Elections, Political Competition and Bank Failure

We model and predict that politicians have incentives to delay bank failure in election years and that this incentive is exacerbated if the election is close. Our empirical application using the US data supports these predictions. At the bank level, we show that bank failure in an election year is four times less likely to occur if the election was among the most competitive (top quartile). At the state level, bank failure is about 1.8 times less likely to occur in an election year. A three point swing in the competitiveness of the election increases this election year bias to 2.2.