Should have? Could have? Would have? The role of electoral pressures in banking crises.

摘要

The recent wave of global banking crises raises important questions about the role of public policy during banking crises. Should governments intervene in the banking sector to prevent a crisis? Could policymakers foresee or prevent a crisis? And if so, would politicians act? This thesis adds to current empirical literature on banking crises by investigating how political institutions and considerations affect banking crisis probability. A banking crisis occurs when banking institutions suffer from a major loss of capital, leading to bank failure, government assistance, and/or government intervention. Banking crises are also often preceded by credit booms and a subsequent bust in asset prices when the economic system can no longer sustain the rampant growth. While economic conditions are typically cited as the immediate cause of banking crises, research indicates that economic, institutional, and political decisions help shape an environment that is vulnerable to crisis. I hypothesize that banking crises are the result of government failure to address mounting vulnerabilities due to political considerations. Although policymakers have monetary, fiscal, and regulatory tools to recognize and act against booms, inaction or misguided policy may contribute to the onset of banking crisis---and rightly or wrongly, political considerations play a significant role. I use fixed-effects and random-effects panel logit models that incorporate macroeconomic, institutional and political conditions during 144 episodes of banking crises in 126 countries between the years of 1970--2010. The empirical analysis adds to current literature by controlling for country-specific variation in the panel dataset. The results confirm the hypothesis as well as enhance current literature on the economic and institutional causes of banking crises. Executive elections are found to have a robust and consistently positive relationship with the probability of banking crisis, while legislative elections, more competitive elections, higher fractionalization, and single party governments have a negative, although largely non-significant, effect on banking crisis probability. Political partisanship may also play a role in shaping banking crisis vulnerability. I posit that policymakers face different motivations and constraints based on the role of their policymaking body, leading to different policy outcomes in the face of banking crises.