摘要：Building up human capital and other complementarities may be important in the link between exports of manufactures and economic growth. On the other hand, managerial strategies that push for export promotion may be important, too. Though both may yield non-linearities in the link between exports and growth, the associated patterns differ. In this paper we take an aseptic, empirical view in the link between these two variables and the possible nonlinear links. Since direct testing for non-linearities in panel data may yield non-significant results although they may actually be present, we propose a very simple method that may serve as a first approximation to uncover such non-linearities. We also take into consideration endogeneity and reverse causality problems (Arellano and Bover, 1995), and definitional problems in our variable of interest. In fact, we use a panel of 96 countries for the period 1960–1995 and find evidence consistent with the presence of non-linearities. We apply formal sensitivity analysis and confirm the results.
