November 11, 2019, Paper, "In 2010, two years prior to Delhi’s municipal elections, we informed a random set of municipal councilors that a newspaper would report on their performance just before the next election. To evaluate this intervention we collected data on the infrastructure spending preferences of slum dwellers and created an index of pro-poor spending for each councilor. In wards dense with slums, the anticipation of future disclosures caused councilors to increase their pro-poor spending by 0.6 standard deviations over the next two years. A cross-cutting intervention that privately provided councilors with information about the state of infrastructure in the slums had no eﬀect, suggesting that only public disclosures incentivize councilors. Party and voter responses support this interpretation: treated councilors were 12 percentage points more likely to receive a party ticket for re-election. The eﬀect was concentrated among councilors who undertook more pro-poor spending in high-slum wards, and translates into a substantially higher vote share."