“Vote with your gut, not with your brain!” declared Beppe Grillo, founder of Italy’s Five Star Movement. Fear, anger, resentment, contempt … emotions seem to pervade our democracies, shaping our political debates both on the supply side (policymakers) and demand side (voters), as notably documented by a recent analysis of political tweets, congressional, and campaign speeches (Algan et al. 2025). 

At the same time, one could expect that, in the digital age, the abundance of readily available information could lead to more reasoned opinions and evidence-based political choices. As Barack Obama put it: “If you give people good information, their instincts are good, and they will make good decisions. … I am a big believer in reason and facts and evidence …” (Sandel 2020). Emotion or cognition – which has the greater influence on our political opinions?

In a recent paper (Manzoni et al. 2026), we study how emotions triggered by media content influence policy views and shape people’s relationship to facts, that is, the way they process factual information.

To explain how people form their policy views, economists have examined a variety of factors related to factual knowledge, understanding of the issue, self-interest or sociotropic concerns, or partisanship (Stantcheva 2020, 2021, 2022). While these approaches describe cold, slow cognitive processes, the role of hot visceral factors in shaping policy views has received little attention in the economic literature. This is somewhat surprising given the growing body of research in political science, psychology, and sociology that document the increasing role of collective emotions in the rise of populism, in the polarisation of politics, and in the spread of fake news in social media (Rosanvallon 2020, Dubet 2019, Illouz 2022, Nabi and Myrick 2023, Webster and Albertson 2022).

A simple framework

The standard framework for thinking about policy views can be summarised in two steps (Figure 1). First, individuals receive information and update their beliefs about relevant facts. Second, these beliefs feed into preferences over policies. This ‘belief based’ view underlies much of the empirical work on information provision (Alesina et al. 2020, 2023, Haaland et al. 2023). We extend this framework by explicitly introducing emotions triggered by political or media communication. 

Figure 1 Conceptual framework

Unlike standard information, emotional stimuli do not necessarily convey informative content. Instead, they activate affective states that may influence policy views directly or alter the way factual information is processed and used.

This yields three questions:

• Do media induced emotions affect policy views independently of factual beliefs?
• Do emotions disrupt or bias the updating of beliefs?
• Do emotions weaken the link between factual beliefs and policy preferences?

Evidence from a large survey experiment

We address these questions by examining policy views on immigration, a contentious issue characterised by widespread misperceptions and emotionally charged media coverage. Using a large scale survey experiment conducted in Italy, we randomly exposed 12,000 participants to different types of content: news articles reporting crimes committed by immigrants varying in emotional intensity, from minor incidents (petty theft) to sexual assault or murder; official statistics on immigration and crime; or a combination of both.

Once statistics are provided, the news stories add no factual information other than the description of a specific realisation of the statistical distribution of crimes. This experimental design allows us to isolate the causal impact of emotions independently of any effect on factual beliefs.

Our results can be summarised in three main findings.

First, emotions have a direct causal effect on policy views.

Highly emotional content – particularly accounts of violent crime – strongly influences attitudes by increasing the support for restrictive immigration policies. Less emotional news has much smaller attitudinal effects, even conditional on providing statistical information that holds posterior factual beliefs constant across different news. This demonstrates that emotional reactions, rather than differences in belief updating, drive the attitudinal response.

Second, emotions do not prevent factual learning, but they slightly disrupt it.

Providing statistical information leads respondents to correct their factual beliefs toward the truth, even when this information is combined with emotional news. However, learning is modestly weaker when respondents are emotionally triggered. Hence, emotions seem to consume cognitive resources, reducing – but not eliminating – factual learning.

Third, emotions make policy views resistant to factual correction.

When presented alone, statistical information, by correcting misperceptions, helps reduce hostility toward immigration, consistent with previous evidence (Grigorieff et al. 2020). However, when combined with emotional news, this effect vanishes. Respondents hold more accurate knowledge, yet express policy views that are just as hostile to immigration as those of respondents exposed only to emotional news (Figure 2). Once emotions are triggered, statistical facts no longer matter for the formation of policy views.

Figure 2 Effect of statistical information and emotional news on attitudes towards immigration

Notes: Estimates and 90% confidence intervals of the treatment effects relative to the control group (neutral news). The dependent variables are: support for government intervention to strongly reduce immigration, willingness to sign a petition to increase the number of residence permits for immigrants, and opinion whether immigration makes Italy a better place to live.

Why emotions dominate facts

These findings help explain why information often fails to change policy views. Emotions shape attention, memory, and evaluation. Emotional stimuli are more vivid, more easily recalled, and more heavily weighted in judgement than abstract statistics (Brosch 2013, Graeber et al. 2024). Neuroscience research further suggests that emotional responses can interrupt and redirect cognitive processing toward high-priority concerns (Armony et al. 1995, 1997, De Becker 1997). As emotions can redirect attention, they may act as a distortive lens through which information is processed and assimilated.

Importantly, this mechanism is distinct from motivated reasoning or ideological filtering. We find little evidence that emotional news polarises individuals by political orientation. Emotional reactions influence people broadly, not only those with strong priors. Immigration is not unique in this respect. The same interplay between facts and emotions is likely present in debates on inequality, climate change, or public health – domains where emotionally charged narratives coexist with abundant statistical evidence.

Policy implications

Our results have important implications. First, they help explain why fact-checking efforts and evidence-based arguments often struggle to influence public debate, particularly when it comes to countering misinformation or emotionally driven populist rhetoric. Second, the media coverage of controversial political issues, such as inequality or immigration, often combines factual data with emotional storytelling. Yet, the latter does not simply inform: it shapes opinions in a way that makes them resistant to factual correction. This helps explain why debates remain deeply polarised, even when reliable information is widely available.

Concluding remarks

A key lesson from this research is simple but uncomfortable. Emotions are not merely noise in the political arena. They are a central force that directly influences policy views and shapes how factual knowledge translates into preferences. This raises a difficult challenge: if facts are not enough, what can be done? While offering an answer to this question is beyond the scope of our paper, our results suggest that effective communication of factual information should engage both cognitive and emotional dimensions, particularly by contextualising data and embedding it in narratives capable of resonating on an emotional level.

References

Alesina, A, A Miano, and S Stantcheva (2020), “The polarization of reality”, AEA Papers and Proceedings 110: 324–28.

Alesina, A, A Miano, and S Stantcheva (2023), “Immigration and redistribution”, Review of Economic Studies 90(1): 1–39.

Algan, Y, E Davoine, T Renault, and S Stantcheva (2025), “Emotions and policy views”, Harvard University working paper.

Armony, J L, D Servan-Schreiber, J D Cohen, and J E LeDoux (1995), “An anatomically constrained neural network model of fear conditioning”, Behavioral Neuroscience 109(2): 246.

Armony, J L, D Servan-Schreiber, J D Cohen, and J E LeDoux (1997), “Computational modeling of emotion: Explorations through the anatomy and physiology of fear conditioning”, Trends in Cognitive Sciences 1(1): 28–34.

De Becker, G, and T Stechschulte (1997), “The gift of fear: Survival signals that protect us from violence”.

Brosch, T, K R Scherer, D Grandjean, and D Sander (2013), “The impact of emotion on perception, attention, memory, and decision-making”, Swiss Medical Weekly 143: w13786.

Dubet, F (2019), Le temps des passions tristes, Seuil.

Graeber, T, C Roth, and F Zimmermann (2024), “Stories, statistics, and memory”, Quarterly Journal of Economics 139(4): 2181–225.

Grigorieff A, C Roth, and D Ubfal (2020), “Does information change attitudes toward immigrants?”, Demography 57: 1117–43.

Haaland, I, C Roth, and J Wohlfart (2023), “Designing information provision experiments”, Journal of Economic Literature 61(1): 3–40.

Illouz, E (2022), Les émotions contre la démocratie, Premier parallèle.

Manzoni, E, E Murard, S Quercia, and S Tonini (2026), “Emotions, beliefs, and policy views”, Journal of the European Economic Association, forthcoming.

Nabi, R L, and J G Myrick (eds) (2023), Emotions in the digital world: Exploring affective experience and expression in online interactions, Oxford University Press.

Rosanvallon, P (2020), Le siècle du populisme: Histoire, théorie, critique, Seuil.

Sandel, M J (2020), The tyranny of merit: What’s become of the common good?, Penguin UK.

Stantcheva, S (2020), “Understanding economic policies: What do people know and learn?”, Quarterly Journal of Economics 135(4): 2309–69.

Stantcheva, S (2021), “Understanding tax policy: How do people reason?”, Quarterly Journal of Economics 136(4): 2309–69.

Webster, S W, and B Albertson (2022), “Emotion and politics: Noncognitive psychological biases in public opinion”, Annual Review of Political Science 25: 401–18.