Expropriation of foreign-owned assets: Theoretical and empirical considerations.

Abstract

This study aims to ascertain factors influencing expropriation as well as the impact of the threat of expropriation on the host country and foreign direct investors. Three methods are used to analyze these questions: (1) a theoretical model of host country and foreign direct investor interaction, (2) an econometric study (logit and tobit) to estimate the probability of expropriation and determine the impact of various economic factors on that probability, and (3) an event study to determine the impact of the loss of an affiliate on the parent company. The empirical studies were carried out for U.S. foreign direct investment in 27 countries, 1970-85. In the model, production uses capital, managers and labor (the first two are mobile internationally). If the host expropriates, it faces a managerial shortage. Uncertainty is introduced with respect to the tastes of the host. The host is concerned with national income, self-determination, or a combination of the two. The foreign direct investors are perfectly competitive (each assumes that s/he has no influence on the probability of expropriation) or a monopolistic investor (who recognizes his/her influence on that probability). From the model, it is concluded: (1) the threat of expropriation causes perfectly competitive and monopolistic investors to restrict the amount of capital they invest--the latter more so than the former, and (2) the equilibrium probability of expropriation is positively related to the domestic managerial supply and inversely related to national income, the domestic capital stock, and penalties. These predicted relationships are tested empirically. The logit estimates the reduced form equation of the model with a binary dependent variable: a country gets a 1 if it seized any U.S. assets in a given year and a 0 otherwise. The independent variables are the factors and penalties. The tobit repeats the estimation with a quantified dependent variable: book value of U.S. assets seized. The relationships with respect to national income, domestic capital stock, and some of the penalties are not rejected by the empirical estimation. However, the hypothesis that the lack of a managerial shortage will trigger an expropriation is rejected by the empirical evidence. Finally, the impact of an expropriation on the U.S. parent company is estimated using an event study methodology. As expected, parent companies have earned statistically significant negative abnormal returns at the event of the loss of a foreign affiliate.