An econometric model of capital flight from developing countries

This paper analyzes capital flight from a group of seventeen developing nations over the period 1978 to 1993. The paper briefly discusses several empirical definitions of capital flight and presents estimates of capital flight for the sample based on some of these measures. In general, the data reveal periodic episodes of dramatic flight through the late 1980s, at which point many nations began to experience strong capital inflows. Anecdotal evidence for the nations in our sample underpins our hypothesis that capital flight is driven by a heightened, pervasive risk which reflects the degree of domestic macroeconomic imbalance which is domestically undiversifiable. Our econometric model of the determinants of capital flight extends previous empirical studies of flight by expanding both the cross section of nations and time horizon of analysis. Given the panel data set, we consider a country specific error component to account for the possibility of unobserved country heterogeneity and employ fixed- effects and random-effects estimation. We instrument for potentially endogenous explanatory variables and in doing so consider a fixed-effects system. The results, based on several different measures of flight, highlight the importance of modelling flight with a country specific error component. While other proxies of the risk associated with macroeconomic imbalance are not significant, the central government surplus is negatively, statistically significantly related to flight. This highlights the motivation of investors to move capital both to escape future taxation directly and indirectly via monetization of deficits. Therefore, even when taking into account other measures of risk, the higher taxation risk, both directly and indirectly through expectations of future inflation, dominates the regressions.