Association for Public Policy Analysis & Management

Corruption is arguably one of the biggest problems in developing young democracies around the world. Elections provide a channel for voters to hold government leaders accountable for their actions. With the probability of losing future elections, political leaders face an increased incentive to be successful. There is, however, a surprising lack of causal evidence on the effectiveness of electoral accountability on improving government performance. In this research, we investigate whether governments perform better in the context of a national emergency when incumbent officials are facing reelection. We test this notion by exploiting the fact that natural disasters in the Philippines regularly occur regardless of whether or not local government officials are facing reelection. We use calamity  fund  release variations after  incidences  of Philippine natural  disasters to  estimate the  effect  of local  government unit (LGU) officials’ responsiveness to natural disasters on the probability of a successful reelection between  2010  and  2013.

The study utilizes a longitudinal dataset of 1,489 municipalities from the Philippine Commission on Elections. We merged the data with information on budget and expenditures from the Philippine Department of Budget and Management. We also used data on tropical storms from the International Best Track Archive for Climate Stewardship (IBTrACS) and data on other natural disasters from the Philippine National Disaster Risk Reduction and Management Council. Calamity responsiveness is measured as the per capita amount of calamity funds released to particular LGUs upon their submission of fund requests. The research hypothesis is that voting constituents in a municipality exercise their democratic power to either reward or punish an LGU official based on their ability to provide substantial relief assistance to their community. We use a difference-in-differences research design that relies on two main sources of variation to identify the effect of electoral accountability. First, we observe variations in the timing of natural disaster incidences as exogenous to whether or not elected officials are facing an upcoming election. Second, elected officials who are no longer eligible for reelection will not have any reelection incentive to perform well even if a natural disaster strikes during the campaign period.

Our expected results will show that if electoral accountability drives government performance, then elected officials not up for reelection will perform worse during disaster relief operations than elected officials facing reelection. We also expect natural disasters immediately before election periods to matter more than the cumulative disaster response of an official, which would then provide evidence of voter myopia. This study emphasizes the influence of political power on natural disaster funding releases. Policy makers should recognize that funding allocations are subjectively managed, so bureaucrats should take into account how political power can affect which communities recover better.