The evidence points to the American electorate being more polarized than ever – but aside from influencing election results, what does this mean for wider society? In new research Shanto Iyengar examines people’s prejudices towards those that identify with their political opponents – or ‘out groups’. By using experiments which test these prejudices, he finds that partisan prejudice and bias outweighs racial biases. He writes that this heightened state of partisan ill will may be being fuelled by the polarizing effects of a news media which allows people to self-select into an audience that shares their partisan views.
Over the last 30 years, American political leaders have moved steadily toward the extremes of the ideological continuum. However, ideological divergence among elected officials does not necessarily imply equal divergence among voters and the extent of mass polarization is very much a subject of scholarly debate.
Ideological extremity is not the only way to think about party polarization. An alternative indicator – grounded in theories of group identity – is the extent to which partisans view their opponents as a disliked group of ‘others’ (an ‘out group’). Classic studies by psychologists have demonstrated that the mere act of identifying with a particular group — no matter how trivial the basis for group assignment — is typically sufficient to trigger positive evaluations of the in group and negative evaluations of out groups. Applying an affective standard, the American electorate is indeed polarized; “fear and loathing” is an apt characterization of the current state of inter-party relations.
An abundance of survey evidence demonstrates that party affiliation elicits in and out group sentiments. Democrats and Republicans express warm feelings toward co-partisans and harsh evaluations of those who support the non-preferred party. What is especially surprising is the extent to which disdain for political opponents has intruded into non-political judgments and inter-personal relations. Partisans frown upon inter-party marriage, hold harsh stereotypes of their opponents, and are willing to discriminate against individuals based on their party affiliation.
Experimental evidence further demonstrates the extent of partisans “fear and loathing” for each other. In one experiment, study participants were given resumes of two high school students and asked to select the one who was more deserving of a college scholarship. The resumes manipulated both the students’ academic prowess (grade point average) and their participation in extra-curricular activities. The set of activities included involvement in student political organizations (e.g. Young Democrats) as well as organizations representing racial groups (e.g. African-American Student Association). Some participants were assigned to the party cue conditions, others to the race cue conditions. As Figure 1 illustrates, we found that in the partisan cue conditions, partisans invariably awarded the scholarship to the student with the “correct” political inclination irrespective of their grades. In contrast, racial bias in scholarship allocation was modest with white participants more often than not awarding the scholarship to the minority student.
Figure 1 – Predicted Probabilities for Partisan and Racial Winner Selection
Note: The horizontal values group the data by the race of the participant, and the vertical facets group the data by the qualifications of the scholarship candidates. The dots are the predicted probabilities of selection in each group, and the bars are 95% confidence intervals.
We also had our participants play online versions of the trust and dictator games which provide behavioral measures of inter-personal trust. In a trust game, Player 1 is given an initial endowment ($10) and instructed that she is free to give some, all, or none to Player 2 (said to be a member of a designated group). She is further informed that the researcher will triple the amount transferred to Player 2, who will have a chance to transfer some amount back to Player 1 (though Player 2 is under no obligation to return any money). The dictator game is an abbreviated version in which there is no opportunity for Player 2 to return funds to Player 1. As in the scholarship study we compared the effects of partisan and racial group cues on amounts allocated to Player 2. Some participants found themselves paired with a co partisan or party opponent, others played with a Player 2 from the same or different race. Once again, our results (shown in Figure 2) showed significant discrimination in the amount players awarded based on partisanship and only trivial distinctions based on race. We further show that players’ allocation strategies are aimed more at penalizing players from the opposing party than rewarding co-partisans.
Figure 2 – Effects of Race and Partisanship on Allocations to Player 2
Note: The dots are the means for each group, and the bars are 95% confidence intervals.
In a final study, we examined the degree to which partisan prejudice has become sufficiently ingrained in voters’ psyches to be expressed sub-consciously. Psychologists have demonstrated that a great deal of what occurs in the mind is independent of active consciousness. In the case of socially sensitive questions, such as race relations, implicit measures elicit a more accurate readings of race bias since it is impossible for subject to deliberately mask their antipathy toward out groups. A standard measure of implicit racial prejudice is the Race Implicit Association Test (IAT). We developed an analogue association test aimed at uncovering implicit bias toward political parties (details on the methodology of these tests is outlined here). The IAT results converged with the scholarship and our behavioral games findings; implicit prejudice toward the out party easily exceeded whites’ implicit bias toward African-Americans. Thus, no matter how it is measured, inter-group animosity based on partisanship exceeds animosity based on race.
What explains the heightened state of partisan ill will? Elite ideological polarization may be implicated as partisans may attribute leaders’ preferences to the rank and file. In fact, partisans systematically under-estimate the extent to which their party has moved away from the center while exaggerating the extremity of the out party. The changing media environment is also a plausible causal factor. Precisely over the same period that partisans’ feelings have polarized, media consumers have achieved a greater degree of control over their exposure to news. In particular, they can now self-select into an audience that shares their partisan views. Two of the major American cable TV networks exhibit overt partisan slant in their reporting and the vast majority of online political commentators are similarly aligned. In these outlets, what passes for news frequently includes ad hominem attacks on political opponents. Although the market share of partisan news sources remains small, there is growing evidence of a large secondary audience because their controversy-filled news reports recirculate extensively through social media networks. In the long run, exposure to information that systematically denigrates political opponents can only intensify ill will across the party divide.
This article is based on the paper, ‘Fear and Loathing across Party Lines: New Evidence on Group Polarization’, in the American Journal of Political Science.
Note: This article gives the views of the author, and not the position of USApp– American Politics and Policy, nor of the London School of Economics.
Shortened URL for this post: http://bit.ly/1xCefzb
Shanto Iyengar – Stanford University
Shanto Iyengar holds a joint appointment as the Harry and Norman Chandler Chair in Communication and Professor of Political Science at Stanford University. Iyengar is also a senior fellow (by courtesy) at the Hoover Institution. Iyengar’s teaching and research addresses the role of the news media and mass communication in contemporary politics. He is the author of several books including Media Politics: A Citizen’s Guide (W. W. Norton, 2007), and Going Negative: How Political Advertisements Shrink and Polarize the Electorate (Free Press, 1995).