Philip Converse and Warren Miller Fellowship in American Political Behavior

The Converse Miller Fellowship in American Political Behavior provides support for a graduate student in the University of Michigan Political Science Department to work with a faculty member in the Center for Political Studies. The Fellowship honors the individual and joint contributions of Phil Converse and Warren Miller to the fields of American Politics, public opinion, and voting behavior as well as the study of the concept of representation.

Application Process

Applications are available starting in February each year. Information about the competition, award amount, and eligibility will be available on this website.

Please contact us with any questions about the competition. 

Award Winners

2020: Behavioral Foundations of Presidential Accountability

By Benjamin Goehring and Kenneth Lowande

Project Description:

Presidents possess vast authority to shape policy outcomes. This project investigates the dynamics of presidential actions by conducting a panel survey and companion media study. The survey will evaluate whether presidents receive a benefit or pay a cost for pursuing unilateral policy change and test whether this relationship depends upon the successful implementation of the order. In the first panel, respondents will receive information regarding a policy issue. They will be told the administration’s position; that the president took unilateral action; that Congress took action; or that both the president and Congress took action. In the second wave, the treatment will consist of information regarding the implementation of the policy: that the policy was successfully implemented, unsuccessfully implemented, or reversed by the Supreme Court. Post-treatment in each wave, they will be asked to report their level of support for the president and level of support for the policy change.

 

This experiment will be complemented with a media study that contextualizes its findings. The research team will examine whether the press provides clear attribution in reporting on executive action and symmetric coverage of issuance and implementation.

 

2019: Physiology of Threat Sensitivity and Policies Impacting Immigrant Youth

by Guadalupe Madrigal/Stuart Soroka

Project Description

Over the last 18 years, public debates about immigration in the U.S. have increasingly focused on one specific group: the Dreamers. In 2001, the DREAM Act was introduced to provide undocumented youth who immigrated to the U.S. before the age of sixteen a path toward legalization. In 2012, the continued failure to pass the DREAM Act led to the introduction of the DACA program, which temporarily gives these immigrants (deemed “Dreamers”) work permits, opportunities in college or the military, and protection from deportation. The program has broad public support: according to a CBS News Poll taken in January 2018, nearly nine in 10 Americans (87 percent) favor allowing young immigrants who entered the U.S. illegally as children to remain in the country. Polling also suggests that a majority of Americans support a roadmap to citizenship for Dreamers. This is even true for two-thirds of Republican voters.

Attitudes about Dreamers stand in contrast to attitudes about undocumented immigrants more generally. Even as support for Dreamers is high, there continues to be opposition to reform that would grant amnesty to the millions of undocumented people in the country, for instance. What accounts for the difference? Soroka and Madrigal contend that support for these immigrant youth is buoyed primarily by two related factors: (1) the perception that Dreamers are fundamentally non- threatening, and (2) the tendency for Dreamers to be seen as individuals rather than in groups.

Our sense of media coverage as having both de-emphasized threat and focused more frequently on individuals is based on an ongoing analysis of media coverage of immigration across 17 newspapers over the past 40 years. This news content is a precursor, but not the focus of the present study. Rather, this collaborative work uses a combination of psychophysiological and survey-based experimentation to test the significance of both threat sensitivity and person- positivity in accounting for the uniquely high levels of public support for Dreamers.

Soroka and Madrigal rely on psychophysiological responses to arrays of photos to capture underlying threat sensitivity as a potential moderator of threat-based cues. They also use a simple vignette-based experiment, embedded in an online survey. The experiment uses Dreamer and non-Dreamer immigrant vignettes, with a simple manipulation of threat cues and individuals versus group cues, through a combination of text and photos. The combination of these two approaches in the context of work on immigration attitudes is unique.

2018: Emotional and Contextual Origins of Bipartisan Trust.

By Hwayong Shin and Christopher Fariss. 

Project description

In contemporary American politics, dislike and distrust have increasingly become defining features of mass polarization. Polarization raises concerns, because it obstructs effective communication, reproduces prejudice, and hinders cooperation. To better understand the sources and obstacles of bipartisan trust, Fariss and Shin will investigate the linkage among emotions, contexts, and trust under partisan polarization. By using behavioral economics experiments, they will study how emotions and contextual cues influence interpersonal trust within and across the party divide.

2018: Mapping Socio-Political Change through the Negro Motorist Green Books.

By Kiela Crabtree and Mai Hassan. 

Project description

The Negro Motorist Green Book, published annually from 1936 through 1966, guided the travels of African-Americans across the entire United States in the era preceding the Civil Rights Act of 1964. As the ubiquity of the automobile expanded travel possibilities for white Americans, the opportunities available to blacks were limited by restrictions imposed through Jim Crow laws and implicit social and racial mores in different regions of the country. At their most basic, the Green Books tried to equalize travel opportunities for the races: each Green Book lists the hotels, private residences, resorts, and other establishments open to (and willing to serve) African-Americans. In this way, the Green Books thus became an invaluable source of information for black travelers. But today, the Green Books represent much more than a basic travel guide for historical America. These texts are an untapped resource about the changing social and political relationship between blacks and whites in the mid-twentieth century. The Green Books are an opportunity to analyze these changing relationships both over time and across geographic space, comparing where political attitudes and social norms were aligned and where they diverged.

Broadly, this project asks – to what extent does exposure of a minority out-group lead to changing political attitudes and behavior in the majority in-group? Hassan and Crabtree situate this question within broad theories of intergroup threat on the one hand and social contact theories on the other. They adjudicate between these two theories by creating new datasets drawn from several underutilized historical data sources, including the Negro Motorist Green Books, that map out social relations between blacks and whites over the continental US. Their intuition is expressed in the epigraph above, the words of Mark Twain and a motivating theme of the Green Books, that exposure to new places and ideas through travel and social contact more broadly breaks down perceived differences among individuals and their respective groups. The importance of this project is not contained wholly within the past, though. They hope that our project will provide insight into the ability (or inability) of ordinary social interactions to drive social and political change for future generations. Moreover, the conclusions of this project are applicable outside racial-bias and discrimination, but to discrimination along other identity cleavages. This partnership brings Mai’s expertise in comparative politics and ethnic conflict to the American subfield, using Kiela’s interest in black political behavior and American political development to situate the project. Further, they intend to make the data for this project publicly available for other scholars.

2017: Riots or Rallies? The Impact of Race on News Coverage of Protest Events

By Mara Ostfeld and Steven Moore. 

Project description

The literature on protest movements in the U.S. has broadly focused on the way that these movements have been able to overcome collective action problems and to meet their goals. This involves examining the conditions that lead up to a protest and the way that protest impacts the behavior of both citizens and elites (Tilly, 1978; McAdam D, Su Y. 2002; Gillion, 2012; Wasow, 2016). However, there is a missing link in this process. Protest movements must be seen and heard to change minds. Though demonstrations are strategically placed in locations that might draw attention, obviously the impact is not limited to those in the immediate vicinity. Actions are covered in the news media and this provides the protest activity a much wider audience of people whose minds can be changed. In short, accounts of the success or failure of protest have failed to account for a very important mediator, how the news media covers protest. To better understand these accounts, it’s necessary to better understand the way that the news media covers protests, and specifically whether there is variation based on the race and racial aims of the protestors.

A number of scholars who’ve taken a critical look at media coverage by race have found that negative stereotypes about African Americans are often exacerbated by these media sources. Particularly, blacks seem to be overrepresented as criminals and welfare recipients, and each of these findings has been demonstrated to have a causal effect on public opinion through experimental work (Gilliam & Iyengar 2000, Gilens, 1999). There is substantial reason to believe that similarly consequential differences may emerge in coverage of protests by the race of the protestors. Vesla Weaver’s frontlash thesis claims that racial conservatives, once their defeat on the issue of civil rights was imminent, changed the scope of the conflict (Weaver, 2007). She suggests that these elites refocused their efforts to retain the status quo by linking the civil rights movement and black political activism in general with the many racially motivated violent civil disturbances of the 1960’s (Weaver, 2007). Weaver claims that starting in the early 1960’s Southern legislators began to try to tie legitimate protest action in their states to crime, explicitly arguing that civil disobedience would lead to selective adherence to authority (Weaver, 2007). There is a litany of work form social psychology demonstrating that implicit and explicit association of African Americans and criminality is quite common (Greenwald, Oakes, & Hoffman, 2003; Payne, 2001). Important research from Davenport, Soule, and Armstrong seems to confirm that this bias can be found in response to protest by demonstrating that there is a differential response from the state based on the race of protestors. They find that police were more likely to be present at black protests from 1960-1990, though the difference varied over time (Davenport, Soule & Armstrong, 2011).

2017: What Makes You 'Like'?: Emotionality and Engagement with News Content in Social Media

By Brian Weeks and Dan Hiaeshutter-Rice. 

Project description

In the last presidential election, more news and political information was shared by individual users on social media than ever before, further cementing the importance of social media in political campaigns. Indeed, there is emerging evidence that political information shared in social media increasingly influences both the volume of citizens’ exposure to political news and the nature of the content they see. In particular, Facebook and Twitter are now instrumental in the production of what we term the political information market, or the scope of information about politics that is available for citizens to consume. Despite the increasingly important role of social media in political information exposure, there are many unanswered questions regarding what information is available on social media as well as how individuals react and interact with that information. This project will leverage a unique and comprehensive data set to address these important, lingering theoretical questions.

2017: The Limits of Religious Tolerance: Identity, Principle, and Pragmatism in Judgments of Religious Discrimination

By Carly Wayne and Ted Brader. 

Project description

Religious tensions in the United States have reached a crescendo in recent months. President Trump’s temporary ban on immigration from seven Muslim-majority nations sparked outrage among a large segment of Americans. However, this ban was also supported by nearly 48% of the American public. At the same time, a recent spate of bomb threats and vandalism against Jewish Community Centers has brought attention to a resurgence of anti-Semitism in America. Meanwhile, concerns about Christian religious freedom on issues like abortion, birth control and gay marriage, which flared repeatedly over the past decade, have returned in debates over Supreme Court nominations, executive orders, and health care reforms.

Tensions over religious diversity and claims of religious discrimination or entitlement have alternatively simmered or raged over the entire course of American history from the colonial era to today. A number of developments—recent patterns of immigration, fears of terrorism, international military conflicts, social movements advancing LGBTQ rights, the rising prominence of white nationalist or “alt right” political groups—have brought new and old conflicts to the fore. These events have involved people from an array of religions or religious perspectives—Christians, Jews, Muslims, Sikhs, Mormons, atheists. But how do Americans perceive such religious conflicts? In their eyes, when does equal treatment require accommodating diverse religious views or practices (e.g., not ‘imposing’ duties to perform services at odds with religious beliefs), and when do they believe the need for equal treatment overrides religious considerations (e.g., requiring identical security procedures even if some of those violate religious restrictions on dress or contact)? Why do Americans sometimes support “religious accommodations” in the enactment or enforcement of public policy and yet other times oppose them?

We seek to shed light on how the American public reacts to these debates over private and public action and, in particular, the factors that shape religious (in)tolerance and related public policies. Drawing on news stories of both well-known and obscure conflicts that have arisen in the U.S. in recent years, our project will feature a series of survey experiments to study variation across individuals and situations. We especially consider the relative bearing three factors have on public reactions: the religious identities of claimants and respondents, principled views (e.g., about the First Amendment, equal treatment, public vs. private spheres), and pragmatic considerations the details of particular situations (e.g., the relative imposition on a religious claimant relative to others). We also investigate the moderating role, if any, played by individual differences personal religiosity and religious literacy.

2016: #New(s)MediaEffects: Journalistic Reliance on Twitter and News Representations of Blacks and Latinos

By Mara Otsfeld and Alejandro Pineda. 

Project description

This research will serve to build on and advance Dr. Ostfeld’s research exploring the relationship between political media and the political dynamics of minority racial and ethnic communities in the U.S. It will also expand her focus into the understudied area of online political media, as well as add to the limited political science literature on this topic. Mr. Pineda shares an interest in the political behavior and attitudes of minority racial and ethnic groups in the U.S., with a specific interest in Latinos. However, he brings unique and important skill sets to this project. He has extensive experience with statistical programming and analysis, and is deeply interested in social media, web-scraping, and the type of “big data” analytics that this study will employ. Drawing from his programming experience, his skills will be particularly valuable in employing existing “sentiment” analysis software to create the dataset of news and Twitter content needed for our inquiry. We plan to meet every week and be in regular contact so that we can routinely check in on each other’s progress, and troubleshoot challenges that arise.

2014: Untangling Constituency Influence in State Legislatures

By Rocio Titiunik and Christopher Skovron. 

Project Description

What influence does public opinion have on legislative outcomes? This question is a central one to theories of representative democracy and empirical investigations of the causes and correlates of legislative outcomes (e.g., Bafumi and Herron 2010; Canes-Wrone, Brady and Cogan 2002; Lax and Phillips 2009, 2012; Miller and Stokes 1963). While pressures from interest groups, party elites, and the like obviously play an essential role in determining legislative behavior, the classic Miller-Stokes theory of constituency influence remains persuasive. We expect that constituents’ preferences will operate through electoral selection and through the “perceptual control” provided by the legislator’s perception of their preferences. We also expect that this kind of constituency control is unlikely to operate perfectly. Empirically, a strong but imperfect relationship between constituents’ preferences and representatives’ actions is often found, although the precise nature of the relationship is a topic of considerable debate.

Testing the Miller-Stokes theory requires data that has, until recently, been difficult to acquire. Valid estimates of constituency opinion in legislative districts have been hard to come by until recently. Advances in statistical modeling of opinion and new national databases make this task easier. Even harder still is collecting data on representatives’ perceptions of district opinion. However, Skovron’s work with colleagues on the National Candidate Study (Broockman, Carnes, Crowder-Meyer and Skovron 2013) provides insight into state legislators’ perceptions of district preferences. In an investigation of the “perceptual” side of the Miller-Stokes theory, Broockman and Skovron (2013) find that candidates for state legislative offices have very inaccurate perceptions of public opinion in their districts and that the candidates tend to overestimate how conservative their constituents are.

2015: The Clue’s in the News: Unpacking the Thermostatic Model’s Mechanism of Policy Responsiveness

By Stuart Soroka and Fabian Neuner. 

Project description

The relationship between government outputs and public preferences is central to political representation and democratic responsiveness. One theory that has recently gained substantial traction in this literature suggests that citizens respond thermostatically to policy change. Critics of this model – drawing in large part on work by Converse himself – are quick to point out that the average citizen has only a very limited grasp of politics and an even hazier understanding of the policy process and ensuing policy outputs. This alleged dearth of knowledge clashes with the normative ideal of democratic responsiveness, which requires citizens to be able to connect policy outputs to their preferences.

It nevertheless appears to be the case that government spending affects the public’s preferences for the amount of future spending. How can we reconcile evidence of thermostatic responsiveness with evidence of uninformed publics? Professor Soroka’s current theorizing suggests that even very basic information in media content provides cues facilitating an understanding of the direction of policy change. Tentative support for this theory comes from an analysis of media content, which reveals that the tone and emphasis of the news environment provides basic cues on the direction in which policy is moving. Media content analyses suggest that the public may be able to glean the information that is necessary for thermostatic responsiveness.

2013: The Evolution of Partisanship and Party Politics in the U. S. Electorate

By John Jackson and Elizabeth Mann. 

Project Description

What factors explain changes in individual and aggregate partisanship over time, and what factors explain temporal stability in partisanship at each level? Political parties and partisanship are central factors in all studies of politics at both the collective and the individual levels. The parties’ actions structure the electorate’s choices, influence the electorate’s partisanship, and affect a government’s likely policy directions after an election. In any given election individual partisanship provides a foundation of support for each party and constrains the parties’ options in the electoral competition. The interactions between the individual in the electorate and the party elites are not temporally stable, as evidenced by the dramatic realignment of “red” and “blue” states over the past several decades.1 In the field of party competition and partisanship, we seek to explain both the connections between mass and elite actors and how these connections contribute to the evolution of party systems. We propose to examine empirically models of both individual and aggregate partisanship changes over time.

Donations

If you would be interested in contributing financial support, you may do so by clicking the “Give Online” button below.

Every gift is important and makes a difference. Thank you!

 

The Converse-Miller Fellowship has been crucial to my development as a scholar at the University of Michigan. The fellowship has allowed me to take my research agenda — and the skills I’ve developed at the University — beyond Ann Arbor.

Working in collaboration with Mai Hassan, I was able to travel to several cities across the United States to photograph un-digitized editions of the Negro Motorist Green Books. The information contained in the Green Books is a rich qualitative dataset, mapping the country, politically and socially, as it was for African-Americans in the mid-20th century.

I am incredibly grateful that the Converse Miller Fellowship has given us the opportunity to contribute to preserving the legacy of the Green Books and analyzing this under-utilized resource. 

Kiela Crabtree

2018 recipient of the Converse Miller fellowship