Exports and Economic Growth in a Simultaneous Equations Model

Abstract

In economic journals since the 1 960s there has been a steady stream of econometric cross-national tests of exports and economic growth. Most of these studies have contemporaneously tested the export growth-economic growth relationship in a single equation, typically finding a positive association between the two.' A few scholars have attempted to capture the relationship in a simultaneous equations model.2 There have also been efforts to measure better the direction of causation between export growth and economic growth by various single equation techniques.3 In total, there are well over 40 studies of these kinds. Econometric cross-national tests of relationships between international trade and economic growth have not been confined to economists. There also exists a rather cohesive body of literature found primarily among sociologists, with contributions from political scientists and international relations scholars as well.4 These studies, at least 20 in number, date from the 1970s and focus on trade structural characteristics-the volume, composition, and concentration of trade (concentration of leading exports and leading trade partners). These characteristics are combined with other presumed determinants of economic growth into a single equation multiple regression model in which the independent variables are typically lagged 10 to 15 years behind the dependent variable. The sociologists provide a much more pessimistic conclusion regarding the role of trade in LDC development than do the economists. Both sets of econometric studies have contributed to a better understanding of the impact of international trade on growth. Yet there has been virtually no interaction between the two groups ofscholars, and, hence, no learning from each other's research. This is not only because the sister professions tend not to interact (the scholars tend to publish in separate journals), but also because each group draws from a very different