A Machine Learning Approach to Analyze and Support Anti-corruption Policy

Abstract : Can machine learning support better governance? In the context of Brazilian municipalities, 2001-2012, we have access to detailed accounts of local budgets and audit data on the associated fiscal corruption. Using the budget variables as predictors, we train a tree-based gradient-boosted classifier to predict the presence of corruption in held-out test data. The trained model, when applied to new data, provides a prediction-based measure of corruption that can be used for new empirical analysis or to support policy responses. We validate the empirical usefulness of this measure by replicating and extending some previous empirical evidence on corruption issues in Brazil. We then explore how the predictions can be used to support policies toward corruption. Our policy simulations show that, relative to the status quo policy of random audits, a targeted policy guided by the machine predictions could detect almost twice as many corrupt municipalities for the same audit rate. Similar gains can be achieved for a politically neutral targeting policy that equalizes audit rates across political parties.

---

Why Governments Grow “lemons” in the Market for Technology

Abstract : Governments around the world spend an enormous amount of money on R&D. A large part of this investment is wasted, flooding the system with patents that never result in any actual innovation. This effect is especially pronounced in countries with the low level of government accountability. To reconcile the growth of “lemon patents” with the genuine desire of the government to spur innovation, I offer a game-theoretic model, in which the government has a significant stake in technological development and invests in R&D, even if this simultaneously encourages the growth of “lemons”. I illustrate the mechanism by demonstrating the causal impact of Russia’s government policy, which resulted in a simultaneous increase in the number of high-quality patents and a decrease in the share of such patents in the patent pool.

---

Subsidies for Sale: Post-government Career Concerns, Revolving-door Channels, and Public Resource Misallocation in China

Abstract : While the existing literature focuses on how revolving-door officials deliver favorable government treatment to firms after leaving public office, this paper theorizes that the post-government career concerns of public officials distort public resource allocation while still in office. To test this theory, I construct a new dataset that links over 98,000 corporate subsidy programs approved by multiple levels of governments with revolving-door officials who joined publicly listed Chinese firms between 2007 and 2019. I show that forward-looking officials provide sizable favorable subsidies to their future employers. To verify the exchange of favors, I document that firms repay public officials who have provided favorable subsidies by hiring and paying them enormous amounts of cash compensation. Finally, I find that the reputation cost is the mechanism through which this quid pro quo relationship is sustained.

---