Quality of local public good provision and rlectoral support

Do voters value the quality of local public goods, such as whether pavements are clean, whether benches in the park need repair, or whether local roads are in good shape? Using almost 150,000 geo-located complaints that were posted on the complaint-platform FixMyStreet.com between 2007 and 2011, I compute quality indicators for all 8,500 local wards that comprise the whole of England. The results provide compelling evidence for both a strong rewarding and a punishment effect: In wards with positive quality indicators, such as a large fraction of complaints that are solved within 30 days, the incumbent party has statistically and economically significantly higher chances of being re-elected. On the contrary, if the local area features a large share of complaints that is solved only after twelve months or never, the probability that the incumbent party is voted out of office is increased by up to seven percentage points. The results also document a considerable short-run memory of voters as suggested by the fact that good and bad performance indicators more than two to three months away from election day have no statistically significant impact anymore on the chances of re-election. The results are robust to the inclusion of various important co-variates at the ward-level taken from the official Census in 2011 and provide new insights into the link between local governmental spending and voting behavior.