Hane Lee
Hi! I am a postdoc at the University of Chicago Data Science Institute, working with Moon Duchin at the Data and Democracy Lab. I recently received my PhD from the Statistics Department at Columbia University, where I was advised by Michael Sobel. My first name is pronounced [Hayn], and my pronouns are they/them.
My research focuses on the intersection of political methodology and American politics, and I am especially motivated by substantive questions in public opinion polarization and racial electoral politics. My research on public opinion develops new measures of opinion divergence using optimal transport and examines how policy disagreement contributes to mass polarization. My work on racial electoral politics studies how electoral systems shape opportunities for representation, including measuring group electoral competitiveness and using simulation methods to study racial representation.
Before Columbia, I was at MIT Media Lab’s Opera of the Future group for an MS where I designed interactive AR musical experiences and helped produce hybrid acoustic+digital musical performances. I also received a BS from MIT in Electrical Engineering with a minor in Music.
News
[Sep 2026] I will be presenting my paper on measuring congressional social ties at APSA 2026.
[Aug 2026] I will be presenting my paper on the Wasserstein Bipolarization Index at JSM 2026.
[Jul 2026] I will be presenting my poster on opinion disagreement and polarization at PolMeth 2026.
[Jun 2026] I will be presenting my paper on the Wasserstein Bipolarization Index at University of Rochester’s Summer Conference in Applied Methods for Political Science.
Research
Job Market Paper. Hane Lee. “Reconsidering the Opinion Pathway to Mass Polarization”.
Abstract
Public opinion researchers have observed ample evidence of mass polarization, including increasing partisan sorting, partisan opinion differentiation, and affective conflict. At the same time, researchers have largely concluded that opinion divergence in the general electorate remains stable, leading theories of mass polarization to de-center policy disagreement and prioritize partisanship and partisan identity as the primary sources of political conflict. We argue in three sequential steps that this theoretical shift rests on an incomplete assessment of the opinion-to-polarization pathway. First, we show that the conclusion of stable opinion divergence results from measures that fail to capture the two distributional features of bipolarization: spread and bi-clustering. Using Wasserstein-based measures of partisan gap and opinion bipolarization that capture these intuitions, we analyze policy opinion items from ANES and GSS between 1992 and 2020, and find increasing opinion divergence in the general electorate, contrary to previous findings. Second, we argue that existing research on partisan opinion differentiation already supports the opinion-to-polarization pathway among partisans and that partisan-centric accounts reinterpret rather than refute this evidence. Finally, we examine non-leaning Independents to test whether the opinion pathway can operate independently of partisan identity. We find that non-leaners exhibit increasing opinion divergence and affective differentiation associated with policy agreement and disagreement. These findings challenge accounts of polarization that treat policy opinions only as a consequence of partisan identity and motivate a substantive-identity dual model in which policy opinions and partisan identity constitute disticnt but interaacting pathways to affective polarization.\\
Hane Lee and Michael Sobel (2026). “Measuring Public Opinion: “The Wasserstein Bipolarization Index”, with Application to Cross-National Attitudes Toward Mandatory Vaccination for COVID-19.”. Annals of Applied Statistics 20 (2) pp. 1719–1735. https://doi.org/10.1214/26-AOAS2162
Abstract
Although the topic of opinion polarization receives much attention from the media, public opinion researchers and political scientists, the phenomenon itself has not been adequately characterized in either the lay or academic literature. To study opinion polarization among the public, researchers compare the distributions of respondents to survey questions or track the distribution of responses to a question over time using ad-hoc methods and measures such as visual comparisons, variances, and bimodality coefficients. To remedy this situation, we build on the axiomatic approach in the economics literature on income bipolarization, specifying key properties a measure of bipolarization should satisfy: in particular, it should increase as the distribution spreads away from a center toward the poles and/or as clustering below or above this center increases. We then show that measures of bipolarization used in public opinion research fail to satisfy one or more of these axioms. Next, we propose a p-Wasserstein polarization index that satisfies the axioms we set forth. Our index measures the dissimilarity between an observed distribution and a distribution with all the mass clustered on the lower and upper endpoints of the scale. We use our index to examine bipolarization in attitudes toward governmental COVID-19 vaccine mandates across 11 countries, finding the U.S and U.K are most polarized, China, France and India the least polarized, while the others (Brazil, Australia, Columbia, Canada, Italy, Spain) occupy an intermediate position.
Working paper. Hane Lee, Andrew Davison, and Zhiliang Ying. “Measuring Social Ties from Roll Call Votes: A Fused Latent Factor and Social Network Approach”.
Abstract
Congressional social ties influence legislative processes and outcomes, but measuring these ties presents a significant challenge. Recent research has adopted social network models to assess congressional social, but existing applications are limited in that they rely on indirect measurements using proxy relations, such as cosponsorship, and fail to account for party influence or ideological preferences of legislators, which may be more decisive factors. In this paper, we aim to directly measure social ties from roll call votes while taking account of partisan-ideological preferences. We combine the partisan-ideological and social approaches to roll call analysis through a fused latent factor and social network model. This model decomposes the variation in votes explained by the partisan-ideological factors from that of the social network, while prioritizing the former. Applying the model to the Senate, we find that the fitted latent factors capture known partisan and ideological patterns from previous literature, while social networks reflect notable friendships and geographical proximity.
Working paper. Yuki Atsusaka, Diana Da In Lee, and Hane Lee. “Quantifying Group-Based Electoral Competitiveness”.
Abstract
How can we measure electoral competitiveness among multiple candidates with varying group affiliations? We propose the concept of group-level electoral competitiveness, which generalizes the conventional margin of victory to any number of candidates and group affiliations while avoiding strong assumptions about perfect coordination among in-group candidates. We demonstrate its applicability in three domains of American politics: racial competitiveness in congressional elections, partisan competitiveness in local elections, and competitiveness across candidates' occupational backgrounds. Our findings reveal patterns of group advantage and disadvantage that are obscured by traditional measures. By extending electoral margins to group-based competition, this paper provides a unified and flexible methodological framework for evaluating long-standing questions in the study of electoral democracy. The proposed measure is implemented via an open-source software R package gmv.
Chris Andrade, Jonathan Auerbach, Icaro Bacelar, Hane Lee, Angela Tan, Mariana Vazquez, and Owen Ward (2023). “Does it pay to park in front of a fire hydrant?”. Significance 20(1), pp. 28–30.