TSE Colloquium Series
TSE Colloquium Series
Date: October 20, 2026 12:00 PM (Taiwan)
Presenter: David Sungho Park, Associate Professor, KDI School
Do misperceptions deter small retailers from stocking profitable products? We apply machine learning to the inventories of 2000 shops in Lusaka, Zambia, to identify clusters of similar retailers and flag a “target" product-one stocked by some but not all shops in a cluster. Retailers that do not stock the product systematically underestimate its profitability. We test whether misperceptions drive this gap using a randomized trial that temporarily subsidized stocking the product. Treated retailers earn similar markups to peers already selling the target product. They attract more customers and earn higher profits. They also continue stocking the product after the subsidy ends, especially when given a longer period to experiment. Because local demand, unlike prices, cannot be reliably inferred without direct experience, temporary experimentation generates persistent learning.
TSE Workshop Series
Date: November 10, 2026 12:00 PM (Taiwan)
Presenter: Sun Young Park, Postdoctoral Fellow, TSE
Social scientists usually place attitudes toward out-groups on a single scale and attempt to sort people according to the intensity of their prejudice. I argue that an intergroup attitude has three dimensions: (1) "stereotyping" (what one believes about an out-group's characteristics), (2) "sympathy" (how much one cares about the out-group's welfare), and (3) "empathy" (how much one understands and shares the out-group's feelings). This framework can reveal the blind spots of a unidimensional conceptualization. For example, a man could hold gender stereotypes while caring a lot about women's welfare, and educated elites could care about working-class welfare while showing little empathy for their feelings. I use this new conceptualization to explain why interventions such as intergroup contact often move one measure of prejudice but not another: treatments may shift people within a three-dimensional space, rather than from prejudiced to unprejudiced.
Date: November 3, 2026 12:00 PM (Taiwan)
Presenter: Doeun Kim, Assistant Professor, TSE
Under what conditions do authoritarian industrial investments generate public loyalty, and when do they instead provoke opposition? We examine industrial complexes established under South Korea’s authoritarian regimes and analyze their effects on local voting in the country’s first democratic presidential election in 1987. We argue that the political consequences of authoritarian industrialization depend on whether it creates conditions that generate collective grievances and facilitate collective action. Using geocoded township-level electoral returns, we find that voters near general industrial complexes, which concentrated workers and fostered collective grievances and organization, were less likely to support the authoritarian successor. By contrast, voters near agro-industrial complexes, which were smaller and more spatially dispersed and generated fewer opportunities for collective grievance formation and mobilization, were more likely to support regime-linked candidates. These findings help reconcile competing theories of authoritarian development by showing how variation in the social and organizational conditions created by industrialization shapes its long-term political consequences.
Date: October 27, 2026 12:00 PM (Taiwan)
Presenter: Wei-Lin Chen, Assistant Professor, TSE
This paper studies how electoral incentives shape frontline implementation of taxes and transfers. Using a three-year panel survey of Indonesian villages, I document electoral cycles in informal tax collection and subsidized rice distribution. In village head election years, households with family ties to village officials (politically connected households) receive more subsidized rice regardless of expenditure level, while high-expenditure politically connected households also pay less in informal taxes. Politically unconnected households exhibit no electoral cycle in either outcome. As a result, the informal tax system becomes more regressive, and overall leakage in the subsidy program falls right before village head elections. The results suggest that the electoral incentives of local agents play an important role in shaping local public finance.
Date: October 13, 2026 12:00 PM (Taiwan)
Presenter: Jonghoon Lee, Assistant Professor, TSE
Democratic stability depends on losers' consent: the willingness of defeated candidates to accept outcomes and to continue competing within electoral rules. Yet recent elections have increasingly featured defeated candidates refusing to concede and alleging electoral fraud. Why do some losing candidates make allegations of electoral fraud and refuse to concede defeat? In this project, we argue that post-electoral fraud allegations are shaped by intraparty competition. Specifically, candidates with precarious standing within their party-due to outsider status, leadership contestation, or expected intraparty challenges-face stronger incentives to delegitimize electoral outcomes and externalize blame for defeat. To test these claims, we construct the first original cross-national, candidate-level dataset of electoral fraud allegations in consolidated democracies since 1990.
Date: September 29, 2026 12:00 PM (Taiwan)
Presenter: Seung-hun Lee, Assistant Professor, TSE
This project examines how bureaucrat training influences tax compliance in Tanzania by focusing on the human capital of frontline tax collectors – a crucial but understudied dimension in the tax compliance literature. While existing studies emphasize institutional design, monitoring, and incentives, much less is known about how the skills and capabilities of individual bureaucrats affect state capacity. We combine administrative data from the Tanzanian Revenue Authority (TRA) with a randomized controlled trial (RCT) of tax collector training implemented by the TRA Training Institute, a well-established institution for building administrative capacity. The RCT experimentally varies training modules to identify which components – technical knowledge, behavioral integrity, or communication skills – most effectively improve post-deployment performance. Administrative data allow us to link trainees’ in-class performance with their subsequent tax collection outcomes, providing a unique opportunity to causally assess the role of bureaucratic human capital in revenue mobilization. By highlighting how investments in bureaucratic skills translate into higher tax compliance and better public revenue performance, this study introduces a novel human-capital perspective to fiscal capacity research and generates actionable evidence for governments seeking to strengthen tax administrations in low-income settings.