So many measures of trade openness and policy: Do any explain economic growth?

摘要

Does trade liberalization lead to economic growth? The question is an important one, particularly for developing countries. Unfortunately, the literature on the effects of trade liberalization is inconclusive. To a large degree this is because there is no consensus as to how the "openness" of a country to trade is measured.; Given that trade policy is multifaceted, employing a wide variety of instruments, and that other domestic policies can affect trade, the creation of a summary measure or index of openness is a challenging task. Researchers have put much thought and energy into trying to find a suitable measure and there now exists a large number and diversity of measures of trade openness and policy (it is, in fact, a growth industry).; The first part of the dissertation focuses on how analysts have measured trade openness and policy. Based on a dataset of 70 measures, I provide a guide to the types of measures available, highlighting the strengths and weaknesses of the different types, and suggest which measures might be most applicable and where further research holds promise.; This analysis of existing measures is not reassuring. They often lack theoretical foundations, are frequently poorly specified, or are simply ad hoc measures driven by data availability. Additionally, the majority focus on trade flows while what researchers really need are direct measures of policies and barriers.; The second part of the dissertation examines the explanatory power of the time-series measures. Using a common framework, a larger number of measures, and longer time period than previous studies, I find a number of useful results. First, the most widely cited measures, trade shares, have no explanatory power under any specification. Tariff measures, on the other hand, have substantial explanatory power and should be a component of composite indices, which offer the best opportunities for future measures. Second, the results show that longer time periods present more of a challenge to these measures, casting doubt on previously reported results. Third, over a longer time period terms of trade shocks appear to matter less to economic growth than factors such as investment.