The idea that fear drives right-wing political attitudes and support for right-wing politicians has largely become accepted as a truism among many scholars (e.g., Altemeyer 1996Jost et al. 2003 2017Robin 2004Wilson 1973) and pundits (e.g., hereherehere). New research, published in Advances in Political Psychologysuggests that this model of conservatism may be misspecified.

Marcus et al. (2019) contend that much of the foundational scholarship on the psychology of conservatism has postulated that a perceived threat triggers a singular emotional response: fear.  This postulation is at odds with findings from neuropsychology and neuroscience suggesting that the brain makes multiple affective appraisals of the environment. These distinct emotional states are often experienced simultaneously and can precede cognitive appraisals (Adolphs 2008Zajonc 19801984).

Hence, in contrast to this well-established model of conservatism, affective intelligence theory (AIT; see Marcus, Neuman & MacKuen 2000) proposes that multiple affective appraisals of a stimulus or situation enable two approaches to judgment in the political domain — for liberals and conservatives alike. The first default approach leads to partisan reliance on habituated practices and traditions.  The second represents a departure from the default and produces nonpartisan deliberation by setting aside traditions and heuristic-based thinking.  The result is more careful deliberation and increased openness to novel solutions.

Importantly, AIT contends that the experience of anger is concerned with noxious threats and will produce default reasoning and a reliance on habit and tradition.  Fear, on the other hand, should lead to a departure from the default mode resulting in a wide-ranging search for novel information and an openness to new possibilities.

Thus, increased authoritarianism in response to threat should depend on whether the threatening stimulus (or stimuli) generate anger, not fear. In three studies, Marcus et al. tested — and found support for — this theory.


For all 3 studies Marcus et al. (2019, p. 119) specified two hypotheses:

  1. Anger serves to launch defenses against challenges to extant core norms by those who threaten them. Given the importance of those norms, people will often disregard the specific benefits or costs of the actions deployed to defend them (i.e. utility, in the language of classic economic theory). Anger will be most potent among those who are most attached to those core norms.
  2. Fear will lead people to disregard extant practices (practices that will likely be ill-suited for novel circumstances).  Anxious people will thus be able to turn away from the habits and assumptions they would otherwise rely on — and concern themselves with securing the best possible outcome.


Study 1 examined how the emotions of anger and fear differentially mobilized right-wing voters in the 2014 European parliamentary election in France — using data from the Making Electoral Democracy Work survey conducted in two different regions in France.  This survey generated a diverse and representative set of responses on a variety of demographic factors.  Respondents reported their vote choice. They also responded to two 5-point Likert scale items — one item assessed fear, the other, anger.  Both employed a “not at all…” to “quite …” scale.  Probability of voting for the Front National party, France’s far-right party, served as the dependent variable.

Study 2 was conducted shortly after the Charlie Hedbo attacks in January of 2015.  Marcus et al. were able to embed a measure of authoritarianism in the CEVIPOF barometer of political confidence (for more details see, Vasilopoulos et al. 2018).  Fear and anger were also assessed.  A total of 1,524 respondents were surveyed in two waves.  The first wave was conducted before the Charlie Hedbo attacks, the second wave was conducted three weeks after the attacks.  The adoption or rejection of authoritarian policies served as the dependent variable.

Study 3 employed data obtained from the first two waves of the French Election Study, which conducted a panel roughly every month between November 2015 and the June 2017 legislative elections.  Wave 1 was conducted from November 20 to November 29, 2015.  Wave 2 was conducted form January 22 to February 3, 2016.  A total of 20,460 respondents participated in both waves.  Vote choice in the first round of elections served as the dependent variable and was measured during wave 2.

In Study 3, all of the independent variables were measured during wave 1.  Fear and anger were measured by asking respondents the extent to which they felt certain emotions, on a 10-point Likert scale ranging from “not at all” to “extremely,” when reflecting on the November 13, 2015 terrorist attacks in France.  Fear was measured with three emotional terms (e.g., anxious) and anger was measured with four (e.g., resentment).  Authoritarianism was measured using Feldman and Stenner’s (1997) child-rearing attitudes scale.


As can be seen below (Figure 1) in Study 1, when French citizens moved from lower to higher levels of fear those on the center-right and the left were more likely to vote for Front National, although the shifts were not statistically significant.  Yet, as anger increased among French citizens so did the probability of those on the far-right voting for Front National.

Figure 1: The effects of fear and voting on ideological identification and the vote for the Front National, 2014 Source: Making Electoral Democracy Work project (MEDW; see Figure 3, Marcus et al. 2019, p. 122).
In Study 2, which took place shortly after the Charlie Hedbo terrorist attacks, the French reported being fearful and angry, although they reported higher levels of anger.  Increased levels of fear increased the probability that those on the left supported authoritarian policies.  In contrast, as anger increased, support for authoritarianism was evident among those on the right.

As can be seen below (Figure 2), in Study 3 both hypotheses were supported.  Fear (left-hand panel) reduced voting for the National Front among those on the center-right, and even more so among those on the far-right.  Anger (right-hand panel), in contrast increased support for the National Front, in particular among those on the far-right.  The same pattern emerged for authoritarian attitudes.

Figure 2: Indirect effects of fear and anger on left-right predispositions and voting for Front National (top) and authoritarianism and voting for Front National (bottom).  Source: Figures 7 and 8, Marcus et al. 2019, p. 127).


Marcus et al. (2019) concluded that the pattern of results obtained in these data suggest that it is anger in response to a noxious threat that drives those predisposed to adopt conservative and far-right attitudes to do so.  Contrary to the well-established model associating fear and conservatism, in Studies 1 and 2 fear had no effect on endorsement of authoritarian attitudes and the probability of voting for Front National, among those on the center-right or far-right.  In Study 3, fear reduced the endorsement of authoritarian attitudes and the probability of voting for Front National, among those on the center-right or far-right.  Thus, the well-established model linking fear and conservatism may be misspecified.

This misspecification could have profound implications.  By treating threat and fear as equivalent discourse, politicians, journalists, pundits, and scholars ignore and obscure the role anger plays in politics.  Attempts to address those who are politically angry via a false understanding that they are not angry but fearful — this is likely to further exacerbate their anger. Furthermore, policy proposals that attempt to address fears are not likely to be viewed as effective because many voters may not be fearful, but are instead angry.

Indeed, as Marcus et al. (2019, p. 131) noted:

“But conventional wisdom here, as it often does elsewhere, keeps us blind to that which lies outside its vision of fear as the essential and singular cause of our discontent.  Anger has long been apparent in the reactions to various progressive projects, such as continued population movement from rural areas to urban, increasing cosmopolitan patterns (cultural exchange, interest in style, fashion and celebrity, as well as trade, travel, and so on), acceptance of previously disparaged groups (such as single women, women in the workplace, gay marriage, atheists and more). ”

Later, Marcus et al. (2019, p. 131) contend that “a political order that cannot address competing claims of angry groups will be more vulnerable to leaders proposing authoritarian governance.  This will be especially so for any regime that does not accurately recognize the angry forces at play.”

Why Are These Findings Important?

Global concerns over rising illiberalism and increased support for authoritarian leaders and policies is growing (see e.g., hereherehere, and here).  Many scholars and pundits however appear to have adopted a narrow lens through which they are interpreting these events.  Attempts to assuage fears over shifting demographics and cultural values appear likely to exacerbate the anger these shifts are likely producing.  The exacerbation of anger may further polarize political parties and voters.  It may also lead people to see the political order in a society as ineffectual, possibly in a state of disarray.  Anger might then lead to a movement to oust the ineffectual elites from power, likely in favor of a demagogue who promises strong leadership and clear solutions.

More narrowly, these results are the latest in a quickly growing body of literature (e.g., Brandt et al. 2014Conway et al. 2016Malka, Lelkes & Holzer 2018Malka et al. 2014) in political psychology that challenges the foundations of the well-established ‘rigidity of the right’ model linking fear in response to perceived threat to political conservatism.  The present findings therefore help broaden our understanding of the psychological underpinnings of political ideology and political behavior — and should prompt research further delineating the different responses that fear and anger, once evoked, can prompt in the political domain.


Marcus, George et al. (2019). “Applying the theory of affective intelligence to support for authoritarian policies and parties.” Advances in Political Psychology 40 (S1): 109-39.