Arguably the most consequential development in American public opinion in the past two decades is the increasingly negative feelings partisans of one party harbor toward their opponents. My book project, "It’s Not Me, It’s You”: How Americans’ Animosity Toward Their Opponents Drives Modern Politics, examines how this negativity influences other aspects of public opinion and, in turn, contributes to the present problems of governance plaguing the U.S. While the past half century of public opinion research has focused on how positive attachments to a party influence individual thought and behavior, I demonstrate the affective basis for partisan opinion has changed fundamentally over the past few decades. I find that negative out-party feelings (not positive in-party feelings) contribute to hyperpartisan thinking in several important ways, including how partisans understand public policy, the way they process information, and the importance they attach to democratic norms and civil liberties. In addition, the book also investigates some of the antecedents of negative out-party feelings, arguing that close party margins at the national level and increasing mental associations between the parties and "disliked" groups contribute to rising levels of partisan hostility over time. These results suggest a need to reconceptualize modern partisanship not as an affinity toward the party with which one identifies, but rather as a product of how Americans feel toward their political opponents. 

My work on this subject was recently featured in reporting by NBC News. You can find a link to that article here


Believing the Worst: Out-Party Negativity and Belief in Conspiracy Theories [PDF]

In recent years, affective polarization—the increase in mutual antipathy between ordinary Democrats and Republicans—has received a great deal of attention both inside and outside the academy. In spite of this, researchers know little about how partisans’ increasingly negative feelings toward their opponents may influence the way that they reason about politics or evaluate information. In this paper, I examine the role that out-party negativity plays in partisans’ willingness to believe a particularly pernicious type of misinformation: political conspiracy theories. Using data from three studies, I demonstrate that negative out-party affect plays a significant role in partisans’ belief in conspiracy theories, above and beyond the effects of strength of party identification, ideology, and positive feelings toward one’s own party. These results persist even after controlling for factors that predispose individuals to believe in conspiracy theories, like interpersonal trust, trust in government, and a propensity to engage in conspiratorial thinking. Moreover, the influence of negative out-party affect cannot simply be attributed to symbolic ideology, a dislike for opposing ideologues, generalized affect, or a propensity to engage in expressive responding. These results have important implications for our understanding of the nature of contemporary partisan bias: partisans appear to traffic in misinformation–no matter how egregious—primarily because they believe the worst of their opponents.

The Micro-Task Market for "Lemons": Collecting Data on Amazon's Mechanical Turk (with Doug Ahler and Gaurav Sood) [PDF]

Amazon's Mechanical Turk (MTurk) has rejuvenated the social sciences, dramatically reducing the costs of collecting original data and freeing researchers from reliance on the "narrow data base" of undergraduate subjects. Recent evidence, however, suggests that a non-trivial portion of data collected on MTurk is "suspicious," either generated by "non-respondents" (bots) or non-serious respondents. Spurred by this concern, we fielded an original survey in August 2018 designed to measure the prevalence of both "cheating" (respondents masquerading as someone else) and "trolling" (respondents intentionally providing misleading humorous or provocative responses) on the platform. We find that about 11% of respondents likely circumvented location requirements or used multiple devices from the same IP address to complete our HIT. More troubling, 16% of responses originated from blacklisted IP addresses. Finally, we estimate that about 5-7% of respondents provided non-serious responses to our questions, either because they were satisificing or trolling. Altogether, we find about a quarter of our responses to be potentially untrustworthy. We conclude by providing some recommendations for researchers to improve data collection and quality on MTurk. 

A Gap in Our Understanding? Reconsidering the Evidence for Partisan Knowledge Gaps (with Gaurav Sood) [PDF]

Conventional wisdom in political science suggests that there are large, frequently occurring gaps between what Democrats and Republicans know. This has long fanned concerns about citizens’ ability to hold elected officials accountable. In this paper, we reconsider the evidence on the frequency and size of such gaps. We assemble a dataset of knowledge items with a “partisan relevance” from three prominent studies on partisan knowledge gaps. Our final dataset includes 152,124 responses to 161 political knowledge items on 39 surveys. We find that partisan knowledge gaps are both less common and are smaller than commonly understood. Tellingly, about 30% of knowledge gaps occur in the “wrong” direction; that is, on nearly one in three knowledge items, partisans are either less likely to know party-congenial facts or more likely to know party-uncongenial facts than their opponents. In addition, despite large sample sizes, more than half of the gaps in the expected direction are not statistically significant at conventional levels. We find that the average knowledge gap between Democrats and Republicans to be a mere four percentage points. These results suggest that we need to revisit our understanding of how partisanship affects information acquisition and retention.


The Dislikability Heuristic: Out-Party Negativity and Partisan Preferences, 1988-2016

While most research on partisanship has focused on how positive attachments to one's own party influences thought and behavior, only recently have scholars turned their attention to how the negativity partisans feel toward their opponents influences how they approach the political world. In this paper, I make use of 30 years of ANES surveys to track how these competing attachments influence partisans' positions on matters of public policy. I find that the affective basis for partisan opinion has changed fundamentally over the past few decades. Prior to the onset of affective polarization, the effect of these feelings was more in line with a traditional understanding of social groups: feelings towards one's own party outweighed negative feelings toward the opposition in guiding issue positions. Since 2000, the reverse is now true: out-party negativity now dominates in-party positivity in preference formation. This trend is strongest among the politically knowledgeable, but even those who know little about politics rely on their negative feelings toward their opponents to inform their stances. These results suggest a need to revisit the nature of contemporary partisanship. 



Explaining Partisan Affect: Partisan Response to Partisan Response (with Doug Ahler & Gaurav Sood)

What causes affective polarization? Past research suggests that the affective gulf between partisans is the result of greater policy disagreement or the increasing alignment of social and political identities. In this paper, we propose a third potential source of interparty animus: partisans' response to their opponents' faulty political reasoning. Specifically, we argue that partisans punish their opponents for engaging in blatant motivated reasoning but fail to hold their own side accountable for similar failures. To test this theory, we rely on two experiments designed to exogenously manipulate perceptions of partisan motivated reasoning among partisans' co-partisans and opponents. Results demonstrate that partisans appear to exhibit this "bias blind spot;" learning about their own side's wishful thinking does not damage evaluations of their co-partisans. In fact, we find some evidence for "partisan cheerleading": Democrats and Republicans actually appear to punish in-party supporters for not sufficiently engaging in biased thinking. On the other hand, learning that one's partisan opponents are less biased than anticipated leads to a lessening of dislike for the other side. 


Receive, Accept, Do Not Sample: Downweighting as a Form of Motivated Reasoning (with Doug Ahler)

Why do partisans punish certain representatives for misdeeds and poor performance but fail to punish others? Motivated reasoning is central to this phenomenon, but existing scholarship fails to account for all of its modes of dissonance reduction. Public opinion research argues that mass partisans either (1) selectively fail to learn or (2) "reinterpret" unfavorable facts to maintain a positive, party-consistent worldview. Although we believe both these explanations have merit, they seem to miss perhaps the most common form of motivated reasoning. Donald Trump's candidacy reminds us that partisans often learn unflattering facts about their side—facts that have no reasonable way of being reinterpreted—but fail to update their evaluations based on this information. In this paper, we demonstrate that partisans engage in a third common form of motivated reasoning. Using both existing survey data and original experimental work, we show that partisans often acknowledge the accuracy of worldview-incongruent information they would typically consider relevant. Instead of rejecting such information, their response is to "downweight" disparaging information and rely more heavily on other considerations that paint their own side in a more flattering light. This explains how partisans continue to hold stable, positive attitudes toward party politicians or policies in the face of negative, but widely-accepted truths. These results have important implications for our understanding of the pathways through which partisan bias occurs. 

The Positive Side of Negative Partisanship: Out-Party Negativity and Democratic Attitudes 


The campaign and election of Donald Trump have pushed the limits of partisanship beyond what many could have imagined, raising concerns about the fundamental health of American democracy. In this paper, I investigate how the trend of rising out-party hostility influences Americans' attitudes toward democratic norms, including civil rights, liberties, and "rules of the game." Using two original, nationally-representative surveys conducted in 2017 and 2018, I demonstrate that out-party negativity plays a dominant role in the degree to which partisans find abstract civil rights and liberties to be important in a democratic society. Surprisingly, out-party negativity appears to bolster individuals' endorsement of democratic norms, while warmer feelings towards one's own party actually reduce support. In this sense, negative partisanship appears to play a normatively beneficial role in promoting democratic values. ​