- Regime Type, Rights, and Foreign Direct Investment in Latin AmericaA Brief Comment
Biglaiser and DeRouen (2006)1 have provided a thorough examination of the effects of different types of economic reforms on flows of foreign direct investment (FDI). Their main finding—that economic reforms were generally unsuccessful in generating inflows of FDI during the time series—will undoubtedly generate further discussion about neoliberalism in Latin America. Although the authors' focus is on economic reform, they also devote considerable attention to the significance of "good governance" variables, including the effects of political regime type. Biglaiser and DeRouen's (2006) paper adds to a growing number of studies that have produced conflicting findings regarding the effects of regime type and/or rights and liberties on FDI.2 Given the conflicting results in the literature, and the policy significance of the issue, a brief commentary on the question of regime type, rights, and FDI in Latin America (and in other developing areas) is warranted.
One problem in the literature is that researchers frequently employ measures of regime type that may not adequately capture cross-sectional differences in rights and civil liberties in Latin America. Similar to previous studies, Biglaiser and DeRouen (2006) use the Polity IV data set (Marshall and Jaggers 2002) as a measure of authoritarianism and democracy in recipient countries. The Polity data focus primarily on constraints faced by the executive, the degree of competition and openness in executive recruitment, and political competition in a [End Page 183] regime. Civil liberties and rights are not assessed separately but rather quite indirectly as part of broader rating of group, party, and electoral competition (Marshall and Jaggers 2002, 78–81). There is a potential drawback here. Decision makers in multinational firms may respond more to the degree to which certain liberties and political rights are present in a recipient country—such as freedom of the press or the associational rights of workers and peasants—than to the degree of overall electoral and political competition.3 To this extent, given the question of how well the Polity data capture cross-sectional differences in rights and liberties, statistical analyses that rely exclusively on the Polity data may be missing an important determinant of FDI flows. To address this issue, researchers using Polity should also conduct sensitivity analyses where the statistical models are estimated with direct measures of civil liberties and political rights.4 At least one study (Tuman 2005) conducting sensitivity analyses found that in fifteen Latin American countries for the period of 1981 to 1996, Polity had no effect while abuse of civil liberties and political rights (as measured by Freedom House  and the Political Terror Scale [Gibney 2004] data) had a positive and statistically significant effect on inflows of U.S. FDI.5 [End Page 184]
A second issue has to do with the measure used for FDI flows. A number of recent studies, including Biglaiser and DeRouen (2006), use data from the World Development Indicators (n.d.) that pool observations of FDI flows from firms based in many different "home" countries. The basic problem with this technique is the failure to control for firm behavior that may be a product of distinctive national differences.6 Multinational firms based in different countries may react in quite varied ways to the type of regime and degree of rights and liberties in recipient countries.7 Indeed, a number of factors in the home country of the multinational firm—ranging from partisan control, a tradition of bureaucratic guidance in investment decisions, the structure of corporate governance, or the relative strength of social movements—can be expected to influence the incentive structure facing corporate decision-makers, possibly making firms from some countries more sensitive to democracy, rights, and civil liberties in recipients than other firms. For these reasons, it is preferable to estimate models using, for example, separate measures of U.S., Japanese, and French FDI in Latin America as opposed to a single measure that pools FDI from firms based in many different countries.
Although many other points could be discussed here, attention to the methodological issues raised in this brief comment will, in my judgment, not only improve the...