Devin’s Research Page


Welcome to my research page. As you can see, my interests vary widely, but one common theme that runs through much of my research is the use of quantitative methodological tools to study U.S. politics in historical and comparative perspective.


Congress, Public Opinion, and Representation in the One-Party South, 1930s–1960s

My dissertation, supervised by Professor Eric Schickler, examines the dynamics of mass preferences and elite representation in the South, focusing on final years before the region’s transition from a restricted one-party regime to a competitive two-party democracy. In particular, I am interested the relationship between ideological patterns among Southern members of Congress—including both cross-sectional differences and trends over time—and the complex and changing policy preferences in the Southern electorate and mass public. I anticipate that my dissertation will not only shed light on this critical period in American political development, but also inform scholarly debates over elite responsiveness to popular preferences when democratic institutions are absent or compromised. A major methodological component of this project is developing optimal statistical techniques for aggregating, weighting, and analyzing early quota-sampled public opinion polls in order to estimate mass opinion for subsets of the population.


“Elections and the Regression-Discontinuity Design: Lessons from Close U.S. House Races, 1942–2008”

(with Jasjeet Sekhon)—Political Analysis 19 (2011)

Following David Lee’s pioneering work, numerous scholars have applied the regression-discontinuity (RD) design to popular elections. Contrary to the assumptions of RD, however, we show that bare winners and bare losers in U.S. House elections (1944–2008) differ markedly on pretreatment covariates. Bare winners possess large ex ante financial, experience, and incumbency advantages over their opponents and are usually the candidates predicted to win by Congressional Quarterly’s pre-election ratings. Covariate imbalance actually worsens in the closest House elections. National partisan tides help explain these patterns. Previous works have missed this imbalance because they rely excessively on model-based extrapolation. We present evidence suggesting that sorting in close House elections is due mainly to activities on or before Election Day rather than post-election recounts or other manipulation. The sorting is so strong that it is impossible to achieve covariate balance between matched treated and control observations, making covariate adjustment a dubious enterprise. Although RD is problematic for post-war House elections, this example does highlight the design’s advantages over alternatives: RD’s assumptions are clear and their implications testable.

The published version of this article can be downloaded here.

Replication files the article are available at the Political Analysis Dataverse.

“Public Opinion, Organized Labor, and the Limits of New Deal Liberalism, 1936–1945”

(with Eric Schickler)—Studies in American Political Development (October, 2011)

Using hundreds of rarely used polls from the 1930s and 1940s, we chart the dynamics of public opinion towards New Deal liberalism in this pivotal era. We find evidence for a broad-based popular reaction against organized labor in this period, creating a difficult terrain for liberal programmatic advances. This anti-labor reaction was particularly virulent in the South but is also prominent among Northern Democrats, providing Republicans with a popular wedge issue in the North as well as a basis for a conservative coalition with Southern Democrats in Congress. More generally, we find that the mass public favored most of the specific programs created by the New Deal but was hardly clamoring for major expansions of the national government’s role. These findings illuminate the role played by the South in constraining New Deal liberalism while also highlighting the tenuousness of the liberal majority in the North.

The published version of this article can be downloaded here.

“Honor and War: Using Southern Presidents to Identify Reputational Effects in International Conflict”

(with Allan Dafoe)—under review.

We exploit a rare source of variation—the fact that a third of U.S. presidents were raised in the American South, a prominent example of a “culture of honor”—to identify the effect of heightened concern for coercive reputation. We analyze a family of formal models of international conflict and deduce multiple predictions for the reputational effects of leaders’ concern for honor, taking into account unobserved selection into disputes. Using matching, permutation inference, and the nonparametric combination of tests—a technique new to political science—we find strong support for our theory. Disputes under Southern presidents are approximately twice as likely to involve U.S. uses of force, last on average twice as long, and are three times more likely to end in victory for the United States.

This paper received the 2011 Kenneth E. Boulding Award for best graduate student paper presented at the ISA annual meeting.

A draft of our paper can be downloaded here.

“Defining, Mapping, and Measuring Bureaucratic Autonomy”

(with Sara Chatfield and Adam Cohon)

In response to the lack of conceptual and semantic coherence in the literature on bureaucratic autonomy, we derive a re-conceptualization that unifies the disparate scholarly uses of the term under a single framework. We emphasize that bureaucratic autonomy has two equally essential components: independent goal formation and capacity to achieve desired outcomes. We suggest measurement and modeling strategies consistent with our conceptualization.

Click here for the version of the paper we presented at the 2009 Meeting of the Midwest Political Science Association.