Working Papers

[Please email me for drafts]

''Illusive Transparency? Evidence on Election Video Monitoring'' [World Bank Development Impact Blog Post]

Authoritarian regimes, like democracies, hold elections and often equip them with transparency-enhancing technologies. Why would autocrats want to hold or appear to hold transparent elections? This paper examines the impacts of broadcast election video monitoring in Russia. I exploit a discontinuity in the assignment of webcams to polling stations during the 2018 presidential election to estimate causal effects on voting. Video monitoring reduces reported voter turnout by 5.2% and votes for the incumbent (autocrat) by 8.3%, suggesting a decrease in fraud. However, that decrease is partially offset by increased votes for the incumbent in neighboring unmonitored polling stations, indicating a displacement of fraud. To explore why autocrats implement video monitoring, I conducted a nationwide survey experiment before the 2019 local elections. Treated respondents were informed of video monitoring, which significantly increased the trust in elections and willingness to vote among those not previously aware of transparency technologies. Overall, these results suggest that video monitoring allows autocrats to improve citizens’ attitudes by creating an illusion of transparency at a low cost in terms of net lost votes.

''Modernizing the State During War: Experimental Evidence from Afghanistan'' (with Joshua Blumenstock, Michael Callen, Stefano Fiorin, and Tarek Ghani)

This paper provides evidence on how to effectively build basic state capacity during an active internal conflict. We report results from a randomized evaluation of a major reform intended to improve the Afghan government's ability to pay its employees, which involved over 30,000 employees of the Ministry of Education between 2018 and 2020. The first element of the program, designed to eliminate `ghost' workers, required employees to register for a mobile money wallet with biometric identification. This helped eliminate `ghosts' from the payroll (1.3% of employees) and allowed us to estimate the true share of ghosts (8.4%-20.4%). The second element transitioned employees from receiving their salary in cash to direct mobile money transfers. This led to a 26 percentage point increase in support for the reform to be scaled nationally, reduced salary delays, and caused employees to dramatically increase activity on the mobile money network, demonstrating a potential pathway toward expanding Afghanistan's formal financial system. Because the experiment spanned both secure cities and contested rural regions, we can examine whether state control complemented the reform. While the reform brought benefits everywhere, improvements materialized faster in cities. Overall, our results highlight the importance of long-term horizons in state-building efforts and provide evidence that progress toward a modern bureaucracy is possible, even in the shadow of war and while the broader state is under threat.

''Community Governance and Local Economic Activity: Evidence from Ukraine''

Decentralization of local governments is one of the most popular governance reforms, actively promoted by major international donors. This reform aims to align government incentives through better information and accountability at the local level. This paper evaluates the effects of fiscal and administrative decentralization of local governments on economic activity in Ukraine. I exploit quasi-experimental variation in the timing of decentralization across communities to estimate causal effects on the universe of firms and self-employed individuals using difference-in-differences methods. I find that decentralization caused a steady reduction in the number of active self-employed individuals by up to 6%. These effects were concentrated in rural areas, for which decentralization was designed to catalyze economic development. There were, however, no changes in the total number of private firms, with only nominal growth in newly established enterprises. These results suggest that successfully applied governance reforms might have unintended economic consequences.