Global Labour Division, Development and Inequality; with Christian Dressel

Studies carried out by Dollar (1992), Sachs and Warner (1995) and Edwards (1998) have demonstrated a long-term significant and positive correlation between openness and economic growth. A study conducted by Greenaway et al. (2002) arrives at the same conclusion but refers to the timelag of effects exerted on economic growth by trade liberalization. Rodriguez and Rodrik (2001) criticize those studies because of their weak empirical foundations and econometric flaws. One critical point is the appropriate operationalization of openness. Because openness entails more aspects than trade alone, it is required to take into account also the openness of financial, capital and labour markets. Measuring openness by the percentage of imports and exports in the GDP implies a problem of endogenousness, according to Rodriguez and Rodrik (2001). In their eyes, the growth of trade may be more a consequence than a cause of economic growth. Frankel and Romer (1999), as well as Irwin and Terviö (2002), address this problem by extending their equations. They instrumentalize, for example, openness with size of population and size of country, as well as distance between trade partners. According to Rodriguez and Rodrik (2001), this strategy entails the problem of the instrument’s correlation with other variables in the equation that exert an influence on economic growth independent of trade.