We welcome all interested to the UiS Business School Seminar Thursday February 6, Elise Ottesen-Jensens hus, EOJ 276/277 at 12:00-13:30.
Seminar guest: Wayne Sandholtz, PhD candidate at the University of California, San Diego.
Title: Do voters reward service delivery? Experimental evidence from Liberia
Do voters provide incentives for politicians to focus on public service delivery rather than patronage? This paper examines the electoral effects of a randomized Liberian school reform, previously shown to have caused increased teacher attendance and student test scores. In the subsequent election, this seemingly successful policy caused a 3pp (10%) reduction in average vote share for the responsible party’s presidential candidate. This may reflect disruptions to the patronage network: treated teachers were less likely to support the incumbent party, staff polling booths, or campaign for candidates. However, heterogeneity in the policy’s effectiveness reveals that voters did perceive and reward school quality improvements. In places where the treatment caused large improvements (reductions) in test scores, it also caused large electoral gains (losses). Electoral effects appear only for presidential, not legislative, candidates – sug- gesting that sophisticated voters attributed credit or blame at the correct level of government. Survey experiments among both candidates and voters confirm the picture of an electorate well informed about the policy’s effectiveness and provenance. This paper highlights the political risks of moving from an electoral strategy based on patronage to one based on public service delivery – but it also suggests that increased policy effectiveness has the potential to counteract the opposition of entrenched interests.