Search

In the run-up to elections, self-interested politicians use privatization opportunistically to buy votes and increase their probability of re-election. When state-owned firms are privatized, politicians use (implicit) subsidies to persuade managers to choose inefficient but politically beneficial strategies such as excess employment. Under plausible assumptions, politicians have a strict preference for privatization over state ownership in the run-up to elections. We test these predictions using a unique dataset covering the full population of former socially owned enterprises in post-Milošević Serbia (2002–2019). We report robust conditional correlations consistent with our theory. Privatization sales and revenues increase significantly in pre-election periods. The firms privatized before elections are sold at a lower price, and exhibit higher total costs after privatization, than otherwise similar privatized firms. They also have a higher probability of bankruptcy and, conditional on surviving, achieve lower profitability than otherwise similar privatized firms. These findings highlight the link between privatization, elections and corruption, and point to the need for monitoring, or even suspending, privatization sales during election periods.

Using data from 64 countries in Eastern Europe and MENA, I study the long-run effects of Ottoman and socialist rule on the incidence of corruption. To proxy Ottoman legacies, I estimate the length of Ottoman rule across all Ottoman successor states. Conditioning on income per capita, I find a robust adverse effect of both socialist and Ottoman legacies on present day corruption a finding which reconciles two rival accounts of post-socialist corruption from the transition literature. The results are robust to controlling for potential confounders and instrumenting for both per capita income and the length of Ottoman rule. However, the explanatory power of long-run historical determinants of corruption is lower than the contribution of short-run factors. While present day income explains about half of the total variation in corruption across post-socialist countries, Ottoman and socialist legacies jointly account for about one third. Although history does matter, these findings suggest that most of the corruption observed in Eastern Europe, Central Asia and the Balkans today may not be inherently 'eastern' or socialist. Rather, the data are consistent with an interpretation of corruption predominantly as a manifestation of persistent economic under-development.