Persuasion in political science and neuroscience
A political scientist take on the neuroscience and cognitive science literature
More recently, political science has studied empirically if, how, or when providing more pieces of information to the public enhances the collective decision process in a democracy. And, naturally, the fake news phenomena has put into question the ability of democracy to produce good decisions if enough voters are ill-informed about the true state of the world.
Unsurprisingly, cognitive science has studied how persuasion occurs and helps to change the behavior of people. This literature, however, has not been cited and used by most political scientists1 except perhaps for the behavioral science literature associated with the work of Tversky and Kahneman. In this note, I will review the neuroscience literature on persuasion, and how their findings and experimental design can help the political science literature overcome the differences found in survey and field experiments studies.
Informed voters and democracy
It is an old idea in political science that democracy is, in general, a good (sometimes the best) way to aggregate disperse information to generate a common good or choose the best alternative (Condorcet, 1785, Galton, 1907).
Modernization theory posited that what caused a country to became a stable and functioning democracy was the economic development of a society (Lipset, 1959, Przeworski, & Limongi, 1996). In this reading, democracy was not suitable for less developed countries, either because would-be voters were not well informed and educated enough to produce good results via democracy or because they would not demand democracy and fight against dictators to establish it. For me, the important point here is that level of education and/or information acquired by citizens would impact how democracy worked (and eventually could make ti descend back to dictatorship).
Political science studies on persuasion are, in general, grounded on this belief that democracy works best when citizens are well informed. Sometimes assumed as true, sometimes put to test, this belief orients most of political science work on persuasion. As a result, a lot of emphases is put on assessing if providing information to citizens leads to change in attitudes, vote behavior, and political and policy outcomes.
Studies on how information changes attitudes have a variety of empirical evidence. Deliberative polls show that if voters engage in discussion about an issue for a relatively short amount of time, it can change their minds about it (Fishkin, 1997; Luskin et. al., 2002; Althaus, 2003). However, it is not easy to disentangle what is the effect of deliberation and of information.
Experiments in which pieces of information about candidates, parties, or issues are provided to voters is a more direct test of the theory that information matters. Assuming that corruption is something that voters would prefer to avoid to happen, for example, several studies have investigated how providing pieces of information change the attitude and behavior of voters about this issue.
Studies about the effect of providing corruption information on voter behavior have mixed results, as a meta-analysis by Incerti (2020) shows. Field experiments found null results, while survey experiments found large and negative effects. As a consequence, there is still much debate about what is the true effect of informing voters about corruption, the role of moderators, and of design of experiments.
Neuroscience 101
The frontal cortex is the most recently evolved brain region and the last to fully mature, going until mid-twenties to be fully developed. The prefrontal cortex (PFC) is the newest part of the frontal cortex. Earlier views thought that only primates have a distinct PFC, but there is still some debates about this (Uylings et. al., 2003).
The PFC is central to executive function (Teffer, & Semendeferi, 2012, Uylings et. al., 2003). According to Sapolsky (2017), the PFC “is essential for categorical thinking, for organizing and thinking about bits of information with different labels” (p. 48). As a result, the prefrontal cortex (PFC) is a very important region for social cognition in humans (Carlén, 2017).
Executive function is related to a distinction “between routine (or ‘automatic’) and non-routine (or ‘controlled’) processing” (Gilbert, & Burgess, 2008). Non-routine processing requires cognitive control in situations in which there is no clear stimulus-response association, for example when trying to avoid error in a novel task, planning for the future, etc.
As a result, the PFC is always involved in social cognition investigations by neuroscientists. However it does not mean that other areas of the brain are not important to executive functions. Higher-level functions involved in executive function control more lower-level ones (such as speech, motor, etc.), but are also influenced by them.
The brain is complicated, and the frontal cortex, more so. This means that there are divisions and subdivisions, with different functions and that are not completely understood. In any case, neuroscientists are currently investigating how subdivisions of the brain and the PFC are related to persuasion and understanding all terms and parts of the brain is important to understand the theories and studies. To keep things easy, a quick primer for direction in the brain, since they are used a lot in its subdivisions. There are three dimensions, organized in pairs. Dorsal versus ventral: dorsal is the top of the brain, ventral is the bottom. Medial versus lateral: medial is the part of the brain close to the middle, and lateral is the region far from the middle. Last, but not least, anterior/posterior: anterior is in the front, and posterior is in the back of the brain.
So, what have we learned in neuroscience about the brain and persuasion?
Most studies of persuasion in neuroscience use functional Magnetic Resonance Image (fMRI) to measure brain activity in the so-called Regions of Interest (ROI). ROI are regions in the brain that are believed to be involved in the mechanism manipulated in an experiment and are chose prior to a experiment to be performed.
Several studies have found that activation of the medial prefrontal cortex (mPFC) in persuasion experiments is predictive of downstream behavior (Falk et al., 2010, 2011). The MPFC and its subdivisions are, thus, the main ROI in persuasion studies.
De la Vega et al. (2016), in a meta-analysis, found that there are at least two processes that can be triggered in persuasion studies and that are predictive of behavior change: settings with a pattern of dorsomedial prefrontal cortex (dmPFC) activation are associated with social cognition, such as the theory of mind (theorizing about internal mental processes and feeling of others).
On the other hand, activation of the ventromedial prefrontal cortex (vmPFC) is associated with affective processing, including valuation reward, empathy, and fear processing. The vmPFC is interconnected with the limbic system and, thus, is related to the impact of emotion on decision making. According to Sapolsky (2017), It activates, for instance, “if the person you’re rooting for wins a game”(p. 55). It is easy to connect the dots here, right? Studies have found that the vmPFC is activated in some settings of a persuasion experiment and is also activated when youŕe rooting for someone who wins something. It is not surprising that political science studies have found partisanship a moderator on the effect of information on political behavior (Anduiza et. al. 2013; Krishna, & Sokolova, 2017).
We can speculate that disinformation and correction are types of messages that activate different regions of the brain and, as a result, are not really addressing the same mechanisms by which attitude and behavior can change. This hypothesis should be tested in experiments combining behavioral and neuroimaging studies to see if indeed this is happening. If true, it could help to explain why corrections often fail and to devise how we should fight disinformation.
Most studies in neurosciences, however, have been atheoretical when correlating neuroimages and predicting behavior changes. One important exception is the work of Vezich and co-authors (2016), which tested a mechanism based on prospect theory and another one based on the importance of action plans in an experiment to persuade people to use sunscreen.
Through what process attitude changes do occur? There are two main lines of thought associates with dual process models of persuasion in neuroscience: the elaboration-likelihood model (ELM) introduced by Petty and Cacioppo in 1981 (Kitchen et. al., 2014) and the heuristic-systematic model (HSM)2. I will focus here only on the ELM, which I am (a bit) more familiar with.
The ELM posits that there are two routes by which attitude can change or be formed: from effortful cognitive processing of a message or lesser effortful heuristic of affective processing. If the person is motivated and has the ability to engage in thinking and elaboration about a message then it is a central route of processing. If, on the other hand, the person is not paying attention, the issue is not relevant or does not have the possibility of devoting high cognitive effort, it will take a peripheral route. According to ELM theory, the more central a route, the more likely the more consequential the attitude will be.
If the attitude is connected to a goal the person has, then the link between behavior and attitude will be fortified (Kruglanski et. al., 2015). For instance, if a person wants to do exercises, then a message explaining how she can achieve it is more likely to strengthen the link between attitude (favorable to do exercise) and behavior (do exercise). This is relevant for some results in the political science literature that found that informing voters about corruption decreased turnout. If the citizen had a favorable view of voting, but did not really want pay the cost to go to vote (getting informed etc.), then showing that politicians are corrupt may decrease the favorability of voting and, in turn, decrease turnout.
An attitude that is accessible, i.e., comes easily to mind as opposed to requiring a high cognitive effort, will be harder to change and will be more predictive of behavior.
Stronger attitudes are more predictive of behavior than weaker ones. In a pioneer study, Sample & Warland (1973) showed in a study with college students that those with higher certainty of their attitude could have behavior predicted better than ones with lower certainty. As Ajzen (2012) notes, things that strengthen attitude, such as being a result of a personal experience, will make it more predictive of behavior. Also, if a message reduces a person belief in self-efficacy, i.e, its belief that can attain a given goal by a given behavior, then it can have an impact on behavior, but diminishing the link between attitude and behavior.
Back to political science
Social science has recently studied how disinformation and corrections change (or not) the beliefs and behavior of people. Most of the studies are interested, first and foremost, in assessing if and how providing pieces of information change the belief of citizens and, secondarily, if it triggers a change in behavior. Experiments that try to understand how and when disinformation and corrections can have an effect should take into account what neuroscience studies on persuasion have found. By comparing this literature with cognitive science studies on persuasion, we can design new avenues for research, methods, and hypothesis in both areas.
Survey social science studies that try to asses how information, misinformation, and/or correction change behaviors in general measure intention to change behavior, not an actual change in behavior.
Nyhan et. al (2014) measured three outcomes in an experiment on messages pro-vaccination: how much respondents agreed that vaccine caused autism, about the likelihood of vaccine causing autism, and how likely parents would be to give the MMR vaccine. They tried to assess if the message persuaded people to change minds about a causal link and intention to change behavior (give the vaccine to future children). In a similar study, Nyhan and Reifler (2015) tested the effect of correcting myths about the flu vaccine and they measured if there was a change in the belief about the flu vaccine and in the intent to vaccinate. Botero and co-authors (2015) investigated if and how sources of information on corruption of political candidates change intention to vote.
These studies did not discuss a well-known criticism in the psychology literature about the lack of reliability of what one thinks about its own mental processing (including how persuasive a message is) and the correspondent real behavior change (Nisbett and Wilson’s (1977), Collins et. al., 1988). Hoeken (2001) found that arguments that people rated as the strongest did not correlated with ratings of quality of argument. Wilson et. al. (1998) in particular studied empirically the correlation between perceived and actual persuasiveness and found it to be poorly correlated. This may explain, at least in part, the difference between survey and filed experiments findings among corruption information studies, as found by Incerti (2020).
Field experiments do provide information to voters, but do not advocate explicitly for a change in behavior (say, do not vote for corrupt politicians, or vote in another candidate). The assumption seems to be that, by providing information (like a journalist), it should be a signal strong enough to produce a change in behavior. Banerjee et al. (2011) and De Figueiredo, Hidalgo, and Kasahara (2011) took this approach and they also measured voter behavior (turnout and vote) and found that information had an effect, although not in completely consistent ways.
Anyways, the theory behind the paper is that more trustworthy sources are more effective cues (especially for citizens with a low level of information about political issues) and would lead to more change. It is worth noting that the message did not present a reason why people should not vote for a political with corruption allegations, or how to follow it through. In this setting, only providing information is supposed to persuade people to change behavior, without the need to explain why or how they should do it. In standard experiments about quitting smoking and the use of sunscreen, the neuroscience literature explicitly calls to action.
The lack of an explicit call to action is odd, from a neuroscience point of view, since being explicit about the intended change is central to persuasion studies. On the other hand, Political science studies seem to think that asking people to change behavior is irrelevant when providing information with the intent to change political behavior. It should be tested if providing explicit call to actions make any difference in behavior, and also with fMRI of the brain.
Conclusion
Political science has studied extensively, mainly through survey and field experiments, about the effect of providing information on changing behavior. The results are not always replicated and there is a lack of theoretical progress on what causes information to have a positive or negative effect. Findings and studies in the neuroscience literature can provide new avenues for research, from the design of experiments, theoretical mechanisms and hypotheses to be testes as well as using brain activities to predict change in behavior.
References:
Ajzen, I. (2012). Attitudes and persuasion. In K. Deaux & M. Snyder (Eds.), Oxford Library of Psychology. The Oxford handbook of personality and social psychology (p. 367–393). Oxford University Press. https://doi.org/10.1093/oxfordhb/9780195398991.013.0015
Althaus, S.L., 2003. Collective Preferences in Democratic Politics: Opinion Surveys and the Will of the People. Cambridge University Press, Cambridge, pp. 33e53.
Anduiza, E., Gallego, A., & Muñoz, J. (2013). Turning a blind eye: Experimental evidence of partisan bias in attitudes toward corruption. Comparative Political Studies, 46(12), 1664-1692.
Carlén, M. (2017). What constitutes the prefrontal cortex?. Science, 358(6362), 478-482.
Collins, R. L., Taylor, S. E., Wood, J. V., & Thompson, S. C. (1988). The vividness effect: Elusive or illusory?. Journal of Experimental Social Psychology, 24(1), 1-18.
Condorcet, M. d., 1785. Essai sur l’application de l’analyse à la probabilité des d́ecisions rendues à la pluralité des voix. Paris
Galton, F., March 1907. Vox populi. Nature 75, 450–451.
Gilbert, S. J., & Burgess, P. W. (2008). Executive function. Current Biology, 18(3), R110-R114.
Falk, E. B., Berkman, E. T., Mann, T., Harrison, B., & Lieberman, M. D. (2010). Predicting persuasion-induced behavior change from the brain. The Journal of Neuroscience, 30, 8421-8424.
Falk, E. B., Berkman, E. T., Whalen, D., & Lieberman, M. D. (2011). Neural activity during health messaging predicts reductions in smoking above and beyond self-report. Health Psychology, 30, 177-185.
Fishkin, J. S. (1997). The voice of the people: Public opinion and democracy. Yale university press.
Hoeken, H. (2001). Anecdotal, statistical, and causal evidence: Their perceived and actual persuasiveness. Argumentation, 15(4), 425-437.
Incerti, T. (2020). Corruption information and vote share: A meta-analysis and lessons for experimental design. American Political Science Review, 114(3), 761-774.
Jost, J. T., & Amodio, D. M. (2012). Political ideology as motivated social cognition: Behavioral and neuroscientific evidence. Motivation and Emotion, 36(1), 55-64.
Jost, J. T., Nam, H. H., Amodio, D. M., & Van Bavel, J. J. (2014). Political neuroscience: The beginning of a beautiful friendship. Political Psychology, 35, 3-42.
Kitchen, P. J., Kerr, G., Schultz, D. E., McColl, R., & Pals, H. (2014). The elaboration likelihood model: review, critique and research agenda. European Journal of Marketing.
Krishna, A., & Sokolova, T. (2017). A focus on partisanship: How it impacts voting behaviors and political attitudes. Journal of Consumer Psychology, 27(4), 537-545.
Kruglanski, A. W., Jasko, K., Chernikova, M., Milyavsky, M., Babush, M., Baldner, C., & Pierro, A. (2015). The rocky road from attitudes to behaviors: Charting the goal systemic course of actions. Psychological Review, 122(4), 598.
Lipset, S. M. (1959). Some social requisites of democracy: Economic development and political legitimacy. The American political science review, 53(1), 69-105.
Luskin, R. C., Fishkin, J. S., & Jowell, R. (2002). Considered opinions: Deliberative polling in Britain. British Journal of Political Science, 455-487.
Nisbett, R. E., & Wilson, T. D. (1977). Telling more than we can know: verbal reports on mental processes. Psychological review, 84(3), 231.
Nyhan, B., Reifler, J., Richey, S., & Freed, G. L. (2014). Effective messages in vaccine promotion: a randomized trial. Pediatrics, 133(4), e835-e842.
Nyhan, B., & Reifler, J. (2015). Does correcting myths about the flu vaccine work? An experimental evaluation of the effects of corrective information. Vaccine, 33(3), 459-464.
Sample, J., & Warland, R. (1973). Attitude and prediction of behavior. Social Forces, 51(3), 292-304.
Sapolsky, R. M. (2017). Behave: The biology of humans at our best and worst. Penguin.
Przeworski, A., & Limongi, F. (1996). Modernization: Theories and facts. World Pol., 49, 155.
Teffer, K., & Semendeferi, K. (2012). Human prefrontal cortex: evolution, development, and pathology. Progress in brain research, 195, 191-218.
Wilson, T. D., Houston, C. E., & Meyers, J. M. (1998). Choose your poison: Effects of lay beliefs about mental processes on attitude change. Social Cognition, 16(1), 114-132.
Uylings, H. B., Groenewegen, H. J., & Kolb, B. (2003). Do rats have a prefrontal cortex?. Behavioural brain research, 146(1-2), 3-17.
Vezich, I. S., Katzman, P. L., Ames, D. L., Falk, E. B., & Lieberman, M. D. (2017). Modulating the neural bases of persuasion: why/how, gain/loss, and users/non-users. Social cognitive and affective neuroscience, 12(2), 283-297.
See Jost & Amodio, 2012; Jost, Nam, Amodio, & Van Bavel, 2014 for examples of political neuroscience studies.
See Chen & Chaiken (1999) for a broad review.