Peer Reviewed Publication

• Cho, Soonhong, Doeun Kim, and Chad Hazlett^*. 2026. Inference at the Data's Edge: Gaussian Processes for Estimation and Inference in the Face of Extrapolation Uncertainty. Political Analysis. [software]
  Abstract

  Many inferential tasks involve fitting models to observed data and predicting outcomes at new covariate values, requiring interpolation or extrapolation. Conventional methods select a single best-fitting model, discarding fits that were similarly plausible in-sample but would yield sharply different predictions out-of-sample. Gaussian Processes (GPs) offer a principled alternative. Rather than committing to one conditional expectation function, GPs deliver a posterior distribution over outcomes at any covariate value. This posterior effectively retains the range of models consistent with the data, widening uncertainty intervals where extrapolation magnifies divergence. In this way, the GP's uncertainty estimates reflect the implications of extrapolation on our predictions, helping to tame the "dangers of extreme counterfactuals" (King and Zeng 2006). The approach requires (i) specifying a covariance function linking outcome similarity to covariate similarity, and (ii) assuming Gaussian noise around the conditional expectation. We provide an accessible introduction to GPs with emphasis on this property, along with a simple, automated procedure for hyperparameter selection implemented in the R package gpss. We illustrate the value of GPs for capturing counterfactual uncertainty in three settings: (i) treatment effect estimation with poor overlap, (ii) interrupted time series requiring extrapolation beyond pre-intervention data, and (iii) regression discontinuity designs where estimates hinge on boundary behavior.


• Kim, Doeun. 2024. Domestic Support for Foreign Aid: Development Firms and Support for Foreign Aid in the U.S. Congress. International Interactions, 50(5), 839-865.
  Abstract

  In this paper, I present a theory of domestic support for foreign aid in the United States, focusing on the role played by American development firms or aid awardees in domestic politics. USAID delivers foreign aid to recipient countries through contracts with private for-profit development firms and non-profit organizations, most of which originate from America. This paper argues that these domestic firms promote domestic support for foreign aid by creating distributional effects and disseminating information through lobbying. This information includes not only the significance of foreign aid but also its distributive benefits to legislators' districts. Analyzing district-level USAID award data and firm-level lobby data from the 111^th to the 115^th Congress, I found that the district-level allocation of contracts is positively associated with the frequency of (co)sponsorship for pro-foreign aid-related bills.

Working Papers

• What Money Can't (or Can) Buy: Inward Foreign Direct Investment and Public Backlash against Globalization in the United States (under review)
  Abstract

  While foreign direct investment (FDI) is often welcomed by host communities, there are instances when even its direct economic beneficiaries oppose foreign investments. In this paper, I examine how foreign companies' investments and their deviations from local norms, such as violations of labor laws or environmental pollution, shape public opinion toward partner countries and foreign investment policies. Drawing on intergroup contact theory, I highlight the significance of foreign firms' interactions with host communities. I argue that exposure to information about foreign companies' business practices influences public views on open foreign economic policies and attitudes toward partner countries. Using both observational and experimental studies with a regional focus on the United States, I analyze how foreign companies' unfair treatment of workers and violations of local regulations trigger public backlash against globalization. The findings from the observational studies suggest that breaking host country's norms and laws suppress public support for open foreign investment policies and diminish the positive effects on public opinion typically associated with FDI. The experimental results further corroborate the causal mechanism.


• Laboring under Global Capital: Work Experience and Political Attitudes toward Foreign Investment
  Abstract

  Classical factor- and industry- specific models of international political economy (IPE) predict that local workers generally support foreign direct investment (FDI) because it creates jobs and raises wages. Yet labor opposition sometimes arises. This study investigates how the behavior of foreign firms, particularly their workplace practices, shapes workers’ attitudes toward investors’ home countries and toward FDI policy itself. Building on intergroup contact theory, I highlight the importance of everyday interactions between foreign investors and local workers. I argue that these interactions influence workers’ views of partner countries and their willingness to admit new foreign investment. I test this argument with an original survey of American workers who have been employed by foreign firms. The evidence shows that when foreign companies treat workers unfairly or fail to meet expected labor standards, they trigger an anti globalization backlash. Such norm violations erode support for liberal FDI rules and negate the usual pro FDI attitudes observed among economically advantaged workers. Mistreatment by foreign firms can turn even the putative economic “winners” of globalization into opponents of further foreign investment.


• Labor Protection-Driven Liberalism: Labor Institutions and FDI Restrictions in OECD Countries
  Abstract

  Foreign direct investment (FDI) delivers jobs and growth but remains politically viable only when domestic labor institutions assure fair treatment, voice, and security for workers. This article develops the concept of labor protection‑driven liberalism, positing that strong employment protections and channels for worker participation are prerequisites for durable support of open investment policies. A multi‑method design evaluates this claim. First, panel analyses of OECD countries show that stricter dismissal regulations and legally backed works councils correlate with less restrictive FDI rules. Second, a paired case study contrasts Spain, where democratization was followed by robust labor reforms, with South Korea, where weak protections persisted. Third, project‑level evidence from the United States probes the underlying mechanisms: manufacturing FDI clusters in states with weaker labor protections (screening effect), and once plants are operating, stronger protections are linked to fewer violations by foreign firms (constraining effect). These findings provide some evidence of strong labor institutions screen out non‑compliant investors and somewhat weaker evidence for discipline those that enter, creating social conditions under which globalization can be sustained. Safeguarding labor rights is therefore not a brake on openness but a foundation for politically resilient integration into the world economy.


• Authoritarian Industrialization and Divergent Electoral Responses: Mobilization versus Legacy Effects -with Junbeom Bahk
  Abstract

  Under what conditions do authoritarian industrial investments generate public loyalty, and when do they provoke opposition? This study examines South Korea's industrial complexes built under two authoritarian leaders, Park Chung-hee (1963–1979) and Chun Doo-hwan (1980–1988). We analyze their influence on local voting in the country's first democratic presidential election in 1987. We argue that political consequences depend on how industrial structure shapes collective action capacity. General manufacturing complexes concentrated workers in dense settings, facilitating collective action and generating electoral opposition to the authoritarian successor. In contrast, agro-industrial complexes were spatially dispersed and less conducive to mobilization. Voters near these areas instead rewarded regime-linked candidates for visible developmental achievements. Using geo-coded township-level electoral returns, we find that proximity to general complexes reduces support for the authoritarian successor, while proximity to agro-industrial complexes increases it. These findings reveal that authoritarian development strategies produce divergent political legacies depending on the type of development strategy they pursue.

Research in Progress

• Electoral Costs of Policy Discontinuity in Green Transitions: Evidence from EV Subsidy Cuts in Germany -with Sung In Kim
• Algorithms and Ideologies: Understanding Political Attitudes in the Tech Sector -with Huimin Li