Poverty, vulnerability and trade policy in general equilibrium

Computable General Equilibrium (CGE) models are increasingly used for global trade policy analyses. This thesis addresses three issues that have arisen in this context. The first is the question of model validation: Do these models capture the kind of behavior actually observed in markets? The second question pertains to the poverty impacts of market price variability versus trade reform. The third issue examined pertains to the incorporation of stockholding behavior in CGE models. Validation of CGE models is a challenging task, complicated by the absence of natural experiments against which to test the model. This thesis exploits the weather-induced supply volatility in agriculture, evaluating the model\u27s ability to reproduce observed price volatility in the wheat market. The GTAP model and a modified version geared to account for specifics of agricultural markets, the GTAP-AGR, are tested. Their performance is found to be improved upon by including a better specification of agricultural trade policies. The assessment of the relative magnitude of poverty impacts with respect to market price variability is addressed by mapping the price and income distributions generated by supply-side volatility to a family of household models of developing countries. Such volatility has a significant impact on poverty and the focus is to assess whether, in the context of volatile commodity markets, it is possible to discern the effects of trade liberalization on poverty. This depends, of course on the extent of trade reforms. For full liberalization, the impacts of trade reform are shown to be discernable in two thirds of the countries examined. However, in the case of the reforms currently being negotiated under the auspices of the WTO, the only measurable poverty effects arise in Latin-America. Finally, this thesis remedies a key shortcoming of CGE models by showing how stockholding can be incorporated. An approach to calibrating and validating the model is demonstrated for the staple grains sector. The stochastic supply shocks framework forms a natural vehicle for reconciling model outcomes with observed behavior. The incorporation of the stockholding structure should prove useful to researchers seeking to conduct short-run analysis of commodity market policy in a CGE context