شرایط سیاسی و بحران ارز در بازارهای نوظهور

This article demonstrates the impact of structural political conditions on the likelihood of currency crises in emerging markets. Controlling for a standard and parsimonious set of macroeconomic variables, I find that: right-wing government is less conducive to currency crises; ‘strong’ governments (those with larger legislative majorities and those which face more fragmented legislative opposition) are also less vulnerable. Democracy also reduced the likelihood of currency crises in emerging markets; yet, in contrast to previous studies, this article does not find a significant impact of elections on the likelihood of currency crises.

Since the publication of Krugman's seminal (1979) study of currency crises, at least 40 empirical articles on this subject have appeared in scholarly journals. The Asian crisis has stimulated a new wave of such studies. Yet, econometric models built to explain currency crises, in keeping with the predominant theories of speculative attacks, have relied almost exclusively on macroeconomic indicators as explanatory variables. Kaminsky, Lizondo and Reinhart's (Kaminsky et al., 1998) extensive review of this literature identifies 60 explanatory variables that have appeared in various models of currency crises, the most commonly used indicators being the real exchange rate, international reserves, real GDP growth and the current account balance. Among those 60 explanatory variables, only four pertain directly to political events. It is curious, however, that so many studies have relied exclusively on macroeconomic indicators, since, as Drazen (2000) notes, decisions to abandon fixed exchange rates may more often reflect a choice among competing policy objectives than a technical incompatibility of monetary and fiscal policy with fixed exchange rates. For instance, Obstfeld (1994) characterizes the government as facing an explicit trade-off between maintaining the fixed exchange rate and other objectives, such as limiting the growth of unemployment. A government's willingness to trade-off inflation for unemployment is inevitably a political decision. Indeed, second-generation currency crisis models typically give rise to the possibility of multiple equilibria in which the realized outcome is a function of market expectations. This article proposes and tests the hypothesis that market expectations regarding a government's possible response to a speculative attack may be a function of political as well as economic conditions. Despite the considerable elegance of second-generation models in depicting optimal government behavior, such models open the door to political explanations of currency crises, but typically do not cross the threshold. While political factors have played a more central explanatory role in several qualitative studies of crises, particularly in Dornbusch et al. (1995), the theoretical and empirical literature has been largely silent on the subject. Drazen (2000) is among the first to address this gap in theory, though his article concentrates specifically on issue of contagion. Part of the theoretical motivations for the present analysis lies specifically in combining second-generation currency crisis models (I take Obstfeld (1994) as the archtype) with insights from partisan political business cycle theory (Hibbs, 1977 and Alesina, 1987). As described in Section 3, this combination of seemingly unrelated theories gives rise to the hypothesis that currency crises are less likely under right-wing governments. Related work (Frieden et al., 2000) suggests the further hypothesis that ‘strong’ governments may be less vulnerable to currency crises, which I confirm with two complementary indicators of strength. There is also reason in theory to hypothesize that the occurrence of elections might create uncertainties in expectations that could give rise to currency crises, and some previous empirical literature (Mishra, 1998, Mei, 1999 and Leblang, 2001) has found such effects. In contrast, I do not find elections to increase the likelihood of currency crises. Democracy itself is also a potentially important political condition that may shape market expectations, and Section 3 describes that the direction of this effect a priori is ambiguous. The article is organized as follows: Section 2 provides a brief review of the empirical literature on explaining and predicting currency crises, emphasizing the few studies that have highlighted political explanations. Section 3 outlines the theoretical motivation for considering the effect of political variables on the likelihood of currency crises and presents the hypotheses to be tested. Section 4 describes the data and empirical strategy used to derive the results presented in 5 and 6 concludes.

Political variables, to date, have been largely ignored in the empirical literature on currency crises. The few articles that have included political variables in explanations of currency crises have tended to concentrate exclusively on the occurrence of elections as the sole proxy for political risk. This article demonstrates that, controlling for a commonly used set of macroeconomic variables; key political variables are indeed statistically significant explanators of crises in emerging markets. However, contrary to previous studies, I find neither elections nor executive change more generally to be robust explanatory variables (though the point estimates are positive). Rather, this article demonstrates that more structural political conditions are much stronger correlates of currency crises in emerging markets. In particular, I find that right-wing government is differentially less vulnerable to currency crises, ‘strong’ governments as measured by both their margin of legislative majority and the fragmentation of opposition parties are also less vulnerable to currency crises, as are democracies. Each of these findings is consistent with the theoretical motivation proposed, in which political conditions add to macroeconomic fundamentals in shaping the expectations of market participants in multiple equilibria models of currency crises. Further theoretical work is required to explain more fully why some of these effects are present. Current second-generation models of currency crises can be interpreted as inviting political economy explanations, and a marriage of those models with models of political business cycles and related political economy models appears promising. Yet, numerous issues remain. As democratic reform increasingly takes hold in emerging markets, clarifying the effect of democracy on currency crises may become increasingly critical. Democracy's benefit may arise because of either greater public accountability leading to better policies or because of greater political stability increasing the likelihood of a good outcome in a multiple equilibrium world. Conversely, if further investigation finds a positive association between elections and crises, it may be because the type of pre-election policy interventions predicted by political business cycles theory are anticipated by markets, thus creating risk premia that make it more costly to defend currency pegs, potentially shifting expectations in a multiple equilibrium world. Similarly, the margin of government legislative majority appears to reduce the likelihood of currency crises. Yet, these benefits may come at a cost of considering competing policy objectives favored by minority parties. These follow-on issues warrant further investigation, both empirical and theoretical. Taken together, the key results from this study also have practical implications for managers and investors working in currency markets as well as public policy makers charged with their prudent regulation. I found that currency crises are less likely when governments are right-wing, ‘strong’ and democratic, thus highlighting greater currency market risks faced by governments lacking those characteristics. As the second-generation crisis models emphasize, negative market expectations can be self-fulfilling in a multiple equilibrium world. If, as these findings suggest, markets perceive ‘weak’ (small majority facing a unified opposition), or left-wing, or less democratic governments in emerging markets to be more vulnerable to speculative attacks, then they may indeed become more vulnerable. Yet, negative market expectations based on these institutional factors might be mitigated by other signals sent by credible government officials. Left-leaning, weak and less democratic governments should clearly communicate their intentions to defend exchange rate pegs and to foster the credibility of those intentions by implementing responsible fiscal and monetary policies. The findings suggest that the benefits of doing so are magnified for such governments, as currency traders may find in these results reason to increase their attentiveness to political conditions in emerging markets.