Skip to main content

Verified by Psychology Today


Who Supports Political Violence?

What is the mindset of people who attack airline attendants over mask mandates?

Key points

  • Support for extreme political action is rising.
  • Being ambivalent about an extreme view enhances support for extreme behavior.
  • The impact of ambivalence is especially likely when it is seen as legitimate.
Source: Gallagher_Photography_Shutterstock

By Richard E. Petty, Ph.D., and Joseph J. Siev, Ph.D.

Nearly one-quarter of the adults contacted in a recent survey agreed that people may have to resort to violence to save the country. Also disturbing is that reports of other sorts of violence such as on airplanes and against members of Congress have risen since 2016.

Who are the individuals who support violence against their political opponents? As social psychologists interested in the link between beliefs and actions, this question has grabbed our attention as well as that of many other social scientists.

Some of the initial predictors of support for political violence were ones that many might have expected. For example, some research has shown that people who tend to be aggressive in general and those with extreme political views are more likely to support violence. That makes sense.

Ambivalence Enhances Support for Extreme Action

In our own research, however, we discovered a surprising new contributor: the extent to which people were ambivalent about their extreme political attitudes. Ambivalence stems from one of three kinds of attitudinal conflict:
(1) having mixed views on an issue
(2) having views that disagree with people you value
(3) holding attitudes different from the ones you wish you had

Each of these forms of ambivalence causes discomfort.

Our research examined the role of attitudinal ambivalence in enhancing support for extreme actions. We analyzed national survey data gathered by others and we also collected our own data in which we measured people’s attitudes, their degree of ambivalence (attitudinal conflict), their support for both moderate and extreme actions, as well as other variables. We have always observed the same thing across a wide variety of topics (for example, abortion, Covid policies, political ideology).

First, in accord with much previous research, more attitudinal conflict was associated with less willingness to support or engage in moderate actions such as voting or advocating for a cause. Second, however, more attitudinal conflict was associated with more support for extreme actions such as sacrificing one’s life for one’s position or physically attacking one’s partisan opponents. The latter was especially pronounced among those with the strongest political views.

These results held even after controlling for other variables that had predicted support for violence in prior research. And, it also held when controlling for the participant’s political ideology, gender, education, and several other variables. Furthermore, we found the same pattern when we examined whether people would be willing to donate their own money to moderate versus extreme advocacy groups.

Ambivalent but Extreme Attitudes Are Like Fragile High Self-esteem

Although these results are clear, why would this be the case? We suspect that our effect is related to another well-known phenomenon in psychology: fragile versus secure self-esteem. To understand the analogy, consider two possible bosses. Both bosses report thinking very highly of themselves, but one boss is secure in that extremely positive self-assessment but the other one is conflicted about it. The latter boss might suffer from imposter syndrome.

Now imagine that you tell the boss with secure high self-esteem that you think a mistake was made in introducing a new company policy. This boss might thank you for pointing it out and then aim to correct it.

However, what happens when you tell the boss with fragile high self-esteem about the mistake? This might instead lead to an extreme reaction, perhaps berating or even punishing you for the disclosure. This boss is taking what might appear to be strong action, but it is largely to compensate for their underlying high self-esteem fragility (that is, their underlying weakness).

We discovered another surprising thing about the link between having polarized but conflicted attitudes and support for extreme behavior. Specifically, this link depended on whether the person believed the ambivalence was justified or not.

If the person came to believe that being ambivalent was not justified, the person might try to become less ambivalent. But, if the person came to believe that being ambivalent was justified, it doesn’t make sense to try and reduce that ambivalence. This person is left to deal with the underlying insecurity by overcompensating for it.

In short, believing that one’s ambivalence is justified closes the door to ambivalence resolution, leaving the conflict intact. This is like telling your boss that their impostor syndrome is valid. It leaves your boss in a fragile (insecure) state in which the boss might feel a need to justify the high self-esteem by acting in a powerful (extreme) way.


If attitudinal ambivalence contributes to support for extreme action, future research should aim to understand the most important causes of that ambivalence. Also, because the deleterious effect of ambivalence is most likely when the ambivalence concerns an extreme attitude, perhaps the best way to combat it is by reducing the extremity of the attitude itself.

Richard E. Petty, Ph.D., and Joseph J. Siev, Ph.D., are with the department of psychology at Ohio State University.


Siev, J. J., & Petty, R. E. (2024). Ambivalent attitudes promote support for extreme political actions. Science Advances, 10(24), eadn2965.

More from Attitudes and Persuasion Lab
More from Psychology Today