Econometric estimation of the demand for meat in South Africa
Loading...
Date
2003-05
Authors
Taljaard, Pieter R.
Journal Title
Journal ISSN
Volume Title
Publisher
University of the Free State
Abstract
In this study the demand relations for meat in South Africa are estimated and interpreted. Two demand model specifications, namely the Rotterdam and Linearized Almost Ideal Demand System (LA/AIDS), were estimated and tested in order to determine which model provide the best fit for South African meat data. Tests for separability included an F and Likelihood ratio version. Both tests rejected the null hypothesis of weak separability between meat, eggs and milk as protein sources, indicating that the demand model for meat products should be estimated separately from eggs and milk. Consequently, separability tests between the four meat products fail to reject the null hypothesis, confirming that the four meat products should be modelled together. According to the Hausman exogeneity test, the expenditure term is exogenous. As a result, a Restricted Seemingly Unrelated Regression (RSUR) was used to estimate both models. Annual time series data from 1970 to 2000 were used. Both models were estimated in first differenced format, whereafter the estimated parameters were used to calculate compensated, uncompensated and expenditure elasticities. In a non-nested test, the Saragan’s and Vuong’s likelihood criterion, selected the LA/AIDS model. In terms of expected sign and statistical significance of the elasticities, the LA/AIDS also proved to be more suitable for South African meat data. Although the magnitudes of most own price and cross-price elasticities were significantly lower than previous estimates of demand relations for meat in South Africa, several reasons, including estimation techniques and time gaps, were offered as explanations for these differences. The uncompensated own price elasticity for beef (-0.7504) is the largest in absolute terms, followed by mutton (-0.4678), pork (-0.36972) and chicken (-0.3502). In terms of the compensated own price elasticities, which contain only the pure price effect, pork (-0.30592) was the most elastic, followed by mutton (-0.27713), chicken (-0.1939) and beef (-0.16111). The expenditure elasticities of beef (1.243) and mutton (1.181) are greater than one, indicating that beef and mutton are luxury goods in South Africa. The expenditure elasticity for beef is the most elastic; indicating that South African consumers as a whole, will increase their beef consumption as the total expenditure on meat products increase.
Description
Keywords
Livestock -- Economic aspects, Meat industry and trade -- South Africa -- Economic models, Dissertation (M.Sc. (Agricultural Economics)--University of the Free State, 2003