New research published in journal Cyculos This suggests that people who value free speech also tend to have higher levels of racial and ethnic tolerance. These findings provide evidence that the social benefits of protecting freedom of expression extend beyond legal rights and promote broader norms of tolerance and acceptance.
Claudia Williamson Kramer, Free Enterprise’s Probasco Distinguished Chair and professor of economics at the University of Tennessee at Chattanooga, designed the study to explore the intersection of two competing perspectives on the social impact of open communication. On the other side of the debate, critics argue that unrestricted speech allows the spread of hateful rhetoric and misinformation, which can harm marginalized minorities. This perspective suggests that it may be necessary to set limits on expression to prevent social exclusion and psychological distress.
On the other side of the issue, advocates of open expression propose that free speech acts as a safety valve for society. This view suggests that allowing people to voice their prejudices publicly creates opportunities for counterspeech, the act of challenging and correcting hateful ideas through open discussion. By making prejudice public, free speech environments tend to reduce prejudice over time through social learning and peaceful dissent.
Kramer initiated this research to address specific gaps in the existing scientific literature. Previous research has focused primarily on the broader effects of press freedom and legal protection of speech at the national level. Kramer wanted to understand how holding free speech as a personal, personal value affects people’s attitudes toward people of different backgrounds.
“Free speech is under real pressure right now, not just in authoritarian states,” Kramer said. “Established democracies prosecute people for what they post online. The usual argument for restricting speech is that it protects vulnerable people from hate. I wanted to test whether that trade-off actually holds at the level of individual attitudes.”
She added that this particular dynamic requires closer investigation. “Are people who value free speech less or more tolerant of others?” Kramer asked. “This issue has been studied at the level of national institutions, but not at the level of individual values. I aimed to fill that gap.”
The researchers hypothesized three potential mechanisms linking these two specific values. First, those who support freedom of expression may believe in the power of open discussion to expose and challenge false stereotypes. Second, a focus on free speech typically includes supporting independent news outlets, which can amplify marginalized voices and expose discrimination. Third, prioritizing free speech may simply reflect a broader cultural worldview rooted in classical liberal ideals that emphasize individual autonomy and equal treatment.
To test these ideas, Kramer analyzed data from the Integrated Values Survey. This large database combines survey responses from 609,552 individuals across 115 countries. The data spanned seven different survey waves conducted between 1981 and 2022. This rich dataset allowed the authors to identify individual differences while accounting for the different political and institutional environments of each specific country.
In this part of the study, Kramer measured free speech priorities by looking at how respondents ranked four competing national goals. Participants had to choose between defending freedom of speech, maintaining domestic order, giving citizens more say in government decisions, and fighting rising prices. Those who selected the protection of free speech as their top priority were classified as valuing freedom of expression highly.
To measure racial tolerance, the survey asked participants to name groups of people they would not want as neighbors. If a respondent did not select a person of a different race from the options provided, that person was mathematically coded as having racially tolerant attitudes. Overall, about 84 percent of global survey respondents expressed racially tolerant views, and about 12 percent said protecting free speech was their absolute top priority.
These results indicate that prioritizing free speech is positively related to racial tolerance. “People who prioritize free speech are more tolerant, not racially tolerant,” Kramer told SciPost. “Our data shows that people who list protecting free speech as a top national priority are 2.3 percentage points more likely to accept a neighbor of a different race.”
Although this proportion may seem modest at first, it represents a large change when applied to a larger population. “The individual effect seems small, but it’s worth explaining why,” Kramer says. “My analysis compares people from the same country and age, which removes the large differences in tolerance that exist between countries.”
She pointed out that mathematical controls leave only local differences between individuals. “What remains is the difference between the two neighbors, and a 2.3 point change is meaningful at that level,” Kramer said. “For comparison, this is nearly half the acceptable difference between people with low and secondary education, and is in the same range as the effects of well-known studies on contact and media exposure.”
This statistical difference has real-world implications. “It’s important to scale up,” she explained. “In a country of 10 million adults, going from no one prioritizing free speech to half the population would mean approximately 115,000 more people expressing racial tolerance.”
This pattern remained stable even after the authors adjusted the statistical model to account for a wide range of individual characteristics. These specific controls included the respondent’s age, gender, marital status, education level, employment status, income, urban or rural residence, religious practices, and political ideology. Level of education was found to be a particularly strong predictor of tolerance, with respondents with higher levels of education having approximately 6.4 percentage points higher rates of tolerance than respondents with the least education level. However, even accounting for this strong educational effect, the mathematical association between respect for free speech and holding tolerant views remained significant.
Kramer also looked at other forms of social acceptance within the survey data. She found that respondents who prioritized free speech were significantly more tolerant of several other marginalized groups. “The same pattern applies to tolerance toward immigrants, religious minorities, Muslims, and Jews,” Kramer said. “Those who value open expression and the tendency to accept people who are different from themselves tend to act together. So we should be wary of assuming that restricting speech will protect minorities, which could erode the very ethos that underpins tolerance.”
Interestingly, the authors found a negative association between free speech values and tolerance of right-wing political extremists. “The most striking result was where the pattern broke,” Kramer said. “Those who value free speech are tolerant of nearly every group I test, but significantly less tolerant of right-wing extremists.”
This detail provides evidence that respondents are not simply providing socially desirable answers to every question. “That’s the opposite of what you would expect if people simply gave socially approved answers and got the results,” she added. “This is consistent with Popper’s paradox of tolerance,” she explained, referring to the philosophical concept that a tolerant society must sometimes reject the tolerance of those who seek to destroy it. “A commitment to open debate does not mean accepting those who seek to shut it down.”
The research options also revealed other specific associations. “Another surprise was the specificity,” Kramer said. “Of the four national priorities people could choose from, only freedom of speech showed a significant positive relationship with tolerance. Choosing order or lower prices was counterproductive.”
To ensure that these findings were not limited to one particular data set, Kramer conducted a second analysis using an Afrobarometer survey. The survey includes responses from 48,206 people in 33 countries in sub-Saharan and North Africa collected between 2014 and 2015. Research in Africa reveals a unique environment characterized by high ethnic diversity and a wide variety of democratic institutions. In this context, ethnic divisions are often the main boundaries along which social conflicts occur, making ethnic tolerance a particularly important measure.
In this second study, the freedom of expression measure focused on media freedom. Participants were asked whether they agreed that the media should have the right to publish their views without government control. The outcome measure focused on ethnic tolerance, asking respondents whether they object to having neighbors of different ethnicities.
Survey data for Africa mirrored the main findings. Those who supported media freedom were 0.9 percentage points more likely to have ethnically tolerant attitudes. This provided supporting evidence from a completely different geographical and cultural context, using slightly different measures of both speech value and tolerance.
“The results hold true in very different settings,” Kramer said. “The main analysis covered more than 600,000 people in 115 countries over 40 years. We then matched it with the Afrobarometer, which covered 33 African countries with different measures of freedom of speech and tolerance, and the association still emerged.” She noted that the relationships shown across such a diverse range of data provide strong evidence of a true connection.
The authors note some limitations. Because this study relies on observational data, it can only show an association between two variables, not a direct cause-and-effect relationship. “This is a correlational study,” Kramer explained. “I use within-country comparisons, omit variable ranges, and use a long list of controls. The result remains with all of them, but you can’t call it causal.”
She acknowledged that human values are difficult to disentangle. “It could lead to more tolerant people valuing free speech instead of the other way around,” Kramer said. “I also measure what people say they prioritize, which is an imperfect proxy for what they really value.”
Additionally, she wants to ensure the public understands the specific scope of the findings. “The last misconception I want to avoid is that speech has no cost,” Kramer stressed. “That’s right. Propaganda fueled the rise of the Nazis and the Rwandan genocide. I discuss both in this paper.”
She declared that her research highlights the far-reaching cultural benefits of freedom of expression. “The bottom line is that not everything will work,” Kramer said. “Cultures that value open expression tend to be more tolerant, and the results for right-wing extremists show that these people are not indiscriminate about it.”
Future studies may investigate these dynamics using different experimental designs to establish cause and effect. “The obvious next step is to determine causality,” Kramer said. “We want to use a setting in which free speech conditions change for reasons unrelated to people’s previous attitudes. That way we can more clearly distinguish between causes and correlations.”
Exploring exactly how open discourse can successfully counter prejudice also remains an active area of interest for social scientists. “I’m also interested in the mechanism,” Kramer said. “This paper offers three options: opposition to speech, support for independent media, and broader liberal culture, but it’s too early to say which one will work. Sorting that out is the next challenge.”
The study, “Do Free Speech Values Influence Norms of Tolerance? Evidence from Personal Preferences,” was authored by Claudia Williamson Kramer.
