Interdisciplinary Workshops on Politics and Policy

This presentation describes challenges faced by, and opportunities available for, communicating science to policymakers and the public. The challenges have several sources. One source is common errors about how people learn. Audiences have far less capacity to pay attention to new information than many communicators appreciate. Another source is differences between communicative dynamics in typical scientific and political settings. When science enters political domains, the manner in which audiences interpret new information can be very different than in scholarly contexts. Collectively, these challenges lead many science communicators to be surprised about how badly their presentations are received. There are also opportunities for better outcomes. The source of these opportunities is scientific research on topics such as attention and source credibility. Research on attention reveals ways to make scientific presentations more memorable. Research on source credibility reveals how to earn and maintain an audience’s trust – even when discussing politically sensitive topics. In sum, science communicators can use insights from the social and cognitive sciences to stay true to their own research while making their findings more relevant and more memorable to more audiences.

Much of the scholarly focus on racial priming has focused on the use of the strategy by White Republicans. However, I argue that this approach may be too narrow. Thus my research revisits the theory of racial priming, by offering a theoretical account of the use of racial appeals by black and white candidates, Democrats and Republicans, in contemporary American politics. Specifically, in the first experiment, I manipulate the message (implicit or explicit) and the race of the candidate (black or white) to test whether explicit appeals have been rejected to the extent suggested by the theory of racial priming. I find that white racial conservatives actually embrace rather than reject explicit racial appeals, regardless of who is making the appeal. Conversely, white racial liberals, reject explicit appeals when the messenger is white, but embrace them when the messenger is black. In the second experiment I test whether the racial composition of the supporters portrayed in campaign materials influences voters’ evaluations of candidates. The results of the experiment indicate that white candidates suffered a loss in electoral support from white voters, when their advertisement featured African Americans as opposed to whites. Conversely, support for African American candidates remained the unchanged, regardless of whether the advertisement featured blacks or whites.

Long periods of one-party rule are more common in parliamentary than presidential regimes. The difference across executive types of the prevalence of these so-called “dominant parties” is poorly explained by existing theories, which emphasize ruling party resource advantages as the main source of prolonged one-party rule. I develop an alternative explanation that highlights the greater variance inherent in presidential elections compared to parliamentary ones. Thus, by making electoral “upsets” more likely, presidentialism reduces the likelihood that one party controls the executive for long periods of time—even a party with large and systematic electoral advantages. I test the argument with an original dataset covering all ruling parties in electorally contested regimes since 1950. I find that (1) ruling party hazard rates begin low but rise substantially over time in presidential regimes, but begin high and fall in parliamentary regimes, and (2) this difference increases with electoral advantage. Thus, presidentialism has the surprising effect of better ensuring regular partisan rotation of the executive.

The ability of citizens to share information is essential to the success of deliberative institutions. This paper builds on game-theoretic models of strategic information transmission to offer a theory of how the interests that deliberators have in the outcome of deliberation can cause some citizens to be unable to share information or influence the deliberative process. Specifically, this paper argues that whether a person is able to share information depends on whether that person is in the majority or the minority in terms of their interest in the outcome of deliberation. Deliberating groups will discount information that is provided by members of the minority, even when this information is an important contribution to deliberation. I offer the first empirical test of this kind of model in realistic deliberative conditions using two experiments, a laboratory experiment and a field experiment, and find that arguments that are made by members of the minority are less influential than the same arguments when they are made by members of the majority. The paper concludes by discussing the implications of these findings for the equality and epistemic quality of deliberative institutions.

Using original local-level data from Berlin, Germany, this project demonstrates that the relationship between center-left political parties and immigrants is more complex than current research suggests. While center-left parties set the agenda on issues that concern Muslims and minorities in some locations, they are largely absent from integration policy-making in other parts of the same city. It is difficult to predict where center-left parties will be active on immigrant integration because party behavior varies in districts with similar demographic and institutional environments. This project shows that political competition is a fundamental factor for center-left behavior on integration. Electoral threat from different left parties has distinct and opposite relationships with center-left policy behavior. Competition from Green parties increases the probability that the center-left will produce integration policies, while contestation from far-left former communist parties decreases this same likelihood. These relationships remain robust even after controlling for demographics and institutions, and across different policy content themes such as education and employment. This project explores the mechanisms underlying these findings and discusses generalizability to other parts of Europe.

Schattschneider and Burnham’s thesis of a ’system of 1896’ has attracted considerable scholarly interest because of the fundamental challenge it presents to a political system in which competition is presumed to serve as the basis for the representation of citizens’ policy concerns. The thesis argues that much like an economic duopoly, the American two-party system may, through elite collusion, privatize conflict and serve the narrow interests of elites rather than the interests of a broader set of consumers. Schattschneider and Burnham contend that conservative Republican and Democratic elites colluded in the 1890s to reduce partisan competition in the South and the non-South, thus eliminating populist challenges and ensuring conservative political and economic ascendancy in both areas of the country as a consequence. Employing data on all counties in the United States, I find no evidence of a collusive 'system of 1896.'

I find that there were not marked declines in partisan competition outside the South after the 1896 election. Instead, partisan competition continued outside the South much as it had prior to the 1896 election and appears to have continued to motivate citizens to vote over the course of the subsequent decades.

This project is a refined and systematic investigation of race and ethnicity and its relation to health disparities. It is an exploration of broader (or more comprehensive) measures of the concept of race. More specifically, we conducted a concentrated and systematic examination of the concepts of race/ethnicity determining how much race is treated as a categorical classification variable as opposed to a multi-dimensional concept and accompanying set of measures. This objective was accomplished in a two staged research process. Part I is a systematic review and analysis of the measurement of race in social science research. Phase one utilized the variable search function in the ICPSR data archives in all studies that have race as a variable in the study. (Use of key word search used race as the initial screener); then additional race-related search terms such as racial, racialization, discrimination, racism and ethnicity were used.

The second phase of our research process builds on the first phase by utilizing our knowledge of the various approaches to conceptualize race to capture comprehensively the concept as a “lived experience”. Using an administration of a monthly Latino Decisions poll, we included several measures of race as well as “self-reported health”. We will therefore attempt to structure our measurement approach around three primary themes: a) In-depth focus on the context surrounding racial identity, or how individuals place themselves within racial categories; b) how ascribed racial status (how others view you regardless of your own racial identification) impacts health status in addition to self-identified race; and c) a measurement approach to capture discriminatory treatment based upon race/ethnicity. We treated the race measures as independent measures as well as a combined scale. A path analytic model was used to test the effects of our race measures for Latinos’ self-reported health status. As a pilot project funded by the UM Robert Wood Johnson Foundation, I was able to involve two faculty and four graduate students at the University of New Mexico and a post-doctoral student at Indiana University. The make-up of our research team represents clear expertise in both the measurement of race and racial health disparities across political science, public policy, and sociology. We are still early in the analysis of the ICPSR database and measurement refinement with the Latino decisions pool.

For certain kinds of governments and organizations, the overwhelmingly dominant trajectory has been towards centralization of power away from subunits, not only toward the central unit but toward the executive in the central unit. When we probe how such centralization occurs, we discover that the processes described are similar in a variety of diverse realms. Using several prominent examples from history, I show how executive authority becomes locked-in. The key to centralization becoming locked-in is the process of change in the way people come to see their interests as linked to different parts of governing institutions. Leaders succeed in building the lasting power of their office by institutionalizing the linked fate of their own policy programs to the ambitions and goals of subunit leaders and those representing subunit interests.

In order to avoid violating the norm of racial equality, some scholars argue that elites must now rely on indirect or implicit appeals, in order to activate Whites’ lingering racial animosities. One unresolved question in the racial priming literature however is whether negative messages about Whites are necessarily viewed as unacceptable among African Americans. We hypothesize that both implicit and explicit racial cues will effectively mobilize anti-White sentiment in the Black community. We rely on two nationally representative survey experiments to address this question. In our first experiment, we present respondents with group-centric political cues that highlight Black economic vulnerability relative to Whites. We find that when the relative status of their racial group appears in jeopardy, Blacks’ anti-White predispositions are powerfully associated with attitudes about this out-group and other policy relevant matters. In our second experiment, we randomly expose respondents to explicit appeals that castigate Whites at varying levels of severity. We hypothesize that Blacks negatively predisposed towards Whites will be most inclined to embrace this message. Our results are consistent with this expectation. In our conclusion we discuss the implications of our findings for Black public opinion research and the theory of racial group priming.

Why does the public support some high profile Supreme Court rulings, grudgingly accept others, and denounce a select few? How can we understand the public response to cases like those involving Obamacare, the Second Amendment, and eminent domain? I explore the ability of the U.S. Supreme Court to frame effectively its decisions in the media. I build on existing models of political communication to show how the media’s reliance on elite sources and its need to report accurately and quickly on decisions causes it to shape coverage in response to voting outcomes on the Court. In particular, newspapers express more criticism of rulings decided by small majority coalitions. In the most closely-divided rulings, the press pays further attention to the makeup of justices in the majority, offering the least favorable coverage for those cases decided by an ideologically-homogenous voting block. A unique portrait of judicial press coverage emerges, suggesting the power of Supreme Court justices in the dissenting coalition to limit support for rulings.

Why do mainstream center-right parties in Western Europe seek the votes of ethnic minorities, non-Western immigrants and their descendants, at some time and not others? Dominant explanations focus on these parties’ use of immigration and race issues to attract anti-immigrant rather than immigrant voters. Yet considerable spatial and temporal variation in center-right party strategy toward ethnic minorities challenges this view. Rather than treat all ethnic minorities the same, center-right parties distinguish between those with citizenship and those without, and this difference drives their outreach strategies. Specifically, changes in the number of ethnic minorities with citizenship to the foreigner population, the citizenship ratio, account for when parties reach out to ethnic minority voters. Cross-sectional time-series analysis of party positions toward ethnic minorities from the Comparative Manifesto Project combined with original data on citizenship acquisitions and the composition of the foreigner population from six Western European countries – Denmark, France, Germany, the Netherlands, Sweden, and the UK – from 1970 and 2010 confirms that changes in the citizenship ratio best explains party positions toward ethnic minorities, despite differences in institutional, competitive, and economic conditions.

Political scientists commonly distinguish issues that are moral from ones that are not. The distinction is taken to be important for understanding persuadability, the stability of opinions, issue salience, and several other factors, but there are inconsistencies in how scholars have conceived it. Drawing insights from psychology, I suggest that it is fruitful to think about moral conviction as a dimension of attitude strength. Using three data sources, I examine how much this perspective contributes to our understanding of politics. I find evidence that moral conviction shapes political opinions and action in surprising ways: it varies across issues, but also within them, including issues usually considered not to be moral. It contributes to participatory zeal, but also may be related to political extremism and hostility. The findings suggest an instrumental approach to conceiving the role of morality in politics.

This paper examines attempts by authorities to undermine overt collective challenges by targeting activities that precede and/or support such behavior. After providing a theory of how repression and resistance develop, the study examines unique data drawn from the confidential records of the Guatemalan National Police to assess the use of repressive action during the years between 1975 and 1985. Empirical tests confirm that 1) government forces anticipate challenger development by identifying the mobilization activities nascent challengers rely on to initiate and sustain overt collective challenges; and 2) that the use of repression to undermine such efforts is specifically designed to contain the spread of radical (i.e., highly transformative) mobilization. Implications are drawn for how we understand and study political conflict and order.

Capital owners are often thought to be more conservative or market oriented than people who rely solely on labor income; democratic societies in which capital ownership is diffused to the median voter are often thought be more conservative or market oriented than societies in which capital ownership is concentrated. Policymakers have often used this idea as the basis for social engineering. Recent history is littered with examples of governments manipulating or attempting to manipulate the distribution of capital for political effect, from Thatcher’s council housing initiatives to under-pricing in post-communist privatization in Eastern Europe to George W. Bush’s attempt to create an “ownership society” through social security reform.

Despite the frequency with which politicians have tried to manipulate the political environment by manipulating the distribution of capital, we don’t really know if it works. Can policymakers actually create an “ownership society”? This talk, and the book project it derives from, looks to shed light on this question by examining the relationships between Latin American pension reform (which has diffused stock and bond ownership to varying degrees in different countries), Latin American public opinion surveys and Latin American election results.

In one of the most economically unequal periods in American history, existing research largely overlooks the possibility that attitudes toward class groups at the two extremes of the economic distribution – the poor and the rich – influence evaluations of candidates for public office, political parties, and social movements. I investigate this possibility through both observational analyses and an original survey experiment on a national sample of adults. In contrast to widespread claims that Americans are either indifferent to class or derogate the poor while admiring the rich; the findings suggest that on balance the poor are viewed with sympathy, while the rich are viewed with resentment. Furthermore, these class group attitudes exert a powerful influence on a wide range of political preferences. The findings have important implications for our understanding of American politics as the gap between the poor and the rich continues to widen.

The study of corruption has largely ignored the question of whether and why it varies across bureaucratic agencies within the same state. In this paper I use existing survey data to establish that within-state variation in agency corruption is significant relative to between-state variation. I then posit two explanations derived from the literature, and a novel one of my own. Existing theory predicts differences in corruption across agencies due to their different functions, and because of different structural constraints. My contribution is to explain how rent-seeking political leaders exploit these differences across agencies, keeping in mind that the opposition has an incentive to discredit the leader by revealing as much corruption as possible. Finally, I offer preliminary tests of some implications of the theory, using data on the purchasing behavior of all federal agencies of the Russian government.

As media coverage of politics has become more ideologically polarized in recent years, it has also become more uncivil. Political blogs, in particular, contain pervasive and often surprisingly offensive political content, with much of the worst content residing in the blog reader comment section. For this reason, blogs have been heavily criticized. However, that criticism often makes an important, unsubstantiated empirical assumption. Researchers assume ill effects, but few have tested whether blog incivility in fact causes normatively undesirable results. This project examines the effects of political blog incivility – particularly incivility expressed by blog commenters – on blog readers’ political opinions. I focus especially on whether incivility polarizes readers’ political views, arguing that uncivil partisan comments are more likely to polarize readers than other types of partisan content. To test this, I conducted an Internet experiment that focused on opinion related to the federal sequester, a salient political topic at the time of data collection. A control group read a neutral blog article about the sequester and two neutral reader comments. Treated participants were then randomly assigned to read additional comments that took a liberal/Democratic or a conservative/Republican perspective. Within those two treatment groups, participants were again randomly assigned to read civil arguments, uncivil comments, or a combination of the two. Statistical analyses demonstrate that the incivility aimed at ideological opponents was unique in its ability to polarize readers’ opinions on policy and evaluations of political leaders.

Skip Lupia, Hal R Varian Collegiate Professor of Political Science and Research Professor for the Center of Political Science, will present research-based techniques for talking about research in ways that non-scientist will find interesting and persuasive.

Decades of scholarship in political behavior has found that the media and political elites can affect the public's attitudes on important political topics. One of the most common methods for doing so involves framing. Yet, even once the media and other elites have chosen to use one frame or another, they still have to decide what type of evidence to use to support their chosen frame. The most common types include indicator-based evidence and human-interest evidence. For example, once one chooses to frame the economy in terms of unemployment, one can discuss objective changes in terms of the unemployment rate (i.e. indicator-based evidence), individuals' experiences with unemployment (i.e. human interest evidence), or both.

In this talk we examine the political implications of these various forms of news evidence, and in particular how they differentially affect people's political attitudes. In addition to enriching our theoretical understanding of framing, we argue that studying news evidence can also help explain what many consider to be a puzzle in the study of Americans' attitudes on income inequality. Past work routinely finds support among a majority of Americans for reducing the gap between the rich and poor, yet at the same time only tepid support for direct political action to achieve that goal. We argue that focusing on news evidence -- in particular, the evidence used by the news media to cover the issue of income inequality -- can help explain this pattern.

In the Regression Discontinuity (RD) design, units are assigned a treatment based on whether their value of an observed covariate (the "score'') is above or below a fixed cutoff. Under the assumption that the distribution of potential confounders changes continuously around the cutoff, the discontinuous jump in the probability of treatment assignment can be used to identify the treatment effect. Although a recent strand of the RD literature advocates interpreting this design as a local randomized experiment, the standard approach to estimation and inference is based solely on continuity assumptions that do not justify this interpretation. In this article, we provide precise conditions in a randomization-inference context under which this interpretation is directly justified, and develop exact finite-sample inference procedures based on them. Our randomization-inference framework is motivated by the observation that only a few observations might be available close enough to the threshold where local randomization is plausible, and hence standard large-sample procedures may be suspect. Our framework is general and applicable to any RD design where small sample sizes constrain researchers' ability to make inferences. We illustrate our methodology with a study of party-level advantages in the U.S. Senate that compares states where the Democratic party barely won a Senate election to states where the Democratic party barely lost. Since the Senate has only one hundred seats up for election every six years, the number of close races is small and our framework is ideally suited for the analysis. We find large effects in the Senate when the incumbent party competes for its seat but, in contrast to results from standard RD estimation methods, negative effects when the party competes for a seat after having won the other seat in the state's Senate delegation.

Deficits of representative democracies, such as declining turnout and growing mistrust in political actors and institutions have made scholars argue in favour of a need for transformation of democracies. The long list of suggested reforms include the expansion of the electoral marketplace and increased citizens participation in electoral processes at different levels as well as access to decision making to new forms such as referenda, popular assemblies or more recently to new forms of deliberative democracy. While those reforms intend to make democratic processes more inclusive there is a great risk that the cure is worse than the disease: New forms of participatory democracy can just as well lead to less inclusiveness, greater inequality and worse representation and make elites less responsive to a public at large while not even increasing legitimacy of democracies.

The empirical relationship between democratic participation beyond elections and equality and substantive representation is explored using data from the European Social Survey round 1-5 covering over 30 countries. Findings from other studies are confirmed that the lower participation rate in various form of participation is associated with much higher inequality in income, education and gender between participants and non-participants. Participants in other forms than elections are more polarised on a number of different issues and also lean slightly more to the left while moderate, centrist views are clearly underrepresented. Those findings question the value of democratic expansion in practice.

How does exposure to partisan media affect polarization in new democracies? Partisan media are often blamed for sowing seeds of discord, intolerance, and even violence in Africa. However, there is little empirical work on the subject, and information-processing theories suggest that extreme position-taking is only one of several possible responses to partisan media. We conducted a field experiment in Ghana in which tro-tros (commuter mini-buses) were randomly assigned to four conditions; passengers ridingtro-tros in the first three conditions heard live talk-radio programming from pro-government, pro-opposition, or neutral stations, while passengers in the fourth condition were not exposed to radio during their morning commute. We find no effect of exposure to like-minded media on partisan polarization. Instead, we find significant evidence of moderation from exposure to cross-cutting broadcasts, indicating that subjects were persuaded by rival arguments. Partisan broadcasts also seemed to expand respect for politicians and encourage displays of national over partisan identity. Rather than fueling extremism, as most observers fear, we argue that partisan media encourage moderation by exposing citizens to alternative perspectives.