Using data from over 2,000 professionals in 24 large corporations, we show that female leaders shape the relational culture in the workplace differently than male leaders. Males form homophilic professional ties under male leadership, but female leadership changes this pattern, creating a less segregated workplace. Female leaders are more likely to establish professional support links with their female subordinates. Under female leadership, female employees are less likely to quit their jobs but no more likely to get promoted. Results suggest that increasing female presence in leadership positions may be an effective way to mitigate toxic relational culture in the workplace.
Who Buys Vote-buying? How, how much, and at what cost?, Journal of Economic Behavior & Organization 193 (2022): 98-124. Paper
In this paper, I estimate the causal effect of a local food-subsidy program on electoral outcomes. I exploit the variation in voters’ walking distances from the program stores to identify their accessibility to the program. I find that a distributive spending of ~5% of GDP per capita buys an additional vote for the incumbent. I then investigate who –based on partisanship– responds to the subsidy, and how much and how they respond. The findings indicate that all types of voters respond to the distributive spending in line with the reciprocity rule; however, they respond through different channels and in different magnitude. Importantly, the salient channel for opposition voters is abstention-buying, whereas incumbent supporters respond by an increased turnout.
Social Norms, Political Polarization, and Vaccination Attitudes: Evidence from a Survey Experiment in Turkey (Forthcoming at the European Economic Review) (with Murat Koyuncu, Sebastian Schneider and Matthias Sutter) Paper
This paper examines vaccination as a descriptive social norm in the context of the Covid-19 pandemic. Using a large-scale survey experiment in Turkey, we first elicit respondents' vaccination attitudes and show that political affiliation is a strong predictor of it. We then use economic games to measure the extent of outgroup discrimination induced by respondents' attitudes towards vaccination. We find that while both pro- and anti-vaxxers discriminate against each other substantially, the pro-vaxxers discriminate more than the anti-vaxxers do. This polarization intensifies when pro- and anti-vaxxers perceive a political difference between them. Using randomized informational treatments, we show that a reminder or priming of external threats, appealing to a broadly shared social identity, might mitigate such outgroup discrimination.
This paper investigates the impact of authoritarian takeovers on the rule of law and economic efficiency in local jurisdictions. Authoritarian takeovers refer to the replacement of elected officials with centrally appointed representatives. Using the universe of state contracts in Turkey and a staggered Difference-in-Differences (DiD) design, we show how authoritarian takeovers deteriorate the rule of law and reduce economic efficiency in public procurement. Notably, centrally appointed mayors use competitive auctions at a rate nearly half that of elected mayors and instead exploit legal provisions almost three times as often. Such practices inflate contract prices by 24% and reduce value for money by 40%, causing waste equivalent to 6% of procurement spending in the affected municipalities. These results are robust to various tests, including Regression Discontinuity (RD) estimation. Probing the underlying mechanisms, we find evidence for diminished local accountability driving these effects. By contrast, we do not find evidence for either the coordination benefits from a more centralized governance or politicians' discretion leading to quality improvements in procurement. In an era of global autocratization, our sub-national evidence on authoritarian takeovers underscores the instrumental value of democracy.
Class Voting and Economic Policy Preferences: A Machine Learning Approach (Draft in preparation for submission) Working Paper
Policy preferences are assumed to have become less anchored in social class due to rising living standards, the broadening reach of education, and increased social mobility. However, there has yet to be a systematic approach to gauging the extent of class-based distinctions in economic policy preferences and their evolution across time and space. In this study, using predictive modeling, I introduce a novel metric for assessing class distinctiveness in economic policy preferences and estimate it for 18 European countries at three different points in time. I then validate this innovative measure and delve into its implications for class-based voting.
Work in Progress
Local Responses to a Global Crisis: Public Service Delivery in Turkish Municipalities Amidst Migrant Influx (with Serkant Adiguzel and Murat Koyuncu)
Is Seeing Believing? How Public Service Visibility Influences Local Government Spending and Procurement Policies (with Serkant Adiguzel, Asli Cansunar and Murat Koyuncu)
Social Identity and Policy Preferences: Evidence from a Large-scale Survey Experiment in the U.S. (with Matthias Sutter)
