Essays on Trade, Growth, and Inequality

Access to files

•   FULL-TEXT.PDF (pdf)

Abstract

This thesis explores the role of openness to international trade on economic development, characterised by economic growth and income inequality, across countries. The existing empirical literature has shown that the effect of trade on economic development may not be well captured by models that impose a completely linear functional form. Thus, the majority of existing literature has ignored the possibility that the effect of trade on economic development may be subject to model mis-specification or be contingent on other factors. This thesis therefore re-examines the effects of openness to international trade on economic development across countries by addressing these gaps in the literature. First, in chapter 2 we contribute to the existing literature as we systematically address shortcomings related to model mis-specification and variable mis-measurement levelled at previous studies examining the effect of trade on economic growth. In order to do so, we apply nonparametric models to a novel dataset that treats trade among countries as a weighted network. Our application on a panel of 107 countries from 1970 to 2012 finds that, compared to the commonly used measure of exports and imports as a share of GDP, a network based perspective on trade is more relevant in explaining economic growth. We also find that this effect is generally linear which suggests that trade does not have a threshold effect on economic growth. We also exploit the fact that our dataset measures trade in raw goods and trade in capital goods to show that being closer to the world trade network in terms of the latter has the stronger effect on economic growth. This follows the suggestion of Hausmann et al. (2007), among others, that, in order to develop industry, developing economies should trade in goods that developed economies also trade in. We also document significant heterogeneities in the effect of trade on economic growth across time and regions. For instance, across time, trade in raw goods significantly increased economic growth in the mid-to-late 1970s while trade in capital goods had a significantly positive effect in the decade after 1985. Across regions, trade had the strongest effect on economic growth in the Middle East & North Africa while it had no significant effect on economic growth in Sub-Saharan Africa. These results, robust to functional form and endogeneity, support the conclusion of Rodriguez and Rodrik (2000) in that the effect of trade on economic growth is contingent on time and geography. Second, in chapter 3, we contribute to the existing literature as we show that, unlike in previous studies, increased trade may not necessarily enhance income inequality across countries. Our approach differs from the existing literature as it explicitly shows that schooling is a threshold variable that conditions the effect of trade on income inequality. Our application on a panel of 118 countries from 1970 to 2014 finds the threshold models of Hansen (2000) and Caner and Hansen (2004) to be superior to interaction models typically used in previous studies. We also document that, ceteris paribus, the inequality enhancing effect of trade found in previous studies can be reversed once (a) re-distributive policies are present and (b) countries accumulate, on average, approximately 6 years of schooling per capita. This is consistent with the model of Cartiglia (1997), among others, that predicts an inverse-U shaped relationship between trade and inequality. We also examine potential heterogeneities in the threshold effects of schooling and show that the strongest threshold effect lies in tertiary schooling per capita with 0.250 years being the estimated threshold value. Comparing our estimated threshold values to raw data on types of schooling across regions, we find that Sub-Saharan Africa is the only region that consistently falls short of these values while at the same time it is the region least open to international trade. Our results also suggest that “Cash for Education” programmes are promising forms of re-distributive policies that can help developing economies achieve these schooling “targets” and reverse the inequality widening effect of increased trade found in previous studies.

Keyword(s)

Economic Growth; Income Inequality; Nonlinearities; Nonparametric Regression; Schooling; Threshold Regression; Trade.

Bibliographic metadata

Institutional metadata

Record metadata