A new study published in Scientific Reports reveals a significant and widespread misunderstanding among Americans regarding their fellow citizens’ views on diversity and inclusion. Researchers discovered that people consistently underestimate the extent of support for diversity in the United States. This misperception, according to the study, can negatively impact inclusive behaviors, but can be corrected by simply informing people about the actual level of public support for diversity.
Researchers at the University of Wisconsin-Madison became interested in this topic because they wanted to understand a puzzling contradiction. On one hand, many people express support for diversity and inclusion. On the other hand, discrimination and exclusion remain persistent problems in society. The researchers wondered if part of the problem might stem from inaccurate perceptions of what others believe.
Specifically, they investigated the idea of “pluralistic ignorance,” which is when people incorrectly assume that their own beliefs are not widely shared, even when they actually are. In the context of diversity, this could mean that people who support diversity might mistakenly believe that most others do not share those views. The researchers hypothesized that this misperception could discourage individuals from speaking out in favor of diversity or acting inclusively, as they might assume they are in the minority. They also wanted to explore if correcting this misperception could encourage more positive attitudes and behaviors related to diversity.
To answer these questions, the research team conducted a series of large-scale studies that together involved more than 5,400 participants in their first five survey studies and an additional 724 participants in two experiments.
The first study involved 1,001 adults recruited online through a platform called Prolific, with the goal of having a sample that roughly matched the U.S. population in terms of age, race, and gender. Participants were presented with fifteen statements expressing positive views about diversity and inclusion. For each statement, they were asked to do two things: indicate whether they personally agreed or disagreed, and estimate what percentage of all U.S. adults would agree. The sequence of these two tasks and the order of the statements were randomized across participants to minimize order effects.
Study 2 was a replication of the first study, using a similar sample size of 997 participants recruited through Prolific with the same demographic quotas. In addition to the measures from Study 1, this study also asked participants about how often they talked about diversity issues, the political leaning of their close friends, how much they thought media influenced others compared to themselves, and how they felt minority groups were portrayed in the media.
Study 3 aimed for a more representative sample of the U.S. population by recruiting 2,010 participants through a different platform called Qualtrics. This time, in addition to quotas for age, race, and gender, they also included quotas to ensure a balanced representation of political affiliations, ranging from strongly Republican to strongly Democrat. This study used a shorter set of six diversity statements from the previous studies. Instead of simple agree/disagree responses, participants indicated their level of agreement on a five-point scale from “strongly disagree” to “strongly agree.”
Similarly, when estimating others’ opinions, they estimated the percentage of Americans who would choose each of these five response options. Beyond the core measures, Study 3 included questionnaires assessing the use of conservative and liberal media, exposure to diversity-related media, perceptions of how seriously discrimination is portrayed in the media, trust in media, and biases in media against diversity initiatives and liberal viewpoints. It also measured participants’ likelihood to confront discrimination and engage in inclusive behaviors, as well as their levels of racism and sexism.
Study 4 modified the approach slightly and involved 497 participants from Prolific. This study used ten new diversity-related statements, half of which were supportive of diversity and inclusion, and half of which were critical or not supportive of diversity and inclusion. Participants again rated their own agreement on a five-point scale and estimated the percentage of Americans who would agree with each statement (specifically, those who would “somewhat agree” or “strongly agree”). This allowed researchers to see if the misperception effect was present for both pro-diversity and anti-diversity statements.
Study 5 focused specifically on political groups. It recruited 943 participants from Prolific who had voted for either Donald Trump or Joe Biden in the 2020 presidential election. Using the same ten statements from Study 4 (five pro-diversity, five anti-diversity), participants were randomly assigned to estimate the opinions of either Trump voters or Biden voters. All participants also indicated their own agreement with the statements. This design allowed the researchers to compare how accurately Trump voters and Biden voters perceived the diversity views within their own group and the opposing group.
Across the first five studies, a clear and consistent pattern emerged: Americans tend to significantly underestimate how much their fellow citizens support diversity and inclusion. Studies 1 and 2 revealed a substantial gap of about 27 percentage points between the actual level of agreement with pro-diversity statements in their samples and what participants estimated that level to be in the broader U.S. population.
In other words, while about 82% of participants in these studies agreed with statements like “Racial diversity benefits the country,” participants believed only around 55% of Americans would agree. This notable underestimation suggested a widespread misperception.
Participants not only underestimated agreement with pro-diversity statements but also overestimated agreement with anti-diversity statements. This suggests a skewed perception where individuals believe pro-diversity views are less common and anti-diversity views are more common than they actually are.
Demographic factors like age, gender, race, and region had little impact on the degree of underestimation. “Pluralistic ignorance related to diversity and inclusion is a general phenomenon that cuts across demographic categories,” the researchers wrote. However, individuals who identified as Democrats or as members of marginalized racial or ethnic groups tended to underestimate support slightly less than others.
Furthermore, a stronger tendency to underestimate support was associated with more conservative political views, lower consumption of media focused on diversity and inclusion, less trust in media generally, and higher scores on measures of racism and sexism. The researchers also found that this misperception had consequences: those who underestimated support for diversity were less likely to discuss diversity-related topics, less inclined to behave inclusively, and less willing to confront discrimination when they witnessed it.
In Studies 6 and 7, the focus shifted from measuring existing perceptions to experimentally assessing whether exposure to accurate information could alter them. Study 6 recruited 367 white adults from Prolific who lived in states that voted for Biden in 2020. Study 7 recruited 356 white adults from Prolific who lived in states that voted for Trump in 2020.
Participants in both studies were told they were taking part in two separate studies. The first was presented as a task to assess their ability to identify trends in public opinion. Participants played a “Two Truths and One Lie” game. They were randomly assigned to one of three conditions: a “descriptive condition,” a “dynamic condition,” or a “control condition.” In the descriptive condition, participants were repeatedly shown statistics from real opinion polls indicating that a majority of Americans support diversity and inclusion. In the dynamic condition, they saw statistics suggesting that support for diversity was increasing over time. In the control condition, they saw statistics about unrelated topics like changes during the COVID-19 pandemic.
In each round of the game, participants had to identify which of three statements was false. After each choice, they were shown the correct answers, reinforcing the intended message about public opinion. After this game, participants moved on to what they believed was a separate study, where they were asked to complete questionnaires measuring their pro-diversity attitudes, intentions to behave inclusively, and willingness to confront discrimination. This allowed the researchers to see if exposure to information about widespread diversity support could influence people’s own attitudes and behaviors related to diversity.
The researchers found that participants in the descriptive norm condition, and to a lesser extent those in the dynamic norm condition, showed improvements across several measures related to diversity and inclusion after playing the game. Specifically, they expressed stronger values for diversity in the workplace, reported increased intentions to confront discrimination, and scored higher on an overall measure of inclusiveness.
This positive effect of social norms messaging was observed in both Study 6, which focused on participants from Democratic-leaning states, and Study 7, which focused on participants from Republican-leaning states, suggesting that this type of intervention can be effective across different political contexts.
“When people learn about the widespread support for diversity and inclusion, they change their own attitudes about these issues,” the researchers explained. “After being exposed to actual survey data from national opinion polls, our participants scored higher on a variety of indicators related to diversity and inclusion. In other words, correcting people’s pluralistic ignorance has important beneficial effects.”
But the researchers pointed out some limitations of their work. They acknowledged that people who participate in online studies may not perfectly represent all Americans. Also, they noted that people might sometimes say they support diversity more than they truly do, which could slightly inflate the reported levels of support. However, they also argued that the consistent underestimation of support across different groups and types of diversity statements suggests that the core finding of pluralistic ignorance is robust. Future research, they suggested, could investigate why this misperception is so common.
“It seems that those with favorable attitudes toward diversity have not succeeded in communicating to the wider public their commitment to inclusion and their support for pro-diversity initiatives,” the researchers concluded. “The individuals with less favorable attitudes—a numerical minority—not only are the most pluralistically ignorant, but they also seem to be more vocal and thus have a disproportionate influence on Americans’ perceptions of their fellow citizens’ attitudes.”
The study, “Diversity and inclusion have greater support than most Americans think,” was authored by Naomi Isenberg and Markus Brauer.
