Devin Caughey

Devin Caughey

Auto-generated profile, based on publicly available information

Devin Caughey

Professor, Political Science, Massachusetts Institute of Technology, Cambridge, USA

Research Focus & Metrics

Academic Career Length

16Years (2011-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
3 August 2018 in Political Analysis
Verified
Journal Article
Closed access
PDF file
ADAllan Dafoe
Allan Dafoe
BZBaobao Zhang
Baobao Zhang
DCDevin Caughey
Devin Caughey
Survey experiments often manipulate the description of attributes in a hypothetical scenario, with the goal of learning about those attributes’ real-world effects. Such inferences rely on an underappreciated assumption: experimental conditions must be information equivalent (IE) with respect to background features of the scenario. IE is often violated because subjects, when presented with information about one attribute, update their beliefs about others too. Labeling a country “a democracy,” for example, affects subjects’ beliefs about the country’s geographic location. When IE is violated, the effect of the manipulation need not correspond to the quantity of interest (the effect of beliefs about the focal attribute). We formally define the IE assumption, relating it to the exclusion restriction in instrumental-variable analysis. We show how to predict IE violations ex ante and diagnose them ex post with placebo tests. We evaluate three strategies for achieving IE. Abstract encouragement is ineffective. Specifying background details reduces imbalance on the specified details and highly correlated details, but not others. Embedding a natural experiment in the scenario can reduce imbalance on all background beliefs, but raises other issues. We illustrate with four survey experiments, focusing on an extension of a prominent study of the democratic peace.

1 January 2011 in Political Analysis
Verified
Journal Article
Closed access
PDF file
DCDevin Caughey
Devin Caughey
JSJasjeet S. Sekhon
Jasjeet S. Sekhon
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 (1942–2008) differ markedly on pretreatment covariates. Bare winners possess largeex antefinancial, experience, and incumbency advantages over their opponents and are usually the candidates predicted to win byCongressional 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 postelection 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 postwar House elections, this example does highlight the design's advantages over alternatives: RD's assumptions are clear and weaker than model-based alternatives, and their implications are empirically testable.

29 November 2017 in American Political Science Review
Verified
Journal Article
Open access
PDF file
DCDevin Caughey
Devin Caughey
CWChristopher Warshaw
Christopher Warshaw
Using eight decades of data, we examine the magnitude, mechanisms, and moderators of dynamic responsiveness in the American states. We show that on both economic and (especially) social issues, the liberalism of state publics predicts future change in state policy liberalism. Dynamic responsiveness is gradual, however; large policy shifts are the result of the cumulation of incremental responsiveness over many years. Partisan control of government appears to mediate only a fraction of responsiveness, suggesting that, contrary to conventional wisdom, responsiveness occurs in large part through the adaptation of incumbent officials. Dynamic responsiveness has increased over time but does not seem to be influenced by institutions such as direct democracy or campaign finance regulations. We conclude that our findings, though in some respects normatively ambiguous, on the whole paint a reassuring portrait of statehouse democracy.

18 September 2015 in American Journal of Political Science
Verified
Journal Article
Open access
PDF file
DCDevin Caughey
Devin Caughey
CWChristopher Warshaw
Christopher Warshaw
Applying a dynamic latent‐variable model to data on 148 policies collected over eight decades (1936–2014), we produce the first yearly measure of the policy liberalism of U.S. states. Our dynamic measure of state policy liberalism marks an important advance over existing measures, almost all of which are purely cross‐sectional and thus cannot be used to study policy change. We find that, in the aggregate, the policy liberalism of U.S. states steadily increased between the 1930s and 1970s and then largely plateaued. The policy liberalism of most states has remained stable in relative terms, though several states have shifted considerably over time. We also find surprisingly little evidence of multidimensionality in state policy outputs. Our new estimates of state policy liberalism have broad application to the study of political development, representation, accountability, and other important issues in political science.

1 October 2017 in The Journal of Politics
Verified
Journal Article
Closed access
DCDevin Caughey
Devin Caughey
CWChristopher Warshaw
Christopher Warshaw
YXYiqing Xu
Yiqing Xu
How much does it matter whether Democrats or Republicans control the government? Unless the two parties converge completely, election outcomes should have some impact on policy, but the existing evidence for policy effects of party control is surprisingly weak and inconsistent. We bring clarity to this question, using regression-discontinuity and dynamic panel analyses to estimate the effects of party control of state legislatures and governorships on a new annual measure of state policy liberalism. We find that throughout the 1936–2014 period, electing Democrats has led to more liberal policies, but that in recent decades the policy effects of party control have approximately doubled in magnitude. We present evidence that this increase is at least partially explained by the ideological divergence of the parties’ office holders and electoral coalitions. At the same time, we also show that party effects remain substantively modest, paling relative to policy differences across states.

1 February 2019 in American Political Science Review
Verified
Journal Article
Closed access
PDF file
DCDevin Caughey
Devin Caughey
TOTom O’grady
Tom O’grady
CWChristopher Warshaw
Christopher Warshaw
Using new scaling methods and a comprehensive public opinion dataset, we develop the first survey-based time-series–cross-sectional measures of policy ideology in European mass publics. Our dataset covers 27 countries and 36 years and contains nearly 2.7 million survey responses to 109 unique issue questions. Estimating an ordinal group-level IRT model in each of four issue domains, we obtain biennial estimates of the absolute economic conservatism, relative economic conservatism, social conservatism, and immigration conservatism of men and women in three age categories in each country. Aggregating the group-level estimates yields estimates of the average conservatism in national publics in each biennium between 1981–82 and 2015–16. The four measures exhibit contrasting cross-sectional cleavages and distinct temporal dynamics, illustrating the multidimensionality of mass ideology in Europe. Subjecting our measures to a series of validation tests, we show that the constructs they measure are distinct and substantively important and that they perform as well as or better than one-dimensional proxies for mass conservatism (left–right self-placement and median voter scores). We foresee many uses for these scores by scholars of public opinion, electoral behavior, representation, and policy feedback.