YALE DEPARTMENT OF ECONOMICS
WORKING PAPER NO. 23

IS FOREIGN DIRECT INVESTMENT PRODUCTIVE
IN THE LATIN AMERICAN CASE?
A PANEL UNIT ROOT AND PANEL COINTEGRATION ANALYSIS, 1980-2001

Miguel D. Ramirez

May 2007

This paper analyzes the theoretical and empirical links between key economic variables such as foreign direct investment (FDI) and private investment spending in Latin America during the 1980-2001 period. The pooled model for nine major Latin American countries tests the complementarity hypothesis which suggests that a ceteris paribus increase in FDI raises the marginal productivity of private capital by increasing the pool of available resources and enhancing the transfer of more advanced technology and better managerial practices. The paper also addresses the issue of whether changes in the real exchange rate (expenditure-switching policies) have a deflationary effect on the economies of Latin America. To test the hypothesis put forth by the critics of FDI that it diverts resources away from financing capital formation, the paper generates a "net" variable for FDI flows by deducting the repatriation of profits and dividends from the gross FDI inflows. In general, the findings suggest that (lagged) gross FDI, public investment spending, and real credit to the private sector have a positive and significant effect on private capital formation, while lagged changes in the real exchange rate, particularly its volatility, as measured by th period standard deviation, have a negative effect. In the case of the net FDI variable, it remains significant but its impact is reduced by more than half when the repatriation of profits and dividends is taken into account. The aforementioned estimates are obtained via the application of recently developed panel unit root and panel (Pedroni) cointegration tests on the included variables in the pooled investment function.

JEL Classifications: O10, O54, C33, H54