Sara Chatfield, Ph.D.
sara.chatfield@du.edu
University of Denver
About Me:
Sara Chatfield is an assistant professor of political science at the University of Denver. Prior to coming to DU, she was a postdoc in the political science department at MIT and taught courses in American politics and law at Tufts University. Her research focuses on the development of married women’s economic rights in U.S. state courts, legislatures, and constitutional conventions in the 1800s and early 1900s. She also conducts research on political behavior (including various aspects of political participation and vote choice) and American political development (including congressional politics and analysis of historical polling data). A current ongoing project examines the gender pay gap for campaign staffers in contemporary congressional election campaigns.
Research Interests
Gender and Politics
Judicial Politics
Political Participation
Elections, Election Administration, and Voting Behavior
State and Local Politics
Countries of Interest
United States
Publications:
Journal Articles:
Using publicly available data on campaign disbursements, we examine gender gaps in staffing and compensation for congressional campaigns between 2010 and 2016. We find significant differences in both staff representation and pay on the basis of gender. Women make up a lower proportion of campaigns’ staffs, on average, and receive less compensation overall than their male counterparts. These disparities are present across a wide variety of campaigns, but are exacerbated by partisanship, candidate gender, and incumbency status. We find that staffing and wage gaps are larger for Republican candidates, male candidates, and challengers. These findings provide new insight into distinctions in the political networks between Democrats and Republicans and the pipeline of future political candidates.
Do voters prefer dominant looking candidates in times of war? By replicating previous survey experiments, we find that respondents do prefer candidates with dominant facial features when war is salient. We then investigate whether these survey results generalize to the real world. Examining US Senate elections from 1990 to 2006, we test whether voters prefer candidates with dominant facial features in wartime elections more than in peacetime elections. In contrast with the survey studies, we find that dominant-looking candidates appear to gain a slight advantage in all elections but have no special advantage in wartime contexts. We discuss possible explanations for the discrepancy between the findings and conduct additional experiments to investigate one possible explanation: additional information about candidates may rapidly erode the wartime preference for dominant looking candidates. Overall, our findings suggest that the dominance-war findings may not generalize to the real world.
Causal inference and American political development (APD) are widely separated and (to some) fundamentally incompatible tendencies within political science. In this paper, we explore points of connection between those two perspectives, while also highlighting differences that are not so easily bridged. We stress that both causal inference and APD are centrally interested in questions of causation, but they approach causation with very different ontological and epistemological commitments. We emphasize how the sort of detailed, contextualized, and often qualitative knowledge privileged by APD can promote credible causal (and descriptive) inferences, but also that scholars of causal inference can benefit from alternate conceptions of causality embraced by APD work. We illustrate with two empirical examples from our own research: devising weights for quota-sampled opinion polls and estimating the political effects of the Tennessee Valley Authority. We conclude that bringing APD and causal inference together on more equal terms may require a broader perspective on causation than is typical of scholarship in the causal-inference tradition.
Ronald Kahn has argued that the social construction process, which embeds social realities outside the court in legal doctrine, can lead to more rights-protective constitutional interpretations, even during conservative eras. However, especially when particular groups are the subject of multiple or competing social constructions, the social construction process may not always lead to rights-expansive outcomes for disadvantaged groups. From the late 1800s through 1937, state and federal courts struggled to fit women workers into changing legal conceptions of the "right to contract." Across numerous cases, courts vacillated between two competing social constructions of women workers: women as vulnerable victims in need of special state protections, and women as independent economic actors who were qualified to make their own workplace decisions. Ultimately, social constructions that provided important workplace protections for women workers before 1937 became embedded in legal doctrine in ways that limited their economic and civil equality for decades afterward.
Beginning in 1839 and continuing through the early twentieth century, the American states passed laws expanding married women's economic rights, including the right to own property and sign contracts. In almost every state, these significant legal changes took place before women had the right to vote. I argue that married women's economic rights reform is best understood as a piecemeal, iterative process in which multiple state-level institutions interacted over time. This rights expansion often occurred as a by-product of male political actors pursuing issues largely unrelated to gender—such as debt relief and commercial development—combined with paternalistic views of women as needing protection from the state. State courts played a crucial role by making evident the contradictions inherent in vague and inconsistent legal reforms. Ultimately, male political actors liberalized married women's economic rights to the extent that they thought it was necessary to allow for the development of efficient and workable property rights in a commercial economy, leaving women's place in the economy partially but not fully liberalized.
Over the past several years, there has been growing use of the draft lottery instrument to study political attitudes and behaviors (see, e.g., Bergan 2009; Erikson and Stoker 2011; Henderson 2012; Davenport 2015). Draft lotteries, held in the United States from 1969 to 1972, provide a potentially powerful design; in theory, they should provide true randomization for the “treatment” of military service or behavioral reactions to the threat of such service. However, the first draft lottery conducted in 1969 was not conducted in a random manner, giving those citizens born in the fourth quarter of the year disproportionately higher chances of being drafted. In this note, we describe the randomization failure and discuss how this failure could in theory compromise the use of draft lottery numbers as an instrumental variable. We then use American National Election Studies data to provide support for the conclusion that individuals most affected by the randomization failure (those born in the fourth quarter of the year) largely do not look statistically distinct from those born at other times of the year. With some caveats, researchers should be able to treat the 1969 draft numbers as if they were assigned at random. We also discuss broader lessons to draw from this example, both for scholars interested in using the draft lottery as an instrumental variable, and for researchers leveraging other instruments with randomization failures. Specifically, we suggest that scholars should pay particular attention to the sources of randomization failure, sample attrition, treatment and dependent variable selection, and possible failure of the exclusion restriction, and we outline ways in which these problems may apply to the draft lottery instrument and other natural experiments.
American federalism is often described as a system that contains “political safety valves,” or institutional mechanisms that ensure that major policy reforms can be created, even during periods of intense political conflict. By granting discretion to the states, for example, scholars claim that Congress can ensure that diverse constituencies receive their preferred policies. In this article, we examine Congress’s pattern of delegating discretion to sub-national institutions in the postwar period, systematically assessing how the political conditions under which a broad sample of landmark legislation passed are related to the delegation of administrative authority to the states. Contrary to the “safety valve” image of federalism, we find very little evidence to suggest that Congress grants more discretion to sub-national governments under periods of intense political conflict.
In a recent study, Kam and Palmer (2008) employ propensity score matching to assess whether college attendance causes participation after reducing selection bias due to pre-adult factors. After matching the authors find no correlation, upending a major pillar in political science. However, we argue that this study has serious flaws and should not be the basis for rejecting the traditional view of an ‘‘education effect’’ on participation. We match on 766,642 propensity scores and use genetic matching to recover better matches with lower covariate imbalances. We consistently find positive effects as covariate balance improves, though no matching approach yields unbiased results. We demonstrate that selection is a serious concern in studying the participatory effects of college attendance and that balance in the covariates and robustness to sensitivity diagnostics should be the ultimate guide for conducting matching analyses.
Books Written:
We elaborate a general workflow of weighting-based survey inference, decomposing it into two main tasks. The first is the estimation of population targets from one or more sources of auxiliary information. The second is the construction of weights that calibrate the survey sample to the population targets. We emphasize that these tasks are predicated on models of the measurement, sampling, and nonresponse process whose assumptions cannot be fully tested. After describing this workflow in abstract terms, we then describe in detail how it can be applied to the analysis of historical and contemporary opinion polls. We also discuss extensions of the basic workflow, particularly inference for causal quantities and multilevel regression and poststratification.