Description:
Public spending often increases at the end of fiscal years. This is undesirable because late spending tends to be inefficient. The causes for these spending spikes are however poorly understood. This paper offers a novel identification strategy that relies on the historic variation in countries' fiscal years to analyze their effect on government disbursements. We show that the end of fiscal years rather than alternative explanations cause spending spikes at the end of fiscal years. Our accounting data includes discretionary contributions of 27 OECD countries to the World Bank from 2002 to 2013 at the daily level. As suggested by the principal-agent theory, we find that the end of year effect is smaller in countries with high administrative quality. We analyze the pertinent budget institutions as possible mechanism. For the first time, we can show that unexpected positive demand shocks decrease year-end spending, a common assumption in the literature. Finally, we revisit the complementary explanations for year-end effects in public spending.