By Jin Woork Chang, Nazli Turan, and Rosalind Chow

According to a new survey, polarization is the defining feature of 21st century American politics. Americans express increasingly extreme political views, prefer to be around those that share their beliefs and values, and show antipathy to those who don’t (Pew Research Center, 2014). Another way in which this polarization is manifested is the rise of political stars such as Sarah Palin, Ted Cruz, or the choice of Paul Ryan, a figure known for his extremism, as the Republican vice presidential nominee in 2012. Importantly, the appeal (and success) of these ideologically extreme individuals runs against the conventional wisdom that political parties should appeal to the middle in order to win general elections. Given the apparent costliness of having such extreme members represent the political party, why are Americans turning to extreme leaders?

In the months leading to the 2012 presidential elections, this was the exact question that puzzled us and led us to explore when and why people choose extreme leaders to represent their groups. We reasoned that one possibility for this desire for extremism might be the fact that the political context has been increasingly portrayed as competitive. When people perceive competition between their group and others, it becomes especially crucial for them that it is clear what their group represents and what sets it apart from other groups (Jetten et al., 2004). This desire for clarity regarding the distinctive qualities and values of the group, in turn, may incline them towards leaders that can best achieve this goal. And, who better to achieve such differentiation than a leader who holds the extreme version of the views and values of the group? We set out to test this idea in three studies.

In the first study, we asked self-identified Democrats to consider an upcoming hypothetical primary in their district. Half of the participants were told that the district was hotly contested between Republicans and Democrats, and the other half were told that the district had always voted Democratic. We then asked participants to choose between three candidates: 1) one that votes along the party lines 93% of the time (or normative), 2) one that votes along the party lines 67% of the time, but when not voting with the party, would vote in a more conservative direction (or anti-normative), and 3) one that votes along party line 65% of the time and other times would vote in a more extreme liberal direction (or extreme). As we expected, participants were more likely to choose the extreme candidate when the primary was described as being highly competitive than when it was described as being a sure win.

With this initial demonstration of our primary effect, in the next two studies, we set out to examine in depth why people preferred extreme leaders when there was high inter-group competition. In our second study, which was conducted in the weeks leading up to the 2012 US Presidential election, we asked Democrats and Republicans for their levels of support for Barack Obama and Mitt Romney, respectively. Importantly, we also led participants to believe that the election was going to be either very close or not. We also described the two candidates as holding political views that either aligned with their party’s norms or as being even more extreme than the average party viewpoint. Again, we found that participants preferred a more extreme version of their respective candidate when the election was described as being close. Moreover, this effect was driven by their belief that the more extreme candidate was a better representative of their party.

Finally, in our last study, we asked participants to consider the upcoming 2016 Presidential elections and gave them information that led them to believe that the election would either be competitive or not. In addition, we provided them with information about a hypothetical Presidential candidate, whose political views either aligned with the party norm or were more extreme than the party norm. We also asked participants to what extent this candidate would effectively differentiate their party from others. As we saw in the prior two studies, participants preferred extreme candidates when they thought the election would be close, and this preference, which was even more pronounced with participants highly identified with their party, was driven by their expectation that such a candidate would successfully differentiate the party from others.

To summarize, our findings demonstrate that when people think that their group is in competition with another group, their desire to ensure that their group’s identity is distinctive from the other group is heightened. This desire, in turn, can lead to a preference for extreme leaders. Importantly, the rise of extreme representatives and leaders are costly to the social system at large. If all groups choose leaders that are at the groups’ ideological extremes, they are less likely to be able to work effectively with one another, despite needing to do so in order for political systems to function successfully.


Jin Wook Chang is a doctoral student at the Tepper School of Business, Carnegie Mellon University. Nazli Turan is an assistant professor at Catolica-Lisbon. Rosalind Chow is an associate professor of organizational behavior and theory at the Tepper School of Business, Carnegie Mellon University.