TDCA and SADC Trade Protocol had created room for
potential increases in all agricultural trade flows between South Africa and EU countries, as
well as between South Africa and SADC countries, for both periods under review. Despite
this possibility, some increases were instead diverted to the other markets, more especially
during the period 2000-2004. On average, during the implementation of the EU-SA TDCA
for the period 2000-2004, 0.44% of agricultural exports, 0.96% of cut flowers exports and
0.77% of wine exports from South Africa ordinarily destined for EU countries were diverted
to other markets. Furthermore, 2.01% of South Africa's wine imports ordinarily soureed from
EU countries came from South Africa's other wine trading partners. Moreover, 0.73% of the
total wine trade from the South Africa-EU markets was diverted to either the South Africa
176
Impacts of trade agreements on the agricultural trade flows between
South Africa and its agricultural trading partners
and other wine trading partner market or to the EU and other wine trading partner market.
The implementation of the SADe Trade Protocol also led to the diversion of agricultural
exports (0.43%), cut flowers exports (0.93%), total cut flowers trade (0.92%), wine exports
(0.73%), wine imports (1.45%) and total wine trade (0.35%) during the period 2000-2004.
However, the implementation of the EU-SA TDeA and SADe Trade Protocol during the
period 2005-2009 led to trade creation in some of the agricultural trade flows between South
Africa and the EU countries, as well as between South Africa and the SADe countries. With
regard to the EU-SA TDeA implementation, trade creation was observed in total agricultural
exports, total agricultural trade, total preserved fruits and nuts trade and total wine trade.
Similarly, for the SADe Trade Protocol implementation, trade creation was also observed in
total agricultural trade, cut flowers exports and preserved fruits and nuts exports.
These findings have clearly shown that tariff reductions alone are not a panacea for improving
agricultural trade between South Africa and its major trading partners, given the fact that the
EU-SA TDeA and the SADe Trade Protocol were mainly characterised by tariff phase down
schedules. These findings are supported by the underutilisation of preferential tariff rate
quotas under the EU-SA TDeA, as shown in Table 5.9, where some of the qualified
agricultural products' tariff rates were discounted by 50% and others were zero-rated (i.e.
100% discount).
177
CHAPTER6
SUMMARY, CONCLUSIONS AND RECOMMENDATIONS
6.1 Introduction
The agricultural sector in South Africa had been characterised during the apartheid era by
trade distorting measures, such as quantitative restrictions, price controls, subsidies directly
related to production quantities, and the like. During the early 1990s, South Africa embarked
on a process of trade liberalisation policy reform with an objective of improving trade with its
trading partners in order to improve economic growth, generate employment, improve welfare
gains, and gain other benefits. In the process, South Africa substantially liberalised the
economy through reform of the import regime and deregulation of the agricultural sector.
South Africa's commitment to move from a protective to a liberal trade regime in the
agricultural sector has been witnessed by its trade diplomacy engagements with the
international community through commitments to the globalisation policies in the context of
multilateralisation, bilateralisation and regionalisation processes.
I ~ ..
At the multilateral level, South Africa has successfully implemented its commitments as
negotiated in terms of the World Trade Organisation Agreement on Agriculture (WTO AoA)
during the Uruguay Round (UR) of the General Agreement on Tariffs and Trade (GATT)
negotiations. At the bilateral level, South Africa has signed a Preferential Trade Agreement
(PTA) with the European Union (EU) called the Trade, Development and Co-operation
Agreement (TDCA) (better known as the EU-SA TDCA) which also includes a Free Trade
Agreement. At the regional level, South Africa is a member of the Southern African Customs
Union (SACU) and Southern African Development Community (SADC). SACU member
states have signed a Protocol on Trade and a Regional Trade Agreement (RTA) with non-
SACU SADC countries to establish a Free Trade Agreement (FTA) in the SADC region.
Some of the agricultural products which have been examined have been given certain
preferential treatments both reciprocally and non-reciprocally, such as in-quota tariff rates and
annual tariff phase-down, effective from implementation of the abovementioned trade
agreements. Given the fact that agricultural trade liberalisation was part of the trade
liberalisation strategy of South Africa, it is necessary to analyse the impacts of these trade
agreements on the agricultural trade between South Africa and its trading partners. Therefore,
178
Summary, conclusions and recommendations
this study has endeavoured to assess the impacts of these multilateral, bilateral and regional
trade agreements which South Africa has signed (WTO AoA, EU-SA TDCA and SADC
Trade Protocol) on agricultural trade flows between South Africa and its trading partners.
These impacts have been examined at both the aggregate level (total agricultural imports,
exports and total trade) and the disaggregate or product level (imports, exports and total trade
of selected agricultural products which benefited from the preferential treatments under these
agreements).
The overall objective of this study has been to measure the impacts of these trade agreements
on the agricultural trade between South Africa and its trading partners. The specific
objectives were as follows:
(i) to provide an overview of the trade agreements that have implications for agricultural
sector trade in South Africa;
(ii) to review the impacts of agricultural trade liberalisation policies in the context of the trade
agreements on the economic growth and welfare of South Africa and the Southern Africa
region as well as of its trading partners;
(iii) to determine whether the trade agreements have a significant influence on agricultural
trade between South Africa and its trading counterparts;
(iv) to investigate whether the trade agreements have caused trade creation or trade diversion;
and
(v) to estimate trade potentials between South Africa and its trading partners owing to the trade
agreements.
Several studies have attempted to 'answer the above questions, but with limited scope. Most
of the South African case studies have focused on the impacts of trade agreements on
economic growth and welfare, but have concentrated on a single agreement without
comparing it to the others which also affect trade between South Africa and its counterparts.
A review of the literature has revealed that only very limited research has been conducted on
this issue over recent years. In addition, the research efforts were mainly conducted early
during the implementation phases of the various agreements and were not always in
agreement on the potential impacts the agreements might have.
After reviewing various models that are used to analyse the impact of trade policies and/or
trade agreements, the gravity model was selected in view of the types of research questions
179
Summary, conclusions and recommendations
that this study attempts to answer. The gravity model was previously used by researchers to
analyse the impact of the factors influencing trade performance in order to examine whether a
trade agreement led to trade creation or trade diversion between trading partners, as well as to
estimate trade potentials resulting from the implementation of such trade agreements.
6.2 Empirical results of this study
The results of this study are divided into three sections as follows:
• the results of the suitable model selection exercise for the agricultural trade flows
(exports, imports and total trade) between SA and its agricultural trading partners;
• the results of the effects of the control explanatory variables on agricultural trade
flows between SA and its agricultural trading partners; and
• the results of the impacts of the trade agreements on the agricultural trade flows
between SA and its agricultural trading partners.
The main findings are summarised below in the following subsections.
6.2.1 Selected suitable models for all the agricultural trade flows datasets
Various statistical tests were undertaken to select the suitable models for the datasets of total
agricultural and selected agricultural products trade flows between South Africa and its
agricultural trading partners. After the statistical tests were undertaken, the results have
shown that 189 models, in total, were found to be efficient and suitable for the datasets of the
selected agricultural trade flows, of which 161 were dynamic models and 28 were static
models. Of the selected dynamic models; 152 FE, 2 RE and 7 pooled OLS estimators were
found to be efficient and suitable; while 14 FE and 14 RE estimators were found to be
efficient and suitable for the selected static models. The dominance of the dynamic models
over static models has shown that past trade is the predictor for current trade. The dominance
of FE estimators over RE and OLS estimators has indicated that the joint country-pair specific
effects were not important in determining the majority of agricultural trade flows between SA
and its agricultural trading partners.
180
Summary, conclusions and recommendations
6.2.2 Effects of the control explanatory variables on agricultural trade flows
The control explanatory variables that were included when estimating models are: per capita
Gross Domestic Products of South Africa and of its trading partners (GDPPCs); real effective
exchange rates (REER); and geographic distances between South Africa and its trading
partners (DIST). The results indicated that 56 agricultural trade flows between South Africa
and its agricultural trading partners were positively affected by the income and S of them were
negatively affected by income. These results showed that the majority of agricultural trade
flows between SA and its agricultural trading partners were elastic to income. This means
that there were significant positive correlations between income and agricultural trade flows
between SA and its agricultural trading partners.
Regarding the exchange rates effects, the results showed that the real effective exchange rates
had positively affected 13 agricultural trade flows between South Africa and its agricultural
trading partners, which were mainly dominated by import flows. Furthermore, the results
showed that the real effective exchange rates had negatively affected 11 agricultural trade
flows between South Africa and its agricultural trading partners, which were mainly
dominated by export flows. These results have shown that the effects of the real effective
exchange rates on the majority of these agricultural trade flows were as expected because the
exchange value of the South African currency (Rand) had been appreciating against a basket
of currencies in real terms during the period under review. As a result, it was cheaper for
South Africa to import such agricultural products from the world markets whereas South
Africa's exports of such agricultural products were more expensive in the world markets
owing to the stronger Rand.
With respect to distance, the results showed that distance had negatively affected 8
agricultural trade flows between South Africa and its agricultural trading partners, whereas
only 1 agricultural trade flow was positively affected. These results were in line with the
expectations because distances between trading partners are regarded as proxies for
transportation costs, proxies for risks associated with the quality of some of the perishable
goods and for the cost of personal contact between managers and customers. In this case,
countries with short distances between each other are expected to trade more than those which
are far apart, owing to lower transaction costs. Given the geographic location of South Africa
in relation to its major trading partners, as well as the fact that most of the agricultural
181
Summary, conclusions and recommendations
products traded are perishable, it was expected that distance would have negative effects on
South Africa's exports and imports of agricultural products.
6.2.3 Impacts of trade agreements on agricultural trade flows
This study has analysed the ex-post impact of the implementation of the WTO AoA, EU-SA
TDeA and SADe Trade Protocol on agricultural trade flows between South Africa and its
agricultural trading partners. The study has further analysed the responses of agricultural
trade flows between South Africa and ROW (non-EU and non-SADe) countries during the
implementation of the EU-SA TDeA and the SADe Trade Protocol in order to compare them
with the responses of agricultural trade flows between South Africa and EU countries as well
as the SADe countries.
The results are presented as follows: Firstly, the joint period effects and the individual yearly
effects are shown for the impacts of all trade agreements on agricultural trade flows between
South Africa and its agricultural trading partners (including the ROW countries even though
they had no trade agreement with South Africa). For the WTO AoA impact analysis, the
results cover the period 1995-1999, whereas for the EU-SA TDeA and SADe Trade Protocol
impacts analysis, the results cover the periods 2000-2004 and 2005-2009. Similarly, for the
ROW response analysis, the results cover the periods 2000-2004 and 2005-2009 as the study
would benefit by comparing the response of the agricultural trade flows between South Africa
and ROW countries with those between the EU and SADe countries.
Secondly, for the EU-SA TDeA and the SADe Trade Protocol, the results go further by
reporting whether the implementation of these trade agreements have created or diverted the
agricultural trade flows between South Africa and the EU as well as between South Africa
and SADe countries for the periods 2000-2004 and 2005-2009. Finally, the results report
the agricultural trade flows' potential estimates arising from the implementation of the EU-SA
TDeA and SADe Trade Protocol, as well as showing whether the capacity for the absorption
of agricultural trade flows between South Africa and the EU countries and South Africa and
SADe countries was exhausted or underachieved for the periods 2000-2004 and 2005-2009.
182
Summary, conclusions and recommendations
6.2.3.1 Impact of the implementation of the WTO AoA
The detailed results of the exchange rates effects on the agricultural trade flows are presented
in Appendix 6.A and a summarised version of the results is reported in Table 6.1 below. The
implementation of the WTO AoA during the period 1995-1999 had significantly improved
four agricultural exports flows from South Africa to its agricultural trading partners, both
periodically and on annual basis, namely: the total agricultural exports, cheese exports, cut
flowers exports and wine exports. Furthermore, one import trade flow (South Africa's wine
imports), two total trade flows, both periodically and annually (total cheese and wine trade
flows) and two total trade flows on annual basis only (total cheese trade and total fruits and
nuts trade) were positively affected by the implementation ofWTO's AoA.
Table 6.1: Impact results for the implementation of WTO AoA on agricultural trade
flows between South Africa and its worldwide agricultural trading partners
Impact types Exports (Xl Imports (M) Total Trade (T = X+ M) TOTAL
Positive joint period effects 4 I I 6
Negative joint period effects 0 0 I 1
No joint period effects 3 6 4 13
Positive individual yearly effects 4 I 3 8
Negative individual yearly effects 0 2 2 4
No individual yearly effects 3 4 2 9
Source: Author's calculations
However, two trade flows (total agricultural and total fruits and vegetable juices) as well as
two import flows (cut flowers and preserved fruits and nuts on annual basis only) were
negatively affected during the WTO AoA implementation. The implementation of the WTO
AoA had no joint periodic and individual effects on nine agricultural trade flows between
South Africa and its agricultural trading partners.
6.2.3.2 Impact of the implementation of the EU-SA TDCA
The detailed results of the selected model for all the agricultural trade flows datasets are
presented in Appendix 6.B and a summarised version of the results is reported in Table 6.2
below. The results show that during the implementation of the EU-SA TDCA over the period
2000-2004, four export flows from South Africa to the EU countries suffered and these are
the total agricultural exports, cheese exports and frozen fruits and nuts exports (both
periodically and on annual basis) and wine exports (only periodically). Similarly, one trade
flow (total cut flowers trade) and one import flow (wines imports) had also decreased
183
Summary, conclusions and recommendations
significantly during the implementation of the EU-SA TDCA over the period 2000-2004,
both periodically and annually. Likewise, the preserved fruits and nuts imports, as well as the
total preserved fruits and nuts trade, were negatively affected by the implementation of the
EU-SA TDCA over the period 2000-2004, on annual basis only. On the other hand, two
trade flows (total agricultural trade and total cheese trade) and one export flow (cut flowers
exports) had significantly improved during the implementation of the EU-SA TDCA over the
period 2000-2004, on annual basis only, while one import flow (cut flowers imports)
improved periodically only. The results also show that the implementation of the EU-SA
TDCA over the period 2000-2004 (both periodically and annually) had no effects on nine
agricultural trade flows (i.e. two export flows; four import flows and three total trade flows)
between South Africa and the EU countries.
Table 6.2: Impact results for the implementation of the EU-SA TDCA on agricultural
trade flows between South Africa and EU countries
Impact types Exports Imports Total Trade
(X) (M) (T=X + M)
2000 2005 2000 2005 2000 2005
- - - - - -
2004 2009 2004 2009 2004 2009
Positive joint period effects 0 0 I 0 0 3
Negative joint period effects 4 0 I.'._".' 0 I 0
No joint period effects 3 7 5 7 6 4
Positive individual yearly effects I 0 5 0 2 4
Negative individual yearly effects 3 2 2 2 2 I
No individual yearly effects 3 5 5 5 3 2
Trade creation effects 0 1 0 0 0 3
Trade diversion effects 3 0 I 0 I 0
No trade creation and diversion effects 4 6 6 0 6 4
Source: Author's calculations
The results also show that the implementation of the EU-SA TDCA over the period 2000-
2004 had led to the diversion of five agricultural trade flows between South Africa and EU
countries. These are:
• total agricultural exports, cut flowers exports and wine exports from South Africa that
would ordinarily have been destined for the EU market, but instead went to other
South African trading partners;
• South African wine imports that would ordinarily have been soureed from the EU
countries but instead came from South Africa's other wine trading partners;
184
Summary, conclusions and recommendations
• the diversion of total wine trade from the South Africa-EU market to either the South
Africa and other wine trading partner market or to the EU and other wine trading
partner market.
However, the results show no proof of trade creation or diversion in 16 agricultural trade
flows between South Africa and EU countries during implementation of the EU-SA TDCA
over the period 2000-2004.
Interestingly, the results show that the implementation of the EU-SA TDCA over the period
2000-2004 had created room for potential increases in all of South Africa's agricultural
exports flows in certain EU countries; in South Africa's agricultural imports flows from
certain EU countries; as well as in South Africa's total agricultural trade flows with certain
EU countries. For example, on the export side, there was room for a potential increase in
South Africa's total agricultural exports in eight EU countries. This indicates that the
absorption capacity of South Africa's total agricultural exports in these eight EU countries
had not been exhausted, meaning that South Africa's exports of agricultural products have
underachieved in these eight EU countries. On the other hand, the results show that there was
no room for a potential increase in South Africa's total agricultural exports in seven EU
countries, meaning that the absorption capacity of South Africa's total agricultural exports in
these seven EU countries had been exhausted and also that South Africa's exporters of
agricultural products and these seven EU countries' importers of South Africa's agricultural
products had outperformed.
On the import side, the results indicate that there was room for a potential increase in South
Africa's total agricultural imports from eleven EU countries. This indicates that the
absorption capacity of South Africa's total agricultural imports in these eleven EU countries
had not been exhausted, meaning that imports of agricultural products from these eleven EU
countries had underachieved in the South Africa market. On the other hand, the results show
that there was no room for a potential increase of South Africa's total agricultural imports in
four EU countries, meaning that the absorption capacity of South Africa's total agricultural
imports from these four EU countries had been exhausted and that South Africa's importers of
agricultural products from these four EU countries and the exporters of agricultural products
from these four EU countries to South Africa had outperformed during the implementation of
the EU-SA TDCA over the period 2000-2004.
185
Summary, conclusions and recommendations
On the total trade (exports plus imports) side, the result indicate that there was room for a
potential increase in total agricultural trade between South Africa and eight EU countries.
This indicates that the absorption capacity of total agricultural trade between South Africa and
these eight EU countries had not been exhausted, meaning that the total trade of agricultural
products had underachieved in both South Africa and these eight EU countries. On the other
hand, the results show that there was no room for a potential increase in total agricultural
trade between South Africa and seven EU countries, meaning that the absorption capacity of
total agricultural trade between South Africa and these eight EU countries had been exhausted
and that the agricultural products traders (exporters and importers) of South Africa and these
seven EU countries had outperformed during the implementation of the EU-SA TDCA over
the period 2000-2004.
With regard to the implementation of the EU-SA TDCA over the period 2005-2009, the
results show that three total trade flows between South Africa and the EU countries had
increased significantly, both periodically and on an annual basis: total agricultural trade, total
cut flowers trade and total preserved fruits and nuts trade. The total wine trade also increased,
but on annual basis only. However, two exports flows (total agricultural exports and wine
exports), two imports flows (cheese imports and frozen fruits and nuts imports) and one total
trade flow (total fruits and vegetable juices trade) decreased significantly during the
implementation of the EU-SA TDCA over the period 2000-2004, on annual basis only.
Furthermore, the results also show that the implementation of the EU-SA TDCA over the
period 2005-2009 (both periodically and annually) had no effects on 12 agricultural trade
flows (i.e. 5 export flows; 5 import flows and 2 total trade flows) between South Africa and
the EU countries.
The results also show proof of trade creation for four agricultural trade flows between South
Africa and EU countries arising from the implementation of the EU-SA TDCA over the
period 2005-2009; and these are total agricultural exports, total agricultural trade, total
preserved fruits and nuts trade and total wine trade. However, the results show no proof of
trade creation or diversion in 17 agricultural trade flows (i.e. 6 exports flows, 7 imports flows
and 4 total trade flows) between South Africa and EU countries during implementation of the
EU-SA TDCA over the period 2005-2009. Despite the EU-SA TDCA's insignificant effects
on the majority of agricultural trade flows between South Africa and EU countries, the
implementation of the EU-SA TDCA over the period 2005-2009 had also created room for
186
Summary, conclusions and recommendations
potential increases in all South Africa's agricultural exports flows in certain EU countries; in
South Africa's agricultural imports flows from certain EU countries; as well as in South
Africa's total agricultural trade flows with certain EU countries. The potential effects results
on all agricultural trade flows between South Africa and EU countries should be interpreted in
a similar manner as above (detailed results for agricultural trade flows between South Africa
and EU countries are presented in Appendix 6B).
6.2.3.3 Impact of the implementation of the SADe Trade Protocol
The detailed results of the selected model for all the agricultural trade flows datasets are
presented in Appendix 6.e and a summarised version of the results is reported in Table 6.3
below. The results show that during the implementation of the SADe Trade Protocol over the
period 2000-2004, total agricultural exports from South Africa to the SADe countries as well
as total agricultural trade between South Africa and the SADe countries had significantly
declined both periodically and annually. Furthermore, on an annual basis only, two exports
flows (wine exports and fruits and vegetable juices exports), one import flow (cut flowers
imports) and one trade flow (total cut flowers trade) between South Africa and the SADe
countries also suffered during the implementation of the SADe Trade Protocol over the
period 2000-2004.
Table 6.3: Impact results for the implementation of the SADe Trade Protocol on
agricultural trade flows between South Africa and SADe countries
Impact types Exports Imports Total Trade
(X) (M) (T=X + M)
2000 2005 2000 2005 2000 2005
- - - - - -
2004 2009 2004 2009 2004 2009
Positive joint period effects 0 I 0 0 0 3
Negative joint period effects I I 0 0 I 0
No joint period effects 6 5 4 4 3 I
Positive individual yearly effects 2 I I 0 I 4
Negative individual yearly effects 3 2 I 2 2 0
No individual yearly effects 2 4 2 2 I 0
Trade creation effects 4 2 I 0 2 I
Trade diversion effects 3 0 I 0 2 0
No trade creation and diversion effects 0 5 2 4 0 3
Source: Author's calculations
On the other hand, total agricultural imports, preserved fruits and nuts exports as well as total
fruits and vegetable juices trade between South Africa and the SADe countries significantly
187
Summary, conclusions and recommendations
increased during the implementation of the SADC Trade Protocol for the period 2000-2004.
The results also show that the implementation of the SADC Trade Protocol over the period
2000-2004 (both periodically and annually) had no effects on five agricultural trade flows
(i.e. five export flows and one total trade flow) between South Africa and the SADC
countries.
During the implementation of the SADC Trade Protocol over the period 2005-2009, there
were significant increases in total agricultural exports from South Africa to the SADC
countries as well as in total trade flows of total agriculture, cut flowers and fruits and
vegetable juices between South Africa and the SADC countries, both periodically and
annually. Wine trade between South Africa and the SADC countries increased significantly
on an annual basis only. However, wine exports from South Africa to the SADC countries
decreased significantly during the implementation of the SADC Trade Protocol over the
period 2005-2009, both periodically and annually. Cut flowers imports and both exports and
imports of fruits and vegetable juices between South Africa and the SADC countries
decreased on an annual basis only. The results also show that the implementation of the
SADC Trade Protocol over the period 2005-2009 (both periodically and annually) had no
effects on six agricultural trade flows (i.e. four export flows and two import flows) between
South Africa and the SADC countries.
The results also show that the implementation of the SADC Trade Protocol over the period
2000-2004 had led to the diversion of six agricultural trade flows between South Africa and
the SADC countries. These are:
• total agricultural exports, cut flowers exports and wine exports from South Africa that
would ordinarily have been destined for the SADC market, but instead went to other
South African trading partners;
• South African wine imports that would ordinarily have been soureed from the SADC
countries but instead came from South Africa's other wine trading partners;
• the diversion of total cut flowers trade and total wine trade from the South Africa-
SADC market to either the South Africa and other cut flowers and wine trading
partner market or to the SADC and other cut flowers and wine trading partner market.
188
Summary, conclusions and recommendations
However, the results show proof of import creation for cut flowers imports from the SADe
countries to South Africa during the same period. Similarly, the results of the implementation
of the SADe Trade Protocol over the period 2005-2009 show proof of trade creation for three
agricultural trade flows between South Africa and the SADe countries, and these are total
agricultural trade, cut flowers exports and preserved fruits and nuts exports. However, the
results show no proof of trade creation or diversion in 8 and 12 agricultural trade flows
between South Africa and the SADe countries during implementation of the SADe Trade
Protocol during the periods 2000-2004 and 2005-2009 respectively.
Furthermore, the results show that the implementation of the SADe Trade Protocol over both
the 2000-2004 and 2005-2009 periods had created room for potential increases in all South
Africa's agricultural exports flows in certain SADe countries; in South Africa's agricultural
imports flows from certain SADe countries; as well as in South Africa's total agricultural
trade flows with certain SADe countries. For example, on the export side, there was room
for a potential increase in South Africa's total agricultural exports in two SADe countries.
This indicates that the absorption capacity of South Africa's total agricultural exports in these
two SADe countries had not been exhausted, meaning that South Africa's exports of
agricultural products had underachieved in these two SADe countries. On the other hand, the
results show that there was no room for a potential increase in South Africa's total
agricultural exports in four SADe countries, meaning that the absorption capacity of South
Africa's total agricultural exports in these four SADe countries had been exhausted and that
South Africa's exporters of agricultural products and these four SADe countries' importers of
South Africa's agricultural products had outperformed.
On the import side, the results show that there was room for a potential increase in South
Africa's total agricultural imports from two SADe countries. This indicates that the
absorption capacity of South Africa's total agricultural imports in these two SADe countries
had not been exhausted, meaning that imports of agricultural products from these two SADe
countries had underachieved in the South Africa market. On the other hand, the results show
that there was no room for a potential increase in South Africa's total agricultural imports in
four SADe countries. This means that the absorption capacity of South Africa's total
agricultural exports from these four SADe countries had been exhausted and that South
Africa's importers of agricultural products from these four SADe countries and the exporters
189
Summary, conclusions and recommendations
of agricultural products from these four SADe countries to South Africa had outperformed
during the implementation of the SADe Trade Protocol over the period 2000-2004.
On the total trade (exports plus imports) side, the results indicate that there was room for a
potential increase in total agricultural trade between South Africa and three SADe countries.
This indicates that the absorption capacity of total agricultural trade between South Africa and
these three SADe countries had not been exhausted, indicating that the total trade of
agricultural products had underachieved in both South Africa and these three SADe
countries. On the other hand, the results show that there was no room for a potential increase
in total agricultural trade between South Africa and three SADe countries. This indicates that
the absorption capacity of total agricultural trade between South Africa and these three SADe
countries was exhausted and that the agricultural products traders (exporters and importers) of
South Africa and of these three SADe countries had outperformed over the implementation
of the SADe Trade Protocol during the period 2000-2004. The potential effects results on
other agricultural trade flows between South Africa and the SADe countries should be
interpreted in a similar manner as above (detailed results for agricultural trade flows between
South Africa and SADe countries are presented in Appendix 6C).
6.2.3.4 Agricultural trade response between South Africa and ROW countries during
the implementation of the EU-SA TDCA and SADC Trade Protocol
The detailed results of the selected model for all the agricultural trade flows datasets are
presented in Appendix 6.D and a summarised version of the results is reported in Table 6.4
below. The results show that during the period 2000-2004, only one export flow (fruits and
vegetable juices on an annual basis only) and three total trade flows (total agricultural, total
preserved fruits and nuts, and total fruits and vegetable juices trade flows, both periodically
and annually) between South Africa and the ROW countries had significantly improved.
Similarly, during the period 2005-2009, three total trade flows (total agricultural trade, total
preserved fruits and nuts trade, and the total fruits and vegetable juices trade, both
periodically and annually) between South Africa and the ROW countries had also increased
significantl y.
Furthermore, during the period 2005-2009, four export flows from South Africa to the ROW
countries had also increased significantly (but periodically only) and these are the total
190
Summary, conclusions and recommendations
agricultural, cheese, cut flowers and wine exports. Surprisingly, on annual basis, the total
agricultural and wine exports had negatively declined in certain years during the period 2005-
2009 while they were jointly significant periodically. In addition, preserved fruits and nuts
exports also declined significantly on an annual basis during the same period.
Table 6.4: Responsiveness results of the agricultural trade flows between South Africa
and ROW countries during the of the implementation EU-SA TnCA and SADC Trade
Protocol
.._--
Impact types Exports Imports Total Trade
{X) (M) (T = X + M)
2000 2005 2000 2005 2000 2005
- - - - - -
2004 2009 2004 2009 2004 2009
Positive joint period effects 0 4 0 '0 3 3
Negative joint Eeriod effects 3 0 I 1 I 0
No joint period effects 4 3 6 6 I 2
Positive individual yearly effects I 0 0 0 3 3
Negative individual yearly effects 3 3 3 3 2 I
No individual yearly effects 3 4 4 4 0 I
Source: Author's calculations
The results also show that, during the period 2000-2004, three exports flows from South
Africa to the ROW countries suffered both periodically and on an annual basis: these are the.
total agricultural exports, cut flowers exports and wine exports. Similarly, both periodically
and annually, one import flow (wine imports) and one trade flow (total wine trade) as well as
two import flows (cheese and preserved fruits and nuts imports on annual basis only) between
South Africa and the ROW countries had significantly declined during the period 2000-2004.
Furthermore, three import flows (cheese and cut flowers imports on an annual basis and wine
imports periodically) and one trade flow (total cut flowers trade on an annual basis) between
South Africa and the ROW countries had also declined significantly during the period 2005-
2009. However, there were no joint periodic and individual yearly effects on seven
agricultural trade flows (for the period 2000-2004) and six agricultural trade flows (for the
period 2005-2009) between South Africa and ROW countries.
6.3 Conclusions of the study
The objective of this study has been to analyse the impacts of the implementation of WTO's
AoA, EU-SA TDCA and SADC Trade Protocol on agricultural trade flows between South
Africa and its agricultural trading partners. The results emphasise the importance of
dynamics in trade analysis because more dynamic models were found to be suitable than
191
Summary, conclusions and recommendations
static models. These findings support many economic arguments which have suggested that
lagged trade is the predictor for current trade. This is true because, historically before the
conclusion and implementation of these trade agreements, South Africa had already been
trading with some of these trading partners. For example, the European Union had been
South Africa's main trading and investment partner, accounting for over 40% of its total
trade, before the EU-SA TDCA was implemented. Likewise, EU foreign investment in South
Africa had accounted for over 70% of its total foreign direct investment (FDI).
The per capita GDPs of South Africa and of its trading partners, the real effective exchange
rates and distances have also all played a significant and expected role in influencing
agricultural trade flows between South Africa and its agricultural trading partners as follows:
• In the cases where income had significant effects, the majority of agricultural trade
flows between South Africa and its agricultural trading partners were positively
affected by percentage changes in per capita GDPs, meaning that these' trade flows
were income-elastic.
• In the cases where effective exchange rates had significant effects, the majority of
South Africa's agricultural export flows suffered whereas all South Africa's
agricultural import flows gained because the exchange value of the South Africa
currency (Rand) had been appreciating against a basket of currencies of major South
African trading partners in real terms over the period under review.
II In the cases where distance had significant effects, the majority of agricultural trade
flows between South Africa and its agricultural trading partners were negatively
affected. This shows that the distances between trading partners can indeed serve as
proxies for both the transportation costs and the risks associated with the quality of
some of the perishable goods. This is evident given the geographic location of South
Africa in relation to its major trading partners and the fact that most of the agricultural
products traded are perishable.
The overall findings indicate that agricultural trade flows between South Africa and its
agricultural trade partners have responded positively to the global implementation of WTO's
AoA, coupled with the implementation of the deregulation policy in South Africa. However,
192
Summary, conclusions and recommendations
the unexpected outcomes of the implementation of the EU-SA TDCA and SADC Trade
Protocol were that the majority of agricultural trade flows between South Africa and EU
countries, as well as between South Africa and SADC countries, were not affected. During
the first five years (2000-2004) of the implementation of both the EU-SA TDCA and SADC
Trade Protocol, the joint period effects on all the affected agricultural trade flows between
South Africa and the EU countries, as well as between South Africa and the SADC countries,
were significantly negative. Compared to the ROW countries for the same period, some of
the agricultural trade flows between South Africa and the ROW countries have improved
significantly even though they did not have a trade agreement with South Africa.
However, during the second five-year term (2005-2009) of the implementation of the EU-SA
TDCA, three agricultural trade flows (total agricultural exports, total agricultural trade and
total fruits and vegetable juices trade) between South Africa and EU countries had responded
positively. Similarly, during the second five-year term (2005-2009) of the implementation of
the SADC Trade Protocol, four out of five of the affected agricultural trade flows (total
agricultural exports, total agricultural trade, total cut flowers trade and total fruits and
vegetable juices trade) between South Africa and SADC countries had improved significantly
with the exception of wine exports from South Africa to SADC countries. Despite these few
improvements, the majority of agricultural trade flows between South Africa and EU
countries, as well as between South Africa and SADC countries, were not affected.
Compared to the ROW countries for the same period, agricultural trade flows between South
Africa and ROW countries were still above those of the EU and SADC countries in terms of
the number of positive significant flows, as eight of them have improved during the same
period. These are: total agricultural exports, total agricultural trade, cheese exports, cut
flowers exports, total preserved fruits and nuts trade, total fruits and vegetable juices trade,
wine exports and wine imports.
However, the implementation of the EU-SA TDCA and SADC Trade Protocol created room
for potential increases in all the agricultural trade flows between South Africa and EU
countries, as well as between South Africa and SADC countries, over both periods. However,
some of these potential increases for the period 2000-2004 were diverted to other markets.
These are: agricultural exports, cut flowers exports, wine exports, wine imports and total wine
trade in the case of the EU-SA TDCA; and agricultural exports, cut flowers exports, total cut
193
Summary, conclusions and recommendations
flowers trade, wine exports, wine imports and total wine trade in the case of the SADC Trade
Protocol.
In conclusion, for the period 2005-2009, the results do show some proof of trade creation for
certain agricultural trade flows between South Africa and EU countries, as well as between
South Africa and SADC countries, owing to the implementation of the EU-SA TDCA and
SADC Trade Protocol. These are: total agricultural exports, total agricultural trade, total
preserved fruits and nuts trade and total wine trade in the case of the EU-SA TDCA; and total
agricultural trade, cut flowers exports and preserved fruits and nuts exports in the case of the
SADC Trade Protocol. However, the results show no proof of trade creation or diversion in
the majority of agricultural trade flows between South Africa and EU countries, nor between
South Africa and the SADC countries, for either period over the implementation of the EU-
SA IDCA and SADC Trade Protocol.
6.4 Recommendations of the study
The results of this study indicate various issues for which recommendations may be made.
These will be set out below.
The trade agreements which South Africa has signed with its major trading partners, i.e. the
EU and SADC countries, were aimed at improving agricultural trade between the trading
partners through the liberalisation of agricultural markets by phasing down tariffs as well as
by the introduction of preferential tariff rate quotas for selected agricultural products. The
overall findings of this study clearly indicate that the implementation of these agreements has
not achieved their intended objectives. This is shown by the facts that the majority of
agricultural trade flows between South Africa and its major trading partners have not
improved significantly and that they have even declined significantly in some cases after the
implementation of these agreements.
Given these findings, it is clear that tariff reductions alone are not a panacea for improving
agricultural trade between South Africa and its major trading partners because even the trade
on zero-rated agricultural products has not improved. It is clear that agricultural trade
between South Africa and its major trading partners is strongly influenced by other factors
such as non-tariff barriers (NTBs) and other technical barriers to trade (TBTs). These factors
194
Summary, conclusions and recommendations
may have negatively affected the competitiveness of South Africa's agricultural products in
the markets under review. Therefore, this study recommends that the agenda for future
bilateral trade negotiations should strongly focus on the other factors that affect agricultural
trade between South Africa and its trading partners. The negotiations should be geared
towards the harmonisation of the standards affecting agricultural trade between South Africa
and its trading partners. This recommendation is proposed in the light of the deteriorating
functionality of the multilateral trade body (WTO) which has developed a framework to
regulate international trade through gradual reduction of trade barriers so as to stimulate
international commerce.
This study further recommends that, while trade agreements between South Africa and its
major trading partners have created a conducive trade environment, trade promotions through
trade fairs should be strongly emphasised as these may help to improve trade. Trade
promotions play significant roles in assisting agricultural traders to better understand various
agricultural markets in the world, as well as to make. contacts among themselves. This is the
area in which South African agri-business companies (exporters and importers), including
industry and commodity organisations, should take the lead in order to continuously promote
South African agricultural products in the world's markets.
This study recommends further similar studies to be undertaken in order to continuously
assess the impact of trade agreements between South Africa and its major trading partners
(i.e. EU and South Africa countries) on the trade flows of selected agricultural products.
These studies are necessary because both the EU-SA TDCA and SADC Trade Protocol are
aimed at becoming Free Trade Agreements (FTAs). Furthermore, these studies are
necessitated by the expansion of these historical South African markets, given the EU's
enlargement from 15 to 26 countries, as well as by the fact that some of the SADC countries
which were not part of the SADC Trade Protocol are planning to join. Finally, the study
recommends that further studies on the impacts of these agreements should also focus on the
implications of NTBs and TBTs on the trade flows of selected agricultural products between
South Africa and its major trading partners, the EU and SADC countries.
195
REFERENCES
Adhikari, R. 2000. Agreement on Agriculture and Food Security: South Asian Perspective.
Journal of Asian Economics Il :43-64.
Aitken, N.D. 1973. The effect of the EEC and EFTA on European trade: a temporal cross-
section analysis, American Economic Review 63 :881-892.
Almon, C. 1991. The INFORUM Approach to iter-industry modeling, Economic Systems
Research 31: 1-7.
Almon, C.; Ruiz-Moncayo, A. and L. Sangines 1991. Simulation of a Mexico-USA Free
Trade Agreement. Economic Systems Research 31 :93-97.
Alston, J.M.; Carter, C.; Green, R., and Pick D. 1990. Whither armington trade models?
American Journal of Agricultural Economics 72:455-67.
Alston, J.M. and Chalfant, A. 1993. The silence of the Lamdas: A test for the almost ideal
demand system and Rotterdam models. American Journal of Agricultural Economics 75:304-
314.
Anderson, 1.E. 1979. A theoretical foundation for the gravity equation. American Economic
Review 69:106-16.
Anderson, G. and Blundell, R. 1983. Testing restrictions m a flexible dynamic demand
system: An application to consumer's expenditure in Canada. Review of Economic Studies
50:397-410.
Anderson, K.; Dimaranan, B.; Francois, F.; Hertel, F.; Hoekman, B. and Martin W. 2001. The
cost of rich and poor country protection to developing countries. Journal of African
Economies 103: 227-257.
196
References
Anderson, J .E. and Van Wincoop, E. 2003. Gravity with gravitas: A solution to the border
puzzle. American Economic Review 93 (1): 170-192.
Anderson, J.E. and Van Wincoop, E. 2004. Trade Costs. Journal of Economic Literature XLII
(September 2004):691-751
Andriamananjara, S. and Hillberry, R. 2001. Regionalism, Trade and Growth: The case of
EU-South Africa Free Trade Agreement. Us. International Trade Commission Working
Paper No. 2001-07-A, July.
Arellano, M. and Bond, S. 1991. Some tests of specification for panel data: Monte Carlo
evidence and an application to employment equations. Review of Economic Studies 58: 277-
297.
Arrnington, P. 1969. A Theory of Demand for Products Distinguished by Place of Production.
International Monetary Fund Staff papers 16(1969): 159-178.
Babarinde, O.A. 1996. Analyzing the proposed African Economic Community: Lessons from
the experience of the European Union, Prepared for the Third ECSA-World Conference on
The European Union in a Changing World, Sponsored by the European Commission, D-G X,
Brussels, Belgium, 19-20 September, 1996.
Bade, J. 1998. The effect of the Uruguay Round Agreement on World Food Production and
Food Security. Proceedings of the 5ih Seminar of the European Association of Agricultural
Economists, Wageningen, The Netherlands.
Baltagi, B.H. 1995. Econometric analysis of panel data, John Wiley & Sons, West Sussex,
England.
Baldwin, R. and Taglioni, D. 2006. Gravity for dummies and dummies for gravity equations.
NEER Working Paper 12516, Cambridge. Available online at:
http://www.nber.org/papers/wI2516
197
References
Banks, J.; Blundell, R. and Lewbel, A. 1997. Quadratic Engel curves and consumer demand.
Review of Economics and Statistics 79:527-39.
Barten, A.P. 1964. Consumer martin demand functions under conditions of almost additive
preferences. Econometrica 1-2:1-38
Bamett, W.A. 1979. Theoretical foundations of the Rotterdam model. Review of Economic
Studies 46: 109-130.
Bamett, W.A. 1983. New indices of money supply and the flexible Laurent demand system.
Journal of Business and Economic Statistics 1:7-23.
Bamett, W.A. 1985. The Minflex Laurent Translog flexible functional form. Journal of
Econometrics 30: 33-44.
Bamett, W.A. and Lee, W.Y. 1985. The global properties of the Minflex Laurent, Generalized
Leontief and Translog flexible functional forms. Econometrica 53: 1421-1437
Bamett, W.A.; Lee, W.Y. and Wolfe, M.D. 1985. The three- dimensional global properties of
the Minflex Laurent, Generalized Leontief, and Translog flexible functional forms. Journal of
Econometrics 30:3-31.
Barnett, W.A. and Seck, O. 2008. Rotterdam model vs. almost ideal demand system: Will the
best demand specification please stand up? Journal 0.[Applied Econometrics 23(6):795-824.
Barten, A.P. 1964. Consumer demand functions under conditions of almost additive
preferences. Econometrica 1-2:1-38.
Barten, A.P. 1968. Estimating demand equations. Econometrica 36(2):213-51.
Bayoumi, T. and Eichengreen, B. 1995. Restraining yourself: The implications of fiscal rules
for economic stabilisation. IMF Staff Paper 42:32-48.
198
References
Beghin, J.C.; Roland-Holst, D. and Van der Mensbrugghe, D. 2002. Global agricultural trade
and the Doha Round: What are the implications for North and South? Center for Agricultural
and Rural Development, Iowa State University, USA.
Bergstrand, I.H. 1986. The gravity equation In international trade: Some microeconomic
foundations and empirical evidence. Review of Economics and Statistics 67:474-481.
Bergstrand, I.H. 1989. The generalized gravity equation, monopolistic competition and the
factor-proportion theory in international trade. Review of Economics and Statistics 71: 143-53
Blake, A.T.; Hubbard, LJ.; Philippidis, G.; Rayner, A.I. and Reed, G.V. 1999. General
Equilibrium Modelling of the Common Agricultural Policy. Report to UK MAFF and HM
Treasury, March 1999.
Blanciforti, L. and Green, R. 1983. An almost ideal demand system incorporating habits: an
analysis of expenditures on food and aggregate commodity groups. Review of Economics and
Statistics 65 :511-515
'BlundelI, R. and Bond, S. 1998. Initial conditions and moment restrictions in dynamic panel
data models. Journal of Econometrics 87: 115-143.
Boumellassa, H.; Decreux, Y. and Fontagné, L. 2006. Economic impact of a Potential Free
Trade Tgreement FTA between the European Union and ASEAN. Final report 2006-05-03,
Commission of the European Union - Directorate-General for Trade. Available online at:
http://trade.ec.europa.eu/doclib/htmI1134017.htm
Britz, W. and Heckelei, T. 1997. Pre-study for a medium-term simulation and forecast model
of the agricultural sector for the EU. Institut fur Agrarpolitik, University of Bonn.
Brown D.K.; Deardorff, A.V. and Stern, R. M. 1992a. A North American Free Trade
Agreement: Analytical issues and a computational assessment. The World Economy 15:15-29.
Brown D.K.; Deardorff, A.V. and Stern R.M. 1992b. North American Economic Integration.
Economic Journal I 02: 1507-1518.
199
References
Brown, M. G. and Lee, J. 1993. Alternative specifications of advertising in the Rotterdam
model. European Review of Agricultural Economics 20(4):419-436.
Brown D.K.; Deardorff, A.V. and Stem, R.M. 1995. Modeling multilateral trade liberalization
in services, School of Public Policy, University of Michigan Discussion Paper No. 378.
Brown D.K.; Deardorff, A.V.; Fox, A.K. and Stem, R.M. 1996. Computational analysis of
goods and services liberalization in the Uruguay Round. In: Martin, W. and Winters, L.A.
eds. The Uruguay Round and the Developing Economies, New York: Cambridge University
Press.
Brown, M. and Lee, J. 1997. Incorporating generic and brand advertising effects in the
Rotterdam demand system. International Journal of Advertising 16:211-220.
Brown, M. and Lee, J. 2002. Restrictions on the effects of preference variables III the
Rotterdam model. Journal of Agricultural and Applied Economics 34(2002):17-26.
Bun, MJ.G. and Klaassen, F.J.G.M. 2002. The importance of dynamics in panel gravity
models of trade. University of Amsterdam, February 6, 2002.
Burfisher, M.E. and Jones E.A. 1998. Regional Trade Agreements and U.S. Agriculture.
Economic Research Service, AER No. 77J, Washington, DC: U.S. Department of Agriculture
Burniaux, J-M. 1997. Le Radeau de la Méduse: Analyse de dilemmes alimentaires. In: Van
Tongeren, F. and Van Meijl, H. eds. Review of applied models of international trade in
agriculture and related resource and environmental modeling. FAIR6 CT 98-4148 Interim
report No.1, December 1999, Agricultural Economics Research Institute LEI, The Hague, The
Netherlands.
Cafferata1, lP. and Segura, lA. 2007. The Peru-US Trade Promotion Agreement TPA:
Possible impact on Peruvian agriculture. Comuniica Second Stage: May - August 2007.
Available online at:
http://webiica.iica.ac.cr/bibliotecas/repiica/B0577i/B0577i.pdf
200
References
Cassim, R. 2000. The determinants of intra-regional trade in southern Africa with specific
reference to South Africa and the rest of the region. Working paper 9671, Development
Policy Research Unit, University of Cape Town, Cape Town, South Africa
Cassim, R. 2001. The determinants of intra regional trade in Southern Africa with specific
reference to South Africa and the Rest of the Region, DPRU Working Paper 01/51,
University of Cape Town.
Cassim, R.; Onyango, D. and Van Seventer, D.E. 2002. The state of trade policy in South
Africa. Trade and Industrial Policy Strategies, South Africa.
Centre for International Economics (CIE) 2004. The Australia- Thailand Free Trade
Agreement: Economic effects. A report prepared for Department of Foreign Affairs and
Trade, Canberra & Sydney, Australia. Available online at:
www.dfat.gov.au/ftaltaftaltafta eco effects cie.pdf
Chant, L.; McDodald, S. and C. Punt 2001. Agricultural trade liberalization, agricultural
productivity growth and employment. Agrekon 40 (4):573-583.
Chauvin, S. and G. Gaulier 2002. Prospects for increasing trade among SADC Countries. A
paper presented at 2002 Annual Forum of the Trade and Industry Policy Strategies, Glenburn
Lodge, Muldersdrift, South Africa.
Cheng, I.-H. and Wall, H. J. 1999. Controlling for Heterogeneity in Gravity Models of Trade.
Federal Reserve Bank of St. Louis Working Paper 1999-010E.
Cheng, I.-H and Wall, H.I. 2005. Controlling for heterogeneity in gravity models of trade and
integration. Federal Reserve Bank ofSt. Louis Working Paper 871 :49-63.
Chitiga, M.; Kandiero, T. and Ngwenya, P. 2008. Agricultural trade policy reform in South
Africa. Agrekon 47(1 ):76-1 01.
Christensen, L.R.; Jorgenson, D.W. and Lau, L.J. 1975. Transcendental logarithmic utility
functions. American Economic Review 65: 367-83.
201
References
Christou, C. and D. Nyhus 1994. Industrial Effects of European Community Integration,
Economic Systems Research. 62: 179-98.
Cooper, R.l. and McLaren, K.R. 1996. A system of demand equations satisfying effectively
global regularity conditions. Review of Economics and Statistics 78 (2): 359-364.
Coulibaly, N. and Brorsen, B.W. 1999. Monte Carlo sampling approach to testing nonnested
hypotheses: Monte Carlo results. Econometric Reviews 18(2): 195-209.
Dascal D.K.; Mattas, K. and Tzouvelekas, V. 2002. An analysis of EU wine trade: A gravity
model approach. International Advances in Economic Research 82: 135-148
Davenport, M.; Hewitt, A. and Koning, A. 1994. The impact of the GATT Uruguay Round on
ACP States. Overseas Development Institute London and European Center for Development
Policy Management Maastricht.
Davies. R. 1996. Promoting regional integration in Southern Africa: an analysis of prospects
and problems from a South African perspective. African Security Review 5(5):20-35.
Davies, R. 1998. The resource allocation effects of European Union-South Africa Trade
Agreements: A general equilibrium analysis using GTAP. A paper presented at Forum 1998,
Industrial Restructuring in South Africa.
Deardorff, A.V. and Stem, R.M. 1994. Multilateral trade negotiations and preferential trading
arrangements. In Deardoff, A.V. and Stem, R.M. eds ..Analytical and negotiating issues in the
Global Trading System. Ann Arbor. University of Michigan press.
Deardorff, A.V. 1998. Determinants of bilateral trade: Does gravity work in a neoclassical
world. In Frankel, lA. ed. The regionalization of the world economy. University of Chicago
Press: 7-32.
Deaton, A. and Muellbauer, J. 1980. An almost ideal demand system. American Economic
Review 70:312-26.
202
References
De Blasi G.; Seccia A.; Carlucci, D. and Santeramo F.G. undated. Effects of political-
economic integration and trade liberalization on exports of Italian quality wines produced in
determined regions, Unpublished manuscript, Available online at:
http://www.fabiosanteramo.netlwp-contentluploads/secciagravityoiv.pdf
De Grauwe, P. and Skudelny, F. 2000. The impact of EMU on trade flows.
Weltwirtschaftliches Archiv 136: 381-402
Department of Trade and Industry (the dti) 2010. A South African trade policy and strategy
framework. Prepared by the International Trade and Economic Development Division of the
Department of Trade and Industry, Pretoria, South Africa, May 2010
Devadoss, S.; Westhoff P.; Helmar, M.; Grundmeier, E.; Skold, K.; Meyers, W.H. and
Johnson, S.R. 1993. The FAPRI modeling system at CARD: A documentation summary. In:
Van Tongeren, F. and Van Meijl, H. eds. Review of applied models of international trade in
agriculture and related resource and environmental modeling. FAIR6 CT 98-4148 Interim
report No.l, December 1999, Agricultural Economics Research Institute LEI, The Hague, The
Netherlands.
Diao, x., Rose, T. and A, Somwatu 2002. Developing Country Interests in Agricultural
Reforms under the World Trade Organization. TMD Discussion Paper No. 85, IFPRI,
Washington D.C.
Diao, X. Somwaru, A. and Roe, T, 200l. A Global analysis of agricultural reform in WTO
Member Countries. In Burfisher, ed. The road ahead: Agricultural policy reform in the WTO,
Economic Research Service, USDA, AER No. 802.
Diao, X. and Robinson, S. 2003. Market opportunities for Southern African agriculture in the
new trade agenda: An economy-wide analysis from a global CGE model. International Food
Policy Research Institute, Washington DC, United States.
Diewert, W.E. 1971. An application of the Shephard duality theorem: A Generalized Leontief
production function. Journal of Political Economy 79:461-507
203
References
Dimaranan, B.; Hertel, T. and R. Keeney 2003. OECD domestic support and the developing
countries. GTAP Working Paper No. 19,22 January 2003
Dixit, P.M. and Roningen, V.O. 1986. Modelling bilateral trade flows-with the static world
policy simulation (SWOPSIM)-modelling framework. In: Van Tongeren, F. and Van Meijl,
H.eds. Review of applied models of international trade in agriculture and related resource and
environmental modeling. FAIR6 CT 98-4148 Interim report No.1, December 1999.
Agricultural Economics Research Institute LEI, The Hague, The Netherlands.
Duffy, M. H. 1987. Advertising and the inter-product distribution of demand: A Rotterdam
model approach. European Economic Review. 31 (July 1987): 1051-1 070
Duffy, M. 1990. Advertising and alcoholic drink demand in the UK: Some further Rotterdam
model estimates. International Journal of Advertising 9(3):247-257.
Eales, J.S. and Unnevehr, LJ. 1988. Demand for beef and chicken products: Separability and
structural change. American Journal of Agricultural Economics 70:521-532.
Eaton, J. and Kortum, S. 2002. Technology, geography and trade. Econometrica, 705:1741-
1779
Edwards, S. 1989. Openness, outward orientation, trade liberalization, and economic
performance in developing countries. World Bank Working Papers, No.191, June.
Eichengreen, B. and Irwin, D.A. 1997. The role of history in bilateral trade flows. NBER
working papers 5565, Cambridge, Massachusetts.
Engel, C. and Rogers J. 1997. Regional patterns in the law of one pnce: The role of
geography vs. currencies. In A.I. Frankel ed. Regionalization of the world economy. NBER
Series, No. 7, University of Chicago Press
European Union Commission EC 1999: Agreement on trade, development and cooperation
between the European Community and its member states, of the one part, and the Republic of
South Africa, of the other part. EC journal, 8731/99, Brussels.
204
References
Evans, D. 2001. "Options for regional integration in Southern Africa." The South African
Journal of Economics, 64 4: 662-92.
Evenett, 1.S. and Keiler, W. 2002. On theories explaining the success of the gravity equation.
Journal of Political Economy 1102: 281-316.
Fabiosa, J.; Beghin, J.; De Cara, S.; Fang, C.; Isik, M. and Matthey, H. 2003. Agricultural
markets liberalization and the Doha Round. A paper presented at the zs" Conference of the
International Association of Agricultural Economics, 16 - 22 August 2003, Durban South
Africa.
Food and Agriculture Organization (FAO) 1993. The world food model - model specification,
FAO Mimeograph, ESC/M/93/1, Rome.
Food and Agriculture Organization (FAO) 1994. Impact of the Uruguay Round on
agriculture. FAO CCP: 95/13, Rome.
Food and Agriculture Organization (FAO) 1998. World food model technical documentation.
Commodity and Trade Division, FAO, Rome.
Francois, J.F.; McDonald, B. and Nordstrom, H. 1995. Assessing the Uruguay Round. In:
Martin, W. and L.A., Winters eds. The Uruguay Round and the Developing Countries,
Cambridge: Cambridge University Press, 1995
Francois, J. F. and Reinert, K.A. 1997. Applied methods for trade policy analysis: A
handbook. Cambridge University Press.
Frankel A.I., and Stein, E. 1994 The welfare implication of continental trading blocs in a
model with transports costs. Pacific Basin Working Paper Series No. PB94-03. San
Francisco, Federal Reserve Bank of San Francisco.
Frankel, J.A., and Wei, S. 1993. Trade blocs and currency blocs. NBER working paper 1335.
Cambridge, Massachusetts.
205
References
Frankel, I.A.; Stein, E. and Wei, S. 1995. Trading blocs and the Americas: The natural, the
unnatural, and the supernatural. Journal of Development Economics 47(1 ):61-95
Frankel, I.A.; Stein, E. and Wei, S. 1996. Regional trading arrangements: Natural or
supernatural? American Economic Review 86(2):52-56.
Frankel, J. A., E. Stein and S-l. Wei 1997. Continental trading blocks: Are they natural or
supernatural? In: Poonyth, D, Esterhuizen, D., Ngqangweni, S and lF. Kirsten 2002. Trade
policies and agricultural trade in the SADC region: Challenges and implications, South Africa
Country Study, University of Pretoria, South Africa.
Frankel, I.A. and Romer, D. 1999. Does trade cause growth? American Economic Review.
89(3):379-399.
Francis, S. 2011. The ASEAN-India Free Trade Agreement: A sectoral impact analysis of
increased trade integration in goods. Available online at:
http://www.networkideas.org/ideasact/dec09/pdf/Smitha Francis Paper.pdf
Fulponi, L.; Shearer, M. and Almeida, J. 2011, Regional trade agreements - Treatment of
agriculture. OECD Food, Agriculture and Fisheries Working Papers, No. 44, OECD
Publishing. doi: 10.1787/5kgg53fmnjxv-en. Available online at:
www.standardsfacility.orgIFiles/News/AGR 45.pdf
GATT 1994. Agreement on Agriculture. Available online at:
http://www .wto .org/ english! docs e/legal el 14-ag. pdf
Gay, S.H. 2004. Free Trade Agreement between South Africa and the EU: Southern African
regional concerns and implications for the fruit trade. Presentations to German ENARPRI
Seminar, Braunschweig, March 15,2004.
Gebrehiwet, Y.; Ngqangweni, S. and Kirsten lF. 2007. Quantifying the trade effect of
sanitary and phytosanitary regulations of OECD countries on South African food exports.
Agrekon,46 1, 23 - 39
206
References
Geda, A. and Kibret, H. 2002. Regional economic integration in Africa: A review of problems
and prospects with a case study of COMESA. Available online at:
http://mercury .soas. ac.uk/economics/workpap/ adobe/wp 125.pdf.
/
Glick, R. and Rose, A.K. 2001. Does a currency union affect trade? The time series evidence.
Unpublished manuscript. Available online at: http://haas.berkely.edu/~arose.
Goldin, I.; Knudsen, 0.; van der Mensbrugghe, D. 1993. Trade liberalisation: Global
economic implications. In: Jooste, A. 2001. Economic implications of trade liberalisation on
the South African Red Meat Industry. Unpublished PhD Thesis, University of the Free State,
Bloemfontein, South Africa.
Goldstein, A. 2004. Regional integration, FDI, and competitiveness III Southern Africa.
Development Centre Studies, Organization for Economic Cooperation and Development,
Paris, France.
Goldstein, M. and Khan, M.S. 1985. Income and Price Effects in Foreign Trade. In: Jones,
R.W. and Kenen, P.B. eds. Handbook of International Economics, Amsterdam: North-
Holland, 1041-1105.
Green, R. and Alston, J.M 1990. Elasticity III AIDS Models. American Journal of
Agricultural Economics 70:442-445.
Greene, W.H. 2000. Econometric Analysis 4th edition, New Jersey, Prentice
Greenaway, D. 1994. Implications of the Uruguay Round for trade and development policies
of developing countries. In: Ingco, M., and Townsend, R.F. 1998. Experience and Lessons
from the Implementation of Uruguay Round Commitments: Policy Options and Challenges
for African Countries. Paper presented at the International Workshop on Agricultural Policy
of African Countries and Multilateral Trade Negotiations: Challenges and Options, 23-26
November, Harare, Zimbabwe.
Gros, D. and Gonciarz, A. 1996. A note on the trade potential of Central and Eastern Europe.
Journal of Political Economy, 12: 709-721
207
References
Guyomard H.; Mahé, L.; Tavéra, C. and Trochet, T. 1988. Some problems of modeling
agricultural trade policy interactions between the EC and US. International Agricultural Trade
Consortium. San Antonio.
Guyomard H.; Mahé, L.; Tavéra, C. and Trochet, T. 1991. Technical change and EC-US
agricultural trade liberalization. Journal of Agricultural Economics 42(2):119-137.
Guyomard H. and Mahé, L. 1994. EC-US trade relations in the context of the GATT
negotiations and of the reform of the Common Agricultural Policy. In: The economics of the
Common Agricultural Policy CAP. European Economy Reports and Studies, no 5, 1994.
Hanoch, G. 1975. Production and demand models with direct or indirect implicit additivity.
Econometrica 43:395-419.
Hansohm, D.; Breytenbach, W.; Hartzenberg, T. and McCarthy, C. 2004. Monitoring regional
integration in Southern Africa. Yearbook Volume 4-(2004), Namibian Economic Policy
Research Unit, Windhoek, Namibia.
Harrison, G.; Rutherford, T. and Tarr, D. 1997. Quantifying the Uruguay Round. Economic
Journal,107:1405-1430.
Harrold, R. 1996. Education production and management reforms. The Australian Economic
Review 116(4):409-415.
Hathaway, D.E. and Ingco, M.D. 1995. Agricultural liberalisation and the Uruguay Round.
Paper presented at "The Uruguay Round and Developing Economies Conference", World
Bank, 25-27 January 2005.
Havrylyshyn, O. and Pritchett, L. 1991. European trade patterns after the transition. Policy,
Research and External Affairs Working Paper Series No. 748, World Bank, Washington DC.
Haveman, J. and Hummels, D. 2004. Alternative hypotheses and the volume of trade: The
gravity equation and the extent of specialization. Canadian Journal of Economics, 371: 199-
218.
208
References
Hayes, D.J.; Wahl, T.I. and Williams, C.W. 1990. Testing restrictions on a model of Japanese
meat demand. American Journal of Agricultural Economics 72: 556-566.
Healy, S.; Pearce, R. and Stockbridge, M. 1998. The implications of the Uruguay Round
Agreement on Agriculture for developing countries. Training Materials for Agricultural
Planning 41. Food and Agriculture Organization of the United Nations, Rome.
Heckelei, T.; Britz, W. and Loehe, W. 1998 Recursive dynamic or comparative static solution
for CAPRI. CAPRI Working Paper 97-13, University of Bonn, Bonn.
Helmers, C. and Pasteels, J-M. 2005. Tradesim third version, a gravity model for the
calculation of trade potentials for developing countries and economies in transition.
Unpublished manuscript. Available online at:
http://www.intracen.org/countries/tsim3/tsim3 paper v6.pdf
Helpman, E. and Krugman, P. R. 1985. Market structure and foreign trade: Increasing
returns, imperfect competition and the international economy. Cambridge, Mass: The MIT
Press, 1985
Helpman, E.; Melitz, M. and Rubenstein, Y. 2007. Estimating trade flows: Trading partners
and trading volumes. NBER working paper series, 12927, Cambridge, Massachusetts.
Henry, G.; Cap, E.; Junker, F.; Britz, W.; Meirelles De Souza Filho, H. and Batalha, M. 2006.
EU-Mercosur agriculture competitiveness and trade agreement impacts: Preliminary results
nd
for Argentina and Brazil. Invited paper presented at the 2 ACRALENOS International
Seminar, CEPAL, Santiago de Chile, 9-10 November 2006. Available online at:
http://www.inta.gov.ar/ies/docs/otrosdoc/ Acralenos 06 GHe et al.pdf
Hertel, T.W.; Martin, W.; Yanagashima, K. and Dimaranan, B. 1995. Liberalizing
manufactures trade in a changing world economy. In: Van Tongeren, F. and H. Van Meijl eds.
Review of applied models of international trade in agriculture and related resource and
environmental modeling. FAIR6 CT 98-4148 Interim report No.I, December 1999,
Agricultural Economics Research Institute LEI, The Hague, The Netherlands.
209
References
Hertel, T.W. 1997. Global trade analysis: Modelling and applications, Cambridge University
Press, 1997.
Hertel, T.W.; Brockmeier, M. and Swaminathan, T. 1997. Sectoral and economywide analysis
of integrating Central and East European Countries into the European Union: Implications of
alternative strategies. European Review of Agricultural Economics 1997:24.
Hertel, T.W.; Masters, W.A. and Elbehri, A. 1998. The Uruguay Round and Africa: A global,
general equilibrium analysis. Journal of African Economies 7(2):203-234.
Hertel, T. and W. Martin 1999. Would developing countries gam from inclusion of
manufactures in the WTO negotiations?, Paper prepared for the World Bank's Conference on
Developing Countries and the Millennium Round, WTO Secretariat, Geneva, 19-20
September.
Hertel, T.W.; Anderson, K.; Francois, J.F. and Martin, W. 2000. Agriculture and non-
Agricultural liberalization in the Millennium Round. Centre for International Economic
Studies Policy Discussion Paper No. 0016, University of Adelaide, Australia
Hertel, T.W.; Maros, 1.; Preckel, P.V. and Cranfield, J.A.L. 2003. The earnings effects of
multilateral trade liberalization: Implications for poverty in developing countries. Revised
GTAP Working Paper, March 2003.
Hoekman, B. 2002. Strengthening the Global Trade Architecture for Development: The Post
Doha Agenda. World Trade Review 1:23-45.
Hoekman, B. and Anderson, K. 2000. Developing country agriculture and the new trade
agenda. Economic Development and Cultural Change 49( 1) 171-196
Holden, M. 1996. Economic integration and trade liberalisation in Southern Africa: Is there a
role for South Africa? World Bank Discussion Paper No 342, World Bank, Washington DC,
United States.
210
References
Holt, M.T. and Goodwin, B.K. 1997. Generalized habit formation in an inverse almost ideal
demand system: An application to meat expenditures in the U.S. Empirical Economics
22:293-320.
Huber, R. and Lehmann, B. 2009. WIO agreement on agriculture: Potential consequences for
agricultural production and land use patterns in the Swiss lowlands. Geograjisk Tidsskrift-
Danish Journal of Geography, 1092: 131-145,2009.
International Food Policy Research Institute (IFPRI) 2003. How much does it hurt? The
impact of agricultural trade policies on developing countries. IFPRI, Washington, D.C.,
United States, August 2003.
International Food Policy Research Institute (IFPRI) 2007. Impact of trade liberalization on
agriculture in the Near East and North African. Washington, D.C., United States.
International Trade Centre (ITC) 2000. Tradesim, the ITC simulation model of bilateral trade
potentials. A background paper prepared by the Market Analysis Section of the ITC.
Available online at: http://www.intracen.org/countries/toolpd99/trs tot.pdf
Ingco, M. 1995. Agricultural trade liberalization in the Uruguay Round: One step forward,
one step back. Policy Research Working Paper No. 1500, World Bank, Washington D.C.
Ingco, M., and Townsend, R.F. 1998. Experience and lessons from tie Implementation of
Uruguay Round commitments: Policy options and challenges for African Countries. Paper
presented at the International "Workshop on Agricultural Policy of African Countries and
Multilateral Trade Negotiations: Challenges and Options, 23-26 November, Harare,
Zimbabwe.
Islam, N. 1996. Implementing the Uruguay Round: Increased food price stability by 2020?
International Food Policy Research Institute (IFPRI) 2020 Brief 34.
211
References
Jansen, H.G.P.; Morley, S. and Torero, M. 2007. The impact of the Central America Free
Trade Agreement on agriculture and the rural sector in five Central American countries.
IFPRI Working Paper. Available online at:
http://www.ruta.org/adminlbibliotecaldocumentos/361 EN.pdf
Jenkins, C.; Leape, J. and Thomas, L. 2000. Gaining from trade in Southern Africa:
Complementary policies to underpin the SADC Free Trade Area. MacMillan Press Ltd.
Jensen, H.G.; Frandsen, S.E. and Bach, C.F. 1998. Agricultural and economy-wide effects of
European enlargement: Modelling the Common Agricultural Policy. In: Van Tongeren, F. and
H. Van Meijl eds. Review of applied models of international trade in agriculture and related
resource and environmental modeling. FAIR6 CT 98-4148 Interim report No.l, December
1999, Agricultural Economics Research Institute LEI, The Hague, The Netherlands.
Jooste, A. 2001. Economic implications of trade liberalisation on the South African Red Meat
Industry. Unpublished PhD Thesis, University of the Free State, Bloemfontein, South Africa.
Jooste A.; Van Schalkwyk, H.D. and Von Lampe, M. 2001. The impact of a reduction in
tariffs on the South African Red Meat Industry: A Spatial Partial Equilibrium Approach.
Agrekon 40(4): 811-820
Jooste, A.; Van Schalkwyk, H. and Groenewald, 1. 2003. South African agriculture and
international trade. In: L. Nieuwoudt and J. Groenewald eds. The challenge of change:
agriculture, land and the South African economy, University of Natal Press, 2003.
Kalaba, M.W. 2001. The effects of Free Trade Area on South African agriculture: Impact on
exports. Unpublished MS Thesis, Oklahoma State University, United States.
Karim, A.I. and Kirschke, D. 2002. The impact of the Uruguay Round Agreement on
agriculture on developing countries: A Case study of Sudan's agriculture. Paper presented in
Deutsche Tropen Tags (Conference on Tropical and Subtropical Agricultural and Natural
Resource Management) (DTT), 8-10 October 2002, Kassel-Witzenhausen, Germany.
212
References
Kastens, T. L. and Brester, G.W. 1986. Model selection and forecasting ability of theory-
constrained food demand systems. American Journal of Agricultural Economics 78:301-312.
Kim, M.; Cho, G.D. and Koo, W.W. 2003. Determining bilateral trade patterns using a
dynamic gravity equation. Agribusiness and Applied Economics Report No. 525, November
2003, Center for Agricultural Policy and Trade Studies, Department of Agribusiness and
Applied Economics, North Dakota State University, Fargo, North Dakota.
Kinnucan, H.W.; Xiao, H.; Hsia, C.-J. and Jackson, J. D. 1997. Effects of health information
and generic advertising on U. S. meat demand. American Journal of Agricultural Economics
79:13-23.
Konate, T.M. 1994. The implications of the Uruguay Round on Africa. In: Ingco, M., and
Townsend, R.F. 1998. Experience and lessons from the implementation of Uruguay Round
Commitments: Policy options and challenges for African Countries. Paper presented at the
International Workshop on Agricultural Policy of African Countries and Multilateral Trade
Negotiations: Challenges and Options, 23-26 November, Harare, Zimbabwe.
Koo, W.W.; Karemera, D. and Taylor, R. 1994. A gravity model analysis of meat trade
policies. Agricultural Economics, 10: 81-88.
Korinek, J. and Melatos, M. 2009. Trade impacts of selected Regional Trade Agreements in
Agriculture. OECD Trade Policy Working Papers, No. 87, OECD publishing, © OECD.
doi:10.1787/225010121752. Available online at:
http://www.oecd.org/dataoecd/42/37/42770785.pdf
Kusi, N.K. 2002. Trade liberalization and South Africa's export performance. A paper
presented at 2002 Annual Forum of the Trade and Industry Policy Strategies, Glenburn
Lodge, Muldersdrift, South Africa.
Laaksonen, K. 2008. Free Trade Agreement TOCA between South Africa and the European
Union - An exemplar for the Economic Partnership Agreements. Pellervo Economic Research
Institute Working Papers 112. P 39. ISBN 978-952-224-006-4 PDF.
213
References
LaFrance, J.T. 1998. The bleating of the Lambdas: Comment. American Journal of
Agricultural Economics 80 (1):221-230.
Lairds, S. and A. Yeats 1990. Quantitative methods for trade barriers analysis. New York
New York University Press.
Landes, D. 1998. The wealth and poverty of Nations, Norton, New York.
Lee, J.Y, Seale, J.L. and Jierwiriyapant, J.L. 1990. Do trade agreements help U.S. exports? A
study by Japanese Citrus Industry. Agribusiness: An International Journal, 6: 505 - 514
Lee, J.H. and Brorsen, W. 1993. Non-nested tests and agricultural trade models. In: Kalaba,
M.W. 2001. The effects of Free Trade Area on South African agriculture: Impact on exports.
Unpublished MS Thesis, Oklahoma State University, United States.
Lewis, J.; Robinson, S. and Thierfelder, K. 1999. After the negotiations: Assessing the impact
of Free Trade Agreements in Southern Africa. Trade and Macroeconomics Division
Discussion Paper No. 46. IFPRl, Washington, DC, United States.
Lewis, J. 2001. Reform and opportunity: The changing role and patterns of trade in South
Africa and SADC. Africa Region Working Paper Series No. 14: The World Bank,
Washington, DC, United States.
Lewis, J.; Robinson, S. and Thierfelder, K. 2002. Free Trade Agreements and the SADC
Economies. Africa Region Working Paper Series No. 27. World Bank, Washington DC,
United States. Available online at: http://www.worldbank.org/afr/wps/wp27.pdf
Liapis, P.S. 1990. Incorporating inputs in the Static World Policy Simulation Model
(SWOPSIM). In: Van Tongeren, F. and H. Van Meijl eds. Review of applied models of
international trade in agriculture and related resource and environmental modeling. FAIR6 CT
98-4148 Interim report No.1, December 1999, Agricultural Economics Research Institute
LEI, The Hague, The Netherlands
214
References
Loehe, W. and Britz, W. 1997. EU's regulation 2078/92 in Germany and experiences of
modelling less intensive production alternatives. Institut mr Agrarpolitik, University of Bonn,
Bonn
Longo, R. and Sekkat, K. 2001. New forms of co-operation and integration in emerging
Africa: Obstacles to expanding intra-African trade. Technical paper 169, OECD Development
Centre. Organization for Economic Cooperation and Development. Paris, France.
Loots, E. 2002. Globalisation and economic growth in South Africa: Do we benefit from trade
and financial liberalisation? A paper presented at 2002 Annual Forum of the Trade and
Industry Policy Strategies, Glenburn Lodge, Muldersdrift, South Africa.
Mahé L. and Moreddu, C. 1987. An illustrative trade model to analyse CAP changes:
Unilateral moves and interaction with USA. In: Van Tongeren, F. and H. Van Meijl eds.
Review of applied models of international trade in agriculture and related resource and
environmental modeling. FAIR6 CT 98-4148 Interim report No.I, December 1999,
Agricultural Economics Research Institute LEI, The Hague, The Netherlands.
Mahé L. and Tavéra, C. 1989. Bilateral harmonization of EC and US agricultural policies.
European Review of Agricultural Economics 15:325-348.
Makhura, M. and Mokoena, M.R. 2003. Market access for small-scale farmers in South
Africa. In: L. Nieuwoudt and J. Groenewald eds. The challenge of change: Agriculture, land
and the South African economy, University of Natal Press, 2003.
Manser, M. 1976. Elasticities of demand for food: An analysis using non-additive utility
functions allowing for habit formation. Southern Economic Journal 43:879-901.
Markusen, LR. 1984. Multinationals, multi-plant economies, and the gains from trade'
Journal of International Economics 16:205-226.
Martin, W. and Winters, L.A. 1995. The Uruguay Round and the developing countries,
Cambridge: Cambridge University Press, 1995.
215
References
Mashabela, T.E. and Vink, N. 2008. Competitive performance of global deciduous fruit
supply chains: South Africa versus Chile. Agrekon 472(47):240-245.
Mattoo, A.; Roy, D. and Subramanian, A. 2002. Africa Growth Opportunity Act and its rules
of origin: Generosity undermined? IMF Working Paper No. WP/021l58. Washington, o.C.
Available online at: www.imf.org/external/pubs/ft/wp/2002/wp02158.pdf
Matusz, S. and Tarr, D. 1999. Adjusting to trade policy reform. Policy Research Working
Paper No. 2142, World Bank, Washington DC, United States.
Mátyás, L. 1997. Proper econometric specification of the gravity model. The World Economy
20:363-369.
McCallum, 1. 1995. National borders matter: Canada-US regional trade patterns. American
Economic Review, 853 615-623.
McDonald, S. and Walmsley T. 2001. Bilateral Free Trade Agreements and Customs Unions:
The impact of the EU Republic of South Africa Free Trade Agreement on Botswana. A paper
presented at the AEASA Annual Conference, September 2001, Drakensberg, South Africa.
McDougall, R. and Tyres, R. 1994. Asian expansion and labour-saving technical change:
Factor market effects and policy reactions. American Journal of Agricultural Economics
76(5): 1111-1119.
McKibbin, W. and Wilcoxen, P. 1999. The theoretical and empirical structure of the G-cubed
model. Economic Modelling 16:123-148.
Meyn, M. 2003. The TDCA and the proposed SACU-USA FTA: Are Free Trade Agreements
with Industrialised Countries beneficial for the SACU. Institute for World Economics and
International Management IWIM, University of Bremen
216
References
Mixon, B.1. and Henneberry, S.R.1996. Competitiveness of U.S. processed meat industries in
the Pacific Rim. In: Kalaba, M.W. 2001. The effects of Free Trade Area on South African
Agriculture: Impact on Exports. Unpublished MS Thesis, Oklahoma State University, United
States.
Mokoena, M.R.; Jooste, A. and Alemu, Z.G. 2007. Agricultural exports response to trade
policy reform and implementation in South Africa. A paper presented at the 45th Annual
Conference of the Agricultural Economics Association of South Africa (AEA SA),
Johannesburg, South Africa, 26 - 28 September 2007.
Mokoena, M.R.; Jooste, A. and Alemu, Z.G. 2008. Impacts of the EU-SA TDCA's reciprocal
preferential tariff quotas on market access for cheese and wines. Quarterly Journal of
International Agriculture 47(4): 343-364.
Moschini, G. and Meilke, K. 1989. Modeling the pattern of structural change in US meat
demand. American Journal of Agricultural Economics 71:253-261.
Moschini, G. 1995. Units of measurement and the stone index in demand system estimation.
American Journal of Agricultural Economics 77: 63-68.
Morrissey, O. and Zgovu, E. undated. The impact of Economic Partnership Agreements on
ACP agriculture imports and welfare. Centre for Research in Economic Development and
International Trade CREDIT, University of Nottingham. Available online at:
http://www.nottingham.ac. uk/credi t/documents/papers/07 -09.pdf
Motsumi, L.; Oldfield, M.; Mokoetla, A.; Swart, P. and De Beer, B.C. 2008. Note on the
revision of South Africa's nominal and real effective exchange rate indices. South African
Reserve Bank Quarterly Bulletin December 2008.
Mountain, D. C. 1988.' The Rotterdam model: An approximation In variable space.
Econometrica 56:477-484.
Mutume, G 1998. What does Clinton hold for Pretoria? In: Nouve, K. and Staatz, J. 2003. The
African Growth and Opportunity Act and the latent agricultural export response in Sub-
217
References
Saharan Africa, a paper presented at the 2003 Annual Meeting of the American Agricultural
Economics Association, Montreal, Quebec, Canada, July 27-30
National Agricultural Marketing Council (NAMC) 1999. Report on the investigation into
market access. Report on Phase 1, February 1999.
National Agricultural Marketing Council (NAMC) 2008. International TradeProbe. March
2008.
Nin-Pratt, A.; Diao, X. and Bahta, Y. 2008. Assessing potential welfare impacts on
agriculture of a Regional Free Trade Agreement in Southern Africa. Regional Strategic
Analysis and Knowledge Support System ReSAKSS Working Paper No. 15, November, 2008.
Available online at: http://pdf.usaid.gov/pdf docs/PNADS617.pdf
Ndirangu, N. 2002. Africa worse off after Agreementon Agriculture. A paper contributed to a
Roundtable on Food and Trade: The WTO Development Change held on 4 - 5 November
2002, Ottawa, Canada.
Nouve, K. and Staatz, 1. 2003. The African Growth and Opportunity Act and the latent
agricultural export response in Sub-Saharan Africa, a paper presented at the 2003 Annual
Meeting of the American Agricultural Economics Association, Montreal, Quebec, Canada,
July 27-30.
Nyhodo, B.; Punt, C. and Vink, N. 2009. The potential impact of the Doha Development
Agenda on the South African economy: Liberalising OECD agriculture and food trade.
Agrekon 48(1):35-39.
Nyirabu, M. 2004. Appraising regional integration III Southern Africa. African Security
Review 13(1):21-32.
Organisation for Economic Co-operation and Development (OECD) 1998a. Agricultural
policies in OECD countries: Monitoring and evaluation 1998. Paris, France.
218
References
Otsuki, T.; Wilson, J.S. and Sewadeh, M. 2001. Saving two in a billion: Quantifying the trade
effect of European food safety standards on African exports. Food Policy 26 (2001):495-514.
Oyewumi, O.A.; Jooste, A.; Britz, W.; Van Schalkwyk, H.D. 2007. Tariff and tariffrate quota
liberalization in the South African livestock industry: Approaches to welfare measurements.
Agrekon 46(1): 1-22.
Panagariya, A. 2002. Trade and Food Security: Conceptualizing the linkages. Paper prepared
for presentation at the Conference on Trade, Agricultural Development, and Food Security:
The Impact of Recent Economic and Trade Policy Reform, Food and Agricultural
Organization, Rome, July 11-12.
Peacemaker-Arrand, B. 2004. The impact of a New WTO Agricultural Agreement on cereals
markets in Sub-Saharan Africa. An Honours Thesis for the Department of Economics, Tufts
University.
Penzhom, N. and Kirsten, J.F. 1999. The impact of the EU FTA on South African agriculture:
A general equilibrium analysis. Agrekon 38(4):788-804.
Perroni, C. and Wigle, R. M. 1997. Environmental policy modelling. In: Hertel, T.W ed.
Global trade analysis: Modeling and applications, Cambridge University Press, Cambridge
Pesaran, M.H. 1974. On the general problem of model selection. Review of Economic Studies
41:153-171.
Pollak, R.A. and Wales, T.J. 1969. Estimation of the linear expenditure system.
Econometricia 37:611-628.
Pollak, R.A. and Wales, T.J. 1992. Demand system specification and estimation. New York
and Oxford: Oxford University Press, 1992.
PooIe, M.S. 2003. African Group comes looking for business. The Atlanta Journal-
Constitution, 2002 (August):E6.
219
References
Poonyth, D.; Esterhuizen, D.; Ngqangweni, S. and Kirsten J.F. 2002. Trade policies and
agricultural trade in the SADC region: Challenges and implications. South Africa Country
Study, University of Pretoria, South Africa.
Poonyth, D. and Sharma, R. 2003. The impact of the WTO negotiating modalities on
Southern African Development Community countries. FAO, May 2003.
Poonyth, D., Sharma, R.; Konandreas, P. 2004. The impact of the WTO agricultural
negotiating modalities on Southern African Development Community SADC countries.
Agrekon, 43(3):276-296.
Pëyhënen, P. 1963. A tentative model for the volume of trade between countries. In: Nouve,
K. and Staatz, J. 2003. The African Growth and Opportunity Act and the latent agricultural
export response in Sub-Saharan Africa, a paper presented at the 2003 Annual Meeting of the
American Agricultural Economics Association, Montreal, Quebec, Canada, July 27-30
Pustovit, N. and Schrnitz, P.M. 2003. Impact of agricultural protection in OECD countries on
South African agriculture. A paper presented at the zs" Conference of the International
Association of Agricultural Economics, 16 - 22 August 2003, Durban South Africa.
Radelet, S. 1997. Regional integration and cooperation in sub-Saharan Africa: are formal
trade agreements the right strategy? Development discussion paper No. 592, Harvard Institute
for International Development, Cambridge, Mass.
Rae, A.N. and Strutt, A. 2003. The current round of agricultural trade negotiations: Should we
bother about domestic support? The Estey Center Journal of International Law and Trade
Policy 4(2):98-122.
Raghavan, C. 2000. Africa: NGOs start campaigns against US AGOA. Third World Network.
Available online at: www.twnside.org.sg/title/agoa.htm
220
References
Reeder, 1.; Torene, 1.A.; Jabara, C. and Babula, R.A. 2005. Regional Trade Agreements:
Effects of the ANDEAN and MERCOSUR pacts on the Venezuelan soybean trade and US
exports. Office of Industries Working Paper. U.S. International Trade Commission, January
2005. Available online at: http://www.usitc.gov/publications/332/working papers/ID-
Ilsoybean.pdf
Richter, 1.1994. Austria and the single market. Economic Systems Research 6(1):77-90.
Roberts, R. 2000. Understanding the effects of trade policy reform: The case of South Africa,
The South African Journal of Economics 68(4):607-638.
Roningen, V.O. 1986. A Static World Policy Simulation (SWOPSIM) Modeling Framework.
In: Van Tongeren, F. and Van Meijl H. eds. Review of applied models of international trade
in agriculture and related resource and environmental modeling. FAIR6 CT 98-4148 Interim
report No.1, December 1999, Agricultural Economics Research Institute LEI, The Hague, The
Netherlands
Roningen, V.O.; Dixit, P.M. and Seeley, R. 1999. Agricultural outlook in the Year 2000:
Some alternatives. In: Maunder, A. and Valdes, A. eds. Agriculture and governments in an
interdependent world. Dartmouth, England, 1990
Rose, K. A. and Van Wincoop, E. 2001. National money as a barrier to trade: The real case
for currency union. American Economic Review, 912: 386-390.
Rose, K. A. 2002. Do we really know that the WTO increases trade? NBER Working Paper
9273, October 2002. Available online at: www.nber.org/papers/w9273
Salarnon, P. 1998. Impacts of different policy options on the EU dairy market: In: Van
Tongeren, F. and Van Meijl, H. eds. Review of applied models of international trade in
agriculture and related resource and environmental modeling. FAIR6 CT 98-4148 Interim
report No.1, December 1999, Agricultural Economics Research Institute LEI, The Hague, The
Netherlands
221
References
Sandrey, R. 2006. Trade creation and trade diversion resulting from SACU trading
agreements. Tra/ac Working Paper no. 11/2006, Tralae, South Africa, August 2009.
Sandrey, R. and Jensen, H.G. 2009. SACU and Mercosur: The implication of a Free Trade
Agreement for Botswana, Lesotho, Namibia and Swaziland. In: Bësl A, G. Erasmus, T.
Hartzenberg and McCarthy, C. eds: Monitoring regional integration in Southern Africa,
Yearbook Volume 9 2009, Tralae, South Africa.
Sandrey, R. and Jensen, H.G. 2009. SACU, China and India: The implication of FTAs for
Botswana, Lesotho, Namibia and Swaziland BLNS Tra/ac Working paper No 1/2009.
Sandrey, R. 2010. An assessment of the Trade, Development and Cooperation Agreement
(TDCA). In: Bësl A.; Du Pisane, A.; Erasmus, G.; Hartzenberg, T. and Sandrey, R. eds:
Monitoring regional integration in Southern Africa, Yearbook Volume 10 2010, Tralae, South
Africa.
Sandrey, R. 2011. Intra-African trade in Southern and Eastern Africa and the role of South
Africa. In: Bësl A.; Du Pisane, A.; Erasmus, G.; Hartzenberg, T. and Sandrey, R. eds:
Monitoring regional integration in Southern Africa, Yearbook Volume 102010, Tralae, South
Africa.
Sanso, M.; Rogelio, C. and Sanz, F. 1993. Bilateral trade flows, the gravity equation, and
functional form. The Review of Economics and Statistics, 75:266-275.
Scheepers, S.; Jooste, A. and Alernu, Z.G. 2007. Quantifying the impact of phytosanitry
standards with specific reference to MRLs on the trade flow of South African avocados to the
EU. Agrekon 46(2):260-273.
Schiff, M. and Valdes, A. 1992. The political economy of agricultural pricing policy: A
synthesis of the economics in developing countries. Baltimore, Johns Hopkins University
Press
Sharma, R.; Konandreas, P. and Greenfield, 1. 1996. An overview of assessments of the impact
of the Uruguay Round on agricultural prices and incomes. Food Policy 21(4/5):351-363.
222
References
Short, C. 2003. The dangers to Doha: The risks of failure in the trade round. Royal Institute of
International Affairs, London, March 25, 2003
Soledad Bos, M. 2003. The impact of maize subsidies on Sub-Saharan Africa. M.P.P. Thesis,
University of California-Berkeley, United States.
Soloaga, 1. and Winters, A.L. 2001. Regionalism in the nineties: What effect on trade. The
North American Journal of Economics and Finance 121:1-29.
Sorsa, P. 1995. The burden of Sub-Saharan African own commitments in the Uruguay Round-
myth or reality? IMF Working Paper WP/95/48-EA, International Monetary Fund
Washington DC, United States.
Sparks, A.L.; Seale, J.L. and Buxton, B.M. 1990. Fresh import demand: Four markets for U.S.
fresh apples. Agricultural Economics. Report No 641, USDA/Economic Research Service,
Washington DC.
Stigier, G. J. and Becker, G. 1977. De gustibus non est disputatum. In: Xiao, H., Kinnucan,
H.W. and Kaiser, H.M. 1998. Advertising, structural change, and U.S. non-alcoholic drink
demand. National Institute for Commodity Promotion Research and Evaluation (NICPRE)
98-01, Cornell University, New York, United States.
Stevens, C.; Greenhill, R.; Kennan, J. and Devereux, S. undated. The WTO Agreement on
Agriculture and Food Security. Available online at:
http://www.observatori.org/documents/The WTO agreement on Agriculture and
Food Security.pdf
Subramanian, A. and Tamiriza, N. 2001. Africa's trade revisited. IMF Working Paper No
01133, March 2001.
Siileyman, T.O. 2010. What determines intra-EU trade? The gravity model revisited.
International Research Journal of Finance and Economics 39 (2010): 244 - 250
223
References
Suparmoko, M. 2002. The impact of the WTO Agreement on Agriculture in the rice sector.
Paper presented at the Workshop on Integrated Assessment of the WTO Agreement on
Agriculture in the Rice Sector, Geneva, Switzerland, 5thApril 2002.
Surry, Y. 1989. The 'Constant Difference of Elasticities' (CDE) functional form: A neglected
alternative. In: Van Tongeren, F. and Van Meijl H. eds. Review of applied models of
international trade in agriculture and related resource and environmental modeling. FAIR6 CT
98-4148 Interim report No.l, December 1999, Agricultural Economics Research Institute
LEI, The Hague, The Netherlands.
Tamirisa, N. 1999. Exchange and capital controls as barriers to trade. IMF staff papers, 461:
69-88
Tangermann, S. and Josling, T.E. 1994. Pre-accession agricultural policies for Central Europe
and the European Union, Study commissioned by DG I of the European Commission,
Gëttingen.
Theil, H. 1965. The information approach to demand analysis. Econometrica 33 :67-87.
Theil, H. 1975. Theory and measurement of consumer demand. Volume 1 and 2. In: Barnett,
W.A. and Seck, O. 2008. Rotterdam model vs. almost ideal demand system: Will the best
demand specification please stand up? Journal of Applied Econometrics 23(6):795-824.
Theil, H. 1980. System-wide explorations in international economics, input-output analysis,
and marketing research .. In: Xiao, H., Kinnucan, H.W. and Kaiser, H.M. 1998. Advertising,
structural change, and U.S. non-alcoholic drink demand. National Institute for Commodity
Promotion Research and Evaluation (NICPRE) 98-01, Cornell University, New York, United
States.
Thomas, V. and J. Nash, J. 1992. Trade policy reform: Recent evidence from theory and
practice. In: R. Adhikar, R.; Kirkpatrick, C. and Weiss, 1. eds. Industrial and trade policy
reform in developing countries. Manchester: Manchester University Press.
224
References
Tinbergen, 1. 1962. Shaping the World Economy: Suggestions for an international economic
policy. New York: The Twentieth Century Fund.
Toosi, M.; Moghaddasi, R.; Yazdani, A. and Ahmadian, M. 2009. Regionalism and its effects
on Iranian agricultural exports: The case of Economic Cooperation Organization. American
Journal of Applied Sciences 67:1380-1384.
Tsigas, M.E. and Ingco, M. 2001. Market access liberalization for food and agricultural
products: A general equilibrium assessment of tariff-rate quotas. Us. International Trade
Commission, Office of Economics Working Paper No. 2001-10-A, October.
Tsikata, Y.M. 1999. Southern Africa: Trade liberalization and implication for a Free Trade
Area. TIPS 1999 Annual Forum, Muldersdrif, South Africa, 19-22 September 1999.
Tsolo M.; Mogotsi, LB. and Motlaleng, G. 2010. The impact of European Union - South
Africa Trade, Development and Cooperation Agreement on Botswana, Lesotho, Namibia and
Swaziland. Review of Economics and Business Studies 3(1): 129-148.
United Nations Conference on Trade and Development (UNCTAD) 1994b. Trade-related
investment measures. In UNCTAD 1994. The outcome of the Uruguay Round: An initial
assessment. Supporting papers to the Trade and Development Report 1994, Geneva,
Switzerland.
Van Schalkwyk, H.D. 1997. Agricultural trade policy. Working document, University of the
Free State, Bloemfontein, South Africa.
Van Tongeren, F. and Van Meijl, H. 1999. Review of applied models of international trade in
agriculture and related resource and environmental modeling. FAIR6 CT 98-4148 Interim
report No.l, Agricultural Economics Research Institute LEI, The Hague, The Netherlands.
Van Tongeren, F.; Van Meijl, H. and Surry, Y. 2001. Global models applied to agricultural
and trade policies: a review and assessment. Agricultural Economics 26: 149-172.
225
References
Victorio, A.G. and Rungswang, M. 2008. The effect of a Free Trade Agreement upon
agricultural imports. International Journal of Human and Social Sciences 3:42008
Vink, N, Tregurtha, N and Kirsten, IF. 2002. South Africa's changing agricultural, food and
beverage imports: Implications for SADC Suppliers, A Report to the World Bank.
Von Lampe, M. 1998. The World Agricultural Trade Simulation System (WATSIM). An
overview. In: Van Tongeren, F. and Van Meijl, H. eds. Review of applied models of
international trade in agriculture and related resource and environmental modeling. FAIR6 CT
98-4148 Interim report No.1, December 1999, Agricultural Economics Research Institute
LEI, The Hague, The Netherlands
Von Lampe, M. 1999. A modelling concept for the long-term projection and simulation of
agricultural world market developments: World Agricultural Trade Simulation Model
(WATSIM). ISBN 3-8265-6725-0: Doktor der Agrarwissenschaften Dissertation. University
of Bonn, Shaker Verlag.
Wainio, J.; Gehlhar, M. and Dyck, J. 2011. Selected trade agreements and implications for
U.S. agriculture. ERR-115, U.S. Department of Agriculture, Economic Research Service.
Apri12011. Available online at: http://www.ers.usda.gov/Publications/ERRl15/ERRl15.pdf
Wall, HJ. 1999. Using the gravity model to estimate the costs of protection. Federal Reserve
Bank ofSt. Louis Working Paper, January/February 1999: 33-40.
Wall, HJ. 2000. Gravity model specification and the effects of the Canada-U.S. border.
Federal Reserve Bank ofSt. Louis Working Paper, 2000-024A
Wang, Z.K.and Winters, L.A. 1998. Africa's role in multilateral trade negotiations: Past and
future. Journal of African Economies, June 1998.
Wei, SJ. and Parsley, D.C. 1995, Purchasing power disparity; exchange volatility, trade
barriers, and other culprits. NBER Working Paper No. 5032, Cambridge, Massachusetts:
National Bureau of Economic Research.
226
References
Winters, L.A. 1984. Separability and the specification of foreign trade function. Journal of
International Economics 17(1984):239-263.
Wooldridge, J.M. 2002. Econometric analysis of cross section and panel data. The MIT Press,
Cambridge, Massachusetts, London, England.
Wooldridge, J.M. 2003. Introductory econometrics: A Modern Approach. 2nd Edition,
Michigan State University, USA.
Weston, A. 1994. The Uruguay Round: Unravelling the implications for the Least-Developed
Countries and Low-Income Countries. UNCTAD.
Whalley, J. 2000. What can developing countries infer from the Uruguay Round models for
future negotiations. Policy Studies in International Trade and Commodities Study Series No.
4, UNCTAD/ITCD/T AB/6.
World Trade Organization (WTO) 2001. Market access: Unfinished business - Post Uruguay
inventory and issues. WTO Special Studies 6 2001, Geneva.
World Trade Organization (WTO) Secretariat 2002. Regional trade integration under
transformation. Seminar Paper, Geneva, 26 April 2002.
World Trade Organization (WTO) 2011. Regional Trade Agreements (RTAs). Available
online at: http://www.wto.org/englishltratope/regione/regione.htm
Yang, S. and Koo, W.W. 1994. Japanese meat import demand estimation with the source
differentiated AIDS model. Journal of Agricultural resource Economics 19:396-408
Xiao, H. 1997. Case studies in advertising effectiveness. In: Xiao, H., Kinnucan, H.W. and
Kaiser, H.M. 1998. Advertising, structural change, and U.S. non-alcoholic drink demand.
National Institute for Commodity Promotion Research and Evaluation (NICPRE) 98-01,
CornelI University, New York, United States.
227
References
Xiao, H.; Kinnucan, H.W. and Kaiser, H.M. 1998. Advertising, structural change, and U.S.
non-alcoholic drink demand. Research Paper No. 98-01, National Institute for Commodity
Promotion Research and Evaluation, Cornell University, Ithaca, New York.
Xiao, H, H.W. Kinnucan, and H.M. Kaiser. Effects of advertising on U.S. non-alcoholic
beverage demand: Evidence from a Rotterdam model. American Agricultural Economics
Association Annual Meetings. Salt Lake City, UT, August 1998.
Yeboah, O-A.; Shaik, S. and Batson, S. 2009. The trade effects of MERCOSUR and the
Andean Community on U.S. cotton exports to CBI countries. Selected paper presented at the
Southern Agricultural Economics Association Annual Meeting, Atlanta, Georgia, January 31-
February 3, 2009. Available online at:
http://ageconsearch.umn.edu/bitstrearn/46028/2/Y eboah Batson SAEA 2009.pdf
Zahniser, S. and Link, 1. 2002. Effects of North American Free Trade Agreement on
agriculture and the rural economy. Economic Research Service, USDA. Available. online at:
http://www.ers.usda.gov/publications/wrs0201/wrs020 1fm.pdf
Zgovu, E.K. and J.P. Kweka 2006. Empirical analysis of tariff line-level trade and welfare
effects of reciprocity under an Economic Partnership Agreement with the EU: Evidence from
Malawi and Tanzania. Paper presented at the African Economic Research Consortium AERC
Biannual Research Workshop, Nairobi, Kenya, 27 May - 01 June 2006.
Zwinkels, R.C.J. and Beugelsdijk, S. 2010.Gravity equations: Workhorse or Trojan horse in
explaining trade and FDI patterns across time and space? International Business Review 19
(2010):102-115
228
APPENDICES
Appendix SA: Selection of the Estimator suitable for Agricultural Exports from
South Africa to the World
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Wald Test Period Impact 1190 5 3.69* OLS No 4 22.44* OLS No
Statistic FEor RE Yes 1309 FE or RE Yes
Yearly Impact 1190 9 3.67* OLS No 1309 8 22.76* OLS No
FE or RE Yes FEor RE Yes
Durbin Period Impact 1190 5 -0.62 FE-no auto Yes 1309 4 1.46 FE-no auto No
Watson - - - FE-auto - 1190 4 1.89** FE-auto Yes
Statistic (for 1190 6 -1.31 RE-no auto Yes 1309 6 1.21 RE-no auto No
static) or - - - RE-auto - 1190 6 1.81 RE-auto No
Durbin-H Yearly Impact 1190 9 -0.56 FE-no auto Yes 1309 8 1.46 FE-no auto No
Statistic (for - - - FE-auto - 1190 8 1.92** FE-auto Yes
dynamic) 1190 10 -1.34 RE-no auto Yes 1309 9 1.22 RE-no auto No
- - - RE-auto - 1190 9 1.84 RE-auto No
Hausman Period Impact N/A 5 471.79" FE Yes - - - FE -
Test RE No RE -
Statistic Yearly Impact N/A 9 464.07" FE Yes - - - FE -
RE No RE -
NB: " •• & ••• denote significance at the I, 5 and 10 percent levels respectively. N & K denote the South Africample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto denote estimation
assuming no autocorrelation problem and estimation with 'the correction of I" order autocorrelation_j)_roblcm re~ective'!y:
Appendix SB: Suitable equations for agricultural exports from South Africa to the
World.
229
Appendices
Appendix SC: Selection of the Estimator suitable for Agricultural Imports from the
World to South Africa
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Wald Test Period Impact 1270 5 4.27" OLS No 1397 4 22.22" OLS No
Statistic FE or RE Yes FE or RE Yes
Yearly Impact 1270 9 4.25* OLS No 1397 8 22.38* OLS No
FE or RE Yes FE or RE Yes
Ourbin Period Impact 1270 5 0.05 FE-no auto Yes 1397 4 1.71 FE-no auto No
Watson - - - FE-auto - 1270 4 1.95** FE-auto Yes
Statistic (jar 1270 6 -1.84*** RE-no auto No 1397 6 I.I3 RE-no auto Nostatic) or
Durbin-H 1143 6 -0.25
RE-auto Yes 1270 6 2. I0** RE-auto Yes
Statistic (jar Yearly Impact 1270 9 0.05 FE-no auto Yes 1397 8 1.70 FE-no auto No
dynamic) - - - FE-auto - 1270 8 1.95** FE-auto Yes
1270 ID - I .86** RE-no auto No 1397 9 1.13 RE-no auto No
1143 ID -0.25 RE-auto Yes 1270 9 2. I I·· RE-auto Yes
Hausman Period Impact N/A 5 904.36* FE Yes N/A 4 1.66 FE No
Test RE No RE Yes
Statistic Yearly Impact N/A 9 931.51* FE Yes N/A 9 4.00 FE No
RE No RE Yes
NB: ", ** & *** denote significance at the 1,5 and 10 percent levels respectively. N & K denote the sample size and the number of regressors respectively.
OlS, FE & RE denote Pooled Ordinary least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto denote estimation assuming no
autocorrelation problem and estimation with the correction of I" order autocorrelation problem respectively.
Appendix SD: Suitable equations for agricultural imports from the World to South
Africa
Variables
Constant
InGDPPCit
098
099
From 1994 to 2004, SA . agricu products 127 coun in the World: AGO, ALB, ARE, A
AUS, AUT, BOl, BEL, BEN, BGO, BGR, BHR, BHS, BOL, BRA, BTN, CAN, CHE, CHL, CHN, CIV, CMR, COG, COL, COM,
CRI, CYP, ClE, DEU, OMA, ONK, DOM, ORC, ECU, EGY, ESP, EST, HH, FIN, FRA, GBR, GHA, GIN, GMB, GRC, GRD,
GTM, GUY, HRV, H1"I, HUN, ION, fNO, IRL, IRN, ISL, ISR, ITA, JAM, JOR, JPN, KEN, KGl, KOR, KWT, LAO, LBN, LKA,
LUX, MAR, MOG, MEX, MU, MOl, MRT, MUS, MWI, MYS, NER, NGA, NIC, NLO, NOR, NPL, NZL, OMN, PAK, PAN, PER,
PHL, POL, PRI, PRT, PRY, ROM, RUS, RWA, SAU, SON, SEN, SGP, SLE, SLV, STP, SUR, SVK, SVN, SWE, SYC, SYR, TCO,
TGO THA UGA URY VCT 2MB and
230
Appendices
Appendix SE: Selection of the Estimator suitable for Agricultural Trade between
South Africa and the World
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Wald Test Period Impact 1090 4 4.60' OLS No 1199 3 37.51 OLS No
Statistic FE or RE Yes FE or RE Yes
Yearly Impact 1090 8 4.54" OLS No 1199 7 37.81 OLS No
FE or RE Yes FE or RE Yes
Durbin Period Impact 1090 4 0.71 FE-no auto Yes 1199 3 1.46 FE-no auto No
Watson - - - FE-auto - 1090 3 1.83 FE-auto No
Statistic (jar 1090 5 0.08 RE-no auto Yes 1199 4 1.33 RE-no auto No
static) or - - - RE-auto - 1090 4 1.87 RE-auto No
Ourbin-H Yearly Impact 1090 8 0.74 FE-no auto Yes 1199 7 1.45 FE-no auto No
Statistic (jar - - - FE-auto - 1090 7 1.84 FE-auto No
dynamic) 1090 9 0.04 RE-no auto Yes 1199 8 1.33 RE-no auto No
- - - RE-auto - 1090 8 1.90 RE-auto ?
Hausman Period Impact N/A 4 462.25· FE Yes - - - FE -
Test RE No RE -
Statistic Yearly Impact N/A 8 453.22* FE Yes - - - FE -
RE No RE -
NB: *, u & ••• denote significance at the 1,Sand lO percent levels respectively. N & K denote the sample size and the number of regressors respectively.
OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto denote estimation assuming no
autocorrelation problem and estimation with the correction of I" order autocorrelation problem respectively. ? means inconclusive.
Appendix SF: Suitable equations for agricultural trade between South Africa and
the World
Variables
Constant
InYijt_1
InGOPPCijt
09599
095
096
1994 to SA traded ral products ( plus with the countries in the World:
ARE, ARG, ATG, AUS, AUT, BOl, BEL, BEN, BGR, BHR, BHS, BRA, CAN, CHE, CHL, CHN, CIV, CMR, COG, COL, COM,
CRl, CYP, ClE, DEU, DOM, ONK, ORC, EGY, ESP, ETH, FIN, FRA, GBR, GHA, GIN, GMB, GRC, HUN, ION, INO, IRL, IRN,
ISL, ISR, ITA, JAM, JOR, JPN, KEN, KOR, KWT, LBN, LKA, LUX, MAR, MOG, MEX, MU, MOl, MRT, MUS, MWI, MYS,
NER, NGA, NLD, NOR, NZL, OMN, PAK, PAN, PER, PHL, POL, PRI, PRT , PRY, ROM, RUS, RWA, SAU, SON, SEN, SGP,
SLE, STP, SUR, SVK, SVN, SWE, SYC, SYR, TCO, TGO, THA, TTO, TUN, TUR, TlA, UGA, URY, USA, VCT, VEN, VNM,
2MB lWE
231
Appendices
Appendix SG: Selection of the Estimator suitable for Cheese Exports from South
Africa to the World
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Wald Test Period Impact 320 5 2.42** ~O~LS~~,_~~~No 352 4 7.68* OLS No
Statistic Yes I-::.:-::::--:----;::-:::;--t--:-:-:'----lFE or RE FEar RE Yes
Yearly Impact 320 9 2.39** OLS No 352 8 7.73* OLS No
Yes I-::.:-:::=--;::-:::;--t--:-:FEar RE FE or RE Y-:e'-s---l
Durbin Period Impact 320 5 -0.10 FE-no auto Yes 352 4 1.49 FE-no auto No
Watson FE-auto 320 4 1.97** FE-auto Yes
Statistic (jar 320 6 -1.28 RE-no auto Yes 352 6 0.83 RE-no auto No
static) or RE-auto 320 6 1.97** RE-auto Yes
Durbin-H Yearly Impact 320 9 -0.06 FE-no auto Yes 352 8 1.46 FE-no auto No
Statistic (jar FE-auto 320 8 1.97** FE-auto Yes
dynamic) 320 10 -1.25 RE-no auto Yes 352 9 0.82 RE-no auto No
RE-auto 320 9 1.97** RE-auto Yes
Hansman Period Impact N/A 5 172.99* FE Yes N/A 4 3.44 ~F=E~R~E------+---N~o~~ -----~~~~NoTest RE Yes
Statistic Yearly Impact N/A 9 168.02* FE Yes N/A 9 -15.38*** FE Yes
~RE~-----r~~~ No ~RE~-----_'--~N~o~~
NB: *, ** & *** denote significance at the I, 5 and 10percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of I SI order autocorrelation problem respectively.
Appendix SH: Suitable equations for cheese exports from South Africa to the
World.
232
Appendices
Appendix 51: Selection of the estimator suitable for cheese imports from the World
to South Africa
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Wald Test Period Impact 240 5 2.81 ** OLS No 264 4 15.08* OLS No
Statistic FE or RE Yes FE or RE Yes
Yearly Impact 240 9 2.85** OLS No 264 8 15.28* OLS No
FE or RE Yes FE or RE Yes
Durbin Period Impact 240 5 0.14 FE-no auto Yes 264 4 1.33 FE-no auto No
Watson - - - FE-auto - 240 4 1.89** FE-auto Yes
Statistic (jar 240 6 -0.45 RE-no auto Yes 264 6 0.68 RE-no auto No
static) or - - - RE-auto - 240 6 1.97** RE-auto Yes
Durbin-H Yearly Impact 240 9 0.22 FE-no auto Yes 264 8 1.31 FE-no auto No
Statistic (jar - - - FE-auto - 240 8 1.93** FE-auto Yes
dynamic) 240 lO -0.50 RE-no auto Yes 264 9 0.68 RE-no auto No
- - - RE-auto - 240 9 1.98** RE-auto Yes
Hausman Period Impact N/A 5 142.99* FE Yes N/A 4 -2.04 FE No
Test RE No RE Yes
Statistic Yearly Impact N/A 9 109.53* FE Yes N/A 9 1.61 FE No
RE No RE Yes
NB: *, ** & *** denote significance at the 1,5 and lO percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of I st order autocorrelation problem respectively.
Appendix SJ: Suitable equations for cheese imports from the World to South Africa
Variables
Constant
InYijt_1
InGDPPCit
InGDPPCjt
D9599
D95
D96
D97
D98
D99
233
Appendices
Appendix 5K: Selection of the estimator suitable for cheese trade between South
Africa and the World
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Wald Test Period Impact 100 4 6.32* OLS No 110 3 3 I .86* OLS No
Statistic FEor RE Yes FE or RE Yes
Yearly Impact 100 8 6.43* OLS No 110 7 33.94* OLS No
FE or RE Yes FE or RE Yes
Durbin Period Impact 100 4 0.77 FE-no auto Yes 110 3 1.95** FE-no auto Yes
Watson - - - FE-auto - - - - FE-auto -
Statistic (for 100 5 -1.33 RE-no auto No 110 4 1.52 RE-no auto No
sialic) or - - - RE-auto Yes 100 4 1.97** RE-auto Yes
Durbin-H Yearly Impact 100 8 1.05 FE-no auto Yes 110 7 1.94** FE-no auto Yes
Statistic (for - - - FE-auto - - - - FE-auto -
dynamic) 100 9 - I .36 RE-no auto No 110 8 1.53 RE-no auto No
- - - RE-auto Yes 100 8 1.98** RE-auto Yes
Hausman Period Impact N/A 4 87.92* FE Yes N/A 3 4.34 FE No
Test RE No RE Yes
Statistic Yearly Impact N/A 8 88.30* FE Yes N/A 7 21.39* FE Yes
RE No RE No
NB: *, ** & *** denote' significance at the 1,5 and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of IS!order autocorrelation problem respectively.
Appendix 5L: Suitable equations for cheese trade between South Africa and the
World
Variables Period Impact Yearly Impact
Dynamic FE Static RE Dynamic FE Static FE
Constant - cI5.·15 - -
(-1.26)
InYijt_t 0.05 - 0.05 -
(045) (046)
InGDPPCijt 3.83 8.74· 5.00 4.61
(096) (2.92) (1.22) ( 1.34)
REERt 0.83 1.50 0.11 0.80
(0.77) (l.48) (0.08) (1.02)
09599 0.35 0.45 - -
(0.68) (113)
095 - - 0.45 -0.02
(0.45) (-0.04)
096 - - 1.50··· 1.11··
(175) (229)
097 - - 1.26 0.93···
(1.52) ( 1.94)
D98 - - 0.29 0.06
(0.43) (0 13)
099 - - 0.29 0.10
(0.51 ) (023)
InDlSTij - -7.43·· -
('2.52)
Adjusted R2 0.83 0.47 0.83 0.83
Observations 100 100 100 110
Cross-Sections 10 10 10 10
., •• & ••• denote significance at the 1.5 and ID percent levels respectively. t-values are in parentheses
From 1994 to 2004, SA traded cheese (imports plus exports) with the following 10 countries in the World: CI'IE, FRA, DEU, ESP,
FRA,GBR, GRC, MOZ, NLD, USA and ZWE
234
Appendices
Appendix SM: Selection of the estimator suitable for cut flowers exports from South
Africa to the World
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Wald Test Period Impact 560 5 3.08' OLS No 616 4 15.90* OLS No
Statistic FEor RE Yes FE or RE Yes
Yearly Impact 560 9 2.99* OLS No 616 8 15.96· OLS No
FE or RE Yes FE or RE Yes
Durbin Period Impact 560 5 -0.32 FE-no auto Yes 616 4 1.49 FE-no auto No
Watson - - - FE-auto - 560 4 2.00*' FE-auto Yes
Statistic (for 560 6 -1.03 RE-no auto Yes 616 6 1.07 RE-no auto No
static) or - - - RE-auto - 560 6 2.05" RE-auto Yes
Durbin-H Yearly Impact 560 9 -0.28 FE-no auto Yes 616 8 1.50 FE-no auto No
Statistic (for - - - FE-auto - 560 8 2.01** FE-auto Yes
dynamic) 560 10 -1.18 RE-no auto Yes 616 9 1.09 RE-no auto No
- - - RE-auto - 560 9 2.05** RE-auto Yes
Hausman Period Impact N/A 5 252.03' FE Yes N/A 4 -3.33 FE No
Test RE No RE Yes
Statistic Yearly Impact N/A 9 235.35' FE Yes N/A 9 -9.54 FE No
RE No RE Yes
NB: " >ti, & * •• denote significance at the I, 5 and 10percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of Ist order autocorrelation problem respectively.
Appendix SN: Suitable equations for cut flowers exports from South Africa to the
World.
Variables
Constant
InYijt_1
InGDPPCit
InGDPPCjt
D9599
D95
096
097
098
099
235
Appendices
Appendix 50: Selection of the estimator suitable for cut flowers imports from the
World to South Africa
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Wald Test Period Impact 260 5 3.14* OLS No 286 4 10.21* OLS No
Statistic FE or RE Yes FE or RE Yes
Yearly Impact 260 9 3.28* OLS No 286 8 10.39* OLS No
FE or RE Yes FEar RE Yes
Durbin Period Impact 260 5 -0.05 FE-no auto Yes 286 4 1.74 FE-no auto No
Watson - - - FE-auto - 260 4 1.90** FE-auto Yes
Statistic (jar 260 6 -1.58 RE-no auto Yes 286 6 0.88 RE-no auto No
static) or - - - RE-auto - 260 6 1.95** RE-auto Yes
Durbin-H Yearly Impact 260 9 -0.06 FE-no auto Yes 286 8 1.74 FE-no auto No
Statistic (jar - - - FE-auto - 260 8 1.93** FE-auto Yes
dynamic) 260 10 -1.54 RE-no auto Yes 286 9 0.89 RE-no auto No
- - - RE-auto - 260 9 1.96** RE-auto Yes
Hausman Period Impact N/A 5 156.70* FE Yes N/A 4 -4.77 FE No
Test RE No RE Yes
Statistic Yearly Impact N/A 9 162.34* FE Yes N/A 9 -0.22 FE No
RE No RE Yes
NB: ", ** & *** denote significance at the 1,5 and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of ISI order autocorrelation problem respectively.
Appendix SP: Suitable equations for cut flowers imports from the World to South
Africa
Variables
Constant
InVijl_1
InGDPPCil
InGOPPCjl
REERI
09599
095
096
097
098
099
236
Appendices
Appendix SQ: Selection of the estimator suitable for cut flowers trade between
South Africa and the World
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Wald Test Period Impact 220 4 4.27" OLS No 242 3 16.95" OLS No
Statistic FE or RE Yes FE or RE Yes
Yearly Impact 220 8 4.44* OLS No 242 7 17.20" OLS No
FE or RE Yes FE or RE Yes
Durbin Period Impact 220 4 -0.76 FE-no auto Yes 242 3 1.62 FE-no auto No
Watson - - - FE-auto - 220 3 2.11·" FE-auto Yes
Statistic (for 220 5 -0.88 RE-no auto Yes 242 4 1.27 RE-no auto No
static) or - - - RE-auto - 220 4 2.06*" RE-auto Yes
Durbin-H Yearly Impact 220 8 -0.47 FE-no auto Yes 242 7 1.64 FE-no auto No
Statistic (for - - - FE-auto - 220 7 2.11 ** FE-auto Yes
dynamic) 220 9 -0.85 RE-no auto Yes 242 8 1.28 RE-no auto No
- - - RE-auto - 220 8 2.07*" RE-auto Yes
Hausman Period Impact N/A 4 134.18* FE Yes N/A 3 0.93 FE No
Test RE No RE Yes
Statistic Yearly Impact N/A 8 134.38" FE Yes N/A 7 1.75 FE No
RE No RE Yes
NB: " ** & *** denote significance at the I, 5 and 10percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of I SI order autocorrelation problem respectively.
Appendix SR: Suitable equations for cut flowers trade between South Africa and the
World
Variables Period Impact Yearly Impact
Dynamic FE Static FE Dynamic FE Static RE
Constant - -2.28 - -2.29
(-043) (-040)
InYij'_1 0.24" - 0.20'
(4.29) (347)
InGDPPCij, 4.32 2.78" 245 2.71'
( 148) (4.20) (0.80) (391 )
REER, -0.23 -0.68 0.66 -0.61
(.-033) (-1.10) (0.71) (-091)
09599 0.36 0.25 - -
( 1.02) (1.03)
095 - -0.82 -0.11
("1.15) (-0.35)
096 - - -0.21 0.34
(_0.36) (091)
097 - - 0.07 0.35
(012) (0.86)
098 - :0.14 0.30
(_0.30) (0.84)
099 - - 0.24 043
(0.63) (IA8)
InOISTij - -1.09'· - -1.05"
(-2.32) (-2.38)
Adjusted R2 0.80 0.51 0.80 0.51
Observations 220 220 220 220
Cross-Sections 22 .. 22 22 22
" ** & ,., denote significance at the I,Sand 10percent levels respectively. t-values are in parentheses
From 1994 to 2004, , SA traded (imports plus exports) cut flowers with the following 22 countries in the World: BGR, CHN, DEU,
ESP, FRA, GBR, IND, ITA, KEN, MOZ, MUS, MWI, NLD, PRT, SGP, SYC, TUR, UGA, USA, 2MB and ZWE
237
Appendices
Appendix SS: Selection of the estimator suitable for frozen fruits and nuts exports
from South Africa to the World
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Wald Test Period Impact 210 5 2.39** OlS No 231 4 6.25* OlS No
Statistic FE or RE Yes FE or RE Yes
Yearly Impact 210 9 2.24** OlS No 231 8 6.40* OlS No
FE or RE Yes FE or RE Yes
Durbin Period Impact 210 5 -0.53 FE-no auto Yes 231 4 1.75 FE-no auto No
Watson - - - FE-auto - 210 4 1.94** FE-auto Yes
Statistic (for 210 6 -1.86*** RE-no auto No 231 6 0.85 RE-no auto No
static) or 189 6 0.31 RE-auto Yes 210 6 2.15** RE-auto Yes
Durbin-H Yearly Impact 210 9 -0.54 FE-no auto Yes 231 8 1.70 FE-no auto No
Statistic (for - - - FE-auto - 210 8 1.93** FE-auto Yes
dynamic) 210 10 -1.69*** RE-no auto No 231 9 0.84 RE-no auto No
189 10 0.39 RE-auto Yes 210 9 2.07** RE-auto Yes
Hausman Period Impact N/A 5 128.18* FE Yes N/A 4 4.59 FE No
Test RE No RE Yes
Statistic Yearly Impact N/A 9 124.24* FE Yes N/A 9 4.52 FE No
RE No RE Yes
NB: ", ** & *** denote significance at the 1,5 and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OlS, FE & RE denote Pooled Ordinary least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of Ist order autocorrelation problem respectively.
Appendix ST: Suitable equations for frozen fruits and nuts exports from South
Africa to the World.
Variables
Constant
InYijl_1
InGDPPCit
InGDPPCjl
D9599
D95
D96
D97
D98
D99
238
Appendices
Appendix 5U: Selection of the estimator suitable for frozen fruits and nuts imports
from the World to South Africa
Selection Models Dynamic Static
Criteria N J( Statistic Estimator Decision N J( Statistic Estimator Decision
Wald Test Period Impact 230 5 1.35 OLS Yes 253 4 3.86* OLS No
Statistic FEor RE No FE or RE Yes
Yearly Impact 230 9 1.41 OLS Yes 253 8 3.83* OLS No
FEor RE No FE or RE Yes
Durbin Period Impact 230 5 -1.31 OLS-no auto Yes 253 4 1.50 FE-no auto No
Watson - - - FE-auto - 230 4 2.04** FE-auto Yes
Statistic (jor - - - RE-no auto - 253 6 0.60 RE-no auto No
static) or - - - RE-auto - 230 6 2.40** RE-auto Yes
Durbin-H Yearly Impact 230 9 1.30 OLS-no auto Yes 253 8 1.43 FE-no auto No
Statistic (jor - - - FE-auto - 230 8 2.04*" FE-auto Yes
dynamic) - - - RE-no auto - 253 9 0.57 RE-no auto No
- - - RE-auto - 230 9 2.37** RE-auto Yes
Hausman Period Impact - - - FE - N/A 4 -60.17· FE Ye.f
Test RE - RE No
Statistic Yearly Impact - - - FE - N/A 9 29.99* FE Yes
RE - RE No
NB: *, ** & *** denote significance at the 1,5 and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of Ist order autocorrelation problem respectively.
Appendix 5V: Suitable equations for frozen fruits and nuts imports from the World
to South Africa
Variables
Constant
InVijt_1
InGDPPCit
InGDPPCjt
REERt
D9599
D95
D96
097
D98
099
239
Appendices
Appendix 5W: Selection of the estimator suitable for frozen fruits and nuts trade
between South Africa and the World
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Wald Test Period Impact 110 4 1l.51 * OLS Yes 121 3 5.67* OLS No
Statistic FE or RE No FE or RE Yes
Yearly Impact 110 8 9.70* OLS Yes 121 7 5.87* OLS No
FE or RE No FE or RE Yes
Durbin Period Impact 110 4 -0.96 FE-no auto Yes 121 3 1.47 FE-no auto No
Watson - - - FE-auto - 110 3 2.13** FE-auto Yes
Statistic (for 110 5 -1.22 RE-no auto - 121 4 0.68 RE-no auto No
static) or - - - RE-auto - 110 4 2.25** RE-auto Yes
Durbin-H Yearly Impact 110 8 -0.92 FE-no auto Yes 121 7 1.44 FE-no auto No
Statistic (for - - - FE-auto - 110 7 2.17** FE-auto Yes
dynamic) 110 9 -1.20 RE-no auto - 121 8 0.68 RE-no auto No
- - - RE-auto - 110 8 2.31** RE-auto Yes
Hausman Period Impact N/A 4 64.30* FE - N/A 3 -0.13 FE No
Test RE - RE Yes
Statistic Yearly Impact N/A 8 62.86* FE - N/A 7 1.67 FE No
RE - RE Yes
NB: ". ** & *** denote significance at the 1,5 and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of l" order autocorrelation problem respectively.
Appendix 5X: Suitable equations for frozen fruits and nuts trade between South
Africa and the World
Variables
Constant
InYijl_1
InGOPPCijl
REERI
09599
095
096
097
098
099
240
Appendices
Appendix 5Y: Selection of the estimator suitable for preserved fruits and nuts
exports from South Africa to the World
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Wald Test Period Impact 850 5 2.75** OLS No 935 4 9.57* OLS No
Statistic FE or RE Yes FE or RE Yes
Yearly Impact 850 9 2.75* OLS No 935 8 9.56* OLS No
FEar RE Yes FE or RE Yes
Durbin Period Impact 850 5 -0.03 FE-no auto Yes 935 4 1.53 FE-no auto No
Watson - - - FE-auto - 850 4 1.98** FE-auto Yes
Statistic (jar 850 6 -1.28 RE-no auto Yes 935 6 0.82 RE-no auto No
static) or - - - RE-auto - 850 6 2.06** RE-auto Yes
Durbin-H Yearly Impact 850 9 -0.06 FE-no auto Yes 935 8 1.53 FE-no auto No
Statistic (jar - - - FE-auto - 850 8 1.97** FE-auto Yes
dynamic) 850 10 -1.32 RE-no auto Yes 935 9 0.82 RE-no auto No
- - - RE-auto - 850 9 2.05** RE-auto Yes
Hausman Period Impact N/A 5 461.11 * FE Yes N/A 4 14.42* FE Yes
Test RE No RE No
Statistic Yearly Impact N/A 9 457.59* FE Yes N/A 9 -8.68 FE No
RE No RE Yes
NB: " ••& .** denote significance at the I, 5 and 10percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of l" order autocorrelation problem respectively.
Appendix 5Z: Suitable equations for preserved fruits and nuts exports from South
Africa to the World.
Variables
InVijt_1
InGOPPCit
InGOPPCjt
09599
095
096
097
098
099
exported preserved and nuts to the I 85 countries in the World: AUT,
BEL, BGO, BI-IR, BHS, BRA, CAN, CHE, CHl, CHN, Cl V, COG, COM, CRI, CYP, CZE, DEU, ONK, DOM, ORC, EGY, ESP,
ETH, FIN, FRA, GAB, GBR, GHA, GRC, HUN, IRl, ISL, ISR, ITA, JOR, JPN, KEN, KOR, KWT, LBN, LBR, LUX, MAR, MOG,
MOV, MU, MLT, MOZ, MUS, MWI, MYS, NGA, NLD, NOR, NZl, OMN, PAK, PER, PHl, POL, PRI, PRT, RUS, SAU, SEN,
SGP STP THA URY 2MB and ZWE
241
Appendices
Appendix 5AA: Selection of the estimator suitable for preserved fruits and nuts
imports from the World to South Africa
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Wald Test Period Impact 380 5 2.05*** OLS No 418 4 8.40* OLS No
Statistic FE or RE Yes FE or RE Yes
Yearly Impact 380 9 2.23** OLS No 418 8 8.79* OLS No
FE or RE Yes FE or RE Yes
Durbin Period Impact 380 5 -0.27 FE-no auto Yes 418 4 1.51 FE-no auto No
Watson - - - FE-auto - 380 4 1.95** FE-auto Yes
Statistic (jar 380 6 -1.66*** RE-no auto No 418 6 0.75 RE-no auto No
sIalic) or 342 6 -0.73 RE-auto Yes 380 6 2.23** RE-auto Yes
Durbin-H Yearly Impact 380 9 -0.18 FE-no auto Yes 418 8 1.52 FE-no auto No
Statistic (jar - - - FE-auto - 380 8 1.97** FE-auto Yes
dynamic) 380 10 -1.75*** RE-no auto No 418 9 0.77 RE-no auto No
342 10 -0.52 RE-auto Yes 380 9 2.24** RE-auto Yes
Hausman Period Impact N/A 5 123.63* FE Yes N/A 4 1.09 FE No
Test RE No RE Yes
Statistic Yearly Impact N/A 9 149.02* FE Yes N/A 9 -6.01 FE No
RE No RE Yes
NB: *, ** & *** denote significance at the I,5 and 10percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of l" order autocorrelation problem respectively.
Appendix 5AB: Suitable equations for preserved fruits and nuts imports from the
World to South Africa
Variables
Constant
InYijt_1
InGDPPCit
InGDPPCjt
D9599
D95
D96
D97
D98
D99
242
Appendices
Appendix SAC: Selection of the estimator suitable for preserved fruits and nuts
trade between South Africa and the World
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Wald Test Period Impact 320 4 5.33* OLS No 352 3 8.61* OLS No
Statistic FE or RE Yes FE or RE Yes
Yearly Impact 320 8 5.18* OLS No 352 7 8.80* OLS No
FE or RE Yes FE or RE Yes
Durbin Period Impact 320 4 -0.31 FE-no auto Yes 352 3 1.94** FE-no auto Yes
Watson - - - FE-auto - - - - FE-auto -
Statistic (jar 320 5 -0.34 RE-no auto Yes 352 4 1.50 RE-no auto No
static) or - - - RE-auto - 320 4 1.99** RE-auto Yes
Ourbin-H . Yearly Impact 320 8 -0.49 FE-no auto Yes 352 7 1.92** FE-no auto Yes
Statistic (jar - - - FE-auto - - - - FE-auto -
dynamic) 320 9 -0.56 RE-no auto Yes 352 8 1.49 RE-no auto No
- - - RE-auto - 320 8 1.99** RE-auto Yes
Hausman Period Impact N/A 4 1764.06* FE Yes N/A 3 -19.94* FE Yes
Test RE No RE No
Statistic Yearly Impact N/A 8 888.36* FE Yes N/A 7 6.13 FE No
RE No RE Yes
NB: *, ** & *** denote significance at the I, 5 and 10percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of l" order autocorrelation problem respectively.
Appendix SAD: Suitable equations for preserved fruits and nuts trade between
South Africa and the World
Variables
Constant
InVijt_1
InGOPPCijt
REERt
09599
095
096
097
098
099
243
Appendices
Appendix 5AE: Selection of the estimator suitable for fruits and vegetable juices
exports from South Africa to the World
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Wald Test Period Impact 860 5 3.74* OLS No 946 4 12.87* OLS No
Statistic FE or RE Yes FE or RE Yes
Yearly Impact 860 9 3.78* OLS No 946 8 12.96* OLS No
FE or RE Yes FE or RE Yes
Durbin Period Impact 860 5 0.00 FE-no auto Yes 946 4 1.59 FE-no auto No
Watson - - - FE-auto - 860 4 1.96*· FE-auto Yes
Statistic (for 860 6 -1.58 RE-no auto Yes 946 6 0.87 RE-no auto No
static) or - - - RE-auto - 860 6 2.10** RE-auto Yes
Durbin-H Yearly Impact 860 9 -0.04 FE-no auto Yes 946 8 1.59 FE-no auto No
Statistic (for - - - FE-auto - 860 8 I.96** FE-auto Yes
dynamic) 860 10 -1.60 RE-no auto Yes 946 9 0.87 RE-no auto No
- - - RE-auto - 860 9 2.10*" RE-auto Yes
Hausman Period Impact N/A 5 587.70· FE Yes N/A 4 1.78 FE No
Test RE No RE Yes
Statistic Yearly Impact N/A 9 592.01 * FE Yes N/A 9 1.32 FE No
RE No RE Yes
NB: *, ** & *** denote significance at the I, 5 and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of 1" order autocorrelation problem respectively.
Appendix 5AF: Suitable equations (or fruit and 'vegetable juices exports from South
Africa to the World.
Variables
Constant
InYijt_1
InCDPPCit
InCDPPCjt
09599
095
096
097
098
099
244
Appendices
Appendix SAG: Selection of the estimator suitable for fruits and vegetable juices
imports from the World to South Africa
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Wald Test Period Impact 410 5 2.87** OLS No 451 4 9.57* OLS No
Statistic FE or RE Yes FE or RE Yes
Yearly Impact 410 9 2.96* OLS No 451 8 9.71* OLS No
FE or RE Yes FE or RE Yes
Durbin Period Impact 410 5 -0.08 FE-no auto Yes 451 4 1.62 FE-no auto No
Watson - - - FE-auto - 410 4 1.97** FE-auto Yes
Statistic (jar 410 6 -1.61 RE-no auto Yes 451 6 0.71 RE-no auto No
static) or - - - RE-auto - 410 6 2.17** RE-auto Yes
Durbin-H Yearly Impact 410 9 -0.02 FE-no auto Yes 451 8 1.62 FE-no auto No
Statistic (jar - - - FE-auto - 410 8 1.95** FE-auto Yes
dynamic) 410 10 -1.64 RE-no auto Yes 451 9 0.72 RE-no auto No
- - - RE-auto - 410 9 2.17** RE-auto Yes
Hausman Period Impact N/A 5 238.85* FE Yes N/A 4 7.27 FE No
Test RE No RE Yes
Statistic Yearly Impact N/A 9 244.12* FE Yes N/A 9 5.96 FE No
RE No RE Yes
NB: *, ** & *** denote significance at the 1,5and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of Ist order autocorrelation problem respectively.
Appendix SAH: Suitable equations for fruit and vegetable juices imports from the
World to South Africa
Variables
Constant
09599
095
096
097
098
and vegetable juices from the following 41 countries in the World: ARE, , AUT,
, CZE, DEU, ONK, ESP, FRA, GBR, GRC, ION, !NO, IRL, ISL, ISR, ITA, lPN, KEN, LKA,
PRT SAU SGP 2MB and ZWE
245
Appendices
Appendix SAl: Selection of the estimator suitable for fruits and vegetable juices
trade between South Africa and the World
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Wald Test Period Impact 400 4 4.52* OLS No 440 3 14.74* OLS No
Statistic FE or RE Yes FE or RE Yes
Yearly Impact 320 8 5.18* OLS No 440 7 14.98* OLS No
FE or RE Yes FE or RE Yes
Durbin Period Impact 400 4 0.39 FE-no auto Yes 440 3 1.89*' FE-no auto Yes
Watson - - - FE-auto - - - - FE-auto -
Statistic (for 400 5 -0.83 RE-no auto Yes 440 4 1.42 RE-no auto No
static) or - - - RE-auto - 400 4 2.00*· RE-auto Yes
Durbin-H Yearly Impact 320 8 -0.49 FE-no auto Yes 440 7 1.89** FE-no auto Yes
Statistic (for - - - FE-auto - - - - FE-auto -
dynamic) 320 9 -0.56 RE-no auto Yes 440 8 1.42 RE-no auto No
- - - RE-auto - 400 8 2.01 ** RE-auto Yes
Hausman Period Impact N/A 4 303.48' FE Yes N/A 3 3.08 FE No
Test RE No RE Yes
Statistic Yearly Impact N/A 8 888.36* FE Yes N/A 7 4.22 FE No
RE No RE Yes
NB: *, ** & **. denote significance at the I, 5 and 10percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of l" order autocorrelation problem respectively.
Appendix SAJ: Suitable equations for fruit and vegetable juices trade between
South Africa and the World
Variables Period Impact Yearly Impact
Dynamic FE Static RE Dynamic .FE Static RE
Constant .~ 24.96· - . 26.90'
(4.39) (4.83)
InYijt_1 0.09*** - 0.09**·
(185) . ( 1.87)
InGOPPCijt 3.83" 0.10 4.27*' 0,11
(1.97) (025) (2.12) (0.30)
REERt 0.71 -0.37 -0.33 -0.89
( 1.12) (-064) (-0.39) (-1.39)
09599 -0.54'" -0.39'" - -
(-1.68) (-1.69)
095 - - 0.19 0.01
(0.31 ) (005)
096 - - _0.08 -0.18
(-0.16) (-0.49)
097 - 0.35 0.22
(068) (0.56)
098 - - -0.11 -0.15
(cO.2S) (-0.43)
099 - - -0.65** -0.59"
(-185) (-2.14)
InDlSTij - -1.39" - -1.36""
(-229) (-2.39)
Adjusted R2 0.65 0.51 0.66 0.51
Observations 400 400 400 400
Cross-Sections 40 40 40 40
" •• & ".* denote significance at the 1,5and 10percent levels respectively. t-values are in parentheses
From 1994 to 2004, SA traded (imports plus exports) fruits and vegetable juices with the following 40 countries in the World: ARE,
ARG, AUS, AUT, BEL, BRA, CAN, CHE, CHl, CHN, DEU, DNK, ESP, FRA, GBR, GRC, ION, INO, IRl, ISl, ISR, ITA, JPN,
KEN, lKA, MOZ, MUS, MYS, NlD, NZl, PHl, pal, PRT, SAU, SGP, SWE, THA, USA, 2MB and ZWE
246
Appendices
Appendix SAK: Selection of the estimator suitable for wine exports from South
Africa to the World
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Wald Test Period Impact 1020 5 3.16" OLS No 1122 4 15.77" OLS No
Statistic FE or RE Yes FE or RE Yes
Yearly Impact 1020 9 3.20* OLS No 1122 8 16.01" OLS No
FE or RE Yes FE or RE Yes
Durbin Period Impact 1020 5 -0.46 FE-no auto Yes 1122 4 1.53 FE-no auto No
Watson - - - FE-auto - 1020 4 2.08*" FE-auto Yes
Statistic (for 1020 6 -1.78**" RE-no auto No 1122 6 0.99 RE-no auto No
static) or 918 6 -0.20 RE-auto Yes 1020 6 2.15** RE-auto Yes
Durbin-H Yearly Impact 1020 9 -0.39 FE-no auto Yes 1122 8 1.54 FE-no auto No
Statistic (for - - - FE-auto - 1020 8 2.10** FE-auto Yes
dynamic) 1020 10 -1.78*** RE-no auto No 1122 9 1.00 RE-no auto No
918 10 -0.26 RE-auto Yes 1020 9 2.15** RE-auto Yes
Hausman Period Impact N/A 5 450.36* FE Yes N/A 4 5.45 FE No
Test RE No RE Yes
Statistic Yearly Impact N/A 9 414.50· FE Yes N/A 9 -4.18 FE No
RE No RE Yes
NB: ", ** & .** denote significance at the I, 5 and 10percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no- auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of I SI order autocorrelation problem respectively.
Appendix SAL: Suitable equations for wine exports from South Africa to the World
Variables
Constant
InYijt_1
InGDPPCit
247
Appendices
Appendix SAM: Selection of the estimator suitable for wine imports from the World
to South Africa
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Wald Test Period Impact 430 5 4.98* OlS No 473 4 21.96* OlS No
Statistic FE or RE Yes FE or RE Yes
Yearly Impact 430 9 4.90* OlS No 473 8 22.43" OLS No
FEar RE Yes FE or RE Yes
Durbin Period Impact 430 5 -0.17 FE-no auto Yes 473 4 1.78 FE-no auto No
Watson - - - FE-auto - 430 4 2.04*- FE-auto Yes
Statistic (for 430 6 - 1.80**- RE-no auto No 473 6 1.08 RE-no auto No
static) or 387 6 -0.28 RE-auto Yes 430 6 2.09*" RE-auto Yes
Durbin-H Yearly Impact 430 9 -0.29 FE-no auto Yes 473 8 1.78 FE-no auto No
Statistic (jar - - - FE-auto - 430 8 2.05"· FE-auto Yes
dynamic) 430 10 - 1.99*- RE-no auto No 473 9 1.08 RE-no auto No
387 10 -0.38 RE-auto Yes 430 9 2.09"* RE-auto Yes
Hausman Period Impact N/A 5 353.56" FE Yes N/A 4 -1.79 FE No
Test RE No RE Yes
Statistic Yearly Impact N/A 9 320.35* FE Yes N/A 9 2.56 FE No
RE No RE Yes
NB: *, •• & •• " denote significance at the 1,5 and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OlS, FE & RE denote Pooled Ordinary least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of I st order autocorrelation problem respectively.
Appendix SAN: Suitable equations for wine imports from the World to South Africa
Variables
Constant
(nVijt_I
(nGOPPCit
InGOPPCjt
09599
095
096
097
098
099
248
Appendices
Appendix SAO: Selection of the estimator suitable for wine trade between South
Africa and the World
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Wald Test Period Impact 420 4 3.56* OLS No 462 3 26.35* OLS No
Statistic FEor RE Yes FE or RE Yes
Yearly Impact 420 8 3.51 * OLS No 462 7 26.64* OLS No
FE or RE Yes FE or RE Yes
Durbin Period Impact 420 4 1.23 FE-no auto Yes 462 3 1.35 FE-no auto No
Watson - - - FE-auto - 420 3 1.74 FE-auto No
Statistic (jar 420 5 0.56 RE-no auto Yes 462 4 1.10 RE-no auto No
static) or - - - RE-auto - 420 4 1.66 RE-auto No
Durbin-H Yearly Impact 420 8 1.10 FE-no auto Yes 462 7 1.35 FE-no auto No
Statistic (jar - - - FE-auto - 420 7 1.71 FE-auto No
dynamic) 420 9 0.40 RE-no auto Yes 462 8 1.10 RE-no auto No
- - - RE-auto - 420 8 1.66 RE-auto No
Hausman Period Impact N/A 4 184.68* FE Yes - - - FE -
Test RE No RE -
Statistic Yearly Impact N/A 8 179.76* FE Yes - - - FE -
RE No RE -
NB: *, ** & *** denote significance at the I, 5 and 10percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of I" order autocorrelation problem respectively.
Appendix SAP: Suitable equations for wine trade between South Africa and the
World
Variables
InYijt-1
InGDPPCijt
09599
095 ?
096
097
098 ?
249
Appendices
Appendix 5AQ: Selection of the estimator suitable for agricultural exports from
South Africa to the EU countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 225 5 4.68* OLS No 240 4 216...30* OLS No
FE or RE Yes FE or RE Yes
Yearly Impact 225 9 4.29* OLS No 240 8 213.56* OLS No
FE or RE Yes FE or RE Yes
Export Direction 1785 6 5.65* OLS No 1904 5 37.35* OLS No
FE or RE Yes FE or RE Yes
225 5 0.87 FE-no auto Yes 240 4 1.32 FE-no auto No
Durbin Period Impact - - - FE-auto - 225 4 2.05** FE-auto Yes
Watson 225 -1.15 RE-no auto Yes 240 6 1.25 RE-no auto No
Statistic - - - RE-auto - 225 6 1.91** RE-auto Yes
(jar static) 225 9 -1.35 FE-no auto Yes 240 8 1.35 FE-no auto No
or Durbin- Yearly Impact - - - FE-auto - 225 8 2.09** FE-auto Yes
H Statistic 225 10 -1.60 RE-no auto Yes 240 9 1.28 RE-no auto No
(jar - - - RE-auto - 225 9 1.94** RE-auto Yes
dynamic) 1785 6 -0.25 FE-no auto Yes 1904 5 1.31 FE-no auto No
Export Direction - - - FE-auto - 1785 5 1.85 FE-auto No
1785 7 -0.51 RE-no auto Yes 1904 6 1.19' RE-no auto No
- - - RE-auto - 1785 6 1.78 RE-auto No
Period Impact N/A 5 50.67* FE Yes N/A 4 -17.73' FE Yes
Hausman RE No RE No
Test Yearly Impact N/A 9 45.54* FE Yes N/A 8 -15.34** FE Yes
Statistic RE No RE No
Export Direction N/A 6 472.87* FE Yes N/A 5 0.03 FE No
RE No RE Yes
NB: *, ** & *** denote significance at the 1,5 and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of I" order autocorrelationgroblem re~ectively.·
250
Appendices
Appendix 5AR: Suitable equations for agricultural exports from South Africa to the
EU countries
MODEL
2004 2009 2004 2009
InYijt_1
InGDPPCit
InGDPPCjt
D0004 / DOS09
DOO/DOS
DOl/D06
D02/D07
D03/D08
D04/D09
PTAyes
PTA no
251
I
Appendices
Appendix SAS: Average actual, simulated and potential value of agricultural
exports from South Africa to EU countries for the periods 2000-2004 and 2005-
2009 in dollars (US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adiusted Potential
Name Code Value Value Value Value Value Value Value Value
(US$) (logs) (US$) (Lol!s) (US$) (Loss) (US$) Cl ogs)
Austria AUT 2000 - 2004 5,404,720 15.50278 6,006,301 15.60832 6,037,724
15.61354 6,022,012 15.61093
2005 - 2009 14,187,847 16.46790 12,233,764 16.31971 16,618,446 16.62602 14,426,105 16.48455
Belgium BEL 2000 - 2004 112,023,658 18.53422 141483627 18.76769 144,101,678 18.78603
142,792653 18.77690
2005 - 2009 246,994,352 19.32488 227,221,770 19.24144 334,009,957 19.62668 280,615,864 19.45250
Denmark DNK 2000 - 2004 16,562,195 16.62263 12,644,508 16.35273 12,415,435
16.33445 12,529,971 16.34363
2005 - 2009 78,445,976 18.17792 61,795,816 17.93935 56,842,138 17.85579 59,318,977 17.89844
FIN 2000 - 2004 5,963,522 15.60117 5,717,855 15.55910 5,702,426 15.55640 5,710,140 15.55775Finland
2005 - 2009 28,688,860 17.17202 22,806,005 16.94253 27,078,655 17.11426 24,942,330 17.03208
France FRA 2000 - 2004 61,144,925 17.92876 59,660,060 17.90417 59,550,712 17.90234 59,605,386 17.903262005 - 2009 145,165,227 18.79338 124,067,453 18.63634 149,332,507 18.82169 136,699,980 18.73330
Germany DEU 2000 - 2004 114,658,832 18.55747 119,993,278 18.60295 120,418,554
18.60648 120,205,916 18.60472
2005 - 2009 391,378,601 19.78519 325,260,369 19.60014 393,878,191 19.79155 359,569,280 19.70042
Greece GRC 2000 - 2004 7,536,694 15.83529 7,230,934 15.79388 7,211,416 15.79118
7,221,175 15.79253
2005 - 2009 29,786,988 17.20958 26,530,393 17.09380 37,544,301 17.44103 32,037,347 17.28241
Ireland IRL 2000 - 2004 11,189,475 16.23048 11,666,519
16.27223 11,699,392 16.27505 11,682,955 16.27364
2005 - 2009 57,930,674 17.87476 48,535,498 17.69781 55,470,006 17.83135 52,002,752 17.76681
Italy ITA 2000 - 2004 72,352,100 18.09706 72,772,746
18.10285 72,804,662 18.10329 72,788,704 18.10307
2005 - 2009 222,330,992 19.21968 189,141,509 19.05801 213,016,848 19.17688 201,079,179 19.11921
Luxembourg LUX 2000 - 2004 570,144 13.25365 618,971 13.33581 621,753
13.34030 620,362 13.33806
2005 - 2009 3,009,079 14.91714 1,353,462 14.11818 2,297,929 14.64752 1,825,696 14.41747
Netherlands 2000 - 2004 288,850,015 19.48142 237,590,970 19.28606 233,866,855 19.27026 235,728,913
19.27819
NLD 2005 - 2009 1,005,638,695 20.72889 842,528,606 20.55192 949,338,524 20.67128 895,933,565 20.61338
Portugal 2000 - 2004 24,361,526 17.00852 26,224,532 17.08221 26,362,066 17.08744 26,293,299
17.08482
PRT 2005 - 2009 79,320,322 18.18900 69,777,044 18.06082 86,505,976 18.27572 78,141,510 18.17403
Spain ESP 2000 - 2004 117,742,263 18.58401 112,890,779' 18.54193 112,523,256 18.53867 112,707,018 18.540302005 - 2009 316,875,702 19.57402 294,755,911 19.50166 359,152,173 19.69926 326,954,042 19.60533
Sweden 2000 - 2004 20,030,105 16.81275 19,821,078 16.80226 19,806,594 16.80153 19,813,836 16.80189SWE 2005 - 2009 158,572,016 .18.88172 119,116,338 18.59561 125,231,980 18.64568 122,174,159 18.62096
United Kingdom GBR 2000 - 2004 362,027,052 19.70723 376,046,068 19.74522 377,233,412 19.74837 376,639,740 19.746802005 - 2009 1,137,822,230 20.85238 979,347,707 20.70240 1,255,209,790 20.95057 1,117,278,748 20.83416
252
Appendices
Appendix SAT: Selection of the estimator suitable for agricultural imports from the
EU countries to South Africa
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 225 5 6.35* OLS No 240 4 61.33- OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 225 9 6.05* OLS No 240 8 61.41 • OLS No
FE or RE Yes FE or RE Yes
Import Direction 1905 6 5.89* OLS No 2032 5 32.38* OLS No
FE or RE Yes FE or RE Yes
225 5 -0.13 FE-no auto Yes 240 4 1.45 FE-no auto No
Durbin Period Impact - - - FE-auto - 225 4 2.02** FE-auto Yes
Watson 225 6 -1.56 RE-no auto Yes 240 6 1.31 RE-no auto No
Statistic - - - RE-auto . 225 6 2.09** RE-auto Yes
(jar static) 225 9 -0.17 FE-no auto Yes 240 8 1.44 FE-no auto No
or Durbin- Yearly Impact - - - FE-auto - 225 8 2.03*" FE-auto Yes
H Statistic 225 10 -1.66*** RE-no auto No 240 9 1.30 RE-no auto No
(jar 210 10 -0.91 RE-auto Yes 225 9 2.09*· RE-auto Yes
dynamic) 1905 6 0.11 FE-no auto Yes 2032 5 1.59 FE-no auto No
Import Direction - - - FE-auto - 1905 5 1.94** FE-auto Yes
1905 7 -1.51 RE-no auto Yes 2032 6 1.18 RE-no auto No
- - - RE-auto - 1905 6 2.08** RE-auto Yes
Period Impact N/A 5 94.68· FE Yes N/A 4 20.38' FE Yes
Hausman RE No RE No
Test Yearly Impact N/A 9 93.41 • FE Yes N/A 8 19.37· FE Yes
Statistic RE No RE No
Import Direction N/A 6 977.19· FE Yes N/A 5 -0.70 FE No
RE No RE Yes
NB: *, ** & **. denote significance at the 1,5 and 10 percent levels respectively. N & K denote the sample size-and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of I" order autocorrelation problem respectively.
253
Appendices
Appendix SAU: Suitable equations for agricultural imports from South Africa to
the EU countries
MODEL
2004 2009
InYijt_1
InGOPPCit
InGOPPCjt
REERt
00004 / 00509
DOO/ 005
DO]/006
002/007
003/008
004/009
PTAyes
PTA no
InDlSTij
Adjusted R2
Observations
254
Appendices
Appendix 5AV: Average actual, simulated and potential value of agricultural
imports from the EU countries to South Africa for the periods 2000-2004 and 2005-
2009 in dollars (US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adjusted Potential
Name Code Value Value Value Value Value Value Value Value
(US$) (Logs) (US$) (Loes) (US$) (Lozs) (US$) (Logs)
Austria 2000 - 2004
14,769,368 16.50807 10,358,398 16.15331 10,193,351 16.13725 10,275,875 16.14531
AUT 2005 - 2009 66,503.044 18.01276 49,106,497 17.70950 47,798,440 17.68250 48,452,468 17.69609
Belgium 2000 - 2004 14,725,647 16.50510 18,054,746 16.70892
18,331,545 16.72413 18,193,145 16.71656
BEL 2005 - 2009 78,875,372 18.18338 78,489,713 18.17848 78,460,993 18.17811 78,475,353 18.17830
2000 - 2004 11,637,128 16.26971 12,733,197 16.35972 12,817,107 16.36629 12,775,152 16.36301
Denmark DNK 2005 - 2009 45,149,774 17.62550 47,418,338 17.67452 47,586,862 17.67807 47,502,600 17.67629
2000 - 2004 1,641,458 14.31110 1,897,850 14.45623 1,915,545 14.46551 1,906,697 14.46088
Finland FIN 2005 - 2009 3.676,276 15.11741 6,447,026 15.67913 6,681,821 15.71490 6,564,423 15.69718
France FRA 2000 - 2004 45,456,959 17.63228 51,860,303 17.76406 52,408,140 17.77457 52,134,222 17.769332005 - 2009 212.594,203 19.17490 198,40 I,319 19.10580 197,325,388 19.10036 197,863,354 19.10309
2000 - 2004 44,546,640 17.61205 46,269,842 17.65000 46,409,083 17.65301 46,339,462 17.65150
Germany DEll 2005 - 2009 265,135,203 19.39575 215,952,506 19.19057 212,477,022 19.17434 214,214,764 19.18249
2000 - 2004 2,835,764 14.85782 2,891,773 14.87738 2,895,502 14.87867 2,893,638 14.87802Greece GRC 2005 - 2009 9.123,344 16.02635 13,596,838 16.42535 13,964,756 16.45205 13,780,797 16.43879
2000 - 2004 15,998,471 16.58800 29,926,675 17.21426 31,406,784 17.26253 30,666,730 17.23869Ireland IRL 2005 - 2009 70,487,247 18.07094 94,956,289 18.36893 97,114,808 18.39140 96,035,549 18.38023
Italy 2000 - 2004
33,722,779 17.33368 32,337,522 17.29174 32,232,605 17.28849 32,285,064 17.29012
ITA 2005 - 2009 163,356,555 18.91145 136,604,533 18.73260 134,735,211 18.71882 135,669,872 18.72574
25,142 10.13230 7,437 8.91425 7,097 8.86737 7,267 8.89108
Luxembourg LlJX 2000 - 20042005 - 2009 38.552 10.55976 21,592 9.98007 21,102 9.95710 21,347 9.96865
Nethcrlands 2000 - 2004 44,755,968 17.61674 45,019,454 17.62261
45,040,317 17.62307 45,029,886 17.62284
NLD 2005 - 2009 245,209,582 19.31762 225,842,491 19.23535 224,374,222 19.22883 225,108,357 19.23209
Portugal 2000 - 2004 4,228,945 15.25746 4,527,435 15.32567
4,548,494 15.33031 4.537,964 15.32799
PRT 2005 - 2009 25,359,201 17.04865 24,023,956 16.99456 23,933,940 16.99081 23,978.948 16.99269
2000 - 2004 16, I06,396 16.59473 15,995.987 16.58785 15,987,829 16.58734 15,991,908 16.58759Spain ESP 2005 - 2009 77,312.232 18.16336 73,035,493 18.10646 72,727,420 18.10223 72,881,456 18.10434
Sweden 2000 - 2004
2,506,931 14.73457 2,823,271 14.85341 2,845,438 14.86123 2,834,354 14.85732
SWE 2005 - 2009 12.887,834 16.37179 13,496,018 16.41791 13,537,700 16.42099 13,516,859 16.41945
117,140,037 141,50 I ,894 18.76782 140,467,357 18.76049
lInited Kingdom 2000 - 2004 18.57888 139,432,819 18.75309GBR 2005 - 2009 441,614,905 19.90595 528,924,189 20.08636 536,910,372 20.10134 532,917,281 2009388
255
Appendices
Appendix SAW: Selection of the estimator suitable for agricultural trade between
South Africa and the EU countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 225 4 5.34* OLS No 240 3 200.88* OLS No
Wald Test FEor RE Yes FE or RE Yes
Statistic Yearly Impact 225 8 4.02* OLS No 240 7 215.90* OLS No
FE or RE Yes FEor RE Yes
Trade Direction 1635 5 6.64* OLS No 1744 4 58.39* OLS No
FE or RE Yes FE or RE Yes
225 4 -0.27 FE-no auto No 240 3 1.46 FE-no auto No
Durbin Period Impact - - - FE-auto Yes 225 3 1.84** FE-auto Yes
Watson 225 5 -0.41 RE-no auto No 240 4 1.40 RE-no auto No
Statistic - - - RE-auto Yes 225 4 1.78 RE-auto ?
(for staiicï 225 8 -0.38 FE-no auto No 240 7 1.39 FE-no auto No
or Durbin- Yearly Impact - - - FE-auto Yes 225 7 1.90** FE-auto Yes
H Statistic 225 9 -0.82 RE-no auto No 240 8 1.34 RE-no auto No
(for - - - RE-auto Yes 225 8 1.83 RE-auto ?
dynamic) 1635 5 0.53 FE-no auto No 1744 4 1.35 FE-no auto No
Trade Direction - - - FE-auto Yes 1635 4 1.90 FE-auto No
1635 6 0.12 RE-no auto No 1744 5 1.28 RE-no auto No
- - - RE-auto Yes 1635 5 1.91 RE-auto No
Period Impact N/A 4 53.19* FE No - - - FE -
Hausman RE Yes RE -
Test Yearly Impact N/A 8 43.95* FE No - - - FE -
Statistic RE Yes RE -
Trade Direction N/A 5 497.07* FE Yes - - - FE -
RE No RE -
NB: *, ** & *** denote significance at the 1,5 and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of ISI order autocorrelation problem respectively. ?
means inconclusive.
256
Appendices
Appendix SAX: Suitable equations for agricultural trade between South Africa and
the EU countries
MODEL
InGDPPCijt
D0004 / D0509
DOO/ D05
DOl / D06
D02/ D07
D03/ D08
D04/ D09
PTAyes
257
Appendices
Appendix SAY: Average actual, simulated and potential value of agricultural trade
between South Africa and the EU countries for the periods 2000-2004 and 2005-
2009 in dollars (US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adjusted Potential
Name Code Value Value Value Value Value Value Value Value
(US$) (logs) (US$) (logs) (US$) (Logs) (US$) (Logs)
Austria AUT 2000 - 2004 20,255,230 16.82392 19,047,241 16.76243 18,999,559 16.75993 19,023,400 16.761182005 - 2009 80,834,738 18.20792 72,666,623 18.10139 72,164,259 18.09446 72,415,441 18.09793
Bclgium BEL 2000 - 2004 126,806,811 18.65818 151,782,395 18.83796 153,895,695 18.85179 152,839,045 18.844902005 - 2009 328,837,486 19.61107 427,543,734 19.87357 436,069,760 19.89331 431,806,747 19.88349
Denmark DNK 2000 - 2004 28,437,283 17.16321 26,710,861 17.10058 26,595,124 17.09624 26,652,992 17.098412005 - 2009 123,902,989 18.63501 109,659,716 18.51289 108,730,175 18.50438 109,194,945 18.50865
Finland FIN 2000 - 2004 7,737,780 15.86163 7,549,229 15.83696 7,537,340 15.83538 7,543,285 15.836172005 - 2009 34,318,028 17.35118 34,603,232 17.35946 34,621,889 17.36000 34,612,561 17.35973
France FRA 2000 - 2004 107,288,667 18.49103 109,956,855 18.51560 110,160,926 18.51745 110,058,891 18.516532005 - 2009 359,835,503 19.70116 361,490,460 19.70575 361,614,271 19.70609 361,552,365 19.70592
Germany DEU 2000 - 2004 160,197,912 18.89192 164,314,857 18.91730 164,637,243 18.91926 164,476,050 18.918282005 - 2009 666,935,558 20.31820 638,324,651 20.27436 636,177,483 20.27099 637,251,067 20.27267
Greece GRC 2000 - 2004 10,686,802 16.18452 9,768,739 16.09470 9,711,881 16.08886 9,740,310 16.091782005 - 2009 39,152,836 17.48298 45,679,529 17.63716 46,148,140 17.64737 45,913,835 17.64228
Ireland IRL 2000 - 2004 27,610,899 17.13372 32,244,780 17.28887 32,597,590 17.29975 32,421,185 17.294322005 - 2009 129,852,493 18.68191 136,964,854 18.73523 137,481,585 18.73900 137,223,219 18.73712
Italy ITA 2000 - 2004 106,155,583 18.48042 106,276,698 18.48156 106,285,789 18.48164 106,281,243 18.481602005 - 2009 386,418,362 19.77243 377,511,820 19.74911 376,854,815 19.74737 377,183,318 19.74824
Luxembourg LUX 2000 - 2004 826,318 13.62473 846,759 13.64917 847,889 13.65050 847,324 13.649842005 - 2009 3,060,762 14.93417 2,496,899 14.73056 2,469,249 14.71942 2,483,074 14.72501
Netherlands NLD 2000 - 2004 333,750,063 19.62590 290,948,361 19.48866 287,779,829 19.47771 289,364 ,095 19.483202005 - 2009 1,265,236,017 20.95852 1,195,645,997 20.90195 1,190,286,337 20.89746 1,192,966,167 20.89971
Portugal PRT 2000 - 2004 28,993,727 17.18259 30,035,583 17.21789 30,109,714 17.22036 30,072,648 17.219132005 - 2009 105,855,325 18.47758 114,458,122 18.55572 115,084,878 18.56118 114,771,500 18.55845
Spain ESP 2000 - 2004 134,214,339 18.71495 124,539947 18.64014 123833,413 18.63445 124,186,680 18.637302005 - 2009 395,124,819 19.79471 438,735,905 19.89941 442,209,833 19.90730 440,472,869 19.90336
Sweden SWE 2000 - 2004 22,672,782 16.93668 22,187,409 16.91504 22,154,525 16.91355 22,170,967 16.914292005 - 2009 171,994,311 18.96297 138,602,626 18.74712 136,504,837 18.73187 137,553,732 18.73953
United 2000 - 2004 481,159,852 19.99171 494,114,863 20.01828 495,193,991 20.02046 494,654,427GBR 20.01937Kingdom 2005 - 2009 1 583,128,436 21.18267 1,760,413,273 21.28881 1,775,617,381 21.29741 1,768;015,327 21.29312
,
I i
258
Appendices
Appendix 5AZ: Selection of the estimator suitable for cheese exports from South
Africa to the EU countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 90 5 3.23* OLS No 96 4 8.08* OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 90 9 3.20* OLS No 96 8 8.32 OLS No
FE or RE Yes FE or RE Yes
Export Direction 435 6 2.52** OLS No 464 5 10.44* OLS No
FE or RE Yes FE or RE Yes
90 6 0.29 FE-no auto Yes 96 4 1.46 FE-no auto No
Durbin Period Impact - - - FE-auto - 90 4 1.98** FE-auto Yes
Watson - - - RE-no auto - 96 6 0.41 RE-no auto No
Statistic - - - RE-auto - 90 6 2.29** RE-auto Yes
(jor static) 90 9 0.02 FE-no auto Yes 96 8 1.47 FE-no auto No
or Ourbin- Yearly Impact - - - FE-auto - 90 8 1.99** fE-auto Yes
H Statistic - - - RE-no auto - 96 9 0.43 RE-no auto No
(jor - - - RE-auto - 90 9 2.30** RE-auto Yes
dynamic) 480 -0.11 FE-no auto Yes 512 5 1.28 FE-no auto No
Export Direction - - - fE-auto - 480 5 1.97** fE-auto Yes
480 -1.44 RE-no auto Yes 512 6 0.79 RE-no auto No
- - - RE-auto - 480 6 2.08** RE-auto Yes
Period Impact - fE - N/A 4 0.37 FE No
Hausman - - RE - RE Yes
Test Yearly Impact - - - fE - N/A 8 1.22 FE NoStatistic RE - RE Yes
Export Direction N/A 6 146.69* FE Yes N/A 5 14.94** FE Yes
RE No RE No
NB: *, ** & *** denote significance at the I, 5 a8.d.:10percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelationproblem and estimation with the correction of ISI order autocorrelation problem respectively.
259
Appendices
Appendix 5BA: Suitable equations for cheese exports from South Africa to the EU
countries
MODEL
2004 2009
InYijt_1
InGDPPCit
InGDPPCjt
REERt
D0004 / DOS09
DOO / DOS
DOl / D06
D02 / D07
D03 / DOS
D04/D09
PTAyes
PTA no
InDISTij
Adjusted R2
Observations
Appendix 5BB: Average actual, simulated and potential value of cheese exports
from South Africa to EU countries for the periods 2000-2004 and 2005-2009 in
dollars (VS$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adjusted Potential
Name Code Value Value Value Value Value Value Value Value
(US$) (Logs) (US$) (Logs) (US$) (Logs) (US$) (logs)
5.84447 282
France 2000 - 2004 2 0.61763 220 5.39154 345 5.64343FRA 2005 - 2009 2 0.81228 I 0.11923 I 0.11474 I 0.11698
2000 - 2004 8 2.04074 229 5.43537 257 5.54792 243 5.49323Germany DEU 2005 - 2009 17 2.86196 II 2.38260 10 2.31179 10 2.34719
2000 - 2004 3 1.02676 6 1.87045 6 1.71510Greece 6 1.79579GRC 2005 - 2009 1,599 7.37691 204 5.31992 89 4.49037 135 4.90515
Netherlands 2000 - 2004 294 5.68434 241 5.48316 109
4.68763 175 5.16250
NLD 2005 - 2009 0 0.00000 0 0.00000 0 0.00000 0 0.00000
Spain 2000 - 2004 3 1.09016 21 3.02602 18 2.89956 19 2.96478ESP 2005 - 2009 5 1.54186 5 1.63657 5 1.64575 5 1.64116
United Kingdom 2000 - 2004 947,613 13.76170 387 5.95945 19 2.93936 203 5.31395GBR 2005 - 2009 372 5.91841 8,545 9.05311 171,701 12.05351 38,304 10.55331
260
Appendices
Appendix 5BC: Selection of the estimator suitable for cheese imports from the EU
countries to South Africa
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 195 5 3.99* OLS No 208 4 20.03* OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 195 9 5.38* OLS No 208 8 25.50* OLS No
FE or RE Yes FE or RE Yes
Import Direction 360 6 3.72* OLS No 384 5 17.86* OLS No
FE or RE Yes FE or RE Yes
195 5 0.15 FE-no auto Yes 208 4 1.46 FE-no auto No
Durbin Period Impact - - - FE-auto - 195 4 1.92** FE-auto Yes
Watson 195 6 -0.86 RE-no auto Yes 208 6 . 0.96 RE-no auto No
Statistic - - - RE-auto - 195 6 1.96** RE-auto Yes
(jar static) 195 9 -0.33 FE-no auto Yes 208 8 1.40 FE-no auto No
or Durbin- Yearly Impact - - - FE-auto - 195 8 2.03** FE-auto Yes
H Statistic 195 lO -1.25 RE-no auto Yes 208 9 1.0 I RE-no auto No
(jar - - - RE-auto - 195 9 2.08** RE-auto Yes
dynamic) 360 - 0.14 FE-no auto Yes 384 5 1.38 FE-no auto No
Import Direction - - - FE-auto 360 5 1.98** FE-auto Yes
360 - -0.95 RE-no auto Yes 384 6 0.78 RE-no auto No
- - - RE-auto - 360 6 2.04** RE-auto Yes
Period Impact N/A 5 74.12* FE Yes N/A 4 0.59 FE No
Hausman RE No RE Yes
Test Yearly Impact N/A 9 92.71 * FE Yes N/A 8 -4.20 FE No
Statistic RE No RE Yes
Import Direction N/A 6 131.24* FE Yes N/A 5 -1.99 FE No
RE No RE Yes
NB: *, ** & *** denote significance at the I, 5 and 10percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of l" order autocorrelation problem respectively.
261
Appendices
Appendix 5BD: Suitable equations for cheese imports from the EU countries to
South Africa
MODEL
InYijt-1
InGDPPCit
InGDPPCjt
REERt
D0004 / DOS09
DOO/ DOS
DOl / D06
D02 / D07
D03 / D08
D04 / D09
PTAyc,
PTA no
InDISTij
Adjusted R2
Observations
262
Appendices
Appendix 5BE: Average actual, simulated and potential value of cheese imports
from the EU countries to South Africa for the periods 2000-2004 and 2005-2009 in
dollars (US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adjusted Potential
Name Code Value Value Value Value Value Value Value Value
(LJS$) (Logs) (US$) (Logs) (US$) (Logs) (lJS$) (Logs)
Austria AUT 2000 - 2004 364,986 12.80761 516,260 13.15437 543,642 13.20605 529,774 13.180212005 - 2009 26,697 10.19230 147,771 11.90342 167,466 12.02854 157,311 11.96598
Belgium BEL 2000 - 2004 88,832 11.39450 23,518 10.06552 21,397 9.97100 22,432 10.018262005 - 2009 31,119 10.34556 41,908 10.64323 42,828 10.66495 42,366 10.65409
Denmark DNK 2000 - 2004 2,982,915 14.90841 3,283,069 15.00429 3,317,850 15.01483 3,300,414 15.009562(1('5- 2009 4,291,176 15.27207 3,064,104 14.93527 2,957,143 14.89973 3,010,149 14.91750
France FRA 2000 - 2004 5,433,388 15.50807 6,611,744 15.70436 6,763,492 15.72705 6,687,188 15.715702005 - 2009 5,775,476 15.56913 4,779,608 15.37987 4,682,038 15.35924 4,730,571 15.36956
Germany DEll 2000 - 2004 3,328,056 15.01790 4,433,065 15.30460 4,578,150 15.33681 4,505,024 15.320702005 - 2009 2,134,408 14.57370 2,703,557 14.81008 2,771,188 14.83479 2,737,164 14.82243
Greece GRC 2000 - 2004 93,859 11.44955 111,098 11.61816 112,664 11.63216 111,878 11.625162005 - 2009 9,246 9.13190 67,798 11.12429 78,964 11.27675 73,168 11.20052
Ireland IRL 2000 - 2004 1,341,607 14.10938 1,226,614 14.01977 1,215,468 14.01064 1,221,028 14.015202005 - 2009 100,370 11.51662 869,999 13.67625 1,069,862 13.88304 964,769 13.77964
Italy ITA 2000 - 2004 3,737,395 15.13390 3,606,950 15.09837 3,592,822 15.09445 3,599,879 15.096412005 - 2009 4,693,112 15.36161 3.362,899 15.02831 3,245,924 14.99291 3,303,893 15.01061
Netherlands NLD 2000 - 2004 1,273,774 14.05749 1,834,019 14.42202 1,905,690 14.46035 1,869,511 14.441192005 - 2009 1,136,217 13.94322 1,525,343 14.23773 1,570,972 14.26721 1,547,990 14.25247
Portugal PRT 2000 - 2004 30,809 10.33557 5,008 8.51878 4,494 8.41060 4,744 8.464692005 - 2009 40,839 10.61739 16,959 9.73853 15,998 9.68021 16,471 9.70937
Spain . ESP 2000 - 2004 4 1.36201 9 2.17463 9 2.18646 9 2.18054
2005.- 2009 3,436 8.14214 309 5.73263 282 5.64101 295 5.68682
Sweden SWE 2000 - 2004 5 1.57507 10 2.28560 10 2.29647 10 2.291032005 - 2009 3,428 8.13979 199 5.29251 180 5.19285 189 5.24268 .
United Kingdom GBR 2000 - 2004 1,112,152 13.92181 1,484,760 14.21076 1,529,820 14.24066 1,507,121 14.225712005 - 2009 889,014 13.69787 1,204,007 14.00117 1,240,427 14.03097 1,222,082 14.01607
263
Appendices
Appendix 5BF: Selection of the estimator suitable for cheese trade between the EU
countries and South Africa
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 90 4 4.25* OLS No 96 3 15.40* OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 90 8 4.67* OLS No 96 7 5.97* OLS No
FE or RE Yes FE or RE Yes
Trade Direction 150 5 3.92* OLS No 160 4 20.55* OLS No
FEar RE Yes FE or RE Yes
90 4 1.53 FE-no auto Yes 96 3 1.52 FE-no auto No
Durbin Period Impact - - - FE-auto - 90 3 1.81** FE-auto Yes
Watson 90 5 -1.94**· RE-no auto No 96 4 1.05 RE-no auto No
Statistic 84 5 0.21 RE-auto Yes 90 4 1.93** RE-auto Yes
(for static) 75 8 0.63 FE-no auto Yes 96 7 1.48 FE-no auto No
or Durbin- Yearly Impact - - - FE-auto - 90 7 1.91** FE-auto Yes
H Statistic 90 9 -2.14** RE-no auto No 96 8 1.03 RE-no auto No
(for 84 9 0.79 RE-auto Yes 90 8 2.02*· RE-auto Yes
dynamic) 150 5 1.10 FE-no auto Yes 160 4 1.66 FE-no auto No
Trade Direction - - - FE-auto - 150 4 1.95** FE-auto Yes
150 6 -1.71*** RE-no auto No 160 5 1.15 RE-no auto No
140 6 0.46 RE-auto Yes 150 5 2.15*' RE-auto Yes
Period Impact N/A 4 28.95· FE Yes N/A 3 -113.23* FE Yes
Hausman RE No RE No
Test Yearly Impact N/A 8 28.49* FE Yes N/A 7 -6.45 FE No
Statistic RE No RE Yes
Trade Direction N/A 5 34.35* FE Yes N/A 4 -2.44 FE No
RE No RE Yes
NB: *, ** & *** denote significance at the I, 5 and 10percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of l" order autocorrelation problem respectively.
264
Appendices
Appendix 5BG: Suitable equations for cheese trade between South Africa and the
EU countries
MODEL
InYijt_1
InGDPPCij,
D0004 / D0509
DOO/ D05
DOl / D06
D02 / D07
D03/ D08
D04/ D09
PT Ay.,
PTA no
Appendix SBH: Average actual, simulated and potential value of cheese trade
between South Africa and the EU countries for the periods 2000-2004 and 2005-
2009 in dollars (US$) and natural logarithms (Logs)
Country Country Period. Actual Simulated Adjusted Potential
Name Code Value Value Value Value Value Value Value Value
(US$) (Logs) (US$) (Logs) (US$) (Logs) (US$) (Logs)
France FRA 2000 - 2004 5,433,393 15.50807 3,794,227 15.14899 3,458,650 15.05639 3,622,555 15.10269
2005 - 2009 5,775,486 15.56913 6,721,196 15.72078 7,059,854 15.76993 6,888,444 15.74536
Germany DEU 2000 - 2004 3,328, I07 15.01791 2,719,004 14.81578 2,584,600 14.76508 2,650,950 14.790432005 - 2009 2.134,780 14.57387 3,613,816 15.10027 4,248,650 15.26211 3,918,397 15.18119
Greece 2000 - 2004 93,866 11.44962 73,094 11.19950 69,900 11.15482 71,479 11.17716GRC 2005 - 2009 66,185 11.10021 110,300 11.61096 123,462 11.72369 116,695 11.66732
Netherlands 2000 - 2004 1,285,061 14.06632 1,162,565 13.96614 1,135,736 13.94279NLD 1,149,072
13.95447
2005 - 2009 0 0.00000 0 0.00000 0 0.00000 0 0.00000
Spain ESP 2000 - 2004 18 2.86913 103 4.63224
116 4.75016 109 4.69120
2005 - 2009 10,654 9.27367 1,419 7.25779 1,098 7.00092 1,248 7.12935
United Kingdom 2000 - 2004 3,209,011 14.98147 1,369,401 14.12988 1,119,664 13.92854GBR 1,238,253 14.029212005 - 2009 890,420 13.69945 2,036,096 14.52654 2,593,041 14.76834 2,297,756 14.64744
265
Appendices
Appendix SBI: Selection of the Estimator suitable for cut flowers exports from
South Africa to the EU countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 210 5 4.17* OLS No 224 4 43.65* OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 210 9 4.30* OLS No 224 8 44.19* OLS No
FE or RE Yes FE or RE Yes
Export Direction 840 6 3.25* OLS No 896 5 22.64* OLS No
FE or RE Yes FE or RE Yes
210 5 -0.81 FE-no auto Yes 224 4 1.11 FE-no auto No
Durbin Period Impact - - - FE-auto - 210 4 2.04** FE-auto Yes
Watson 210 6 -1.75*** RE-no auto No 224 6 0.96 RE-no auto No
Statistic 196 6 -0.33 RE-auto Yes 210 6 2.03** RE-auto Yel'
(jar static) 210 9 -0.92 FE-no auto Yes 224 8 1.12 FE-no auto No
or Durbin- Yearly Impact - - - FE~auto - 210 8 2.04** FE-auto Yes
H Statistic 210 lO -1.71** RE-no auto No 224 9 0.96 RE-no auto No
(jar 196 10 -0.36 RE-auto Yes 210 9 2.01 ** RE-auto Yes
dynamic) 840 6 0.09 FE-no auto Yes 896 5 1.11 FE-no auto No
Export Direction FE-auto - 840 5 1.89 FE-auto ?
880 7 -0.79 RE-no auto Yes 896 6 0.90 RE-no auto No
- - - RE-auto - 840 6 1.93** RE-auto Yes
Period Impact N/A 5 51.04* FE Yes N/A 4 1.14 FE No
Hausman RE No RE Yes
Test Yearly Impact N/A 9 -32.82* FE Yes N/A 8 -1.81 FE No
Statistic RE No RE Yes
Export Direction N/A 6 235.37* FE Yes N/A 5 -27.88* FE Yes
RE No RE No
NB: *, ** & *** denote significance at the 1,5 and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of I st order autocorrelation problem respectively. ?
means inconclusive.
266
Appendices
Appendix 5BJ: Suitable equations for cut flowers exports from South Africa to the
EU countries
MODEL
2009
InGDPPCil
InGDPPCjt
REERt
D0004 / 00509
DOO/ DOS
DOl /006
D02 / D07
D03/008
D04 / D09
PTAyes
PTA no
InDISTij
Adjusted R2
Observations
267
Appendices
Appendix 5BK: Average actual, simulated and potential value of cut flowers exports
from South Africa to EU countries for the periods 2000-2004 and 2005-2009 in
dollars (US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adjusted Potential
Name Code Value Value Value Value Value Value Value Value
(US$) (Logs) (US$) (Logs) (US$) (Logs) (US$) (Logs)
Austria 2000 - 2004 5,404,720 15.50278 4,131,725 15.23421 15,987,516 16.58732 10,059,620 16.12404AUT 2005 - 2009 127,206 11.75356 69,925 11.15518 45,758 10.73113 56,566 10.94315
Belgium 2000 - 2004 I 12,023,658 18.53422 18,042,749 16.70825 81,551,720 18.21675 49,797,234 17.72347BEL 2005 - 2009 1,239,669 14.03035 678,236 13.42725 405,420 12.91268 524,376 13.16996
Denmark 2000 - 2004 16,562,195 16.62263 2,991,085 14.91115 11,455,521 16.25398 7,223,303 15.79282DNK 2005 - 2009 5,232 8.56255 7,064 8.86276 5,398 8.59384 6,175 8.72830
Finland 2000 - 2004 5,963,522 15.60117 814,539 13.61038 2,704,480 14.81042 1,759,509 14.38055FIN 2005 - 2009 2 0.78798 15 2.70783 14 2.66029 15 2.68406
France 2000 - 2004 61,144,925 17.92876 12,485,262 16.34006 55,470,813 17.83137 33,978,038 17.34122FRA 2005 - 2009 398,972 12.89665 378,457 12.84386 244,687 12.40773 304,308 12.62580
2000 - 2004 I 14,658,832 18.55747 15,914,818 16.58276 70,218,671 18.06712 43,066,744 17.57826Germany DEll 2005 - 2009 4,551,479 15.33096 3,401,234 15.03965 1,967,837 14.49245 2,587,097 14.76605
Greece 2000 - 2004 7,536,694 15.83529 14,364,281 16.48026 77,268,040 18.16279 45,816,160 17.64015GRC 2005 - 2009 154,964 11.95095 173,173 1206205 116,995 11.66988 142,339 11.86597
Ireland 2000 - 2004 11,189,475 16.23048 1,629,978 14.30408 5,802,790 15.57385 3,716,384 15.12826IRL 2005 - 2009 35 3.56064 154 5.03448 138 4.92845 146 4.98147
Italy 2000 - 2004 72,352,100 18.09706 49,535,625 17.71820 279,924,329 19.45003 164,729,977 18.91982ITA
2005 - 2009 713,071 13.47734 653,744 13.39047 412,664 12.93039 519,400 13.16043
Netherlands 2000 - 2004 288,850,015 19.48142 25,923,155 17.07065 115,676,395 18.56631 70,799,775 18.07537NLD 2005 - 2009 13,342,124 16.40644 8,211,636 15.92106 4,473,475 1531368 6,060,903 15.61737
Portugal 2000 - 2004 24,361,526 17.00852 7,543,362 15.83618 32,922,931 1730968 20,233,147 16.82283PRT 2005 - 2009 531,378 13.18323 441,895 12.99883 280,368 12.54386 351,985 12.77134
Spain 2000 - 2004 117,742,263 18.58401 32,093,626 17.28417 160,371,591 18.89300 96,232,608 18.38228ESP 2005 - 2009 4 1.27989 215 5.37060 212 535861 214 5.36461
Sweden 2000 - 2004 20,030,105 16.81275 2,242,555 14.62313 8,079,280 15.90481 5,160,918 15.45662SWE 2005 - 2009 0 0.00000 0 0.00000 0 0.00000 0 0.00000
United Kingdom . 2000 - 2004 362,027,052 19.70723 31,39S,246 17.26226 142,272,871 18.77326 86,835,558 18.27953GBR 2(105 - 2009 12,736,224 1635996 3,861,460 15.16656 1,998,644 14.50798 2,778,072 14.83727
268
Appendices
Appendix 5BL: Selection of the estimator suitable for cut flowers imports from the
EU countries to South Africa
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 105 5 11.94· OLS No 112 4 25.45· OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 105 9 12.16· OLS No 112 8 26.91 • OLS No
FE or RE Yes FE or RE Yes
Import Direction 390 6 4.57" OLS No 416 5 19.20" OLS No
FE or RE Yes FE or RE Yes
105 0.80 FE-no auto Yes 112 4 1.77"" FE-no auto Yes
Durbin Period Impact - - - FE-auto - - - - FE-auto -
Watson 105 -1.81 """ RE-no auto No 112 5 1.0 I RE-no auto No
Statistic 98 0.36 RE-auto Yes 105 5 2.17 RE-auto Yes
(for static) 105 0.90 FE-no auto Yes 112 8 1.75 FE-no auto No
or Durbin- Yearly Impact - - - FE-auto - 105 8 1.99"" FE-auto Yes
H Statistic 105 -1.32 RE-no auto Yes 112 9 1.02 RE-no auto No
(for - - - RE-auto - 105 9 2.08·" RE-auto Yes
dynamic) 390 -0.11 FE-no auto Yes 416 5 1.58 FE-no auto No
Import Direction - - - FE-auto - 390 5 1.98"· FE-auto Yes
390 -1.68""" RE-no auto No 416 6 1.0 I RE-no auto No
364 -0.36 RE-auto Yes 390 6 2.04** RE-auto Yes
Period Impact N/A 5 101.93" FE Yes N/A 4 -2.13 FE No
Hausman RE No RE Yes
Test Yearly Impact N/A 9 530.39" FE Yes N/A 8 -1.99 FE No
Statistic RE No RE Yes
Import Direction N/A 6 925.91 * FE Yes N/A 5 -0.30 FE No
RE No RE Yes
NB: ., .* & .** denote significance at the I, 5 and 10percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of ISI order autocorrelation problem respectively.
269
Appendices
Appendix 5BM: Suitable equations for cut flowers imports from the EU countries to
South Africa
MODEL
InYijt_1
InGOPPCit
InGOPPCjt
REERt
00004 / 00509
DOO/005
'001/006
/007
003/008
004/009
PTAyes
270
Appendices
Appendix 5BN: Average actual, simulated and potential value of cut flowers imports
from the EU countries to South Africa for the periods 2000-2004 and 2005-2009 in
dollars (US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adjusted Potential
Name Code Value Value Value Value Value Value Value Value
(liS$} (Logs) (US$) (Logs) (lIS$) (Logs) (US$) (Logs)
France 2000 - 2004 1,199 7.08962 114 4.73194 178 5.17991
142 4.95592
FRA 2005 - 2009 561 0.00000 78 0.00000 138 0.00000 104 0.00000
Germany DEll 2000 - 2004 23 3.14150 137
4.92297 580 6.36292 282 5.64295
2005 - 2009 4 1.32150 55 4.00086 432 6.06942 154 5.03514
Italy ITA 2000 - 2004 10 2.30704 17 2.83525 31
3.43579 23 3.13552
2005 - 2009 0 0.00000 0 0.00000 0 0.00000 0 0.00000
Netherlands 2000 - 2004 15,131 9.62452 161 5.08270 154 5.03966
158 5.06118
NLD 2005 - 2009 46,153 10.73973 81 4.39337 27 3.30338 47 3.84837
0.91973 2 0.86170
Portugal PRT 2000 - 2004 5 1.53888 2 0.80366 3
2005 - 2009 0 0.00000 0 0.00000 0 0.00000 0 0.00000
Spain 2000 - 2004 25 3.21464 8 2.04351 10
2.31668 9 2.18010
ESP 2005 - 2009 3 1.01911 24 3.17451 97 4.57901 48 3.87676
United Kingdom 2000 - 2004 30 3.38650 249 5.51746 1,440
7.27271 599 6.39508
GBR 2005 - 2009 0 0.00000 0 0.00000 0 0.00000 0 0.00000
Appendix 5BO: Selection of the estimator suitable for cut flowers trade between
South Africa and the EU countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 105 4 1.35 OLS Yes 112 3 24.53* OLS No
Wald Test FE or RE No FE or RE Yes
Statistic Yearly Impact 105 8 1.0 I OLS Yes 112 7 23.92* OLS No
FE or RE No FE or RE Ye,~
Trade Direction 330 5 4.63" OLS No 352 4 20.91* OLS No
FE or RE Yes FE or RE Yes
105 4 -0.36 OLS-no auto Ye,~ 112 3 1.16 FE-no auto Yes
Durbin Period Impact - - - FE-auto - 105 3 1.93** FE-auto -
Watson - - - RE-no auto - 112 4 1.10 RE-no auto No
Statistic - - - RE-auto - 105 4 1.85*" RE-auto Yes
(jar static) 105 8 -0.40 OLS-no auto Yes 112 7 1.15 FE-no auto No
or Durbin- Yearly Impact - - - FE-auto - 105 7 1.96** FE-auto Yes
H Statistic - - - RE-no auto - 112 8 1.09 RE-no auto No
(jar - - - RE-auto - 105 8 1.87** RE-auto Yes
dynamic) 330 5 0.38 FE-no auto Yes 352 4 1.42 FE-no auto No
Trade Direction - - - FE-auto - 330 4 1.87** FE-auto Yes
330 6 -0.24 RE-no auto Yes 352 5 1.25 RE-no auto No
- - - RE-auto - 330 5 1.92** RE-auto Yes
Period Impact - - - FE - N/A 3 -1.32 FE No
Hausman RE - RE Yes
Test Yearly Impact - - - FE - N/A 7 -11.56 FE No
Statistic RE - RE Yes
Trade Direction N/A 5 77.72* FE Yes N/A 4 5.26 FE No
RE No RE Yes
NB: ., ** & *** denote significance at the 1,5and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of Ist order autocorrelation problem respectively.
271
Appendices
Appendix 5BP: Suitable equations for cut flowers trade between South Africa and
the EU countries
MODEL
InYjjt_1
InGDPPCjjt
REERt
Appendix 5BQ: Average actual, simulated and potential value of cut flowers trade
between South Africa and the EU countries for the periods 2000-2004 and 2005-
2009 in dollars (US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adjusted Potential
Name Code Value Value Value Value Value Value Value Value
(US$) (Logs) (US$) (Logs) (US$) (Logs) (US$l (Logs)
France 2000 - 2004 325,063 12.69177 409,947 12.92378 403,432 12.90776 406,690 12.91581FRA
2005 - 2009 399,969 0.00000 557,603 0.00000 618,370 0.00000 587,201 0.00000
Germany DEU 2000 - 2004
2,197654 14.60290 2,223,144 14.61443 2,221,142 14.61353 2,222,143 14.61398
2005 - 2009 4,551,610 15.33099 4,528,499 15.32590 4,519,374 15.32388 4,523,934 15.32489
Italy 2000 - 2004 472,912 13.06666 405,918 12.91391 410,238
12.92449 408,078
ITA 12.919212005 - 2009 0 0.00000 0 0.00000 0 0.00000 0 0.00000
Netherlands NLD 2000 - 2004
4,899,872 15.40472 4,089,394 15.22391 4,150,266 15.23868 4,119,830 15.23132
2005 - 2009 13,463,241 16.41547 11,475,958 16.25576 10,744,012 16.18986 11,103,956 16.22281
Portugal 2000 - 2004 567,995 13.24987 462401
13.04419 469,110 13.05859 465,756 13.05142
PRT 2005 - 2009 0 0.00000 0 0.00000 0 0.00000 0 0.00000
Spain 2000 - 2004 46,884 10.75543 87,726 11.38198 84,460 11.34403 86,093 11.36318ESP 2005 - 2009 63,425 11.05762 53,658 10.89038 51,518 10.84968 52,577 10.87003
United Kingdom GBR 2000 - 2004
781,708 13.56924 550,619 13.21880 564,507 13.24371 557,563 13.23133
2005 - 2009 0 0.00000 0 0.00000 0 0.00000 0 0.00000
272
Appendices
Appendix 5BR: Selection of the Estimator suitable for Frozen Fruits and Nuts
Exports from South Africa to the EU countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 105 5 1.97*** OLS Yes 112 4 3.78* OLS No
Wald Test FE or RE No FE or RE Yes
Statistic Yearly Impact 105 9 1.89*** OLS Yes 112 8 3.79" OLS No
FE or RE No FE or RE Yes
Export Direction 300 6 3.01 * OLS No 320 5 12.71* OLS No
FE or RE Yes FE or RE Yes
105 6 -0.81 FE-no auto Yes 112 4 1.55 FE-no auto No
Durbin Period Impact - - - - - 105 4 2.03** FE-auto Yes
Watson - - - - - 112 5 0.22 RE-no auto No
Statistic - - - - - 105 5 2.60** RE-auto Yes
(jor static) 105 10 -0.85 FE-no auto Yes 112 8 1.50 FE-no auto No
or Durbin- Yearly Impact - - - - - 105 8 2.04** FE-auto Yes
H Statistic - - - - - 112 9 0.21 RE-no auto No
(jor - - - - - 105 9 2.63** RE-auto Yes
dynamic) 315 -0.44 FE-no auto Yes 336 5 1.31 FE-no auto No
: Export Direction - - - FE-auto - 315 5 2.04*" FE-auto Yes
315 -1.64 RE-no auto Yes 336 6 0.80 RE-no auto No
- - - RE-auto - 315 6 2.17*" RE-auto Yes
Period Impact - - - - - N/A 4 9.39*** FE Yes
Hausman - - RE No
Test Yearly Impact - - - - - N/A 8 -2.56 FE No
Statistic - - RE Yes
Export Direction N/A 6 128.08* FE Yes N/A 5 -1.70 FE No
RE No RE Yes
NB: " ** & *** denote significance at the 1,5 and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of 1st order autocorrelation problem respectively.
273
Appendices
Appendix 5BS: Suitable equations for frozen fruits and nuts exports from South
Africa to the EU countries
MODEL Period Impact Yearly Impact Export Direction
Dynamic Static Dynamic Static Dynamic Static
ESTIMATORS FE FE FE FE FE FE RE RE FE FE RE RE
PERIOD 2000 2005 2000 2005 2000 2005 2000 2005 2000 2005 2000 2005
- - - - - - - - - - - -
VARIABLES 2004 2009 2004 2009 2004 2009 2004 2009 2004 2009 2004 2009
Constant - - - - - -936.2' -936.2' - - -34.62 -34.62
(-3.32) (-3.32) (-0.58) (-0.58)
InYijt_1 0.28' 0.28' - - 0.29' 0.29· - - DAD· 0.40· - -
(2.70) (2.70) (2.86) (286) (1,44) (7.44)
InGDPPCit -14.70 -14.70 -13.72 -13.72 -21..98 -21.98. -36.22 -36.22 3.00 3.00 7.46 7.46
(-0.89) (-089) (-081) (-0.81) (-0.67) (_0.67) (-1.05) (-1.05) (0.31) (0.31) (1.06) (106)
InGDPPCjt 20.32 20.32 19.27 19.27 17.16 17.16 15.2··' 15.2·" 0.32 0.32 0.69 0.69
(1.47) (1.47) ( 1.17) ( 1.17) (1.16) (1.16) (I 95) (1.95) (0.26) _(0.26) (1.50) (1.50)
REERt 1.64 1.64 1.43 1.43 0.71 0.71 2.40 2.40 -1.86··· -1.86"· -2.21 -2.21
(0.51) (0.51 ) (041) (0.41 ) (0.21 ) (0.21) (0.67) (0.67) (-1.93) (~1.93) (-151) (-1.51 )
D0004 / D0509 -3.56*" 0.24 0.34 1.15 - - - - - - - -
(-1.68) (0.07) (0.19) (0.37)
DOO/ D05 - - - ~1O.55'" . 209 0.53 3.12 - - -H.53}: (043) (0.27) (0.69)
"'l
DOl / D06 - - - - ;17.98· 1.68 0.01 3.72 - - - -
H.16) (0.28) (0.00) (0.60)
D02 / D07 - - - - ,26.85* . 2.57 3.29 5.25 - - - -
lC3.86) .lj .. . . (0.35) • (0.91) (068)
D03 / D08 - - - '18.31:· 4.01 2.50 744 - " - -
(-3:30) (OSI) (0.94) (0.87)
D04 / D09 - - - -25.85' . 058 045 3.82 - - -
. ("3.3:2) (0:08) (0.12) (049)
PTAycs - - - - - - - - ~()}7 -0.60 . -0.80 0.73
.t:.'. ('0.62) . _(-0.46) (-0.69) (047)
PTA no - - - - - - o.ez 0.77 0.67 0.35
.: J020L (061) (0.68) (0.25)
InDISTij - - - ,. - - - - - -1.57 -1.57
(-1.40) (-1.40)
Adjusted R' .0.55 0.55 0.34 0.34 0.54 0.54 0.93 0.93 057 0.57 0.33 0.33
Observations' 105 105 105 105 105 105 105 105 315 315 315 315
Cross-Sections 7 7 7 7 7 7 7 7 21 21 21 21
*, ** & *** denote significance at the 1,5 and 10 percent levels respectively. t-values are in parentheses
From 1994 la 2009, SA exported preserved fruits and nuts to the following 7 EU countries: AUT, BEL, DEU, FRA, GBR, NLD and SWE as well as to the following 14
non-EU countries that were added under the export direction model: AGO, AUS, CHE, DRC, lPN, KEN, MOZ, MUS, MWI, NZL, SYC, USA, 2MB and ZWE
Appendix 5BT: Average actual, simulated and potential value of frozen fruits and
nuts exports from South Africa to EU countries for the periods 2000-2004 and
2005-2009 in dollars (US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adjusted Potential
Name Code Value Value Value Value Value Value Value Value
(USS) (Logs) (USS) (Logs) (US$) (Logs) (US$) (Logs)
Austria AUT 2000 - 2004 7 1.98172 43 3.76274
45 3.81530 44 3.78936
2005 - 2009 0 0.00000 0 0.00000 0 0.00000 0 0.00000
Belgium BEL 2000 - 2004 780 6.65943 758 6.63099
790 6.67142 774 6.65141
2005 - 2009 38,639 10.56201 22,927 10.04009 8,555 9.05426 14,005 9.54718
France 2000 - 2004 309,654 12.64321 82,438 11.31981 66,171 11.10000 74,305 11.21593FRA 2005 - 2009 80,003 11.28982 133,286 11.80025 51,939 10.85783 83,203 11.32904
Germany 2000 - 2004 5,275 8.57071 20,934 9.94912 29,383 10.28816DElf 25,158
10.13294
2005 - 2009 98 4.58254 1,056 6.96182 820 6.70925 930 6.83553
Netherlands 2000 - 2004 1,737,491 14.36795 378,754 12.84464 277,604 12.53395 328,179 12.70131NLD 2005 - 2009 4,004,633 15.20296 8,600,989 15.96739 2,333,120 14.66272 4,479,636 15.31505
Sweden SWE 2000 - 2004 63
4.14749 273 5.61095 329 5.79579 301 5.70764
2005 - 2009 0 0.00000 0 0.00000 0 0.00000 0 0.00000
United Kingdom 2000 - 2004 417,503 12.94205 50,750 10.83467 34,781 10.45684 42,766 10.66349GBR 2(J(15- 2009 366,784 12.81253 1,129,029 13.93687 425,832 12.96180 693,380 13.44933
274
I
~-------------------------------------------------------------------------------------------------------------,
Appendices
Appendix 5BU: Selection of the estimator suitable for frozen fruits and nuts imports
from the EU countries to South Africa
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 105 5 2.01 *** OLS No 112 4 3.64* OLS No
Wald Test FEar RE Yes FE or RE Yes
Statistic Yearly Impact 105 9 1.95*** OLS No 112 8 3.62* OLS No
FEar RE Yes FE or RE Yes
Import Direction 345 6 2.70* OLS No 368 5 6.05* OLS No
FE or RE Yes FE or RE Yes
105 5 -0.72 FE-no auto Yes 112 4 1.48 FE-no auto No
Durbin Period Impact - - - - - 105 4 2.15*· FE-auto Yes
Watson - - - - - 112 6 0.14 RE-no auto No
Statistic - - - - - 105 6 2.55** RE-auto Yes
(jar static) 105 9 -1.0 I FE-no auto Yes 112 8 1.48 FE-no auto No
or Durbin- Yearly Impact - - - - - 105 8 2.13** FE-auto Yes
H Statistic - - - - - 112 9 0.13 RE-no auto No
(jar - - - - - 105 9 2.58** RE-auto Yes
dynamic) 345 6 -0.56 FE-no auto 368 5 1.51 FE-no auto No
Import Direction - - - - - 345 5 2.03*· FE-auto Yes
345 7 -1.84*** RE-no auto No 368 6 0.73 RE-no auto No
322 7 -0.67 RE-auto Yes 345 6 2.35** RE-auto Yes
Period Impact - - - - - N/A 4 16.13* FE Yes
Hausman - - RE No
Test Yearly Impact - - - - - N/A 8 8.57 FE No
Statistic - - RE Yes
Import Direction N/A 6 157.96* FE Yes N/A 5 -7.43 FE No
RE No RE Yes
NB: *, ** & *** denote significance at the 1,5 and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of ISI order autocorrelation problem respectively.
275
Appendices
Appendix 5BV: Suitable equations for frozen fruits and nuts imports from the EU
countries to South Africa.
MODEL
2004 2009
InYijl_!
InGDPPCil
InGDPPCjl
REERI
D0004 I D0509
Appendix 5BW: Average actual, simulated and potential value of frozen fruits and
nuts imports from the EU countries to South Africa for the periods 2000-2004 and
2005-2009 in dollars (US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adjusted Potential
Name Code Value Value Value
Belgium BEL 2000 - 2004
775
2005 - 2009 13,302
Denmark 7,443DNK 2000 - 20042005 - 2009 9.57451 17,266
France 2000 - 2004 4.17700 6.83613 1,914FRA 2005 - 2009 0.00000 0.00000 0
Germany 2000 - 2004 3.87805 149 5.00390 210 5.34534DEU 177 5.174622005 - 2009 77 4.33775 263 5.57292 302 5.71007 282 5.64150
Greece 2000 - 2004 676 6.51613 76 4.32518 75 4.31147GRC 75 4.318322005 - 2009 1,082 6.98658 3,078 8.03212 3,670 8.20808 3,361 8.12010
Netherlands NLD 2000 - 2004
285,945 12.56355 37,614 10.53513 37,742 10.53853 37,678 10.53683
2005 - 2009 1,340,080 0.00000 573,406 0.00000 439,997 0.00000 502,292 0.00000
United Kingdom GBR 2000 - 2004
6,074 8.71177 12,499 9.43337 23,252 10.05414 17,047 9.74376
2005 - 2009 3,886 8.26513 9,117 9.11793 10,771 9.28464 9,910 9.20128
276
Appendices
Appendix 5BX: Selection of the estimator suitable for frozen fruits and nuts trade
between South Africa and the EU countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 90 4 0.93 OLS Yes 96 3 2.47*** OLS No
Wald Test FE or RE No FE or RE Yes
Statistic Yearly Impact 90 8 0.94 OLS Yes 96 7 2.77** OLS No
FE or RE No FE or RE Yes
Trade Direction 165 5 2.03**' OLS No 176 4 4.17· OLS No
FE or RE Yes FE or RE Yes
90 4 -8.96' OLS-no auto No 96 3 1.23 FE-no auto No
Durbin Period Impact - - - FE-auto - 90 3 2.11 ** FE-auto Yes
Watson - - - RE-no auto - - - - RE-no auto -
Statistic - - - RE-auto - - - - RE-auto -
(for static) 90 8 00 OLS-no auto ? 96 7 1.21 FE-no auto No
or Durbin- Yearly Impact - - - FE-auto - 90 7 2.14** FE-auto Yes
H Statistic - - - RE-no auto - - - - RE-no auto -
(for - - - RE-auto - - - - RE-auto -
dynamic) 165 5 -1.24 FE-no auto Yes 176 4 1.23 FE-no auto No
Trade Direction - - - FE-auto - 165 4 2.22** FE-auto Yes
- - - RE-no auto - 176 5 0.41 RE-no auto No
- - - RE-auto - 165 5 2.46** RE-auto Yes
Period Impact - - - FE - - - - FE -
Hausman RE - RE -
Test Yearly Impact - - - FE - - - - FE -
Statistic RE - RE -
Trade Direction - - - FE - N/A 4 5.11 FE No
RE - RE Yes
NB: " •• & *** denote significance at the 1,5 and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of l" order autocorrelation problem respectively. 00
means infinity. ? means inconclusive.
277
Appendices
Appendix 5BY: Suitable equations for frozen fruits and nuts trade between South
Africa and the EU countries
MODEL
2004 2004
?
InYijt_1 ?
InGDPPCijt ?
REERt ?
D0004 / DOS09 ?
DOO /DOS
DOl f.D06
D02/D07 ?
D03/DOS ?
D04/D09 ?
PTAyes ?
PTAn.
Appendix 5BZ: Average actual, simulated and potential value of frozen fruits and
nuts trade between South Africa and the EU countries for the periods 2000-2004
and 2005-2009 in dollars (US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adjusted Potential
Name Code Value Value Value Value Value Value Value Value
(US$) (L02S) (US$) (L02S) (US$) (102S) (US$) .(Logs)_
Austria AUT 2000 - 2004 13 2.53625 25 3.21950 27 3.29910 26 3.25930
2005 - 2009 0 0.00000 0 0.00000 0 0.00000 0 0.00000
Belgium BEL 2000 - 2004 2,895 7.97059 4,614 8.43692 5,605 8.63135 5,109 8.534142005 - 2009 1,136,839 13.94376 43,563 10.68197 19,371 9.87151 29,049 10.27674
France 2000 - 2004 337,804FRA 12.73022 59,207 10.98880 47,480 10.76806 53,344 10.878432005 - 2009 0 0.00000 0 0.00000 0 0.00000 0 0.00000
Germany DEU 2000 - 2004 37,419 10.52994 61,439 11.02580 80792 11.29964 71,115 11.16272
2005 - 2009 1,453 7.28135 25,633 10.15163 58,532 10.97732 38,734 10.56448
Netherlands NLD 2000 - 2004 2,170,532 14.59048 1,260,265 14.04683 1,337,080 14.10600 1,298,673 14.076422005 - 2009 5,816,648 15.57623 4,213,757 15.25387 4,215,225 15.25421 4,214,491 15.25404
United Kingdom GBR 2000 - 2004 537,199 13.19412 485,597 13.09314 578,760 13.26864 532,179 13.18089
2005 - 2009 414,957 12.93593 613,330 13.32666 798,336 13.59028 699,745 13.45847
278
Appendices
Appendix SeA: Selection of the estimator suitable for preserved fruits and nuts
exports from South Africa to the EU countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 225 5 5.44* OLS No 240 4 38.48* OLS No
Wald Test FE or RE Yes FEor RE Yes
Statistic Yearly Impact 225 9 5.48* OLS No 240 8 38.74* OLS No
FE or RE Yes FE or RE Yes
Export Direction 1155 6 2.83* OLS No 1360 5 13.36* OLS No
FE or RE Yes FEor RE Yes
225 5 0.42 FE-no auto Yes 240 4 1.27 FE-no auto No
Durbin Period Impact - - - - - 225 4 1.82** FE-auto Yes
Watson 225 6 -1.23 RE-no auto Yes 240 5 1.07 RE-no auto No
Statistic - - - - - 225 5 1.80** RE-auto Yes
(jar static) 225 5 0.40 FE-no auto Yes 240 8 1.28 FE-no auto No
or Durbin- Yearly Impact - - - - - 225 8 1.80 FE-auto ?
H Statistic 225 lO -1.25 RE-no auto Yes 240 9 1.08 RE-no auto No
(jar - - - - - 225 9 1.79 RE-auto ?
dynamic) 1275 6 -0.22 FE-no auto Yes 1360 5 1.13 FE-no auto No
Export Direction - - - - - 1275 5 1.95** FE-auto Yes
1275 7 -1.48 RE-no auto Yes 1360 6 0.73 RE-no auto No
- - - - - 1275 6 2.05** RE-auto Yes
Period Impact N/A 5 92.08* FE Yes N/A 4· 22.39* FE Yes
Hausman RE No RE No
Test Yearly Impact N/A 9 92.54* FE Yes N/A 8 13.06 FE No
Statistic RE No RE Yes
Export Direction N/A 6 387.54* FE Yes N/A 5 18.12* FE Yes
RE No RE No
NB: *, ** & *** denote significance at the 1,5 and lO percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OlS, FE & RE denote Pooled Ordinary least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of I" order autocorrelation problem respectively. ?
means inconclusive.
279
Appendices
Appendix 5CB: Suitable equations for preserved fruits and nuts exports from South
Africa to the EU countries
MODEL
2009
InYijt_1
InGDPPCit
InGDPPCjt
REERt
00004 / 00509
DOO/005
001/006
002/007
003/008
004/009
PTAyes
PTA no
280
Appendices
Appendix SCC: Average actual, simulated and potential value of preserved fruits
and nuts exports from South Africa to EU countries for the periods 2000-2004 and
2005-2009 in dollars (US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adjusted Potential
Name Code Value Value Value Value Value Value Value Value
(USS) (Logs) (USS) (Logs) (USS) (Logs) (USS) (Logs)
15.13369
Austria AlIT 2000 - 2004 3.884,273 15.17245 3,762,815 15.14068 3,710,384
15.12665 3,736,599
2005 - 2009 7.843,104 15.87515 3,775,602 15.14407 3,080,465 14.94059 3,428,034 15.04750
Belgium BEL 2000 - 2004 5,308,462 15.48481 4,784,949 15.38099 4,745,347 15.37268
4,765,148 15.37684
2(l()S - 2009 9,657,804 16.08328 4,145,049 15.23743 3,551,463 15.08287 3,848,256 15.16313
331,094 12.71016
Denmark 2000 - 2004 317,245 12.66743 330,647 12.70881 331,541 12.71151DNK 2005 - 2009 117,862 11.67727 1,865,217 14.43889 2,101,267 14.55805 1,983,242 14.50024
Finland FIN 20UO- 2004 20,396 9.92310 98,587 11.49869 108,133 11.59112
103,360 11.54597
2005 - 2009 592,135 13.29149 2,235,348 14.61991 2,266,998 14.63397 2,251,173 14.62696
France FRA 2000 - 2004 2.260,406 14.63106 1,952,045 14.48439 1,930,681 14.47338 1,941,363 14.478902005 - 2009 8,727,920 15.98204 4,037,147 15.21105 3,480,195 15.06260 3,758,671 15.13958
Germany DEll 2UOO- 2004 18,023,605 16.70719 16,275,208 16.60515
16,131,414 16.59628 16,203,311 16.60073
2UOS- 2009 40,176,037 17.50878 5,173,891 15.45914 4,020,728 15.20697 4,597,310 15.34098
Greece GRC 2000 - 2004 675,339 13.42297 521,102 13.16370 512,031
13.14614 516,567 13.15496
2005 - 2009 379,746 12.84726 3,203,222 14.97967 3,455,739 15.05555 3,329,480 15.01833
Ircland IRL 2000 - 2004 81,484 11.30816 255,880 12.45246 275,271 12.52551 265,576
12.48966
2(l()S - 2009 435,545 12.98435 1,929,346 14.47269 1,979,545 14.49838 1,954,446 14.48562
Italy 2000 - 2004 1,764,877 14.38359 1,783,613 14.39415 1,785,018 14.39494 1,784,315 14.39455ITA 2005 - 2009 3,115,837 14.95201 3,754,072 15.13835 3,487,400 15.06467 3,620,736 15.10219
Luxembourg LUX 2000 - 2004 4,967 8.51064 480 6.17362 447
6.10197 463 6.13844
2005 - 2009 0 0.00000 0 0.00000 0 0.00000 0 0.00000
Netherlands 2000 - 2004 7,601,306 15.84383 7,465,484 15.82580 7,454,383 15.82431 7,459,933 15.82506NLD 2005 - 2009 22,773,727 16.94112 4,577,859 15.33674 3,691,851 15.12164 4,134,855 15.23496
Portugal 2000 - 2004 2,882,330 14.87411 2,451,996 14.71241 2,421,929 14.70007 2,436,962 14.70626PRT 2005 - 2009 932,190 13.74529 4,581,349 15.33750 4,746,835 15.37299 4,664,092 15.35540
Spain 2000 - 2004 1,568,917 14.26590 1,730,990 14.36420 1,743,692 14.37152 1,737,341 14.36787ESP 2005 - 2009 3,297,577 15.00870 4,327,818 15.28057 4,044,311 15.21282 4,186,064 15.24727
Sweden 2000 - 2004 906,854 13.71774 1,202,568 13.99997 1,227,375 14.02039 1,214,972 14.01023SWE 2005 - 2009 3,514,296 15.07235 2,851,448 14.86334 2,574,406 14.76113 2,712,927 14.81354
United Kingdom 2000 - 2004 16,361,609 16.61045 16,491,988 16.61839 16,503,394 16.61908 16,497,691 16.61873GBR 2005 - 2009 57,680,100 17.87042 5,177,018 15.45974 3,909,891 15.17902 4,543,454 15.32920
281
Appendices
. Appendix SeD: Selection of the estimator suitable for preserved fruits and nuts
imports from the EU countries to South Africa
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 135 5 5.17* OLS No 144 4 14.12* OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 135 9 5.09* OLS No 144 8 14.05* OLS No
FE or RE Yes FE or RE Yes
Import Direction 570 6 2.91 * OLS No 608 5 10.29* OLS No
FE or RE Yes FEar RE Yes
135 5 -0.62 FE-no auto Yes 144 4 1.67 FE-no auto No
Durbin Period Impact - - - - - 135 4 1.97** FE-auto Yes
Watson 135 6 -1.89*** RE-no auto No 144 5 1.19 RE-no auto No
Statistic 126 6 -0.89 RE-auto Yes 135 5 2.19** RE-auto Yes
(for staticï 135 9 -0.65 FE-no auto Yes 144 8 1.67 FE-no auto No
or Durbin- Yearly Impact - - - - - 135 8 1.98** FE-auto Yes
H Statistic 135 10 -1.97** RE-no auto No 144 9 1.19 RE-no auto No
(for 126 10 -0.92 RE-auto Yes 135 9 2.22** RE-auto Yes
dynamic) 570 6 0.06 FE-no auto Yes 608 5 1.33 FE-no auto No
Import Direction - - - - - 570 5 1.96** FE-auto Yes
570 7 -1.42 RE-no auto Yes 608 6 0.78 RE-no auto No
- - - - - 570 6 2.12** RE-auto Yes
Period Impact N/A 5 289.55* FE Yes N/A 4 8.01 *** FE Yes
Hausman RE No RE No
Test Yearly Impact N/A 9 191.08* FE Yes N/A 8 4.69 FE No
Statistic RE No RE Yes
Import Direction N/A 6 224.38* FE Yes N/A 5 -0.97 FE No
RE No RE Yes
NB: *, ** & *** denote significance at the I, 5 and 10percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of I" order autocorrelation problem respectively.
282
Appendices
Appendix SCE: Suitable equations for preserved fruits and nuts imports from the
EU countries to South Africa
MODEL
2004 2009
InYijl_1
InGDPPCi,
InGDPPCj,
D0004 1D0509
DOOID05
DOl/D06
D02/D07
D03/D08
D04/D09
PTAyes
.PTA no
283
Appendices
Appendix 5CF: Average actual, simulated and potential value of preserved fruits
and nuts imports from the EU countries to South Africa for the periods 2000-2004
and 2005-2009 in dollars (US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adiusted Potential
Name Code Value Value Value Value Value Value Value Value
(llS$) (Logs) russi (Lozs) russ: (Loss) russ: (Logs)
2000 - 2004 288 5.66296 0 0.00000 0 0.00000 0 0.00000Austria Al!T 2005 - 2009 2,199 7.69579 1,976 7.58903 1,655 7.41169 1,809 7.50036
Belgium BEL 2000 - 2004 0 0.00000 93 4.53569 158,476 11.97336
3,845 8.25452
2005 - 2009 14,233 9.56332 333,877 12.71853 356,993 12.78547 345,241 12.75200
France 2000 - 2004 33,884 10.43069 12 2.46711 619 6.42862
85 4.44786
FRA 2005 - 2009 60,803 11.01540 578,638 13.26843 672,887 13.41933 623,986 13.34388
Germany DEl! 2000 - 2004 565 6.33704 155 5.04077 751,083
13.52927 10,775 9.28502
2005 - 2009 55,260 10.91980 1.019.173 13.83450 1,111,866 13.92155 1,064,511 13.87803
Greece GRC 2000 - 2004 0 0.00000 137 4.92279
518,785 13.15924 8,442 9.04102
2005 - 2009 62,490 11.04277 69,104 11.14336 62,581 11.04421 65,761 11.09379
Italy 2000 - 2004 246,368 12.41458 271 5.60043 4,086,958
15.22331 33,252 10.41187
ITA 2005 - 2009 51,684 10.85291 1,374,541 14.13363 1,717,897 14.35661 1,536,659 14.24512
Netherlands 2000 - 2004 318,578 12.67162 37 3.60249 9,403 9.14873 587 6.37561NLD 2005 - 2009 47,243 10.76306 514,114 13.15020 513.392 13.14880 513,753 13.14950
Spain ESI) 2000 - 2004 193,221 1'2.17159 355 5.87106 9,803,253 16.09822 58,962 10.98464
2005 - 2009 286,215 12.56450 637,880 13.36591 678,055 13.42698 657,661 13.39644
United Kingdom 2000 - 2004 21,678 9.98403 703 6.55549 91,558,857 18.33249 253,721 12.44399GBR 2005 - 2009 376,709 12.83923 794,923 13.58600 700,960 13.46021 746,465 13.52310
284
Appendices
Appendix 5CG: Selection of the estimator suitable for preserved fruits and nuts
trade between South Africa and the EU countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 135 4 3.99* OLS No 144 3 18.41 * OLS No
Wald Test FE or RE Yes FEor RE Yes
Statistic Yearly Impact 135 8 3.59* OLS No 144 7 18.50* OLS No
FE or RE Yes FE or RE Yes
Trade Direction 480 5 6.86* OLS No 512 4 14.34* OLS No
FE or RE Yes FE or RE Yes
135 4 0.60 FE-no auto Yes 144 3 1.33 FE-no auto No
Durbin Period Impact - - - - - 135 3 1.86** FE-auto Yes
Watson 135 5 0.53 RE-no auto Yes 144 4 1.32 RE-no auto No
Statistic - - - RE-auto Yes 135 4 1.86** RE-auto Yes
(jar static) 135 8 0.26 FE-no auto Yes 144 7 1.36 FE-no auto No
or Durbin- Yearly Impact - - - - - 135 7 1.85** FE-auto Yes
H Statistic 135 9 0.25 RE-no auto Yes 144 8 1.34 RE-no auto No.
(jar - - - RE-auto Yes 135 8 1.85** RE-auto Yes
dynamic) 480 5 0.09 FE-no auto Yes 512 4 1.80 FE-no auto No
Trade Direction - - - - - 480 4 1.66 FE-auto No
480 6 -0.08 RE-no auto Yes 512 5 1.60 RE-no auto No
- - - - - 480 5 1.74 RE-auto No
Period Impact N/A 4 17.87* FE Yes N/A 3 0.40 FE Yes
Hausman RE No RE No
Test Yearly Impact N/A 8 16.82** FE Yes N/A 7 0.56 FE Yes
Statistic RE No RE No
Trade Direction N/A 5 215.36* FE Yes - - - FE Yes
RE No RE No
NB: *, ** & *** denote significance at the I,5and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of Ist order autocorrelation problem respectively.
285
Appendices
Appendix SCH: Suitable equations for preserved fruits and nuts trade between
South Africa and the EU countries
MODEL
2009
286
Appendices
Appendix 5CI: Average actual, simulated and potential value of preserved fruits
and nuts trade between South Africa and the EU countries for the periods 2000-
2004 and 2005-2009 in dollars (US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adiusted Potential
Name Code Value Value Value Value Value Value Value Value
(US$) (Loes) (US$) (Loes) (lIS$) (Loes) (US$) (Loes)
Austria AUT 2000 - 2004 3,893,219 15.17475 4,091,915
15.22452 4,117,406 15.23073 4,104,660 15.22763
2005 - 2009 7,879,929 15.87983 8,250,062 15.92573 8,297,296 15.93144 8,273,679 15.92859
Belgium 2000 - 2004 5,50 1,856 15.52060 5,409,165 15.50361 5,397,474 15.50144
5,403,319 15.50252
BEL 2005 - 2009 10,151,577 16.13314 10,810,053 16.19599 10,896,515 16.20395 10,853,284 16.19998
France 2000 - 2004 2,268,381 14.63458 2,147,273 14.57971 2,133,316 14.57319
2,140,294 14.57645
FRA 2005 - 2009 8,934,059 16.00538 7,050,085 15.76855 6,847,633 15.73941 6,948,859 15.75409
2000 - 2004 18,119,341 16.71249 18,170,118 16.71529 18,177,170 16.71568 18,173,644 16.71548Germany OEU 2005 - 2009 40,733,00 I 17.52255 40,646,052 17.52041 40,633,958 17.52011 40,640,005 17.52026
Greece 13.44993 707,593 13.46962GRC 2000 - 2004 1,031,739 13.84676 721,389 13.48893 693,7972005 - 2009 587,252 13.28321 767,256 13.55058 788,872 13.57836 778,064 13.56456
Italy ITA 2000 - 2004 1,940,168 14.47829 2,142,592 14.57753
2,168,009 14.58932 2,155,30 I 14.58344
2005 - 2009 3,500,229 15.06834 4,233,151 15.25846 4,329,654 15.28100 4,281,403 15.26979
Netherlands NLO 2000 - 2004 7,836,483 15.87430 8,399,680 15.94370 8,476,651
15.95283 8,438,166 15.94828
2005 - 2009 23,366,308 16.96681 20,850,597 16.85289 20,537,983 16.83779 20,694,290 16.84537
Spain ESP 2000 - 2004 1,759,926 14.38078 2,009,691 14.51349 2,041,477
14.52918 2,025,584 14.52137
2005 - 2009 3,863,375 15.16705 3,961,100 15.19203 3,972,788 15.19498 3,966,944 15.19351
United Kingdom 2000 - 2004 16,665,918 16.62888 17,977,444 16.70463 18,167,152 16.71513 18,072,298 16.70989GBR 2005 - 2009 60,553,801 17.91904 47,713,245 17.68072 46,147,160 17.64735 46,930,203 17.66417
287
Appendices
Appendix 5CJ: Selection of the estimator suitable for fruits and vegetable juices
exports from South Africa to the EU countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 210 5 9.32* OLS No 224 4 27.17* OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 210 9 9.11 * OLS No 224 8 27.23* OLS No
FE or RE Yes FEar RE Yes
Export Direction 1290 6 4.71 * OLS No 1376 5 17.16* OLS No
FE or RE Yes FE or RE Yes
210 5 0.23 FE-no auto Yes 224 4 1.76 FE-no auto No
Durbin Period Impact - - - FE-auto - 210 4 1.99** FE-auto Yes
Watson 210 6 -0.81 RE-no auto Yes 224 6 1.12 RE-no auto No
Statistic - - - RE-auto - 210 6 2.06** RE-auto Yes
(jar static) 210 9 0.20 FE-no auto Yes 224 8 1.75 FE-no auto No
or Durbin- Yearly Impact - - - FE-auto - 210 8 2.00** FE-auto Yes
H Statistic 210 10 -0.92 RE-no auto Yes 224 9 l.l2 RE-no auto No
(jar - - - RE-auto - 210 9 2.07** RE-auto Yes
dynamic) 1290 6 -0.02 FE-no auto Yes 1376 5 1.53 FE-no auto No
Export Direction - - - FE-auto . - 1290 5 1.97** FE-auto Yes
1290 7 -1.50 RE-no auto Yes 1376 6 0.96 RE-no auto No
- - - RE-auto - 1290 6 2.09** RE-auto Yes
Period Impact N/A 5 220.39* FE Yes N/A 4 4.51 FE No
Hausman RE No RE Yes
Test Yearly Impact N/A 9 210.85 FE Yes N/A 8 -34.70* FE Yes
Statistic RE No RE No
Export Direction N/A 6 644.88* FE Yes N/A 5 37.17* FE Yes
RE No RE No
NB: ", ** & *** denote significance at the I, 5 and 10percent levels respectively. N & K denote the sample size and the number of regressors
respectively. aLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assumiOK no autocorrelation problem and estimation with the correction of I st order autocorrelation problem respectively.
288
Appendices
Appendix SCK: Suitable equations for fruits and vegetable juices exports from
South Africa to the EU countries.
MODEL
2009
InYijt-1
InGOPPCit
InGDPPCjt
REERt
D0004 / D0509
DOO/D05
DOl / D06
D02/D07
D03/D08
004/009
PTAyes
PTA no
289
Appendices
Appendix SeL: Average actual, simulated and potential value of fruits and
vegetable juices exports from South Africa to EU countries for the periods 2000-
2004 and 2005-2009 in dollars (US$) and natural logarithms (Logs)
Country Count!")' Period Actual Simulated Adjusted Potential
Name Code Value Value Value Value Value Value Value Value
(llS$) (Logs) (US$) (Logs) (US$) (Logs) (US$) (Lo_gs)
Austria AUT 2000 - 2004 162 5.08583 249 5.51580 303 5.71453 276
5.62009
2005 - 2009 39 3.67367 43,101 10.67130 153,980 11.94458 98,541 11.49822
Belgium BEL 2000 - 2004 587,876 13.28427 670,988 13.41651 1,020,915 13.83621 845,952 13.648222005 - 2009 2,049,427 14.53307 112,923 11.63446 258,152 12.46130 185,537 12.13101
Denmark DNK 2000 - 2004 496,237 13.11481 131,351 11.78563 166,676 12.02381
149,013 11.91179
2005 - 2009 213,939 12.27345 134,006 11.80564 377,608 12.84161 255,807 12.45218
Finland FIN 2000 - 2004 2 0.68837 0 -2.21302 0 -2.23048 0
-2.22171
2005 - 2009 289 5.66796 53,938 10.89559 212,557 12.26697 133,248 11.79997
France FRA 2000 - 2004 156,298 11.95952 286,386 12.56510 442,190 12.99950 364,288 12.805702(105- 2009 1,079,265 13.89179 96,050 11.47263 225,157 12.32455 160,604 11.98669
Germany DEU 2000 - 2004 982,468 13.79782 1,194,089 13.99289 1,865,517 14.43905 1,529,803
14.24065
2005 - 2009 797,673 13.58945 105,452 11.56601 256,947 12.45663 181,200 12.10736
Greece GRC 2000 - 2004 74,243 11.21510 10,515 9.26053 12,120 9.40265 11,318 9.334112005 - 2009 158,049 11.97066 38,590 10.56074 90,426 11.41228 64,508 11.07454
Ireland IRL 2000 - 2004 247 5.51116 496 6.20594 611 6.41479 553 6.315802005 - 2009 332,254 12.71365 121,796 11.71010 325,854 12.69421 223,825 12.31862
Italy ITA 2000 - 2004 6,937 8.84457 3,142 8.05265 3,797 8.24189
3,469 8.15174
2005 - 2009 262,642 12.47855 57,061 10.95187 137,120 11.82861 97,090 11.48340
Netherlands NLD 2000 - 2004 18,036,986 16.70794 I 1,928,505 16.29444 18,773,433 16.74795 15,350,969 16.546692005 - 2009 52,035,291 17.76743 175,981 12.07813 330,545 12.70850 253,263 12.44218
Portugal PRT 2000 - 2004 468,700 13.05772 622,912 13.34216 959,673 13.77435 791,292 13.581422005 - 2009 1,294,297 14.07348 34,509 10.44898 68,480 11.13429 51,494 10.84923
Spain ESP 2000 - 2004 2,627,827 14.78167 3,470,974 15.05995 6,907,489 15.74812 5,189,231 15.462102005 - 2009 7,517,189 15.83270 67,535 11.12040 129,425 11.77086 98,480 11.49761
Sweden SWE 2000 - 2004 30,472 10.32456 37,100 10.52137 18,994 9.85190 28,047 10.241642005 - 2009 4,364 8.38121 96,077 11.47290 346,505 12.75565 221,291 12.30723
United Kingdom 1,487,491 14.21260GBR 2000 - 2004 1,312,675 14.08758 2,129,651 14.57147 845,331 13.647482005 - 2009 2,436,290 14.70599 139,656 11.84693 325,561 12.69330 232,608 12.35711
290
Appendices
Appendix 5CM: Selection of the estimator suitable for fruits and vegetable juices
imports from the EU countries to South Africa.
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 195 5 4.24 OLS No 208 4 16.96* OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 195 9 4.18* OLS No 208 8 17.05· OLS No
FE or RE Yes FEor RE Yes
Import Direction 645 6 4.25* OLS No 688 5 14.46* OLS No
FE or RE Yes FE or RE Yes
195 5 0.63 FE-no auto Yes 208 4 1.29 FE-no auto No
Durbin Period Impact - - - FE-auto - 195 4 1.77 FE-auto ?
Watson 195 6 -0.77 RE-no auto Yes 208 6 0.71 RE-no auto No
Statistic - - - RE-auto - 195 6 1.86*· RE-auto Yes
(jar static) 195 9 0.60 FE-no auto Yes 208 8 1.30 FE-no auto No
or Durbin- Yearly Impact - - - FE-auto - 195 8 1.79 FE-auto ?
H Statistic 195 10 -0.86 RE-no auto Yes 208 9 0.71 RE-no auto No
(jar - - - RE-auto - 195 9 1.87** RE-auto Yes
dynamic) 645 6 0.10 FE-no auto Yes 688 5 1.56 FE-no auto No
Import Direction - - - FE-auto - 645 5 1.96** FE-auto Yes
645 7 -1.59 RE-no auto Yes 688 6 0.87 RE-no auto No
- - - RE-auto - 645 6 2.06** RE-auto Yes
Period Impact N/A 5 95.38* FE Yes N/A 4 -0.10 FE No
Hausman RE No RE Yes
Test Yearly Impact N/A 9 95.05· FE Yes N/A 8 0.28 FE No
Statistic RE No RE Yes
Import Direction N/A 6 351.39 FE Yes N/A 5 -3.07 FE No
RE No RE Yes
NB: ., ** & *** denote significance at the 1,5 and 10 percent levels respectively. N'& K denote the sample size and the number of regressors
respectively. aLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of ISI order autocorrelation problem respectively. ?
means inconclusive.
291
Appendices
Appendix SeN: Suitable equations for fruits and vegetable juices imports from the
EU countries to South Africa.
MODEL
InYijl_1
InGOPPCit
InGOPPCjl
00004 / 00509
DOO/ 005
001 /006
002/007
003/008
004/009
PTAycs
PTA no
292
Appendices
Appendix seo: Average actual, simulated and potential value of fruits and
vegetable juices imports from the EU countries to South Africa for the periods
2000-2004 and 2005-2009 in dollars (US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adjusted Potential
Name Code Value Value Value Value Value Value Value Value
(llS$) (Logs) (US$) (Logs) (US$) (Logs) (US$) (Logs)
Austria AUT 2000 - 2004 36,987 10,51832 8,184 900990
79,896 11.28848 25,570 10.14919
2005 - 2009 2,940 7.98627 16,657 9.72058 18,637 9.83288 17,619 9.77673
Belgium BEL 2000 - 2004 28,666 10.26345 3,814 8.24656 31,397 10.35448
10,944 9.30052
2005 - 2009 75,316 11.22945 69,657 11.15133 69,136 11.14383 69,396 11.14758
291 5.67381
Denmark 574 6.35278DNK 2000 - 2004 131 4.87713 148 4.994852005 - 2009 8,096 8.99917 9,071 9.11283 9,151 9.12160 9,111 9.11721
France FRA 2000 - 2004 5,916 8.68537 2,861 7.95875 24,797
10.11850 8,422 9.03862
2005 - 2009 65,936 11.09644 45,445 10.72425 43,914 10.69000 44,673 10.70712
Germany DEll 2000 - 2004 67,425 11.11878 15,900 9.67409 212,557
12.26697 58,135 10.97053
2005 - 2009 204,409 12.22788 260,639 12.47089 267,613 12.49730 264,103 12.48409
Greece GRC 2000 - 2004 396 5.98241 3 0.96459 3
1.15547 3 1.06003
2005 - 2009 19 2.95971 142 4.95353 154 5.03429 148 4.99391
Ireland IRL 2000 - 2004 555 6.31942 32 3.46206
71 4.25959 48 3.86083
2005 - 2009 3,939 8.27866 1,575 7.36202 1,489 7.30573 1,531 7.33387
Italy ITA 2000 - 2004 55,172 10.91821 12,297 9.41708 151,068
11.92548 43,100 10.67128
2005 - 2009 287,067 12.56747 160,330 11.98499 150,908 11.92443 155,548 11.95471
Netherlands NLD 2000 - 2004 528,395 13.17760 59,406
10.99215 1,067,550 13.88088 251,830 12.43651
2005 - 2009 1,452,817 14.18901 1,649,576 14.31603 1,676,351 14.33213 1,662,909 14.32408
Portugal PRT 2000 - 2004 3,072 8.02999 257 5.54870 979
6.88662 502 6.21766
2005 - 2009 85,760 11.35931 11,550 9.35448 9,852 9.19541 10,667 9.27494
Spain ESI) 2000 - 2004 87 4.47048 214 5.36739 978 6.88581 458
6.12660
2005 - 2009 35,903 10.48858 17,696 9.78107 16,683 9.72216 17,182 9.75161
Sweden SWE 2000 - 2004 2 0.78798 I -0.49086
I -0.60776 I -0.54931
2005 - 2009 32 3.48074 32 3.45457 32 3.45384 32 3.45421
United Kingdom GBR 2000 - 2004 123,602 11.72482 23,022 10.04422
333,706 12.71802 87,651 1138112
2005 - 2009 239,837 12.38772 386,093 12.86383 407,329 12.91738 396,569 12.89061
293
Appendices
Appendix SCP: Selection of the Estimator suitable for Fruits and Vegetable Juices
Trade between South Africa and the EU countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 195 4 5.29· OLS No 208 3 19.84· OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 195 8 5.46* OLS No 240 7 79.61· OLS No
FE or RE Yes FE or RE Yes
Trade Direction 630 5 5.75· OLS No 672 4 20.26· OLS No
FE or RE Yes FEor RE Yes
195 4 0.03 FE-no auto Yes 208 3 1.74 FE-no auto ?
Durbin Period Impact - - - FE-auto - 195 3 1.88·· FE-auto Yes
Watson 195 5 -1.85*·· RE-no auto No 208 4 1.35 RE-no auto No
Statistic 182 5 -0.41 RE-auto Yes 195 4 2.14** RE-auto Yes
(for static) 195 8 0.09 FE-no auto Yes 240 7 1.05 FE-no auto No
or Durbin- Yearly Impact - - - FE-auto - 225 7 I.71 FE-auto No
H Statistic 195 9 -1.81··* RE-no auto No 240 8 0.88 RE-no auto No
(for 182 9 -0.58 RE-auto Yes 225 8 1.61 RE-auto No
dynamic) 630 5 0.26 FE-no auto Yes 672 4 1.69 FE-no auto No
Trade Direction - - - FE-auto - 630 4 1.90·· FE-auto Yes
630 6 -0.58 RE-no auto Yes 672 5 1.42 RE-no auto No
- - - RE-auto - 630 5 1.98·* RE-auto Yes
Period Impact N/A 4 118.03· FE Yes N/A 3 -27.79· FE Yes
Hausman RE No RE No
Test Yearly Impact N/A 8 130.52* FE Yes N/A 7 32.72* FE Yes
Statistic RE No RE No
Trade Direction N/A 5 290.70* FE Yes N/A 4 32.80· FE Yes
RE No RE No
NB: ., •• & * •• denote significance at the I, 5 and 10percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of ISI order autocorrelation problem respectively. ?
means inconclusive.
294
Appendices
Appendix 5CQ: Suitable equations for fruits and vegetable juices trade between
South Africa and the EU countries
MODEL
InGDPPCijt
REER,
00004 / 00509
DOO/ 005
001/006
002/007
003/008
004/009
PTAyes
295
Appendices
Appendix SeR: Average actual, simulated and potential value of fruits and
vegetable juices trade between South Africa and the EU countries for the periods
2000-2004 and 2005-2009 in dollars (US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adiusted Potential
Name Code Value Value Value Value Value Value Value Value
(US$) (L02S) (US$) (Logs) (US$) (Locs) WSS) (Logs)
201,856 12.21531 198,496 12.19853
Austria 2000 - 2004 128,018 11.75993 195,137 12.18146AUT 2005 - 2009 4,495 8.41065 50,273 10.82522 59,080 10.98664 54,676 10.90918
13.43100
Belgium 13.43468 680,781BEL 2000 - 2004 624,415 13.34457 678,268 13.42730 683,2952005 - 2009 2,150,196 14.58107 1,539,164 14.24675 1,493,599 14.21670 1,516,381 14.23184
2000 - 2004 502,594 13.12754 229,241 12.34253 215,041 12.27859 222,141 12.31107Denmark DNK 2005 - 2009 228,331 12.33855 313,123 12.65435 321,037 12.67931 317,080 12.66691
2000 - 2004 202,767 12.21981 404,971 12.91157 429,656 12.97074 417,314 12.94159France FRA
2005 - 2009 1,206,932 1400359 716.532 13.48218 685,593 13.43804 701,063 13.46035
Germany DEU 2000 - 2004 1,070,226 13.88338 1,504,268 14.22382 1,553,741
14.25618 1,529,004 14.24013
2005 - 2009 1,082,767 13.89503 1,634,006 14.30655 1,695,861 14.34370 1,664,934 14.32530
2000 - 2004 79,719 11.28626 13,285 9.49441 11,899 9.38421 12,592 9.44083Greece GRC 2005 - 2009 159,883 11.98220 97,574 11.48837 94,201 11.45319 95,888 I 1.47093
Ireland fRL 2000 - 2004 2,154 7.67509 2,591 7.85974 2,615
7.86904 2,603 7.86440
2005 - 2009 412,320 12.92955 46,675 10.75096 40,392 10.60638 43,533 10.68128
2000 - 2004 93,574 11.44651 74,748 11.22187 73,523 11.20536 74,136 11.21365Italy ITA 2005 - 2009 747,543 13.52455 227,652 12.33557 207,760 12.24414 217,706 12.29090
2000 - 2004 18,646,700 16.74118 12,868,255 16.37027 12,349,270 16.32911 12,608,763 16.34990Netherlands NLD 2005 - 2009 53,597,301 17.79701 38,017,437 17.45356 36,577,872 17.41495 37,297,655 17.43444
Portugal PRT 2000 - 2004 473,713 13.06836 630,958 13.35499
647,212 13.38043 639,085 13.36779
2005 - 2009 1,384,548 14.14088 1,276,095 14.05932 1,266,906 14.05209 1,271,501 14.05571
Spain 2000 - 2004 2,632,794 14.78356 3,523,211· 15.07488 3,628,751 15.10440 3,575,981
15.08975
ESf' 2005 - 2009 8,335,671 15.93605 7,556,725 15.83795 7,482,257 15.82804 7,519,491 15.8330 I
Sweden SWE 2000 - 2004 30,477
10.32474 39,589 10.58632 40,311 10.60438 39,950 10.59539
2005 - 2009 4,469 8.40486 26,335 10.17867 29,429 10.28972 27,882 10.23574
United Kingdom GBR 2000 - 2004
1,475,855 14.20475 2,720,977 14.81650 2,891,581 14.87731 2,806,279 14.84737
2005 - 2009 2,693,069 14.80619 3,725,762 15.13078 3,843,677 15.16194 3,784,720 15.14648
296
Appendices
Appendix ses: Selection of the estimator suitable for wine exports from South
Africa to the EU countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 225 5 5.70· OLS No 240 4 101.21* OLS No
Wald Test FE or RE Yes FEor RE Yes
Statistic Yearly Impact 225 9 6.26* OLS No 240 8 107.88* OLS No
FE or RE Yes FEor RE Yes
Export Direction 1530 6 4.64* OLS No 1632 5 22.01* OLS No
FE or RE Yes FE or RE Yes
225 5 -0.64 FE-no auto Yes 240 4 1.52 FE-no auto No
Durbin Period Impact - - - - - 225 5 1.95** FE-auto Yes
Watson 225 6 -1.23 RE-no auto Yes 240 5 1.34 RE-no auto No
Statistic 135 6 -0.10 RE-auto Yes 150 5 2.06*· RE-auto Yes
(jar static) 225 9 -0.75 FE-no auto Yes 240 8 1.55 FE-no auto No
or Durbin- Yearly Impact - - - - - 225 8 2.03*· FE-auto Yes
H Statistic 225 10 -1.22 RE-no auto Yes 240 9 1.37 RE-no auto No
(jar - - - - - 225 9 1.96** RE-auto Yes
dynamic) 1530 6 -0.37 FE-no auto Yes 1632 5 1.44 FE-no auto No
Export Direction - - - - - 1530 5 2.04*· FE-auto Yes
1530 7 -1.70**" RE-no auto No 1632 6 1.05 RE-no auto No
1428 7 -0.17 RE-auto Yes 1530 6 2.14*' RE-auto Yes
Period Impact N/A 5 73.42· FE Yes N/A 4 11.65** RE-no auto Yes
Hausman RE No RE-auto No
Test Yearly Impact N/A 9 79.79" FE Yes N/A 9 3.97 FE No
Statistic RE No RE Yes
Export Direction N/A 6 627.17· FE Yes N/A 5 5.78 FE No
RE No RE Yes
NB: *, ** & *** denote significance at the 1,5 and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of l" order autocorrelationjJroblem re~ectively.
297
Appendices
Appendix SeT: Suitable equations for wine exports from South Africa to the EU
countries.
MODEL
InYijt-1
InGDPPCit
InGDPPCjt
.00004 I .00509
298
Appendices
Appendix seu: Average actual, simulated and potential value of wine exports from
South Africa to EU countries for the periods 2000-2004 and 200~2009 in dollars
(US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adjusted Potential
Name Code Value Value Value Value Value Value Value Value
(USS) (t.oas) (USS) (Loesï (l1SS) (Lozs) (l1SS) (L«>gs)
Austria AUT 2000 - 2004 831,620 13.63113 769,578 13.55360
769,277 13.55321 769,427 13.55340
2005 - 2009 3,363,087 15.02837 3,574,385 15.08930 3,599,548 15.09632 3,586,967 15.09282
Belgium BEL 2000 - 2004 11,637,146 16.26971 12,331,094 16.32763
12,451,229 16.33733 12,391,162 16.33249
2005 - 2009 41,324,231 17.53696 49,057,254 17.70850 49,726,739 17.72205 49,391,996 17.71530
Denmark DNK 2000 - 2004
13,370,037 16.40853 10,836,532 16.19843 10,704,897 16.18621 10,770,714 16.19234
2005 - 2009 67,616,214 18.02936 ,57,790,230 . 17.87233 57,070,491 17.85980 57,430,361 17.86608
Finland FIN 2000 - 2004 4,315.060 15.27762 3,583,205 15.09177
3,549,347 15.08227 3,566,276 15.08703
2005 - 2009 22,385,682 16.92393 24,260,964 17.00438 24,409,037 17.01046 24,335,001 17.00743
France FRA 2000 - 2004 4,941,366 15.41315 4,999,332 15.42481 5,027,121 15.43036
5,013,227 15.42759
2005 - 2009 15,611,795 16.56354 16,321,472 16.60799 16,375,064 16.61127 16,348,268 16.60963
Germany 2000 - 2004 29,652,168 17.20505 26,744,435 17.10184
26,644,939 17.09811 26,694,687 17.09998
DEU 2005 - 2009 151,321,044 18.83491 140,320,137 18.75944 139,432,179 18.75309 139,876,158 18.75627
Greece 2000 - 2004 6,951 8.84657 13,527 9.51246 13,989 9.54606
13,758 9.52940
GRC 2005 - 2009 298,420 12.60626 256,918 12.45651 254,837 12.44838 255,878 12.45245
Ireland 2000 - 2004 9,068,109 1602027 10,212,811 16.13915 10,362,188 16.15367 10,287,499
16.14644
IRL 2005 - 2009 27,482,419 17.12906 50,394,991 17.73540 52,872,502 17.78339 51,633,746 17.75969
2000 - 2004 372,390 12.82770 444,833 13.00545 451,696 13.02076 448,265 13.01314Italy ITA 2005 - 2009 1,863,273 14.43784 1,827,785 14.41862 1,825,557 14.41740 1,826,671 14.41801
Luxembourg 2000 - 2004 7,561 8.93082 4,930 8.50317 4,856 8.48803
4,893 8.49563
Ll1X 2(U15- 2009 270,253 12.50711 99,341 11.50631 94,495 11.45631 96,918 11.48162
Netherlands 54,851,184NLD 2000 - 2004 17.82013 49,213,089
17.71167 48,999,783 17.70733 49,106,436 17.70950
2005 - 2009 141,237,888 18.76596 20 I,807, 157 19.12282 208,087,255 19.15347 204,947,206 19.13826
Portugal 'PRT 2000 - 2004 30,060 10.31095 17,736 9.78333 17,345 9.76105
17,540 9,77225
2005 - 2009 113,699 11.64131 116,634 11.66680 116,785 11.66809 116,710 11.66745
Spain ESI) 2000 - 2004 54,861 10.91256 81,920 11.31350 84,044 11.33910 82,982 I 1.32638
2005 - 2009 626,646 13.34814 483,517 13,08884 476,395 13,07400 479,956 13.08145
Sweden SWE 2000 - 2004 14,962,045 16.52103 16,317,792
16,60777 16,519,534 16.62005 16,418,663 16.61393
2005 - 2009 135,415,133 18.72386 130,286,253 18.68524 129,865,628 18.68201 130,075,941 18.68363
United Kingdom 2000 - 2004 132,344,790 18.70092 140,607,300 18.76148 142,238,007 18.7730 I 141,422,653 18.76726GBR 2005 - 2009 371,786,719 19.73383 529,763,960 20,08794 547,035,417 20.12002 538,399,688 20.10411
299
Appendices
Appendix 5CV: Selection of the estimator suitable for wine imports from the EU
countries to South Africa.
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 225 5 10.12* OLS No 240 4 45.04* OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 225 9 9.65* OLS No 240 8 45.53* OLS No
FE or RE Yes FE or RE Yes
Import Direction 675 6 7.05* OLS No 720 5 31.36* OLS No
FE or RE Yes FE or RE Yes
225 5 0.48 FE-no auto Yes 240 4 1.96** FE-no auto Yes
Durbin Period Impact - - - FE-auto - - - - FE-auto -
Watson 225 6 -1.66*** RE-no auto No 240 6 1.42 RE-no auto No
Statistic 210 6 -0.54 RE-auto Yes 225 6 2.08** RE-auto Yes
(for static) 225 9 0.28 FE-no auto Yes 240 8 1.95** FE-no auto Yes
or Durbin- Yearly Impact - - - FE-auto - - - - - -
H Statistic 225 10 -1.85*** RE-no auto No 240 9 1.43 RE-no auto No
(for 210 10 -0.85 RE-auto Yes 225 9 2.07** RE-auto Yes
dynamic) 675 6 0.00 FE-no auto Yes 720 5 1.77 FE-no auto No
Import Direction - - - - - 675 5 1.98** FE-auto Yes
675 7 -1.75**' RE-no auto No 720 6 1.22 RE-no auto No
630 7 -0.65 RE-auto Yes 675 6 2.05" RE-auto Yes
Period Impact N/A 5 141.52* FE Yes N/A 4 -36.32· FE Yes
Hausman RE No RE No
Test Yearly Impact N/A 9 184.57* FE Yes N/A 8 68.47* FE Yes
Statistic RE No RE No
Import Direction N/A 6 414.44* FE Ye,~ N/A 5 -2.95 FE No
RE No RE Ye.~
NB: *, ** & *** denote significance at the I, 5 and la percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of ISI order autocorrelation problem respectively.
300
Appendices
Appendix SCW: Suitable equations for wine imports from the EU countries to
South Africa.
MODEL
InGDPPCit
InGDPPCjl
D0004 / D0509
DOO/D05
DOI/D06
D02/D07
D03/D08
D04 / D09
PTAyes
301
Appendices
Appendix sex: Average actual, simulated and potential value of wine imports from
the EU countries to South Africa for the periods 2000-2004 and 2005-2009 in
dollars (US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adjusted Potential
Name Code Value Value Value Value Value Value Value Value
(USS) (Logs) (USS) (Logs) (USS) (Logs) (USS) (Logs)
Austria AUT 2000 - 2004 1,704 7.44090 1,665 7.41760 1,644 7.40508
1,655 7.41134
2005 - 2009 15,100 9.62245 8,345 9.02937 7,635 8.94052 7,982 8.98495
Belgium BEL 2000 - 2004 8,170 9.00818 5,590 8.62882 5,443 8.60211 5,516 8.615462005 - 2009 15,827 9.66948 19,682 9.88745 19,979 9.90245 19,830 9.89495
Denmark DNK 2000 - 2004 621 6.43071 710 6.56508
715 6.57216 712 6.56862
2005 - 2009 876 6.77516 1,706 7.44185 1,765 7.47579 1,735 7.45882
Finland FIN 2000 - 2004 Il 2.41177 Il 2.43482 Il 2.43525 Il
2.43504
2005 - 2009 8 2.02993 33 3.48647 34 3.52028 33 3.50338
France FRA 2000 - 2004 3,147,653 14.96217 2,613,472 14.77619 2,552,507 14.75259 2,582,810 14.764392005 - 2009 23,947,825 16.99139 10,356,830 16.15316 9,385,543 16.05468 9,859,233 16.10392
Germany 2000 - 2004 21,213 9.96238 25,222 10.13547 25,590 10.14997 25,406 10.14272DEU 2005 - 2009 165,711 12.01800 110,578 11.61347 106,980 11.58040 108,764 11.59693
Greece 2000 - 2004 7,361 8.90389 6,498 8.77924 6,440 8.77031 6,469 8.77477GRC 2005 - 2009 6,579 8.79166 34,026 10.43489 38,354 10.55462 36,126 10.49476
Ireland IRL 2000 - 2004 30 3.40323 66 4.19049 68 4.21647 67 4.20348
2005 - 2009 909 6.81221 434 607386 421 6.04347 428 6.05867
Italy ITA 2000 - 2004 556,239 13.22895 495,514 13.11335 489,194 13.10052
492,344 13.10693
2005 - 2009 2,658,863 14.79341 1,659,135 14.32181 1,580,624 14.27333 1,619,404 14.29757
Luxembourg LUX 2000 - 2004 8 2.12336 8 2.10474 8 2.10444 8
2.10459
2005 - 2009 620 6.42924 115 4.74324 109 4.68954 112 4.71639
Netherlands 2000 - 2004 21,711 9.98556 23,760 10.07577 23,939 10.08328 23,850 10.07952NLD 2005 - 2009 36,478 10.50446 81,826 11.31235 87,252 11.37655 84,495 11.34445
Portugal PRT 2000 - 2004 621,434 13.33978 625,794 13.34678 626,293 13.34757 626,044 13.34718
2005 - 2009 1,959,380 14.48814 2,484,318 14.72551 2,547,617 14.75067 2,515,768 14.73809
Spain ESP 2000 - 2004 342,637 12.74443 425,765 12.96164 436,029 12.98546 430,866 12.97355
2005 - 2009 362,284 12.80018 I, I03,455 13.91396 1,232,873 14.02486 1,166,371 13.96941
Sweden 2000 - 2004 187 5.23372 105 4.64933 102 4.62785 103
4.63859 .
SWE 2005 - 2009 8,525 9.05078 1,362 7.21663 1,244 7.12624 1,302 7.17143
United Kingdom GBR 2000 - 2004 162,144 11.99624 162,092 11.99592 162,088 11.99589 162,090 11.99591
2005 - 2009 390,021 12.87396 595,064 13.29642 619,363 13.33645 607,092 13.31644
302
Appendices
Appendix 5CY: Selection of the estimator suitable for wine trade between South
Africa and the EU countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 225 4 3.08* OLS No 240 3 77.05* OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 225 8 2.88* OLS No 240 7 79.61 * OLS No
FE or RE Yes FE or RE Yes
Trade Direction 660 5 3.75* OLS No 704 4 40.37* OLS No
FE or RE Yes FE or RE Yes
225 4 0.79 FE-no auto Yes 240 3 1.10 FE-no auto Yes
Durbin Period Impact - - - FE-auto - 225 3 1.66 - -
Watson 225 5 0.47 RE-no auto Yes 240 4 0.92 RE-no auto No
Statistic - - - RE-auto - 225 4 1.57 RE-auto Yes
(for static) 225 8 0.78 FE-no auto Yes 240 7 1.05 FE-no auto No
or Durbin- Yearly Impact - - - FE-auto - 225 7 I.71 - No
H Statistic 225 9 0.40 RE-no auto Yes 240 8 0.88 RE-no auto No
(for - - - RE-auto - 225 8 1.61 . RE-auto No
dynamic) 660 5 1.13 FE-no auto Yes 704 4 1.10 FE-no auto No
Trade Direction - - - - - 660 4 1.72 FE-auto No
660 6 0.67 RE-no auto Yes 704 5 0.96 RE-no auto No
- - - RE-auto - 660 5 1.69 RE-aulO No
Period Impact N/A 4 40.28* FE Yes N/A 3 71.09* FE Yes
Hausman RE No RE No
Test Yearly Impact N/A 8 38.92* FE Yes - - - FE -
Statistic RE No RE -
Trade Direction N/A 5 144.25* FE Yes - - - FE -
RE No RE -
NB: ", ** & *** denote significance at the 1,5 and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto &.-auto
denote estimation assuming no autocorrelation problem and estimation with the correction of l" order autocorrelation_E_l"oblemrespectively.
303
Appendices
Appendix SCY: Suitable equations for wine trade between South Africa and the EU
countries.
MODEL
1994
2009
InGDPPCijt
00004 / D0509
DOO/ D05
001 / D06
002/ D07
003/ D08
004/ D09
PTAyes
304
Appendices
Appendix 5DA: Average actual, simulated and potential value of wine trade
between South Africa and the EU countries for the periods 2000-2004 and 2005-
2009 in dollars (US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adjusted Potential
Name Code Value Value Value Value Value Value Value Value
(USS) (Loss) (USS) (Loes) (lJSS) (Lozs) (USS) (Logs)
Austria 2000 - 2004
834,051 13.63405 685,623 13.43808 679,791 13.42954 682,707 13.43382
AUT 2005 - 2009 3,385,950 15.03515 3,319,003 15.01517 3,283,755 15.00450 3,301.379 15.00985
Belgium 2000 - 2004 11,671,337 16.27265 11,277,107 16.23829
11,246,709 16.23559 11,261,908 16.23694
BEL 2005 - 2009 41,350,667 17.53760 43,953,334 17.59864 44,159,466 17.60332 44,056,400 17.60098
2000 - 2004 13,371,932 16.40867 10,870,689 16.20158 10,695,739 16.18536 10,783,214 16.19350Denmark DNK 2005 - 2009 67,625,540 18.02950 58,152,341 17.87858 57,472,012 17.86681 57,812,177 17.87271
2000 - 2004 4,315,428 15.27771 3,562,336 15.08593 3,513,107 15.07201 3,537,721 1507899Finland FIN
2005 - 2009 22,385,740 16.92393 23,532,557 16.97390 23,619,400 16.97758 23,575,978 16.97574
France FRA 2000 - 2004 8,152,780 15.91387 8,681,107 15.97666 8,723,269 15.98150 8,702,188 15.979092005 - 2009 39,622,227 17.49490 34,266,992 17.34969 33,893,393 17.33873 34,080,193 17.34423
2000 - 2004 29,680,166 17.20599 25,220,869 17.04318 24,883,348 17.02971 25,052,108 17.03647
Germany DEU 2005 - 2009 151,503,491 18.83612 135,159,173 18.72196 133,900,549 18.71261 134,529,861 18.71730
2000 - 2004 22,937 10.04051 28,726 10.26555 29,039 10.27641 28,883 10.27099Greece GRC 2005 - 2009 317,303 12.66761 331,239 12.71060 332,012 12.71292 331,625 12.71176
2000 - 2004 9,068,198 16.02028 12,193,449 16.31641 12,481,844 16.33979 12,337,646 16.32817Ireland !RL
2005 - 2009 27,492,795 17.12943 43,187,744 17.58107 44,707,472 17.61565 43,947,608 17.59851
2000 - 2004 944,859 13.75879 1,165,378 13.96856 1,181,827 13.98257 1,173,603 13.97559Italy ITA 2005 - 2009 4,542,182 15.32892 4,076,889 15.22084 4,048,077 15.21375 4,062,483 15.21730
Luxembourg 2000 - 2004
7,622 8.93875 4,395 8.38828 4,302 8.36677 4,348 8.37758
LUX 2005 - 2009 272,984 12.51717 107,756 11.58762 102,940 11.54190 105,348 11.56502
Netherlands 2000 - 2004 54,892,286 17.82088 47,148,547 17.66881
46,535,573 17.65573 46,842,060 17.66229
NLD 2005 - 2009 141,317,030 18.76652 175,933,518 18.98562 179,170,077 19.00385 177,551,797 18.99477
2000 - 2004 659,429 13.39913 717,525 13.48356 721,434 13.48900 719,480 13.48628Portugal PRT 2005 - 2009 2,095,240 14.55518 2,720,894 14.81647 2,766,608 14.83313 2,743,751 14.82484
2000 - 2004 483,542 13.08889 933,417 13.74661 974,631 13.78981 954,024 . 13.76844Spain ESP 2005 - 2009 1,076,489 13.88922 1,329,132 14.10004 1,346,192 14.11279 1,337,662 14.10643
2000 - 2004 14,962,374 16.52105 15,014,248 16.52451 15,018,397 16.52479 15,016,323 16.52465Sweden SWE 2005 - 2009 135,466,476 18.72423 132,253,531 18.70023 131,993,893 18.69827 132,123,712 18.69925
2000 - 2004 132,517,202 18.70222 126,714,290 18.65745 126,197,198 18.65336 126,455,744 18.65540United Kingdom GBR 2005 - 2009 372,209,548 19.73497 452,373,306 19.93002 460,178,338 19.94712 456,275,822 19.93861
305
Appendices
Appendix 5DB: Selection of the estimator suitable for agricultural exports from
South Africa to the SADe countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 90 5 2.26*** OLS No 96 4 3.56* OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 90 9 2.41 ** OLS No 96 8 4.43* OLS No
FE or RE Yes FE or RE Yes
Export Direction 1785 6 5.71· OLS No 1904 5 38.04· OLS No
FE or RE Yes FE or RE Yes
90 5 0.86 FE-no auto 96 4 1.48 FE-no auto No
Durbin Period Impact - - - FE-auto - 90 4 1.85** FE-auto Yes
Watson - - - RE-no auto - 96 5 1.15 RE-no auto No
Statistic - - - RE-auto - 90 5 1.68 RE-auto ?
(for static) 90 9 0.67 FE-no auto Yes 96 8 1.34 FE-no auto No
or Durbin- Yearly Impact - - - FE-auto - 90 8 177 FE-auto ?
H Statistic - - - RE-no auto - 96 9 1.16 RE-no auto No
(for - - - RE-auto - 90 9 1.77 RE-auto ?
dynamic) 1785 6 -0.25 . FE-no auto Yes 1904 5 1.31 FE-no auto No
Export Direction - - - FE-auto - .r-' 1785 5 1.85 FE-auto No
1785 7 -0.51 RE-no auto Yes 1904 6 1.19 RE-no auto No
- - - RE-auto - 1785 6 I.78 RE-auto No
Period Impact - - - FE - - - - FE -
Hausman RE - RE -
Test Yearly Impact - - - FE - - - - FE -
Statistic RE - RE -
Export Direction N/A 6 478.54* FE Yes - - - FE -
RE No RE -
NB: *, ** & *** denote significance at the 1,5 and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of Ist order autocorrelation problem respectively. ? means
inconclusive.
306
Appendices
Appendix 5DC: Suitable equations for agricultural exports from South Africa to the
SADC countries.
MODEL
InGDPPCit
InGDPPCjt
REERt
D0004 / D0509 ? ?
DOO/DOS ? ?
DOl/D06 ? ?
D02/D07 ? ?
D03 / D08 ? ?
D04/D09
PTAyes
Appendix 5DD: Average actual, simulated and potential value of agricultural
exports from South Africa to SADC countries for the periods 2000-2004 and 2005-
2009 in dollars (US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adjusted Potential
Name Code Value Value Value Value Value Value Value Value
(US$) (lC!ill _ill_S~ __(!,Q_g& _1US~ _i_LC!ill J!:!S~ (Logs)
Malawi MWI 2000 - 2004
32,427,760 17.29453 31,830,759 17.27594 31,715,417 17.27231 31,773,036 17.27413
2005 - 2009 63,292,647 17.96328 85,729,847 18.26671 90,878,864 18.32504 88,266,818 18.29587
Mauritius MUS 2000 - 2004 20,436,235 16.83282 20,187,054 16.82055 20,140,245 16.81823 20,163,636 16.819392005 - 2009 126,988,429 18.65961 97,326,693 18.39358 92,435,807 18.34202 94,849,731 18.36780
Mozambique MOZ 2000 - 2004 153,009,627 18.84601 141,842,571 18.77023 139,540,387 18.75386 140,686,770 18.762052005 - 2009 392,399,583 19.78779 436,779,254 19.89494 446,862,615 19.91776 441,792,168 19.90635
Tanzania TZA 2000 - 2004 17,745,405 16.69164 19,895,532 16.80601 20,330,431 16.82763 20,111,806 16.81682
2005 - 2009 62,843,336 17.95616 66,942,419 18.01934 67,747,235 18.03129 67,343,625 18.02532
Zambia 2MB 2000 - 2004 54,000,515 17.80450 49,047,772 17.70831 48,107,316 17.68894 48,575,268 17.698632005 - 2009 194,366,332 19.08526 187,389,036 19.04870 186,009,624 19.04131 186,698,056 19.04500
Zimbabwe 2000 - 2004 82,317,542ZWE 18.22609 89,931,210 18.31456 91,614,211 18.33310 90,768,810 18.323832005 - 2009 415,185,094 19.84423 349,861,031 19.67305 337,499,697 19.63708 343,624,784 19.65506
307
Appendices
Appendix 5DE: Selection of the estimator suitable for agricultural imports from the
SADe countries to South Africa.
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 90 5 3.76* OLS No 96 4 45.78* OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 90 9 4.00* OLS No 96 8 47.95* OLS No
FE or RE Yes FE or RE Yes
Import Direction 1905 6 5.88* OLS No 2032 5 32.36* OLS No
FE or RE Yes FE or RE Yes
90 5 0.06 FE-no auto Yes 96 4 1.15 FE-no auto No
Durbin Period Impact - - - FE-auto - 90 4 2.01** FE-auto Yes
Watson 90 6 -0.89 RE-no auto Yes 96 5 1.04 RE-no auto No
Statistic - - - RE-auto - 90 5 2.06** RE-auto Yes
(jar static) 90 9 -0.72 FE-no auto Yes 96 8 1.18 FE-no auto No
or Durbin- Yearly Impact - - - FE-auto - 90 8 1.99** FE-auto Yes
H Statistic 90 10 -1.20 RE-no auto Yes 96 9 1.06 RE-no auto No
(jar - - - RE-auto - 90 9 2.05** RE-auto Yes
dynamic) 1905 6 0.11 FE-no auto Yes 2032 5 1.59 FE-no auto No
Import Direction - - - FE-auto - 1905 5 1.94** FE-auto Yes
1905 7 -1.52 RE-no auto Yes 2032 6 1.18 RE-no auto No
- - - RE-auto - 1905 6 2.08** RE-auto Yes
Period Impact N/A 5 21.39* FE Yes N/A 4 -3.08 FE No
Hausman RE No RE Yes
Test Yearly Impact N/A 9 21.54* FE Yes N/A 8 5.40 FE No
Statistic RE No RE Yes
Import Direction N/A 6 1063.89* FE Yes N/A 5 -0.64 FE No
RE No RE Yes
NB: *, ** & *** denote significance at the 1,5 and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of Ist order autocorrelation_Qroblem respectively.
308
Appendices
Appendix 5DF: Suitable equations for agricultural imports from SADC countries to
South Africa
MODEL
InYijt_1
InGDPPCit
InGDPPCjt
REERt
00004 / 00509
DOO/DOS
001/006
002/ D07
003/008
004/009
PTAyes
PTA no
Appendix 5DG: Average actual, simulated and potential value of agricultural
imports from the SADC countries to South Africa for the periods 2000-2004 and
2005-2009 in dollars (US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adjusted Potential
Name Code Value Value Value Value Value Value Value Value
~ (US$) (Logs) (US$) (Logs) (US$) (Logs) (US$) (Logs)
Malawi MWI 2000 - 2004 19,445172 16.78311 887035 13.69564 3,099,069 14.94661 1,658,006 14.321132005 - 2009 70,312,360 18.06846 76,602,454 18.15414 78,011,346 18.17236 77,303,690 18.16325
Mauritius MUS 2000 - 2004
385,685 12.86278 348,220 12.76059 1,863,424 14.43793 805,532 13.59926
2005 - 2009 3,518,699 15.07360 2,215,997 14.61121 2,053,808 14.53521 2,133,362 14.57321
Mozambique 16.22143 17,078,506 196,646,432 19.09692MOZ 2000 - 2004 11,088,659 16.65333 57,951,939 17.875122005 - 2009 46,147,529 17.64735 50,786,388 17.74314 51,802,891 17.76296 51,292,122 17.75305
Tanzania TZA 2000 - 2004 4,256178 15.26388 299801 12.61087 1,005,641 13.82114 549,083 13.216002005 - 2009 22,953,923 16.94900 21,238349 16.87132 20,919,450 16.85619 21,078,296 16.86375
Zambia 2MB 2000 - 2004 7,372,201 15.81323 1,862,034 14.43718 10,014,634 16.11956 4,318,285 15.27837
2005 - 2009 35,540,616 17.38619 35,267,602 17.37848 35,212,750 17.37692 35,240,165 17.37770
Zimbabwe ZWE 2000 - 2004 50,061,291 17.72876 2,984,697 14.90901 12,625,842 16.35126 6,138,754 15.63013
2005 - 2009 95,528,998 18.37494 137,792,148 18.74126 149,417,387 18.82225 143,487,082 18.78176
309
Appendices
Appendix 5DH: Selection of the estimator suitable for agricultural trade between
South Africa and the SADe countries.
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 90 4 5.32* OLS No 96 3 15.49* OLS No
Wald Test FEor RE Yes FEor RE Yes
Statistic Yearly Impact 90 8 4.62* OLS No 96 7 17.15* OLS No
FEor RE Yes FE or RE Yes
Trade Direction 1635 5 4.61* OLS No 1744 4 58.84* OLS No
FEor RE Yes fE or RE Ye.5
90 4 0.80 FE-no auto Yes 96 3 1.39 FE-no auto No
Durbin Period Impact - - - fE-auto - 90 3 1.80** fE-auto Yes
Watson 90 5 0.75 RE-no auto Yes 96 4 1.37 RE-no auto No
Statistic - - - RE-auto - 90 4 1.80** RE-auto Yes
(for static) 90 8 0.68 fE-no auto Yes 96 7 1.32 fE-no auto No
or Durbin- Yearly Impact - - - FE-auto - 90 7 1.95** fE-auto Yes
H Statistic 90 9 0.60 RE-no auto Yes 96 8 1.30 RE-no auto No
(for - - - RE-auto 90 8 1.93** RE-auto Yes
dynamic) -1635 5 0.53 fE-no auto Yes 1744 4 1.35 FE-no auto No
Trade Direction - - - fE-auto - 1635 4 1.90 fE-auto No
1635 6 0.12 RE-no auto Yes 1744 5 1.28 RE-no auto No
- - - RE-auto - 1635 5 1.91 RE-auto No
Period Impact N/A 4 -40.12* fE Yes N/A 3 -0.12 fE No
Hausman RE No RE Yes
Test Yearly Impact N/A 8 8.04 FE No N/A 7 0.46 fE No
Statistic RE Yes RE Yes
Trade Direction N/A 5 497.54* FE Yes - - - fE -
RE No RE -
NB: " ** & *** denote significance at the I, 5 and 10percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of ISI order autocorrelationproblem respectively.
310
Appendices
Appendix 5DI: Suitable equations for agricultural trade between South Africa and
the SADC countries.
MODEL
2009
InYijt-t
InGDPPCijt
D0004 / DOS09
DOO/DOS
DOI/D06
D02/D07
D03/D08
D04/D09
PTAyes
PTA no
Appendix 5DJ: Average actual, simulated and potential value of agricultural trade
between South Africa and the SADC countries for the periods 2000-2004 and 2005-
2009 in dollars (US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adjusted Potential
Name Code Value Value Value Value Value Value Value Value
(US$) (Logs) (US$) (Logsl_ (US$) (Logs) (US$) (Logs)
Malawi MWI 2000 - 2004 55,210,277 17.82666 51,384,670 17.75485 53,353,276 17.79245 52,368,973 17.77382
2005 - 2009 134,638,785 18.71811 160,202,757 18.89195 165,772,514 18.92613 162,987,635 18.90918
Mauritius 2000 - 2004 20,928,296 16.85661 22,770,894 16.94099 24,298,502 17.00593 23,534,698MUS 16.973992005 - 2009 132,213,327 18.69993 103,596,835 18.45602 98,876,901 18.40939 101,236,868 18.43297
Mozambique 2000 - 2004 164,246,809 18.91688 161,642,565 18.90090 170,359,109 18.95342MOZ 166,000,837 18.927502005 - 2009 439,245,922 19.90057 503,657,216 20.03741 518,412,884 20.06628 511,035,050 20.05195
Tanzania TZA 2000 - 2004
23,101,606 16.95541 26,340,085 17.08660 28,377,134 17.16109 27,358,610 17.12454
2005 - 2009 86,213,154 18.27233 88,722,134 18.30102 89,205,219 18.30645 88,963,677 18.30374
Zambia 2MB 2000 - 2004 61,862,742 17.94043 59,959,472 17.90918 62,790,955 17.95532 61,375,213 17.932522005 - 2009 242,756,314 19.30757 225,265,497 19.23279 221 ,907,638 19.21777 223,586,568 19.22531
Zimbabwe ZWE 2000 - 2004 139937,404 18.75671 97,895,327 18.39941 95,952,280 18.37936 96,923,804 18.38944
2005 - 2009 532,427,831 20.09296 521,529,179 20.07228 519,253,500 20.06790 520,391,340 20.07009
311
Appendices
Appendix 5DK: Selection of the Estimator suitable for cheese exports from South
Africa to the SADe countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 90 5 0.30 OLS Yes 96 4 3.05* OLS No
Wald Test FE or RE No FE or RE Yes
Statistic Yearly Impact 90 9 0.29 OLS Yes 96 8 3.28* OLS No
FE or RE No FE or RE Yes
Export Direction 480 6 2.48** OLS No 512 5 10.49* OLS No
FE or RE Yes FE or RE Yes
90 5 0.01 OLS-no auto Yes 96 4 0.77 FE-no auto No
Durbin Period Impact - - - - - 90 4 2.18** FE-auto Yes
Watson - - - - - - - - - -
Statistic - - - - - - - - - -
(jar static) 90 9 -0.05 OLS-no auto Yes 96 8 0.80 FE-no auto No
or Durbin- Yearly Impact - - - - - 90 8 2.16** FE-auto Yes
H Statistic - - - - - - - - - -
(jar - - - - - - - - - -
dynamic) 480 6 -0.15 FE-no auto Yes 512 5 1.28 FE-no auto No
-,
Export Direetion - - - - - 480 5 1.98** FE-auto Yes
480 7 -1.48 RE-no auto Yes 512 6 0.79 RE-no auto No
- - - - - 480 6 2.08** RE-auto Yes
Period Impact - - - - - - - - - -
Hausman - - - -
Test Yearly Impact - - - - - - - - - -
Statistic - - - -
Export Direction N/A 6 143.91 * FE Yes N/A 5 17.92* FE Yes
RE No RE No
NB: *, ** & *** denote significance at the 1,5 and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of Ist order autocorrelation problem respectively.
312
Appendices
Appendix 5DL: Suitable equations for cheese exports from South Africa to the
SADC countries.
MODEL
2004 2009
InYijt-1
InGDPPCit
InGDPPCjt
REERt
D0004 / DOS09
DOO/DOS
DOl/D06
D02/D07 -
D03/D08
D04/D09
PTAyes
PTA no
Appendix 5DM: Average actual, simulated and potential value of cheese exports
from South Africa to SADC countries for the periods 2000-2004 and 2005-2009 in
dollars (US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adjusted Potential
Name Code Value Value Value Value Value Value Value Value
(US$) (LOgs) (US$) (Logs) (US$) (Logs) (US$) (Logs)
Malawi MWI 2000 - 2004 188,798 12.14843 157,222 11.96541 151,760 11.93006 154,467 11.947742005 - 2009 556,692 13.22977 528,426 13.17766 522,943 13.16723 525,677 13.17244
2000 - 2004 172,776 12.05975 171,349 12.05146 171,074 12.04985 17l,212 12.05066Mauritius MUS 2005 - 2009 106,636 11.57718 136,702 11.82556 142,811 11.86928 139,723 11.84742
Mozambique MOZ 2000 - 2004 1,055,420 13.86945 973,072
13.78821 955,116 13.76959 964,052 13.77890
2005 - 2009 ·3,531,692 15.07729 3,058771 14.93352 2,957,983 14.90002 3,007,955 14.91677
Tanzania 2000 - 2004 89,991 11.40746 85,134 11.35199 84,282 11.34192 84,707TZA 11.34695
2005 - 2009 302,353 12.61935 276,934 12.53154 272,388 12.51498 274,652 12.52326
Zambia 2MB 2000 - 2004
215,364 12.28009 220,206 12.30232 221,189 12.30677 220,697 12.30455
2005 - 2009 1,012,027 13.82747 802,886 13.59597 765,169 13.54785 783,801 13.57191
Zimbabwe ZWE 2000 - 2004 184,491 12.12536 250470 12.43109 266,451 12.49294 258,337 12.46202
2005 - 2009 319795 12.67543 417,581 12.94223 439,988 12.99450 428,638 12.96837
313
Appendices
Appendix 5DN: Selection of the estimator suitable for cut flowers exports from
South Africa to the SADe countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 90 5 3.79* OLS No 96 4 6.17· OLS No
Wald Test FEor RE Yes FE or RE Yes
Statistic Yearly Impact 90 9 16.03· OLS No 96 8 6.29* OLS No
FEor RE Yes FE or RE Yes
Export Direction 840 6 3. I 1* OLS No 896 5 2l.56* OLS No
FE or RE Yes FEor RE Yes
90 5 l.26 FE-no auto Yes 96 4 l.89** FE-no auto Yes
Durbin Period Impact - - - - - - - - - -
Watson 90 6 0.47 RE-no auto Yes 96 5 l.21 RE-no auto No
Statistic - - - - - 90 5 2.06** RE-auto Yes
(jar static) 90 9 l.13 FE-no auto Yes 96 8 l.83** FE-no auto Yes
or Durbin- Yearly Impact - - - - - - - - - -
H Statistic 90 10 0.58 RE-no auto Yes 96 9 I.I9 RE-no auto No
(jar - - - - - 90 9 2.12·* RE-auto Yes
dynamic) 840 6 0.05 FE-no auto Yes 896 5 1.11 FE-no auto No
Export Direction - - - - - 840 5 l.9I** FE-auto Yes
840 7 -0.82 RE-no auto Yes 896 6 0.89 RE-no auto No
- - - - - 840 6 l.94** RE-auto Yes
Period Impact N/A 5 85.89· FE Yes N/A 4 -8.57*** FE Ye.~
Hausman RE No RE No
Test Yearly Impact N/A 9 86.44* FE Yes N/A 8 -22.9 I* FE Yes
Statistic RE No RE No
Export Direction N/A 6 209.56* FE Yes N/A 5 - I4.44** FE Yes
RE No RE No
NB: *, ** & *** denote significance at the I,5and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of 1st order autocorrelationJ~roblem re~ectively.
314
Appendices
Appendix SDO: Suitable equations for cut flowers exports from South Africa to the
SADC countries.
MODEL
InGDPPCit
InGOPPCjl
00004 1 00509
OOO/DOS
D01/D06
D02/D07
D03/DOS
D04/D09
PTAyes
Appendix SDP: Average actual, simulated and potential value of cut flowers exports
from South Africa to SADC countries for the periods 2000-2004 and 2005-2009 in
dollars (US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adjusted Potential
Name Code Value Value Value Value Value Value Value Value
(VS$) (Lo~) (US$) (l.02S) (US$) (Logs) _(VS$) (Logs)
Malawi 2000 - 2004 1,156 7.05312 734 6.59783 685 6.52935 709MWI 6.563592005 - 2009 24,400 10.10233 6,194 8.73135 4,809 8.47820 5,458 8.60478
Mauritius 2000 - 2004 31,918 10.37092 14,027 9.54876 11,578 9.35687 12,744 9.45282MUS 2005 - 2009 103,484 11.54717 64,175 11.06936 57,051 10.95169 60,508 11.01053
Mozambique MOZ 2000 - 2004 54,111 10.89879
50,354 10.82683 49,374 10.80718 49,861 10.81700
2005 - 2009 314,886 12.65997 253,293 12.44230 238,029 12.38015 245,543 12.41123
Tanzania 2000 - 2004 233 5.44900 416 6.03107 450 6.11010TZA 433 6.070592005 - 2009 39 3.67430 256 5.54344 314 5.74872 283 5.64608
Zambia 2MB 2000 - 2004 4,496 8.41088 4,990 8.51524 5,097 8.53643 5,043 8.52583
2005 - 2009 8,067 8.99558 8,494 9.04714 8,579 9.05707 8,536 9.05210
Zimbabwe 2000 - 2004 3,926 8.27549 7,619 8.93842 8,788 9.08115ZWE 8,183 9.009792005 - 2009 8,419 9.03829 9,741 9.18405 10,023 9.21263 9,881 9.19834
315
Appendices
Appendix 5DQ: Selection of the estimator suitable for cut flowers imports from the
SADe countries to South Africa.
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 90 5 5.02* OLS No 96 4 37.61* OLS No
Wald Test rs or RE Yes fE or RE Yes
Statistic Yearly Impact 90 9 5.97* OLS No 96 8 41.1 0* OLS No
fE or RE Yes fEar RE Yes
Import Direction 285 6 4.98* OLS No 304 5 19.88* OLS No
fE or RE Yes rs or RE Yes
90 5 0.22 fE-no auto Yes 96 4 1.39 fE-no auto No
Durbin Period Impact - - - - - 90 4 1.88** FE-auto Yes
Watson 90 6 -1.75*** RE-no auto N() 96 5 0.81 RE-no auto N()
Statistic 84 6 -0.66 RE-auto Yes 90 5 2.00*" RE-auto Yes
(for SIalic) - - - - - 96 8 1.45 FE-no auto N()
or Durbin- Yearly Impact 84 9 0.67 FE-auto Yes 90 8 1.96*· fE-auto Yes
H Statistic 90 10 -1.88**" RE-no auto N() 96 9 0.87 RE-no auto N()
(for 84 10 -0.73 RE-auto Yes 90 9 2.07** RE-auto Yes
dynamic) 390 6 -0.11 fE-no auto Yes 416 5 1.65 fE-no auto N()
Import Direction - - - - - 390 5 1.97** fE-auto Yes
390 7 -1.59 RE-no auto Yes 416 6 1.06 RE-no auto N()
- - - - - 390 6 2.04*" RE-auto Yes
Period Impact N/A 5 31.86* fE Yes N/A 4 -0.21 fE N()
Hausman RE N() RE Yes
Test Yearly Impact N/A 9 38.65* fE Yes N/A 8 2.96 FE N()
Statistic RE N() RE Yes
Import Direction N/A 6 153.98* fE Yes N/A 5 1.83 fE N()
RE N() RE Yes
NB: ", ** & *** denote significance at the 1, 5 and 10percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, fE & RE denote Pooled Ordinary Least Squares, fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of l" order autocorrelation problem respectively.
316
Appendices
Appendix 5DR: Suitable equations for cut flowers imports from SADC countries to
South Africa
MODEL
InYijt_1
InGDPPCit
cut as to
were added under the import direction model: BGR, BRA, CHN, DEU, ESP, FRA, GBR, IND, ISR, ITA, KEN, NLD, PHL, PRT, SGP, SYC, THA, TUR,
Appendix 5DS: Average actual, simulated and potential value of cut flowers imports
from the SADC countries to South Africa for the periods 2000-2004 and 2005-2009
in dollars (US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adjusted Potential
Name Code Value Value Value Value Value Value Value Value
(US$) (Logs) (US$) (Logs) (US$) (Logs) (US$) (Logs)
Malawi MWI 2000 - 2004
16,587 9.71637 9,638 9.17343 9,735 9.18348 9,686 9.17845
2005 - 2009 48,149 10.78206 33,909 10.43142 31,379 10.35389 32,619 10.39266
Mauritius 2000 - 2004 121 4.79565 331 5.80112 417 6.03208 371 5.91660MUS 2005 - 2009 999 6.90633 1,757 7.47151 1,912 7.55573 1,833 7.51362
Mozambique MOZ 2000 - 2004 22 3.09797 35 3.54366 38 3.63014 36 3.586902005 - 2009 7 1.93246 33 3.49466 36 3.59554 35 3.54510
Tanzania 2000 - 2004 110 4.69885 81 4.39001 83 4.41914 82 4.40457TZA 2005 - 2009 26,381 10.18039 3,663 8.20601 2,639 7.87806 3,109 8.04204
Zambia 2MB 2000 - 2004 54,313 10.90252 38,472 10.55769 41,326 10.62924 39,873 10.593462005 - 2009 722,536 13.49052 571,361 13.25578 532,650 13.18562 551,666 13.22070
Zimbabwe 2000 - 2004 474254ZWE 13.06950 205,254 12.23200
187018 12.13896 195924 12.18548
2005 - 2009 1,649,091 14.31573 2,541,109 14.74811 2,948772 14.89690 2,737,362 14.82251
317
Appendices
Appendix 5DT: Selection of the estimator suitable for cut flowers trade between
South Africa and the SADe countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 90 4 10.95* OLS No 96 3 21.72* OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 90 8 11.17" OLS No 96 7 22.08* OLS No
FE or RE Yes FEor RE Yes
Trade Direction 330 5 5.17* OLS No 352 4 21.85* OLS No
FE or RE Yes FE or RE Yes
90 4 -0.54 FE-no auto Yes 96 3 2.09** FE-no auto Yes
Durbin Period Impact - - - - - - - - - -
Watson 90 5 -0.38 RE-no auto Yes 96 4 1.74** RE-no auto Yes
Statistic - - - - - - - - - -
(for static) 90 8 -0.61 FE-no auto Yes 96 7 2.10** FE-no auto Yes
or Durbin- Yearly Impact - - - - - - - - - -
H Statistic 90 9 -0.29 RE-no auto Yes 96 8 1.73 RE-no auto ?
(for - - - - - 90 8 2.13** RE-auto Yes
dynamic) 330 5 0.24 FE-no auto Yes 352 4 1.45 FE-no auto No
Trade Direction - - - - - 330 4 1.84** FE-auto Yes
330 6 -0.25 RE-no auto Yes 352 5 1.27 RE-no auto No
- - - - - 330 5 1.87*" RE-auto Yes
Period Impact N/A 4 -654.77" FE Yes N/A 3 4.22 FE No
Hausman RE No RE Yes
Test Yearly Impact N/A 8 -465.95* FE Yes N/A 7 1.72 FE No
Statistic RE No RE Yes
Trade Direction N/A 5 90.69· FE Yes N/A 4 -0.18 FE No
RE No RE Yes
NB: ", ** & .** denote significance at the I, 5 and 10percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of ISI order autocorrelation problem respectively. ?
means inconclusive.
318
Appendices
Appendix 5DU: Suitable equations for cut flowers trade between South Africa and
the SADC countries
MODEL
InVijt_1
InGDPPCijt
REERt
00004 / 00509
DOO/ 005
DO] /006
002/007
003/008
004/009
PTA yes
PTA no
as well as to the
BGR, CHN, DEU, ESP, FRA, GBR, IND, ITA, KEN, NLD, PRT, SGP, SYC,
Appendix 5DV: Average actual, simulated and potential value of cut flowers trade
between South Africa and the SADC countries for the periods 2000-2004 and 2005-
2009 in dollars (US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adjusted Potential
Name Code Value Value Value Value Value Value Value Value
(lIS$) (Logs) (llS$) (Logs) (US$) (Logs) (lIS$) (Logs)
Malawi MWI 2000 - 2004 18.113 9.80438 10,677 9,27589 5,574 8,62583 7,715 8,95086
2005 - 2009 73,363 11.20317 73,410 11.20381 73,418 11.20393 73,414 11,20387
Mauritius MlIS 2000 - 2004 33,787 10.42782 38,434 10,55670 20,460 9,92623 28,042 10,24147
211115"2009 107,499 11,58524 289,637 12,57639 355,376 12,78093 320,827 12,67866
Mozambique MOZ 2000 - 2004 56,110 10,93507 8,100,922 15,90749 13,427,856 16.41284 10,429,669 16,16017
2005 - 2009 328,253 12,70154 2,970,860 14,90436 5,201,801 15.46452 3,931,135 15,18444
Tanzania 2000 - 2004 800 6,68494 817 6,70597 555 6,31944 674 6,51271TZA
2005 - 2009 26,934 10,20115 20,235 9,91515 19,352 9,87057 19,789 9,89286
Zambia 2MB 2000 - 2004 60,751 11.01454 51,980 10,85861 25,585 10,14977 36,468 10,50419
2005 - 2009 732,240 13,50386 182,232 12,11304 138,506 11,83867 158,872 11.97585
Zimbabwe ZWE 2000 - 2004 482,447 13,08663 171,146 12,05027 62,757 11.04702 103,636 11,54864
2005 - 2009 1,658,373 14,32135 363,508 12,80356 263,939 12.48347 309,748 12,64351
319
Appendices
Appendix 5DW: Selection of the estimator suitable for frozen fruits and nuts
exports from South Africa to the SADe countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 90 5 3.79* OLS No 96 4 2.61 ** OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 90 9 3.76* OLS No 96 8 3.54* OLS No
FE or RE Yes FE or RE Yes
Export Direction 315 6 2.93* OLS No 336 5 12.33* OLS No
FE or RE Yes FE or RE Yes
90 5 -1.14 FE-no auto Yes 96 4 1.21 FE-no auto No
Durbin Period Impact - - - - - 90 4 2.38** FE-auto Yes
Watson 90 6 -1.95*** RE-no auto No 96 5 0.63 RE-no auto No
Statistic 84 6 -0.80 RE-auto Yes 90 5 2.40** RE-auto Yes
(jar static) 90 9 -1.15 FE-no auto Yes 96 8 1.22 FE-no auto No
or Durbin- Yearly Impact - - - - - 90 8 2.33** FE-auto Yes
H Statistic 90 lO -1.94*** RE-no auto No 96 9 0.57 RE-no auto No
(jar 84 lO -0.57 RE-auto Yes 90 9 2.34** RE-auto Yes
dynamic) 315 6 -0.42 FE-no auto Yes 336 5 1.32 FE-no auto No
Export Direction - - - - - 315 5 2.04** FE-auto Yes
315 7 -1.55 RE-no auto Yes 336 6 0.80 RE-no auto No
- - - - - 315 6 2.15" RE-auto Yes
Period Impact N/A 5 58.48* FE Yes N/A 4 0.24 FE No
Hausman RE No RE Yes
Test Yearly Impact N/A 9 142.75* FE Yes N/A 8 -1.23 FE No
Statistic RE No RE Yes
Export Direction N/A 6 118.90* FE Yes N/A 5 0.40 FE No
RE No RE Yes
NB: *, ** & *** denote significance at the I, 5 and 10percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of IS!order autocorrelation_Q!'oblem re~ectively.
320
Appendices
Appendix 5DX: Suitable equations for frozen fruits and nuts exports from South
Africa to the SADC countries
MODEL
InYijt_1
InGDPPCit
InGDPPCjt
D0004 / DOS09
DOO/DOS
DOI/D06
D02/D07
D03/D08
D04/D09
PTAyes
PTA no
Appendix 5DY: Average actual, simulated and potential value of frozen fruits and
nuts exports from South Africa to SADC countries for the periods 2000-2004 and
2005-2009 in dollars (US$) and natural logarithms (Logs)
Country Country Period Actual. Simulated Adjusted Potential
Name Code Value Value Value Value Value Value Value Value
(US$) (logs) (US$) (logs) (US$) (Logs) (US$) (Logs)
Malawi MW! 2000 - 2004 480 6.17377 714 6.57021 761 6.63481 737 6.602512005 - 2009 3,855 8.25703 5,666 8.64223 6,084 8.71341 5,871 8.67782
Mauritius 2000 - 2004 50,925 10.83810 26,761 10.19470 21,081MUS 9.95612 23,752 10.075412005 - 2009 49,612 10.81199 68,061 11.12816 73,690 11.20762 70,820 11.16789
Mozambique MOZ 2000 - 2004 2,576 7.85389 3,574 8.18155 3,822 8.24855 3,696 8.21505
2005 - 2009 12,743 9.45275 20,512 9.92878 22,759 10.03270 21,606 9.98074
Tanzania TZA 2000 - 2004 0
0.00000 0 0.00000 0 0.00000 0 0.00000
2005 - 2009 488 6.19073 91 4.50956 78 4.36061 84 4.43509
Zambia 2MB 2000 - 2004 3,793 8.24086 2,921 7.97968 2,697 7.89975 2,807 7.939722005 - 2009 19,313 9.86852 30,241 10.31694 33,508 10.41954 31,832 10.36824
Zimbabwe ZWE 2000 - 2004 1,395 7.24090 1,781 7.48510 1,856 7.52613 1818 7.50562
2005 - 2009 50342 10.82659 53,206 10.88194 53,931 10.89546 53,568 10.88870
321
Appendices
Appendix 5DZ: Selection of the estimator suitable for preserved fruits and nuts
exports from South Africa to the SADe countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 90 5 2.74** OLS No 96 4 12.51* OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 90 9 2.86* OLS No 96 8 12.75* OLS No
FE or RE Yes FE or RE Yes
Export Direction 1275 6 2.85· OLS No 1360 5 13.34" OLS No
FE or RE Yes FE or RE Yes
90 5 -0.20 FE-no auto Yes 96 4 1.33 FE-no auto No
Durbin Period Impact - - - - - 90 4 1.90*' FE-auto Yes
Watson - - - - - 96 5 1.24 RE-no auto No
Statistic - - - - - 90 5 1.94". RE-auto Yes
(for static) 90 9 -0.31 FE-no auto Yes 96 8 1.36 FE-no auto No
or Durbin- Yearly Impact - - - - - 90 8 2.00" FE-auto Yes
H Statistic - - - - - 96 9 1.29 RE-no auto No
tfor - - - - - 90 9 2.06** RE-auto Yes
dynamic) 1275 6 -0.22 FE-no auto Yes 1360 5 1.14 FE-no auto No
Export Direction - - - - - 1275 5 1.95*' FE-auto Yes
1275 7 -1.49 RE-no auto Yes 1360 6 0.74 RE-no auto No
- - - - - 1275 6 2.05" RE-auto Yes
Period Impact - - - - - N/A 4 7.57 FE No
Hausman - - RE Yes
Test Yearly Impact - - - - - N/A 8 8.15 FE No
Statistic - - RE Yes
Export Direction N/A 6 384.82" FE Yes N/A 5 -14.84** FE Yes
RE No RE No
NB: ", *' & '** denote significance at the I, 5 and 10percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -110 auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of I st order autocorrelation_rroblem re~ectively.
322
Appendices
Appendix SEA: Suitable equations for preserved fruits and nuts exports from South
Africa to the SADC countries.
MODEL
InYijt_1
InGDPPCit
InGDPPCj,
D0004 / DOS09
DOO/DOS
DOl/D06
D02/D07
D03/D08
D04/D09
PTAyes
PTA no
Appendix 5EB: Average actual, simulated and potential value of preserved fruits
and nuts exports from South Africa to SADC countries for the periods 2000-2004
and 2005-2009 in dollars (US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adjusted Potential
Name Code Value Value Value Value Value Value Value Value
(US$) (Lozs) (US$) (Logs) (US$) (Logs) (US$) (Logs)
Malawi 2000 - 2004 50,D30 10.82037 35,035 10.46409 32,928 10.40209MWI 33,965 10.433092005 - 2009 164,221 12.00897 143,514 11.87419 140,095 11.85008 141,795 11.86213
Mauritius 752,447MUS 2000 - 2004 13.53109 771,405 13.55597 775,983
13.56189 773,691 13.55893
2005 - 2009 2,042,383 14.52963 2,474,215 14.72143 2,586,664 14.76588 2,529,815 14.74366
Mozambique 2000 - 2004 267,437MOZ 12.49664 333,867 12.71850 350,550 12.76726 342,107 12.742882005 - 2009 624,886 13.34532 776,325 13.56233 812,459 13.60782 794,187 13.58507
Tanzania 2000 - 2004 19,217 9.86354 19,831 9.89499 19,933 9.90012TZA 19,882 9.897552005 - 2009 79,752 11.28668 69,808 11.15351 68,280 11.13137 69,040 11.14244
Zambia 2MB 2000 - 2004 215,575 12.28106 181,782 12.11056 175,476 12.07526 178,601 12.092912005 - 2009 1,045,258 13.85977 976,976 13.79222 962,955 13.77776 969,940 13.78499
Zimbabwe ZWE 2000 - 2004 109,009 11.59919 139,772 11.84777 146,959 11.89791 143,321 11.87284
2005 - 2009 551,471 13.22034 512,495 13.14705 504,964 13.13224 508,716 13.13964
323
Appendices
Appendix SEC: Selection of the estimator suitable for fruits and vegetable juices
exports from South Africa to the SADC countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 90 5 1.34 OLS Yes 96 4 4.23* OLS No
Wald Test FE or RE No FE or RE Yes
Statistic Yearly Impact 90 9 1.50 OLS Yes 96 8 4.66* OLS No
FE or RE No FE or RE Yes
Export Direction 1290 6 4.67* OLS No 1376 5 16.99* OLS No
FE or RE Yes FE or RE Yes
90 5 0.46 OLS-no auto Yes 96 4 1.05 FE-no auto No
Durbin Period Impact - - - - - 90 4 1.86** FE-auto Yes
Watson - - - - - 96 5 0.95 RE-no auto No
Statistic - - - - - 90 5 1.74 RE-auto ?
(jor static) 90 9 0.74 OLS-no auto Yes 96 8 0.91 FE-no auto No
or Durbin- Yearly Impact - - - - - 90 8 1.86** FE-auto Yes
H Statistic - - - - - 96 9 0.84 RE-no auto No
(jor - - - - - 90 9 1.72 RE-auto ?
dynamic) 1290 6 -0.03 FE-no auto Yes 1376 5 1.53 FE-no auto No
Export Direction - - - - 1290 5 1.97** FE-auto Yes
1290 7 -1.50 RE-no auto Yes 1376 6 0.96 RE-no auto No
- - - - 1290 6 2.09** RE-auto Yes
Period Impact - - - - - - - - - -
Hausman - - - -
Test Yearly Impact - - - - - - - - - -
Statistic - - - -
Export Direction N/A 6 605.04* FE Yes N/A 5 51.59* FE Yes
RE No RE No
NB: *, ** & *** denote significance at the I, 5 and 10percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of I" order autocorrelation problem respectively. ?
means inconclusive.
324
Appendices
Appendix SED: Suitable equations for fruits and vegetable juices exports from
South Africa to the SADC countries.
MODEL
InVijt_1
InGDPPCit
InGDPPCjt
REERt
00004 / 00509
DOO/ 005
001/006
002/007
003/008
004/009
PTAyes
PTA no
to 2004, SA exported fruits and vegetable juices to the following 6 SADC countries: MOZ, MWI, as to the following
80 non-SADC countries that were added under the export direction model: AGO, ARE, ARG, AUS, AUT, BEL, BOl, BEN, BGD, BHR, BRA, CAN, CHE, CHL,
CHN, CIY, CMR, COG, COL, COM, CPY, CYP, DEU, DNK, DRC, EGY, ESP, ETH, FIN, FRA, GAB, GBR, GHA, GIN, GRC, HUN, lDN, IND, ISL, IRL, ISR,
ITA, JOR, JPN, KEN, KOR, KWT, LBN, LBR, LKA, MAR, MOG, MDY, MU, MLT, MYS, NGA, NLD, NOR, NZL, OMN, PAK, PHL, POL, PRT, RUS, RWA,
SAU SEN SGP SYC URY and USA
Appendix SEE: Average actual, simulated and potential value of fruits and
vegetable juices exports from South Africa to SADC countries for the periods 2000-
2004 and 2005-2009 in dollars (US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adiusted Potential
Name Code Value Value Value Value Value Value Value Value
(US$) (Logs) (US$) (Logs) (US$) (Logs) (US$) (Logs)
MW) 2000 - 2004 480 6.17377 714 6.57021 761 6.63481Malawi 737 6.60251
2005 - 2009 3,855 8.25703 5,666 8.64223 6,084 8.71341 5,871 8.67782
Mauritius 2000 - 2004 50,925 10.83810 26,761 10.19470
21,081 9.95612 23,752 10.07541
MUS 2005 - 2009 49,612 10.81199 68,061 11.12816 73,690 11.20762 70,820 11.16789
2000 - 2004 2,576 7.85389 3,574 8.18155 3,822 8.24855 3,696 8.21505Mozambique MOZ 2005 - 2009 12,743 9.45275 20,512 9.92878 22,759 10.03270 21,606 9.98074
Tanzania TZA 2000 - 2004 0 0.00000
0 0.00000 0 0.00000 0 0.00000
2005 - 2009 488 6.19073 91 4.50956 78 4.36061 84 4.43509
2000 - 2004 3,793 8.24086 2,921 7.97968 2,697 7.89975 2,807 7.93972Zambia 2MB 2005 - 2009 19,313 9.86852 30,241 10.31694 33,508 10.41954 31,832 10.36824
Zimbabwe 2000 - 2004 1,395 7.24090 1,781
7.48510 1,856 7.52613 1,818 7.50562
ZWE
2005 - 2009 50,342 10.82659 53,206 10.88194 53,931 10.89546 53,568 10.88870
325
Appendices
Appendix 5EF: Selection of the estimator suitable for fruits and vegetable juices
imports from the SADe countries to South Africa.
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 90 5 8.10* OLS No 96 4 24.47* OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 90 9 8.16* OLS No 96 8 27.94* OLS No
FE or RE Yes FE or RE Yes
Import Direction 645 6 4.27* OLS No 688 5 14.01* OLS No
FE or RE Yes FE or RE Yes
90 5 0.92 FE-no auto Yes 96 4 1.94** FE-no auto Yes
Durbin Period Impact - - - - - - - - - -
Watson 90 6 -1.91*** RE-no auto No 96 5 1.03 RE-no auto No
Statistic 84 6 -0.14 RE-auto Yes 90 5 2.14** RE-auto Yes
(for static) 90 9 0.77 FE-no auto Yes 96 8 1.83** FE-no auto Yes
or Durbin- Yearly Impact - - - - - - - - - -
H Statistic 90 10 -1.89*** RE-no auto No 96 9 1.01 RE-no auto No
(for 84 10 -0.82 RE-auto Yes 90 9 2.10** RE-auto Yes
dynamic) 645 6 0.10 FE-no auto Yes 688 5 1.53 FE-no auto No
Import Direction - - - - - 645 5 1.97** FE-auto Yes
645 7 -1.58 RE-no auto Yes 688 6 0.84 RE-no auto No
- - - - - 645 6 2.06** RE-auto Yes
Period Impact N/A 5 49.60* FE Yes N/A 4 0.39 FE No
Hausman RE No RE Yes
Test Yearly Impact N/A 9 57.89* FE Yes N/A 8 -0.78 FE No
Statistic RE No RE Yes
Import Direction N/A 6 334.71* FE Yes N/A 5 -0.41 FE No
RE No RE Yes
NB: ", ** & *** denote significance at the 1,5 and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of 1st order autocorrelation problem respectively.
326
Appendices
Appendix SEG: Suitable equations for fruits and vegetable juices imports from
SADC countries to South Africa
MODEL
InYijl_]
InGDPPCil
Appendix SEH: Average actual, simulated and potential value of fruits and
vegetable juices imports from the SADC countries to South Africa for the periods
2000-2004 and 2005-2009 in dollars (US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adjusted Potential
Name Code Value Value Value Value Value Value Value Value
(US$) (logs) (US$) (logs) (US$) (Logs) (US$) (logs)
Malawi MWI 2000 - 2004 9 2.22624 5 1.65396 5 1.60688 5 1.630422005 - 2009 3 1.18537 6 1.71776 14 2.66880 9 2.19328
Mauritius 2000 - 2004
294 5.68352 804 6.68984 1,257 7.13676 1,006 6.91330
MUS 2005 - 2009 1,064 6.96969 151 5.01676 1 877 7.53721 532 6.27699
Mozambique 2000 - 2004 0 0.00000 0 0.00000 0 0.00000MOZ 0 0.000002005 - 2009 26 3.25125 177 5.17475 11,929 9.38671 1,452 7.28073
Tanzania 2000 - 2004 0 0.00000 0 0.00000 0 0.00000TZA 0 0.000002005 - 2009 5 1.68473 7 1.91598 19 2.95378 11 2.43488
Zambia 4.22564 3.60462 3.481302MB 2000 - 2004 68 37 33 35 3.542962005 - 2009 0 0.00000 0 0.00000 0 0.00000 0 0.00000
Zimbabwe ZWE 2000 - 2004 15,053 9.61933
18,146 9.80622 21,156 9.95968 19,593 9.88295
2005 - 2009 536,572 13.19296 57 4.04710 46 3.83218 51 3.93964
327
Appendices
Appendix SEl: Selection of the estimator suitable for fruits and vegetable juices
trade between South Africa and the SADe countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 90 4 1.39 OLS Yes 96 3 2.85** OLS No
Wald Test FE or RE No FEar RE Yes
Statistic Yearly Impact 90 8 1.46 OLS Yes 96 7 3.01* OLS No
FE or RE No FE or RE Yes
Trade Direction 630 5 5.64* OLS No 672 4 20.89* OLS No
FE or RE Yes FE or RE Yes
90 4 0.33 OLS-no auto Yes 96 3 0.85 FE-no auto No
Durbin Period Impact - - - - - 90 3 1.73** FE-auto Yes
Watson - - - - - - - - - -
Statistic - - - - - - - - - -
(jar static) 90 8 0.48 OLS-no auto Yes 96 7 0.75 FE-no auto No
or Durbin- Yearly Impact - - - - - 90 7 1.84** FE-auto Yes
H Statistic - - - - - - - - - -
(jar - - - - - - - - - -
dynamic) 630 5 0.21 FE-no auto Yes 672 4 1.66 FE-no auto No
Trade Direction - - - - - 630 4 1.90** FE-auto Yes
630 6 -0.69 R.E-no auto Yes 672 5 1.40 R.E-no auto No
- - - - - 630 5 1.97** R.E-auto Yes
Period Impact - - - - - - - - - -
Hausman - - - -
Test Yearly Impact - - - - - - - - - -
Statistic - - - -
Trade Direction N/A 5 286.49* FE Yes N/A 4 39.13* FE Yes
RE No R.E No
NB: *, ** & *** denote significance at the 1,5 and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of Ist order autocorrelation_Qroblem re~ectively.
328
Appendices
Appendix 5EJ: Suitable equations for fruits and vegetable juices trade between
South Africa and the SADC countries
MODEL
2009
InYijl_t
InGDPPCij,
00004 / DOS09
DOO/DOS
DOI/D06
D02/D07
D03/D08
D04/D09
PTAyes
PTA no
Appendix 5EK: Average actual, simulated and potential value of fruits and
vegetable juices trade between South Africa and the SADC countries for the periods
2000-2004 and 2005-2009 in dollars (US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adjusted Potential
Name Code Value Value Value Value Value Value Value Value
(US$) (Logs) (US$) (Logs)_ (US$) (Logs) (US$) (Lozs)
MWI 2000 - 2004 2,187,294 14.59818 9,133,807 16.02749 2,313,582 14.65431 4,596,935 15.34090Malawi 2005 - 2009 2,043,085 14.52997 11,204,487 16.23182 5,031,380 15.43120 7,508,265 15.83151
Mauritius 2000 - 2004 3,303,426
15.01047 13,364,846 16.40814 3,202,823 14.97954 6,542,571 15.69384
MUS 2005 - 2009 8,376,812 15.94098 26,365,567 17.08757 9,720270 16.08972 16,008,761 16.58865
Mozambique MOZ 2000 - 2004 0 0.00000 0
0.00000 0 0.00000 0 0.00000
2005 - 2009 13,749,504 16.43651 27,765,350 17.13930 9,084,311 16.02206 15,881,721 16.58068
Tanzania 2000 - 2004 0 0.00000 0 0.00000 0 0.00000 0 0.00000TZA 2005 - 2009 6,567,164 15.69759 12,083,916 16.30739 4,134,550 15.23489 7,068,349 15.77114
Zambia 2MB 2000 - 2004 1,399,899
14.15191 5,441,263 15.50952 1,427,914 14.17173 2,787,410 14.84062
2005 - 2009 0 0.00000 0 0.00000 0 0.00000 0 0.00000
2000 - 2004 712,053 13.47591 2,460,955 14.71606 681,233 13.43166 1,294791 14.07386Zimbabwe ZWE 2005 - 2009 4,636,538 15.34948 10,648,991 16.18098 3,884,629 15.17254 6,431,747 15.67676
329
Appendices
Appendix SEL: Selection of the estimator suitable for wine exports from South
Africa to the SADe countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 90 5 4.72' OLS No 96 4 18.78' OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 90 9 4.23* OLS No 96 8 21.30* OLS No
FE or RE Yes FE or RE Yes
Export Direction 1530 6 4.71 * OLS No 1632 5 22.48* OLS No
FE or RE Yes FE or RE Yes
90 5 0.03 FE-no auto Ye.~ 96 4 1.20 FE-no auto No
Durbin Period Impact - - - - - 90 4 1.97** FE-auto Yes
Watson 90 6 -0.01 RE-no auto Yes 96 5 1.16 RE-no auto No
Statistic - - - - - 90 5 1.95** RE-auto Yes
(jar statiet 90 9 -1.33 FE-no auto Yes 96 8 1.17 FE-no auto No
or Durbin- Yearly Impact - - - - - 90 8 1.89** FE-auto Yes
H Statistic 90 10 -1.26 RE-no auto Yes 96 9 1.13 RE-no auto No
(jar - - - - - 90 9 1.88** RE-auto Yes
dynamic) 1530 6 -0.36 FE-no auto Yes 1632 5 1.44 FE-no auto No
Export Direction - - - - - 1530 5 2.04** FE-auto Yes
1530 7 -1.70 RE-no auto Yes 1632 6 1.06 RE-no auto No
- - - - - 1530 6 2.13** RE-auto Yes
Period Impact N/A 5 12.60** FE Yes N/A 4 0.42 FE No
Hausman RE No RE Yes
Test Yearly Impact N/A 9 10.93 FE No N/A 8 3.54 FE No
Statistic RE Yes RE Yes
Export Direction N/A 6 626.39* FE Yes N/A 5 7.08 FE No
RE No RE Yes
NB: *, ** & *** denote significance at the 1,5 and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of Ist order autocorrelation problem respectively.
330
Appendices
Appendix 5EM: Suitable equations for wine exports from South Africa to the SADC
countries.
MODEL
Appendix 5EN: Average actual, simulated and potential value of wine exports from
South Africa to SADC countries for the periods 2000-2004 and 2005-2009 in dollars
(US$) and natural logarithms (Logs) \
Country Country Period Actual Simulated Adjusted Potential
Name Code Value Value Value Value Value Value Value Value
(US$) (Logs) (US$) (Logs) (US$) (Logs) (US$) (Logs)
Malawi 2000 - 2004 359,872 12.79350 474,781 13.07061 499,952 13.12227 487,204 13.09644MWl 2005 - 2009 1,047,039 13.86148 1,108,430 13.91845 1,119,887 13.92874 1,114,143 13.92360
Mauritius 2000 - 2004 1,848,108 14.42967 1,780,787 14.39257 1,767,012 14.38480 1,773,886 14.38868MUS 2005 - 2009 8,148,743 15.91337 7,322,097 15.80641 7,159,378 15.78393 7,240,281 15.79517
Mozambique 2000 - 2004 2,972,795 14.90501 2,581,937 14.76405
2,504,573 14.73363
MOZ 2,542,961
14.74884
2005 - 2009 8,496,797 15.95520 10,527,850 16.16953 11,026,633 16.21582 10,774,356 16.19268
Tanzania 2000 - 2004 937,366 13.75083 562,352 13.23988 510,505 13.14316 535,802 13.19152TZA 2005 - 2009 6,097,274 15.62335 4,698,339 15.36272 4,456,214 15.30981 4,575,675 15.33626
Zambia 2000 - 2004 652,700 13.38887 857,788 13.66211 905,108 13.715812MB 881,130 13.688962005 - 2009 2,337,933 14.66478 2,262,981 14.63219 2,248,903 14.62595 2,255,931 14.62907
Zimbabwe 2000 - 2004 1,100,538 13.91131 1,264,250 14.04999 1,300,389 14.07817 1,282,192 14.06408ZWE 2005 - 2009 3,339,469 15.02132 3,798,792 15.15019 3,897,793 15.17592 3,847,974 15.16306
331
Appendices
Appendix SEO: Selection of the estimator suitable for wine imports from the SADe
countries to South Africa.
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 90 5 2.55** OLS No 96 4 8.40* OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 90 9 6.28* OLS No 96 8 8.63* OLS No
FE or RE Yes FEor RE Yes
Import Direction 675 6 6.95* OLS No 720 5 32.01* OLS No
FE or RE Yes FEor RE Yes
90 5 -0. I7 FE-no auto Yes 96 4 2.23** FE-no auto Yes
Durbin Period Impact - - - - - - - - - -
Watson 90 6 -3.96* RE-no auto No 96 5 0.92 RE-no auto No
Statistic 84 6 -1.50 RE-auto Yes 90 5 2.49** RE-auto Yes
(jar static) 90 9 0.26 FE-no auto Yes 96 8 2.20** FE-no auto Yes
or Durbin- Yearly Impact - - - - - - - - - -
H Statistic 90 10 -3.91 * RE-no auto No 96 9 0.95 RE-no auto No
(jar 84 10 -1.42 RE-auto Yes 90 9 2.51 ** RE-auto Yes
dynamic) 675 6 -0.06 FE-no auto Yes 720 5 1.76 FE-no auto No
Import Direction - - - - - 675 5 1.99** FE-auto Yes
675 7 -1.81*** RE-no auto No 720 6 1.2 I RE-no auto No
630 7 -0.69 RE-auto Yes 675 6 2.06** RE-auto Yes
Period Impact N/A 5 84.18* FE Yes N/A 4 2.96 FE No
Hausman RE No' RE Yes
Test Yearly Impact N/A 9 71.72* FE Yes N/A 8 1.34 FE No
Statistic RE No RE Yes
Import Direction N/A 6 395.2 I* FE Yes N/A 5 -3.40 FE No
RE No RE Ye.~
NB: ", ** & *** denote significance at the I, 5 and 10percent levels respectively. N & K denote the sample size and the number of regressors
respectively. aLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of ISI order autocorrelation_l)l'oblem re~flectively.
332
Appendices
Appendix SEP: Suitable equations for wine imports from SADC countries to South
Africa
MODEL
InYijl_]
InGDPPCit
InGDPPCjl
D0004 I DOS09
DOO IDOS
DOI/D06
D021 D07
D03/D08
D04/D09
PTAyes
PTA no
Appendix SEQ: Average actual, simulated and potential value of wine imports from
the SADC countries to South Africa for the periods 2000-2004 and 2005-2009 in
dollars (US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adjusted Potential
Name Code Value Value Value Value Value Value Value Value
(US$) (logs) (US$) (Logs) (US$) (Logs) (US$) (Logs)
Malawi MWl 2000 - 2004 3 1.06895 41 3.71532 67
4.19750 52 3.95641
2005 - 2009 221 5.39921 148 4.99436 47 3.84358 83 4.41897
Mauritius 2000 - 2004 20 2.99060 12 2.49939MUS 8
2.04748 10 2.27343
2005 - 2009 195 5.27081 81 4.39943 27 3.30879 47 3.85411
Mozambique MOZ 2000 - 2004 0 0.00000 0 0.00000 0 0.00000 0 0.000002005 - 2009 10 2.34148 75 4.32368 50 3.90840 61 4.11604
Tanzania TZA 2000 - 2004 0 0.00000 0 0.00000 0 0.00000 0 0.000002005 - 2009 24 3.16617 144 4.96999 84 4.42722 110 4.69861
Zambia 2MB 2000 - 2004 5
1.59174 31 3.42282 35 3.56553 33 3.49417
2005 - 2009 0 0.00000 0 0.00000 0 0.00000 0 0.00000
Zimbabwe ZWE 2000 - 2004 416 6.03096 32 3.47191 8 2.03587 16 2.753892005 - 2009 26 3.26204 95 4.55119· 51 3.93995 70 4.24557
333
Appendices
. ,
Appendix SER: Selection of the estimator suitable for wme trade between South
Africa and the SADe countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 90 4 5.56* OLS No 96 3 24.46" OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 90 8 5.03" OLS No 96 7 27.53' OLS No
FE or RE Yes FE or RE Yes
Trade Direction 660 5 3.82* OLS No 704 4 41.27" OLS No
FE or RE Yes FE or RE Yes
90 4 0.15 FE-no auto Yes 96 3 1.15 FE-no auto No
Durbin Period Impact - - - - - 90 3 1.99** FE-auto Yes
Watson 90 5 -0.06 RE-no auto Yes 96 4 1.02 RE-no auto No
Statistic - - - - - 90 4 2.05" RE-auto Yes
(for static) 90 8 -1.56 FE-no auto Yes 96 7 1.18 FE-no auto No
or Durbin- Yearly Impact - - - - - 90 7 2.05** FE-auto Yes
H Statistic 90 9 -1.56 RE-no auto Yes 96 8 1.02 RE-no auto No
(for - - - - - 90 8 2.11 "' RE-auto Yes
dynamic) 660 5 1.12 FE-no auto Yes 704 4 1.11 FE-no auto No
Trade Direction - - - - - 660 4 1.71 FE-auto No
660 6 0.68 RE-no auto Yes 704 5 0.97 RE-no auto No
- - - - - 660 5 1.68 RE-auto No
Period Impact N/A 4 31.08" FE Yes N/A 3 6.22 FE No
Hausman RE No RE Yes
Test Yearly Impact N/A 8 23.62' FE . Yes N/A 7 -46.72' FE Yes
Statistic RE No RE No
Trade Direction N/A 5 146.61* FE Yes - - - - -
RE No - -
NB: " *" & '** denote significance at the 1,5 and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of Ist order autocorrelation problem respectively.
334
Appendices
Appendix SES: Suitable equations for wine trade between South Africa and the
SADC countries.
MODEL
21109
InVijl_.
InGDPPCijl
REERI
D0004 / D0509
DOO / D05
DOl / D06
D02 / D07
D03 / D08
D04/ D09
PTA yes
PTA no
Appendix SET: Average actual, simulated and potential value of wine trade between
South Africa and the SADC countries for the periods 2000-2004 and 2005-2009 in
dollars (US$) and natural logarithms (Logs)
Country Country Period Actual Simulated Adjusted Potential
Name Code Value Value Value Value Value Value Value Value
(lIS$) (Logs) (lJS$) (Logs) russi (Logs) (lJS$) (Logs)
Malawi MW) 2000 - 2004 359,918 12.79363 282,517 12.55150 262,784 12.47909 272,472 12.51529
2005 - 2009 1,047,786 13.86219 1,147,627 13.95321 1,171,215 13.97355 1,159,361 13.96338
Mauritius 2000 - 2004 1,849,108 14.43021 1,866,674 14.43967 1,873,042 14.44307 1,869,855 14.44137MlJS 2005 - 2009 8,154,548 15.91409 7,336,922 15.80843 7,137,364 15.78085 7,236,455 15.79464
Mozambique MOZ 2000 - 2004 0 0.00000 0 0.00000 0 0.00000 0 0.000002005 - 2009 8,497,060 15.95523 10,273,520 16.14508 10,809,868 16.19597 10,538,282 16.17053
Tanzania TZA 2000 - 2004 0 0.00000 0 0.00000 0
0.00000 0 0.00000
2005 - 2009 6,097,486 15.62339 4,608,899 15.34350 4,295,754 15.27314 4,449,572 15.30832
Zambia 2MB 2000 - 2004 652,729 13.38892
697,631 13.45545 713,003 13.47724 705,275 13.46634
2005 - 2009 0 0.00000 0 0.00000 0 0.00000 0 000000
Zimbabwe 2000 - 2004 1,101,089 13.91181 1,300,115 14.07796 1,377,518 14.13579 1,338,257 14.10688ZWE
2005 - 2009 3,341,342 15.02188 3,710,059 15.12656 3,807,220 15.15241 3,758,326 15.13948
335
Appendices
Appendix SED: Selection of the Estimator suitable for agricultural exports from
South Africa to the ROW countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 1470 4 5.38* OLS No 1568 3 31.96* OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 1470 8 5.41 * OLS No 1568 7 32.37* OLS No
FE or RE Yes FE or RE Yes
Durbin 1470 4 -0.25 FE-no auto 1568 3 1.31 FE-no auto No
Watson Period Impact - - - - - 1470 3 1.85 FE-auto No
Statistic 1470 5 -0.55 RE-no auto 1568 4 1.18 RE-no auto No
(for static) - - - - - 1470 4 1.78 RE-auto No
or Durbin- 1470 8 -0.31 FE-no auto 1568 7 1.31 FE-no auto No
H Statistic Yearly Impact - - - - - 1470 7 1.87 FE-auto No
(for 1470 9 -0.60 RE-no auto 1568 8 1.18 RE-no auto No
dynamic) - - - - - 1470 8 1.80 RE-auto No
Period Impact N/A 4 390.56* FE Yes - - - - -
Hausman RE No - -
Test Yearly Impact N/A 8 389.54* FE Yes - - - - -
Statistic RE No - -
NB: *, ** & *** denote significance at the I, 5 and IOpercent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of I st order autocorrelation problem respectively.
336
Appendices
Appendix 5EV: Suitable equations for agricultural exports from South Africa to the
ROW countries.
MODEL
2004 2009 2004 2009
InYijt_1
InGDPPCit
InGDPPCjt
00004 / 00509
DOO/ 005
001/006
002/007
003/008
From 1994 to 2009, SA exported agricul uCISto cou (i.e. non-EU and non-SADC countries): AGO,
ARE, ARG, ATG, AUS, BOl, BEN, BGR, BI-IR, BHS, BRA, CAF, CAN, CHE, CHL, CHN, CIY, CMR, COG, COL, COM, CPY, CRI, CVP,
CZE, DOM, ORC, EGY, ETH, GAB, GHA, GIN, GMB, GNQ, HUN, ION, INO. lRN, ISL, ISR, JAM, JOR, JPN, KEN, KOR, KWT, LBN,
LBR, LKA, MAR, MOA, MOG, MOY, MEX, MU, MLT, MRT, MYS, NER, NGA, NOR, NZL, OMN, PAK, PAN, PER, PHL, PNG, POL,
PRI, PRY, ROM, RUS, RWA, SAU, SON, SEN, SGP, SLE, STP, SUR, SYK, SVN, SYC, SYR, TCO, TGO, THA, TTO, TUN, TUR, UGA,
URY YCT YEN VNM and YEM
337
Appendices
Appendix SEW: Selection of the estimator suitable for agricultural imports from the
ROW countries to South Africa
Selection Models Dynamic Static
Criteria N K Statistic Estimator Oecision N K Statistic Estimator Oecision
Period Impact 1590 4 5.72* OLS No 1696 3 28.95* OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 1590 8 5.82* OLS No 1696 7 29.25* OLS No
FE or RE Yes FE or RE Yes
Durbin 1590 4 0.12 FE-no auto Yes 1696 3 1.60 FE-no auto No
Watson Period Impact - - - - - 1590 3 1.94*- FE-auto Yes
Statistic 1590 5 -1.55 RE-no auto Yes 1696 4 1.13 RE-no auto No
(for static) - - - - - 1590 4 2.09** RE-auto Yes
or Durbin- 1590 8 0.13 FE-no auto Yes 1696 7 1.60 FE-no auto No
H Statistic Yearly Impact - - - - - 1590 7 1.94** FE-auto Yes
(for 1590 9 -1.54 RE-no auto Yes 1696 8 1.14 RE-no auto No
dynamic) - - - - - 1590 8 2.09** RE-auto Yes
Period Impact N/A 4 841.79* FE Yes N/A 3 -0.01 FE No
Hausman RE No RE Yes
Test· Yearly Impact N/A 8 855.33* FE Yes N/A 7 0.19 FE No
Statistic RE No RE Yes
NB: ", ** & *** denote significance at the 1,5 and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of Ist order autocorrelation problem respectively.
338
Appendices
Appendix SEX: Suitable equations for agricultural imports from the ROW
countries to South Africa
MODEL
InGDPPCi,
InGDPPCj,
D0004 / D0509
DOO/ D05
DOl / D06
D02/ D07
D03 / D08
D04/ D09
From 1994 to 2009, SA imported agricultu ucts from 106 countries (i.e, non-EU and non-SADe countries): AGO,
ALB, ARE, ARG, ATG, AUS, BOl, BEN, BGO, BGR, BI'IR, BHS, BOL, BRA, BTN, CAN, CHE, CHL, CHN, CIV, CMR, COG, COL, COM,
CRI, CYP, CZE, OMA, DOM, ORC, ECU, EGY, EST, ETH, Gi-IA, GIN, GMB, GRO, GTM, GUY,'HRV, HTI, HUN, ION, INO, IRN, ISL,
ISR, JAM, JOR, JPN, KEN, KGZ, KOR, KWT, LAO, LBN, LKA, MAR, MOG, MEX, ML!, MRT, MYS, NER, NGA, NIC, NOR, NPL, NZL,
OMN, PAK, PAN, PER, PI-IL, POL, PRI, PRY, ROM, RUS, RWA, SAU, SON, SEN, SGP, SLE, SLV, STP, SUR, SVK, SVN, SYC, SYR,
TC UK URY USA VCT VEN and VNM
339
Appendices
Appendix 5EY: Selection of the estimator suitable for agricultural trade between
South Africa and the ROW countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 1320 4 121.60' OLS No 1408 3 50.50' OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 1320 8 119.27* OLS No 1408 7 50.66' OLS No
FE or RE Yes FE or RE Yes
Durbin 1320 4 0.56 FE-no auto Yes 1408 3 1.35 FE-no auto No
Watson Period Impact - - - - - 1320 3 1.99** FE-auto Yes
Statistic 1320 5 0.12 RE-no auto Yes 1408 4 1.28 RE-no auto No
(jar static) - - - - - 1320 4 1.98*" RE-auto Yes
or Durbin- 1320 8 0.50 FE-no auto Yes 1408 7 1.34 FE-no auto No
H Statistic Yearly Impact - - - - - 1320 7 1.98** FE-auto Yes
(jar 1320 9 0.04 RE-no auto Yes 1408 8 1.27 RE-no auto No
dynamic) - - - - - 1320 8 1.97*" RE-auto Yes
Period Impact N/A 4 414.84' FE Yes N/A 3 -1.36 FE No
Hausman RE No RE Yes
Test Yearly Impact N/A 8 398.92' FE Yes N/A 7 -2.12 FE No
Statistic RE No RE Yes
NB: ',** & *** denote significance at the 1,5and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no.autocorrelation problem and estimation with the correction of I" order autocorrelation_2roblem re~ectively.
340
Appendices
Appendix SEZ: Suitable equations for agricultural trade between South Africa and
the ROW countries
MODEL
InYjjt_1
InGOPPCjjt
00004 / 00509
DOO/ 005
.,.','.
001/006
002/007
003/008
004/009
341
Appendices
Appendix SFA: Selection of the estimator suitable for cheese exports from South
Africa to the ROW countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 300 4 2.46** OLS No 320 3 10.71* OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 300 8 2.50** OLS No 320 7 10.84* OLS No
FE or RE Yes FE or RE Yes
Durbin 300 4 -0.14 FE-no auto Yes 320 3 1.26 FE-no auto No
Watson Period Impact - - - - - 300 3 1.92** FE-auto Yes
Statistic 300 5 -1.47 RE-no auto Yes 320 4 0.78 RE-no auto No
(for static) - - - - - 300 4 2.08** RE-auto Yes
or Durbin- 300 8 -0.11 FE-no auto Yes 320 7 1.26 FE-no auto No
H Statistic Yearly Impact - - - - - 300 7 1.94** FE-auto Yes
(for 300 9 -1.47 RE-no auto Yes 320 8 0.78 RE-no auto No
dynamic) - - - - - 300 8 2.09** RE-auto Yes
Period Impact N/A 4 91.63* FE Yes N/A 3 7.88*** FE Yes
Hausman RE No RE No
Test Yearly Impact N/A 8 92.12* FE Yes N/A 7 6.49 FE No
Statistic RE No RE Yes
NB: *, ** & *** denote significance at the I, 5 and ID percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of I SI order autocorrelation problem respectively.
342
Appendices
Appendix SFB: Suitable equations for cheese exports from South Africa to the
ROW countries.
MODEL
2004 2009
InYijt_1
InGDPPCit
InGDPPCjt
D0004 / D0509
DOO / D05
DOl / D06
D02/ D07
D03/ D08
D04/ D09
non-EU and non-SADe countries): AGO, ARE, BOl, CHE,
SYC UGA and USA
343
Appendices
Appendix 5FC: Selection of the estimator suitable for cheese imports from the ROW
countries to South Africa
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 135 4 3.78* OLS No 144 3 15.20* OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 135 8 3.84* OLS No 144 7 15.90* OLS No
FE or RE Yes FE or RE Yes
Durbin 135 4 0.01 FE-no auto Yes 144 3 1.35 FE-no auto No
Watson Period Impact - - - - - 135 3 2.07** FE-auto Yes
Statistic 135 5 -0.98 RE-no auto Yes 144 4 0.56 RE-no auto No
(jor static) - - - - - 135 4 2.16** RE-auto Yes
or Durbin- 135 8 -0.54 FE-no auto Yes 144 7 1.39 FE-no auto No
H Statistic Yearly Impact - - - - - 135 7 2.11 ** FE-auto Yes
(jor 135 9 -1.38 RE-no auto Yes 144 8 0.59 RE-no auto No
dynamic) - - - - - 135 8 2.17** RE-auto Yes
Period Impact N/A 4 62.60* FE Yes N/A 3 0.94 FE No
Hausman RE No RE Yes
Test Yearly Impact N/A 8 62.61 * FE Yes N/A 7 0.89 FE No
Statistic RE No RE Yes
NB: *, ** & *** denote significance at the 1,5and 10percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of l" order autocorrelation problem respectively.
344
Appendices
Appendix 5FD: Suitable equations for cheese imports from the ROW countries to
South Africa
MODEL
InYijt_1
InGDPPCit
InGOPPCjt
00004 / 00509
DOO/ 005
001 /006
002/007
003/008
004/009
imported cheese from the following 9 ROW countries (i.e. non-EU and non-SADe countries): ARG, AUS, BGR,
POL and USA
345
Appendices
Appendix 5FE: Selection of the estimator suitable for cut flowers exports from
South Africa to the ROW countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 540 4 2.62** OLS No 576 3 18.63* OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 540 8 2.69* OLS No 576 7 18.97* OLS No
FE or RE Yes FE or RE Yes
Durbin 540 4 0.09 FE-no auto Yes 576 3 1.07 FE-no auto No
Watson Period Impact - - - - - 540 3 1.83 FE-auto No
Statistic 540 5 -0.84 RE-no auto Yes 576 4 0.84 RE-no auto No
(for static) - - - - - 540 4 1.91** RE-auto Yes
or Durbin- 540 8 0.11 FE-no auto Yes 576 7 1.09 FE-no auto No
H Statistic Yearly Impact - - - - - 540 7 1.84 FE-auto ?
(for 540 9 -0.81 RE-no auto Yes 576 8 0.85 RE-no auto No
dynamic) - - - - - 540 8 1.92** RE-auto Yes
Period Impact N/A 4 126.77* FE Yes - - - - -
Hausman RE No - -
Test Yearly Impact N/A 8 129.36* FE Yes - - - - -
Statistic RE No - -
NB: *, ** & *** denote significance at the I, 5 and IOpercent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of l" order autocorrelation problem respectively. ?
means inconclusive.
346
Appendices
Appendix 5FF: Suitable equations for cut flowers exports from South Africa to the
ROW countries.
MODEL
InYijl_1
InGDPPCiI
InGDPPCjt
00004 / 00509
DOO/ DOS
DOl /006
D02 / D07
D03 / 008
D04 / 009
347
Appendices
Appendix 5FG: Selection of the estimator suitable for cut flowers imports from the
ROW countries to South Africa
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 195 4 4.13* OLS No 208 3 10.44* OLS No
Wald Test FE or RE Yes FEor RE Yes
Statistic Yearly Impact 195 8 3.90* OLS No 208 7 10.74* OLS No
FE or RE Yes FE or RE Yes
Durbin 195 4 -0.06 FE-no auto Yes 208 3 1.75 FE-no auto ?
Watson Period Impact - - - - - 195 3 1.96 FE-auto Yes
Statistic 195 5 -1.70*** R.E-no auto No 208 4 1.05 RE-no auto No
(jar static) 182 5 -0.70 RE-auto Yes 195 4 2.15** RE-auto Yes
or Durbin- 195 8 0.00 FE-no auto Yes 208 7 1.69 FE-no auto No
H Statistic Yearly Impact - - - - - 195 7 1.97** FE-auto Yes
(jar 195 9 -1.71*** RE-no auto No 208 8 1.03 RE-no auto No
dynamic) 182 9 -0.98 R.E-auto Yes 195 8 2.14** RE-auto Yes
Period Impact N/A 4 110.86* FE Yes N/A 3 2.10 FE No
Hausman RE No RE Yes
Test Yearly Impact N/A 8 74.35* FE Yes N/A 7 1.86 FE No
Statistic RE No RE Yes
NB: *, ** & *** denote significance at the 1,5 and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of ISI order autocorrelation problem respectively. ?
means inconclusive. .
348
Appendices
Appendix 5FH: Suitable equations for cut flowers imports from the ROW countries
to South Africa
MODEL
InYijt_1
InGDPPCit
InGDPPCjt
REERt
00004 / 00509
DOO/005
001/006
002/007
003/008
004/009
349
Appendices
Appendix SFI: Selection of the estimator suitable for cut flowers trade between
South Africa and the ROW countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 135 4 2.77** OLS No 144 3 13.06· OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 135 8 3.46" OLS No 144 7 14.24· OLS No
FE or RE Yes FE or RE Yes
Durbin 135 4 1.23 FE-no auto Yes 144 3 1.24 FE-no auto No
Watson Period Impact - - - - - 135 3 1.71 FE-auto ?
Statistic 135 5 -0.36 RE-no auto Yes 144 4 1.0I RE-no auto No
(for static) - - - - - 135 4 1.74 RE-auto ?
or Durbin- 135 8 1.07 FE-no auto Yes 144 7 1.29 FE-no auto No
li Statistic Yearly Impact - - - - - 135 7 1.75 FE-auto ?
(for 135 9 -0.19 RE-no auto Yes 144 8 1.08 RE-no auto No
dynamic) - - - - - 135 8 1.72 RE-auto ?
Period Impact N/A 4 37.28· FE Yes - - - - -
Hausman RE No - -
Test Yearly Impact N/A 8 36.68" FE Yes - - - - -
Statistic RE No - -
NB: ., •• & ••• denote significance at the 1,5and 10percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of l" order autocorrelation problem respectively. ?
means inconclusive.
350
Appendices
Appendix 5FJ: Suitable equations for cut flowers trade between South Africa and
the ROW countries
MODEL
InGDPPCijt ? ? ? ?
REERt ? ? ? ?
00004 / 00509 ? ? ? ?
DOO/005 ? ? ? ?
001/006 ? ? ? ?
002/007 ? ? ? ?
003/008 ? ? ? ?
351
Appendices
Appendix 5FK: Selection of the estimator suitable for frozen fruits and nuts exports
from South Africa to the ROW countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 135 4 2.18*** OLS No 144 3 9.03* OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 135 8 2.18** OLS No 144 7 9.19* OLS No
FE or RE Yes FE or RE Yes
Durbin 135 4 0.13 FE-no auto Yes 144 3 1.14 FE-no auto No
Watson Period Impact - - - - - 135 3 1.93** FE-auto Yes
Statistic 135 5 2.52** RE-no auto No 144 4 0.61 RE-no auto No
(jar static) 126 5 -1.25 RE-auto Yes 135 4 2.17** RE-auto Yes
or Durbin- 135 8 0.03 FE-no auto Yes 144 7 1.11 FE-no auto No
H Statistic Yearly Impact - - - - - 135 7 2.00** FE-auto Yes
(jar 135 9 2.56** RE-no auto No 144 8 0.61 RE-no auto No
dynamic) 126 9 -1.46 RE-auto Yes 135 8 2.21 ** RE-auto Yes
Period Impact N/A 4 49.22* FE Yes N/A 3 1.09 FE No
Hausman RE No RE Yes
Test Yearly Impact N/A 8 45.98* FE Yes N/A 7 -51.27* FE Yes
Statistic RE No RE No
NB: *, ** & *** denote significance at the I, 5 and ID percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects-and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of I st order autocorrelation problem respectively.
352
Appendices
Appendix 5FL: Suitable equations for frozen fruits and nuts exports from South
Africa to the ROW countries.
MODEL
2004 2009
InVijl_1
InGDPPCil
InGDPPCjl
00004 / 00509
DOO/ 005
001/006
002/007
003/008
004/009
353
Appendices
Appendix 5FM: Selection of the estimator suitable for frozen fruits and nuts
imports from the ROW countries to South Africa
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 210 4 2.72* OLS No 224 3 5.74* OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 210 8 2.79* OLS No 224 7 6.03* OLS No
FE or RE Yes FE or RE Yes
Durbin 210 4 -0.51 FE-no auto Yes 224 3 1.52 FE-no auto No
Watson Period Impact - - - - - 210 3 1.97** FE-auto Yes
Statistic 210 5 -2.13** RE-no auto No 224 4 0.66 RE-no auto No
(jar static) 196 5 -0.81 RE-auto Yes 210 4 2.43** RE-auto Yes
or Durbin- 210 8 -0.59 FE-no auto Yes 224 7 1.51 FE-no auto No
H Statistic Yearly Impact - - - - - 210 7 1.99** FE-auto Yes
(jar 210 9 -2.09** RE-no auto No 224 8 0.67 RE-no auto No
dynamic) 196 9 -0.73 RE-auto Yes 210 8 2.39** RE-auto Yes
Period Impact N/A 4 108.39* FE Yes N/A 3 8.05** FE Yes
Hausman RE No RE No
Test Yearly Impact N/A 8 157.98* FE Yes N/A 7 -4.90 FE No
Statistic RE No RE Yes
NB: *, ** & *** denote significance at the 1,5and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of Ist order autocorrelation problem respectively.
354
Appendices
Appendix 5FN: Suitable equations for frozen fruits and nuts imports from the ROW
countries to South Africa
MODEL
2004 2009
InYijt_1
InGDPPCit
InGDPPCjt
REERt
00004 / 00509
DOO/ 005
001/006
002/007
003/008
004/009
355
Appendices
Appendix 5FO: Selection of the estimator suitable for preserved fruits and nuts
exports from South Africa to the ROW countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 960 4 2.49** OLS No 1024 3 10.73* OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 960 8 2.55** OLS No 1024 7 10.92* OLS No
FE or RE Yes FE or RE Yes
Durbin 960 4 -0.22 FE-no auto Ye.~ 1024 3 1.14 FE-no auto No
Watson Period Impact - - - - - 960 3 1.95** FE-auto Yes
Statistic 960 5 -1.39 RE-no auto Yes 1024 4 0.68 RE-no auto No
(jar static) - - - - - 960 4 2.10** RE-auto Yes
or Durbin- 960 8 -0.21 FE-no auto Yes 1024 7 1.16 FE-no auto No
H Statistic Yearly Impact - - - - - 960 7 1.95** FE-auto Yes
(jar 960 9 - 1.40 RE-no auto Yes 1024 8 0.69 RE-no auto No
dynamic) - - - - - 960 8 2.10** RE-auto Yes
Period Impact N/A 4 280.16* FE Yes N/A 3 -3.16 FE No
Hausman RE No RE Yes
Test Yearly Impact N/A 8 283.97* FE Yes N/A 7 2.83 FE No
Statistic RE No RE Yes
NB: *, ** & *** denote significance at the I, 5 and IOpercent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of Ist order autocorrelation problem respectively.
356
Appendices
Appendix 5FP: Suitable equations for preserved fruits and nuts exports from South
Africa to the ROW countries.
MODEL
InGDPPC;t
InGDPPCjt
REERt
00004 / 00509
DOO/005
001/006
002/007
003/008
004/009 .
357
Appendices
Appendix 5FQ: Selection of the estimator suitable for preserved fruits and Duts
imports from the ROW countries to South Africa
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 390 4 3.84* OLS No 416 3 8.48* OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 390 8 3.89* OLS No 416 7 8.53* OLS No
FE or RE Yes FE or RE Yes
Durbin 390 4 -0.09 FE-no auto Yes 416 3 1.30 FE-no auto No
Watson Period Impact - - - - - 390 3 1.96** FE-auto Yes
Statistic 390 5 -1.95*** RE-no auto No 416 4 0.68 RE-no auto No
(for static) 308 5 -0.48 RE-auto Yes 390 4 2.07** RE-auto Yes
or Durbin- 390 8 -0.14 FE-no auto Yes 416 7 1.30 FE-no auto No
H Statistic Yearly Impact - - - - - 390 7 1.96** FE-auto Yes
(for 390 9 -1.96** RE-no auto No 416 8 0.68 RE-no auto No
dynamic) 308 9 -0.57 RE-auto Yes 390 8 2.06** RE-auto Yes
Period Impact N/A 4 166.03* FE Yes N/A 3 3.25 FE No
Hausman RE No RE Yes
Test Yearly Impact N/A 8 165.83* FE Yes N/A 7 3.46 FE No
Statistic RE No RE Yes
NR: *, ** & *** denote significance at the 1,5and 10percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of l" order autocorrelation ...2_roblemre~ectively.
358
Appendices
Appendix 5FR: Suitable equations for preserved fruits and nuts imports from the
ROW countries to South Africa
MODEL
InYijt_1
InGDPPCil
InGDPPCjl
D0004 / 00509
DOO/ D05
DOl /006
D02 / 007
D03 / 008
D04/D09
359
Appendices
Appendix SFS: Selection of the estimator suitable for preserved fruits and nuts
trade between South Africa and the ROW countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 300 4 6.87* OLS No 320 3 12.64" OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 300 8 6.82" OLS No 320 7 12.70" OLS No
FE or RE Yes FE or RE Yes
Durbin 300 4 0.11 FE-no auto Yes 320 3 1.84*· FE-no auto Yes
Watson Period Impact - - - - - - - - - -
Statistic 300 5 -0.15 RE-no auto Yes 320 4 1.56 RE-no auto No
(for static) - - - - - 300 4 1.82 RE-auto ?
or Durbin- 300 8 0.09 FE-no auto Yes 320 7 1.86** FE-no auto Yes
H Statistic Yearly Impact - - - - - - - - - -
(for 300 9 -0.15 RE-no auto Yes 320 8 1.58 RE-no auto No
dynamic) - - - - - 300 8 1.79 RE-auto ?
Period Impact N/A 4 247.53' FE Yes - - - - -
Hausman RE No - -
Test Yearly Impact N/A 8 234.80· FE Yes - - - - -
Statistic RE No - -
NB: ., •• & *** denote significance at the I, 5 and 10percent levels jespectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of I" order autocorrelation problem respectively. ?
means inconclusive.
360
Appendices
Appendix 5FT: Suitable equations for preserved fruits and nuts trade between
South Africa and the ROW countries
MODEL
InGDPPCijt
00004 / 00509
DOO/ 005
001/006
002/007
003/008
361
Appendices
Appendix 5FU: Selection of the estimator suitable for fruits and vegetable juices
exports from South Africa to the ROW countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 990 4 4.05· OLS No 1056 3 14.87* OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 990 8 4.07* OLS No 1056 7 14.90· OLS No
FE or RE Yes FE or RE Yes
Durbin 990 4 -0.06 FE-no auto Yes 1056 3 1.49 FE-no auto No
Watson Period Impact - - - - - 990 3 1.96** FE-auto Yes
Statistic 990 5 -1.62 RE-no auto Yes 1056 4 0.90 RE-no auto No
(for static) - - - - - 990 4 2.10·· RE-auto Yes
or Durbin- 990 8 -0.06 FE-no auto Yes 1056 7 1.49 FE-no auto No
H Statistic Yearly Impact - - - - - 990 7 1.96*· FE-auto Yes
(for 990 9 -1.61 RE-no auto Yes 1056 8 0.91 RE-no auto No
dynamic) - - - - - 990 8 2.10** RE-auto Yes
Period Impact N/A 4 442.68" FE Yes N/A 3 1138.06" FE Yes
Hausman RE No RE No
Test Yearly Impact N/A 8 443.98* FE Yes N/A 7 24.74" FE Yes
Statistic RE No RE No
NB: *, ** & *** denote significance at the 1,5 and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of I SI order autocorrelation problem respectively.
362
Appendices
Appendix 5FV: Suitable equations for fruit and vegetable juices exports from South
Africa to the ROW countries.
MODEL
InGDPPCit
InGDPPCjt
REERt
00004 / 00509
DOO/ 005
001/006
002/007
D03 / 008
363
Appendices
Appendix 5FW: Selection of the estimator suitable for fruits and vegetable juices
imports from the ROW countries to South Africa
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact OLS No OLS No
Wald Test 360 4 3.84" FE or RE Yes 384 3 9.92* FE or RE Yes
Statistic Yearly Impact OLS No OLS No
360 8 3.89* FE or RE Yes 384 7 10.00* FE or RE Yes
Durbin 360 4 -0.09 FE-no auto Yes 384 3 1.66 FE-no auto No
Watson Period Impact - - - - - 360 3 1.98*- FE-auto Yes
Statistic 360 5 -1.95*** RE-no auto No 384 4 0.84 RE-no auto No
(for static) 308 5 -0.48 RE-auto Yes 360 4 2.21·· RE-auto Yes
or Durbin- 360 8 -0.14 FE-no auto Yes 384 7 1.68 FE-no auto No
H Statistic Yearly Impact - - - - - 360 7 1.98** FE-auto Yes
(for 360 9 -1.96** RE-no auto No 384 8 0.85 RE-no auto No
dynamic) 308 9 -0.57 RE-auto Yes 360 8 2.20·· RE-auto Yes
Period Impact FE Yes FE No
Hausman N/A 4 166.03* RE No N/A 3 -3.02 RE Yes
Test Yearly Impact FE Yes FE No
Statistic N/A 8 165.83· RE No N/A 7 -2.83 RE Yes
NB: *, ** & *** denote significance at the 1,5and 10percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of I" order autocorrelation problem respectively.
364
Appendices
Appendix 5FX: Suitable equations for fruit and vegetable juices imports from the
ROW countries to South Africa
MODEL
InYijt_1
InGDPPCit
InGOPPCjt
REERt
r D0004 / D0509
DOO / D05
DOl / D06
D02/007
D03 / 008
D04/ D09
365
Appendices
Appendix 5FY: Selection of the estimator suitable for fruits and vegetable juices
trade between South Africa and the ROW countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 345 4 6.29* OLS No 362 3 19.30* OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 345 8 6.35* OLS No 362 7 19.21* OLS No
FE or RE Yes FEar RE Yes
Durbin 345 4 0.65 FE-no auto Yes 362 3 1.67 FE-no auto No
Watson Period Impact - - - - - 345 3 1.91** FE-auto Yes
Statistic 345 5 0.21 RE-no auto Yes 362 4 1.38 RE-no auto No
(jar static) - - - - - 345 4 1.86** RE-auto Yes
or Durbin- 345 8 0.68 FE-no auto Yes 362 7 1.67 FE-no auto No
H Statistic Yearly Impact - - - - - 345 7 1.90** FE-auto Yes
(jar 345 9 0.25 RE-no auto Yes 362 8 1.39 RE-no auto No
dynamic) - - - - - 345 8 1.86** RE-auto Yes
Period Impact N/A 4 191.88* FE Yes N/A 3 14.29* FE Yes
Hausman RE No RE No
Test Yearly Impact N/A 8 190.62· FE Yes N/A 7 26.98* FE Yes
Statistic RE No RE No
NB: *, .* & *** denote significance at the I, 5 and 10percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of I" order autocorrelation problem respectively.
366
Appendices
Appendix SFZ: Suitable equations for fruit and vegetable juices trade between
South Africa and the ROW countries
MODEL Period Impact Yearly Impact
Dynamic Static Dynamic Static
ESTIMATORS FE FE FE FE FE FE FE FE
PERIOD 2000 2005 2000 2005 2000 2005 2000 2005
- - - - - - - -
VARIABLES 2004 2009 2004 2009 2004 2009 2004 2009
Constant - - - - - - - -
InYijt_1 0.21* 0.21* - - 0.21* 0.21· -
(4.34) (4.34) (4.17) (4.17)
InGDPPCijt 4.74* 4.74· 9.16* 9.16* 4.52* 4.52* 9.00* 9.00*
(3.54) (3.54) (549) (549) (3.27) (3.27) (4.77) {_Ull
REERt -0.25 -0.25 -0.03 -0.03 -0.21 -0.21 0.11 0.11
(-0.64) (-0.64) (-0.05) (-0.05) (-054) (-0.54) (0.22) (0.22)
D0004 / D0509 0.81*** 0.66·· 0.97** 0.12 - - - -
(189) (2.16) (2.48) (0.34)
DOO / D05 - - - - 0.95** 0.63· ... 0.82·" 0.05
(2.01) (171) (190) (0.14)
001/ D06 - - - - 0.71 0.52 0.73 -0.17
(J .16) (1.31 ) (1.08) (-0.39)
D02 / D07 - - - - 2.22* 0.56 1.88*" -0.13
(2.61) ( 1.30) (181) (-0.26)
D03 / D08 - - - - 2.23* 0.93** 2.31* 0.26
(3.54) (213) (307) (050)
D04 / D09 - - - - 1.43** 0.93** 2.13* 0.51
(251) (2.18) (3.18) (1.00)
InDISTj' - - - - - - - -
Adjusted Rl 0.73 0.73 0.88 0.88 0.73 0.73 0.87 0.87
Observations 345 345 345 345 345 345 345 345
Cross-Sections 23 23 23 23 23 23 23 23
*, ** & *** denote significance at the 1,Sand IOpercent levels respectively. t-values are in parentheses
From 1994 to 2009, SA traded (imports plus exports) fruits and vegetable juices with the following 23 ROW countries (i.e. non-EU and non-
SADe countries); ARE, ARG, AUS, BRA, CAN, CHE, CHL, CHN, IDN, TND, ISL, ISR, jPN, KEN, LKA, MYS, NZL, PHL, POL, SAU,
SGP, THA and USA
/'
367
Appendices
Appendix 5GA Selection of the estimator suitable for wine exports from South
Africa to the ROW countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 1215 4 4.30* OLS No 1296 3 17.46* OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 1215 8 4.34* OLS No 1296 7 17.67* OLS No
FEor RE Yes FE or RE Yes
Durbin 1215 4 -0.35 FE-no auto Yes 1296 3 1.45 FE-no auto No
Watson Period Impact - - - - - 1215 3 2.04** FE-auto Yes
Statistic 1215 5 -1.71*** RE-no auto No 1216 4 1.0I RE-no auto No
(for static) 1134 5 -0.17 RE-auto Yes 1140 4 2.16** RE-auto Yes
or Durbin- 1215 8 -0.42 FE-no auto Yes 1216 7 1.46 FE-no auto No
H Statistic Yearly Impact - - - - - 1140 7 2.04** FE-auto Yes
(for 1215 9 -I. 76*** RE-no auto No 1216 8 1.02 RE-no auto No
dynamic) 1134 9 -0.24 RE-auto Yes 1140 8 2.17** RE-auto Yes
Period Impact N/A 4 496.22* FE Yes N/A 3 1.89 FE No
Hausman RE No RE Yes
Test Yearly Impact N/A 8 539.72* FE Yes N/A 7 0.99 FE No
Statistic RE No RE Yes
NB: *, ** & *** denote significance at the 1,5 and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of Ist order autocorrelation _BJ'oblemresgectively.
368
Appendices
Appendix 5GB: Suitable equations for wine exports from South Africa to the ROW
countries.
MODEL
InYijt_1
InGDPPCit
InGDPPCjt
D0004 / D0509
DOO/ DOS
DOl / D06
D02 / D07
D03/ D08
D04/ D09
369
L- _
Appendices
Appendix SCC: Selection of the estimator suitable for wine imports from the ROW
countries to South Africa
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 360 4 5.83* OLS No 384 3 23.39* OLS No
Wald Test FE or RE Yes FE or RE Yes
Statistic Yearly Impact 360 8 5.87* OLS No 384 7 23.56* OLS No
FE or RE Yes FE or RE Yes
Durbin 360 4 0.15 FE-no auto Yes 384 3 1.68 FE-no auto No
Watson Period Impact - - - FE-auto - 360 3 1.97** FE-auto Yes
Statistic 360 5 -1.31 RE-no auto Yes 384 4 1.03 RE-no auto No
(jar static) _. - - RE-auto - 360 4 2.05** RE-auto Yes
or Durbin- 360 8 0.10 FE-no auto Yes 384 7 1.69 FE-no auto No
H Statistic Yearly Impact - - - FE-auto - 360 7 1.97** FE-auto Yes
(jar 360 9 -1.34 RE-no auto Yes 384 8 1.05 RE-no auto No
dynamic) - - - RE-auto - 360 8 2.05** RE-auto Yes
Period Impact N/A 4 190.08* FE Yes N/A 3 -18.81 * FE No
Hausman RE No RE Yes
Test Yearly Impact N/A 8 191.11* FE Yes N/A 7 30.90* FE No
Statistic RE No RE Yes
NB: *, ** & *** denote significance at the 1,5and 10 percent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, FE & RE denote Pooled Ordinary Least Squares, Fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of I" order autocorrelation problem respectively.
370
Appendices
Appendix 5GD: Suitable equations for wine imports from the ROW countries to
South Africa
MODEL
InGDPPC;t
InGDPPCjt
00004 / D0509
DOO / DOS
001/ D06
002/007
003/008
371
Appendices
Appendix 5GE: Selection of the estimator suitable for wine trade between South
Africa and the ROW countries
Selection Models Dynamic Static
Criteria N K Statistic Estimator Decision N K Statistic Estimator Decision
Period Impact 345 OLS No 362 OLS No
Wald Test 4 3.25' fE or RE Yes 3 22.79* fEor RE Yes
Statistic Yearly Impact 345 OLS No 362 OLS No
8 3.22* fE or RE Yes 7 22.88* fE or RE Yes
Durbin 345 4 1.25 fE-no auto Yes 362 3 1.13 fE-no auto No
Watson Period Impact - - - - - 345 3 1.73 fE-auto No
Statistic 345 5 0.65 RE-no auto Yes 362 4 0.95 RE-no auto No
(jar static) - - - - - 345 4 1.71 RE-auto No
or Durbin- 345 8 1.29 fE-no auto Yes 362 7 1.10 fE-no auto No
H Statistic Yearly Impact - - - - - 345 7 1.75 fE-auto No
(jar 345 9 0.65 RE-no auto Yes 362 8 0.92 RE-no auto No
dynamic) - - - - - 345 8 1.76 RE-auto No
Period Impact fE Yes - - - - -
Hausman N/A 4 84.40* RE No - -
Test Yearly Impact fE Yes - - - - -
Statistic N/A 8 83.17* RE No - -
NB: *, ** & *** denote significance at the I, 5 and IOpercent levels respectively. N & K denote the sample size and the number of regressors
respectively. OLS, fE & RE denote Pooled Ordinary Least Squares, fixed Effects and Random Effects models respectively. -no auto & -auto
denote estimation assuming no autocorrelation problem and estimation with the correction of l" order autocorrelation problem respectively.
372
Appendices
Appendix 5GF: Suitable equations for wine trade between South Africa and the
ROW countries
MODEL
InGDPPCijt
00004 / 00509
DOO / DOS
001/006
002/007
003/008
373
Appendices
Appendix SGG: Lists of countries and their codes
Code Country Name Code Country Name Code Country Name
AGO Angola GIN Guinea OMN Oman
ALB Albania GMB Gambia, The PAK Pakistan
ARE United Arab Emirates GRC Greece PAN Panama
ARG Argentina GRD Grenada PER Peru
ATG Antigua and Barbuda GTM Guatemala PHL Philippines
AUS Australia GUY Guyana POL Poland
AUT Austria HRV Croatia PRI Puerto Rico
BDI Burundi HTI Haiti PRT Portugal
BEL Belgium HUN Hungary PRY Paraguay
BEN Benin ION Indonesia ROM Romania
BGD Bangladesh IND India RUS Russian Federation
BGR Bulgaria IRL Ireland RWA Rwanda
BHR Bahrain IRN Iran, Islamic Rep. SA South Africa
BHS Bahamas, The ISL Iceland SAU Saudi Arabia
BOL Bolivia ISR Israel SON Sudan
BRA Brazil ITA Italy SEN Senegal
BTN Bhutan JAM Jamaica SGP Singapore
CAN Canacja JOR Jordan SLE Sierra Leone
CHE Switzerland JPN Japan SLV El Salvador
CHL Chile KEN Kenya STP Sao Tome and Principe
CHN China KGZ Kyrgyz Republic SUR Suriname
CIV Cote d'lvoire KOR Korea, Rep. SVK Slovak Republic
CMR Cameroon KWT Kuwait SVN Slovenia
COG Congo Republic LAO Lao POR SWE Sweden
COL Colombia LBN Lebanon SYC Seychelles
COM Comoros LKA Sri Lanka SYR Syrian Arab Republic
CRI Costa Rica LUX Luxembourg TCD Chad
CYP Cyprus MAR Morocco TGO Togo
CZE Czech Republic MOG Madagascar THA Thailand
DEU Germany MEX Mexico ITO Trinidad and Tobago
OMA Dominica MU Mali TUN Tunisia
DNK Denmark MOZ Mozambique TUR Turkey
DOM Dominican Republic MRT Mauritania TZA Tanzania
DRC Democratic Republic of Congo MUS Mauritius UGA Uganda
ECU Ecuador MWl Malawi UKR Ukraine
EGY Egypt, Arab Rep. MYS Malaysia URY Uruguay
ESP Spain NER Niger USA United States
EST Estonia NGA Nigeria VCT St. Vincent and the Grenadines
ETH Ethiopia NIC Nicaragua VEN Venezuela, RB
FIN Finland NLD Netherlands VNM Vietnam
FRA France NOR Norway 2MB Zambia
GBR United Kingdom NPL Nepal ZWE Zimbabwe
GHA Ghana NZL New Zealand
374
Appendices
Appendix 6.A: Detailed results of WTO AoA impacts on agricultural trade flows
Trade Flows Joint period effects Individual yearly effects
(1995 - 1999) (1995 - 1999)
Agric X + + (I)
Agric M 0 0
Agric T - - (2)
HS0406 X + + (1)
HS0406 M 0 0
HS0406 T 0 + (I)
HS0603 X + + (1)
HS0603 M 0 - (2)
HS0603 T 0 0
HS0811 X 0 0
HS0811 M 0 0
HS0811 T 0 + (1)
HS2008 X 0 0
HS2008 M 0 - (3)
HS2008 T 0 0
HS2009 X 0 0
HS2009 M 0 0
HS2009 T - - (I)
HS2204 X + + (I)
HS2204 M + + (5)
HS2204 T + + (4)
NB: X, M and T stand for exports, imports and trade respectively. For Joint period effects: + means
positive effects and - means negative effects). For Individual yearly effects: + (NY) means positive
effects (number of years) and - (NY) means negative effects (number of years). na means not
applicable
375
Appendices
Appendix 6.B: Detailed results of the EU-SA TDCA impacts on agricultural trade
flows
Trade Flows Joint period effects Individual yearly effects Direction effects Potential effects
2000 2005 2000 2005 2000 2005 2000 2005
- - - - - - - -
2004 2009 2004 2009 2004 2009 2004 2009
Agric - - + (8) + (8)X 0 - (1) - (3) + - (7) - (7)
Agric M 0 0 0 0 0 0 + (11) + (6)
- (4) - (9)
Agric T 0 + + 0 + (8) + (8)+ (2) (5) + - (7) - (7)
+ (4) + (2)
HS0406 X - 0 - (2) 0 0 0 - (2) - (3)
HS0406 M 0 0 0 0 0 + (9) + (7)- (5) - (4) - (6)
T 0 + + (1) + (4)HS0406 0 (I) 0 0 0 - (5) - (1)
0 0 + - 0 + (1) + (4)HS0603 X (2) 0 - (13) - (9)
+ (3) + (2)
HS0603 M + 0 0 0 0 0 - (4) - (2)
T - 0 + (3) + (1)HS0603 + - (1) + (5) 0 - (4) - (3)
HS0811 X - 0 - (5) 0 0 0 + (3) + (5)- (4) - (1)
+ (4)
HS0811 M 0 0 0 - (3) 0 0 + (3)- (4) - (2)
T 0 0 0 0 + (3) + (2)HS0811 0 0 - (3) - (2)
HS2008 X 0 0 0 0 0 0 + (7) + (7)- (8) - (7)
+ (5)
HS2008 M 0 0 - (2) 0 0 0 + (I)- (8) - (4)
HS2008 T 0 - (2) + 0 + (6) + (5)+ (5) + - (3) - (4)
HS2009 X 0 0 0 0 0 0 + (8) + (3)
- (6) - (11)
HS2009 M 0 0 0 0 0 0 + (3) + (6)- (10) - (7)
HS2009 T 0 0 0 + (9) + (5)-(I) 0 0 - (4) - (8)
HS2204 X - 0 0 - (5) - 0 + (8) + (8)- (7) - (7)
HS2204 M - 0 - (2) 0 - 0 + (7) + (8)- (8) - (7)
HS2204 T 0 0 0 + (I) - + (7) + (8)+ - (8) - (7)
NB: X, M and T stand for exports, imports and trade respectively. For .Joint period effects: + means positive effects and - means
negative effects. For Individual yearly effects: + (NY) means positive effects (number of years) and - (NY) means negative
effects (number of years). For Direction effects: + means trade flow creation and - means trade flow diversion. For potential
effects: + (NC) means potential underscored (number of countries) and - (NC) means potential exhausted (number of countries).
In all cases 0 means no effects and na means not applicable
376
Appendices
Appendix 6.C: Detailed results of the SADC Trade Protocol impacts on agricultural
trade flows
Trade Flows Joint period effects Individual yearly effects Direction effects Potential effects
2000 2005 2000 2005 2000 2005 2000 2005
- - - - - - - -
2004 2009 2004 2009 2004 2009 2004 2009
Agric X - + (2) + (3)+ - (5) + (3) - 0 - (4) - (3)
Agric M 0 0 + 0 0 0 + (2) + (3)(2) - (4) - (3)
Agric T - - (2) + 0 + (3) + (3)+ (5) + - (3) - (3)
HS0406 X 0 + (2) + (2)0 0 0 0 0 - (4) - (4)
HS0406 M na na na na na na na na
HS0406 T na na na na na na na na
HS0603 X 0 0 + + (3) + (3)(I) 0 - + - (3) - (3)
HS0603 M 0 0 - (5) - (5) 0 + (2) + (3)+ - (4) - (3)
HS0603 T 0 + - 0 + (I) + (3)+ - (I) (5) - (5) - (3)
HS0811 X 0 0 0 0 0 + (2) + (2)0 - (3) - (4)
HS0811 M na na na na na na na na
HS0811 T na na na na na na na na
HS2008 X 0 0 + (4) o. 0 + (4) + (2)+ - (2) - (4)
HS2008 M na na na na na na na na
HS2008 T na na na na na na na na
HS2009 X 0 0 - (I) - (3) 0 0 + (2) + (4)
- (4) - (2)
HS2009 M 0 0 0 0 0 + (2) + (3)- (3) - (2) - (2)
HS2009 T 0 + + (I) + (5) 0 0 + (4) + (5)
HS2204 X 0 - - (I) - (4) - 0 + (3) + (3)- (3) - (3)
HS2204 M 0 0 0 0 - 0 + (2) + (3)- (2) - (2)
HS2204 T 0 0 0 + - 0 + (3) + (3)(I)
- (I) - (2)
NB: X, M and T stand for exports, imports and trade respectively. For Joint period effects: + means positive effects and - means
negative effects. For Individual yearly effects: + (NY) means positive effects (number of years) and - (NY) means negative effects
(number of years). For Direction effects: + means trade flow creation and - means trade flow diversion. For potential effects: + (NC)
means potential underscored (number of countries) and - (NC) means potential exhausted (number of countries). In all cases 0
means no effects and na means not applicable
377
Appendices
Appendix 6.D: Detailed results of the agricultural trade flows response between
South Africa and ROW countries
Trade Flows Joint period effects Individual yearly effects
2000 - 2004 2005 - 2009 2000 - 2004 2005 - 2009
Agric X - + - (5) - (4)
Agric M 0 0 0 - (5)
Agric T + + + (2) + (5)
HS0406 X 0 + 0 0
HS0406 M 0 0 - (2) - (2)
HS0406 T na na na na
HS0603 X - + - (5) 0
HS0603 M 0 0 0 - (I)
HS0603 T 0 0 - (3) - (2)
HS08Jl X 0 0 0 0
HS0811 M 0 0 0 0
HS0811 T na na na na
HS2008 X 0 0 0 - (2)
HS2008 M 0 0 - (2) 0
HS2008 T + + + (I) + (4)
HS2009 X 0 0 + (2) 0
HS2009 M 0 0 0 0
HS2009 T + + + (4) + (3)
HS2204 X - + - (5) - (2)
HS2204 M - - - (I) 0
HS2204 T - 0 - (I) 0
NB: X, M and T stand for exports, imports and trade respectively. For .Joint period effects: + means positive
effects and - means negative effects). For Individual yearly effects: + (NY) means positive effects (number of
years) and - (NY) means negative effects (number of years). na means not applicable
Q.I-lV •~!.IFS .:J 378
lo ~il..O~M~Oi\Sl'Eflii\J
• ~gl8.!:I~:ne~Il(_•.ll~~IRY