Trade Liberalisation and Poverty in a Computable General Equilibrium (CGE) Model: The Sri Lankan Case

Abstract

Many trade and development economists, policy makers and policy analysts around the world believe that globalisation promotes growth and reduces poverty. There exists a large body of theoretical and empirical literature on how trade liberalisation helps to promote growth and reduce poverty. However, critics of globalisation argue that, in developing countries, integration into the world economy makes the poor poorer and the rich richer. The most common criticism of globalisation is that it increases poverty and inequality. Much of the research related to the link between openness, growth and poverty has been based on cross-country regressions. Dollar and Kraay (2000; 2001), using regression analysis, argue that growth is pro poor. Moreover, their study suggests that growth does not affect distribution and poor as well as rich could benefit from it. Later, they demonstrate that openness to international trade stimulates rapid growth, thus linking trade liberalisation with improvements in wellbeing of the poor. Several other cross-country studies demonstrate a positive relationship between trade openness and economic growth (see for example Dollar, 1992; Sach and Warner, 1995 and Edward, 1998). In contrast, Rodriguez and Rodrik (2001) question the measurements related to trade openness in economic models, and suggest that generalisations cannot be made regarding the relationship between trade openness and growth. Several other studies also criticise the pro poor growth argument based upon the claim of weak econometrics and place more focus on the distributional aspect (see, for example, Rodrik, 2000). Ultimately, openness and growth have therefore become an empirical matter, and so has the relationship between trade and poverty. These weaknesses of cross-country studies have led to a need to provide evidence from case studies. Systematic case studies related to individual countries will at least complement cross-country studies such as that of Dollar and Kraay. As Chen and Ravallion (2004, p.30) argue, 'aggregate inequality or poverty may not change with trade reform even though there are gainers and losers at all levels of living'. They further argue that policy analysis which simply averages across diversities may miss important matters that are critical to the policy debate. In this study, Sri Lanka is used as a case study and a computable general equilibrium (CGE) approach is adopted as an analytical framework. Sri Lanka was selected as an interesting case in point to investigate this linkage for the following reasons: although Sri Lanka was the first country in the South Asian region to liberalise its trade substantially in the late seventies, it still experiences an incidence of poverty of a sizeable proportion that cannot be totally attributed to the long-standing civil conflict. Moreover, trade poverty linkage within the Sri Lankan context has hardly received any attention, while multi-sectoral general equilibrium poverty analysis within the Social Accounting Matrix (SAM) based CGE model has never been attempted. In order to examine the link between globalisation and poverty, a poverty focussed CGE model for the Sri Lankan economy has been developed in this study. As a requirement for the development of such a model, a SAM of the Sri Lankan economy for the year 1995 has been constructed. Moreover, in order to estimate the intra group income distribution in addition to the inter group income distribution, income distribution functional forms for different household groups have been empirically estimated and linked to the CGE model in 'top down' mode: this will compute a wide range of household level poverty and inequality measurements. This is a significant departure from the traditional representative agent hypothesis used to specifying household income distributions. Furthermore, as the general equilibrium framework permits endogenised prices, an attempt was made to endogenise the change in money metric poverty line within the CGE model. Finally, a set of simulation experiments was conducted to identify the impacts of trade liberalisation in manufacturing and agricultural industries on absolute and relative poverty at household level. The results show that, in the short run, trade liberalisation of manufacturing industries increases economic growth and reduces absolute poverty in low-income household groups. However, it is observed that the potential benefits accruing to the rural low-income group are relatively low compared to other two low-income groups. Reduction in the flow of government transfers to households following the loss of tariff revenue may be blamed for this trend. In contrast, long run results indicate that trade liberalisation reduces absolute poverty in substantial proportion in all groups. It further reveals that, in the long run, liberalisation of the manufacturing industries is more pro poor than that of the agricultural industries. Overall simulation results suggest that trade reforms may widen the income gap between the rich and the poor, thus promoting relative poverty. This may warrant active interventions with respect to poverty alleviation activities following trade policy reforms.