Massive study upends our understanding of the psychology of collectivism

A global study of more than 12,000 people across 100 cultures suggests that standard psychology surveys have struggled to measure collectivism because they confuse obligation to close family and friends with general friendliness toward everyone. By redefining collectivism as “responsibilism,” which focuses on binding duties in specific close relationships rather than warm feelings toward people in general, researchers identified large cultural differences between East Asian and Western societies that earlier scales missed. The findings, published in Nature Human Behaviour, also indicate that this duty-based measure aligns closely with real-world behaviors ranging from multigenerational living arrangements to blood donation patterns.

In psychology and across the social sciences, collectivism refers to a cultural system in which people view themselves as interconnected with others and prioritize social relationships and group obligations over personal desires. It is commonly contrasted with individualism, a cultural orientation that emphasizes personal independence, autonomy, and looking after oneself. When a wave of self-report questionnaires appeared in the 1990s, scholars of East Asia in fields such as history, anthropology, and political science widely expected East Asian societies to score higher on collectivism than Western nations.

Once those questionnaires were administered around the world, however, the results frequently contradicted expert expectations. Earlier reviews of the literature, including an analysis of 76 separate studies, found that standard collectivism surveys were no better than a coin flip at detecting expected differences between East Asian and Western participants, and one meta-analysis even concluded that Japan was less collectivistic than the United States. Some researchers argued that the popular view of Japanese collectivism was simply an inaccurate stereotype. Many others concluded that self-report surveys were inherently flawed for cross-cultural comparisons, suggesting that respondents in different countries use rating scales differently or interpret abstract words like “independence” in incompatible ways.

The new research, led by Thomas Talhelm, the Robert B. Cialdini Endowed Professor of Social Influence at Arizona State University, alongside Liuqing Wei of Hubei University and Alexander S. English of Wenzhou-Kean University, proposes that the flaw lies in how psychologists define collectivism rather than in the survey method itself. According to the authors, classic questionnaires mixed together two opposing ideas: “warm fuzzy” collectivism and what the team calls “responsibilism.” Warm fuzzy collectivism describes feeling positive emotions toward people in general, enjoying cooperation, and extending friendliness broadly without drawing a sharp line between close family and strangers.

“Researchers created scales to measure collectivism, but they kept finding weird results,” Talhelm told PsyPost. “Were places like the U.S. and U.K. really more collectivistic than cultures like China and Japan? Researchers came up with all sorts of sophisticated methodological explanations, but when I looked at the actual survey items, it seemed clear to me. There were ‘warm fuzzy’ items right there in plain sight, and collectivism isn’t about good feelings!”

Free daily newsletter

Responsibilism, by contrast, sets aside warm feelings and focuses on behavioral duties, social roles, and binding obligations within specific, close relationships, paired with weaker obligations toward strangers or acquaintances. In the new study, Talhelm and his colleagues argue that when collectivism scales include warm fuzzy statements like “To me, pleasure is spending time with others,” Western respondents tend to agree more than East Asian respondents, canceling out the items that actually measure interpersonal duty.

To test whether experts still conflate generalized warmth with collectivism, the researchers first surveyed 40 scholars who had published articles in the Journal of Cross-Cultural Psychology. The scholars were shown 14 warm fuzzy statements and one duty-based statement from classic collectivism scales and were asked to guess whether survey respondents agreed with each item more in East Asia, more in the United States, or about equally. The researchers then compared the scholars’ predictions against actual survey responses from 195 participants in China and the United States.

Many of the scholars held warm fuzzy expectations that the survey data did not bear out. For example, 83% predicted that people in China would agree more with the statement “My relationships with others are more important than my own accomplishments,” but American participants agreed more. Similarly, 78% guessed that Americans would agree more with the competitive statement “Winning is everything,” when Chinese participants did.

Across all 14 warm fuzzy items, the most popular prediction matched the actual data only twice. The scholars’ intuitions about duty were far more accurate: 95% correctly predicted that East Asian participants would agree more with “We should keep our aging parents with us at home.”

Next, the authors built a 14-item responsibilism scale, plus a five-item short version that closely tracked the full scale. The questionnaire was designed around concrete, everyday scenarios and specific relationships rather than vague references to “others” or “groups.” The team adapted three items from a widely used 1995 collectivism scale and wrote new items that contrasted obligations to close friends and relatives with openness toward strangers or unfamiliar group members.

In one scenario, participants imagined getting a 60% grade on a midterm exam and being harshly told by a good friend, “It’s your fault. You need to study harder,” then rated whether they would look for more supportive friends. In another scenario, participants rated whether they would feel comfortable telling an unfamiliar coworker from a different department about a major sales account they were developing. The researchers expected people in responsibilistic cultures to stand by the critical friend, since close relationships are meant to last even when they are unpleasant. They also expected those people to be warier of sharing with the unfamiliar coworker, since belonging to the same company is not enough to earn trust.

One question produced a result that challenged Talhelm’s own cultural assumptions. “I think one of the most interesting findings that cut against my expectations was a question about lending money to a friend who asked for it,” Talhelm recalled. “I expected people in responsibilistic cultures to say ‘yes,’ but they were actually less likely to agree. So I followed up with people from responsibilistic cultures.”

“They had plenty of stories to tell me about giving money to friends or family, but they usually didn’t involve directly asking for it,” he continued. “How American of me! I assumed people would ask directly. But the stories people told me clued me in that that’s often not how it plays out.”

The researchers recruited 16,237 adults, mostly college students, so that participants would be similar in age, education, and stage of life across countries. After removing people who failed data-quality checks or were not born and raised in the local culture, the final sample included 12,247 participants in 100 cultures that together cover 90% of the world’s population. Country scores were nearly identical when everyone was included.

To keep English idioms from distorting the questions, the team wrote the scale in English and Chinese at the same time, and local collaborators then translated it into 45 languages. The main study was not preregistered, but the materials, data, and analysis code are publicly available on the Open Science Framework. The team also shares its global map, data, and survey on a responsibilism website.

The new scale picked up large differences between East Asian and Western cultures, the kind of gap that earlier scales had repeatedly failed to find. Regions in Africa, the Middle East, and several parts of Asia scored highest, Latin America and the Caribbean fell in the middle, and North America, Europe, and Australia scored lowest.

“Differences between how people answer these questions in Western cultures versus cultures in East Asia are over a full standard deviation different!” Talhelm explained. “The effect size is in the same ballpark as the difference in height between men and women. Now, that still leaves room for exceptional cases–a woman taller than most men and an American who scores more responsibilistic than the average in China. But these differences are large enough that people should notice them when they travel across cultures.”

These differences held up when the researchers accounted for participants’ age, gender, and socioeconomic status. To rule out translation as the cause, the team also gave the English version to participants in Hong Kong, where English is widely used in education, and still found a large gap with the United States. An independent team’s preregistered replication likewise found that participants in Singapore scored higher than participants at Western sites.

“If people can walk away with the understanding that there are these two competing visions of collectivism–warm fuzzy versus responsibilism, I’ll be happy,” Talhelm said. “Remember that the warm fuzzies are NOT good descriptions of cultures that people think of as collectivistic, like China, Turkey, or Ghana. But responsibilism (concrete behavioral duties in close relationships) does describe these cultures well.”

Scores on the new scale had little relationship with scores on eight of the most widely cited older collectivism scales. On average, the relationship was slightly negative, suggesting that responsibilism captures something different. One of the strongest overlaps was with part of the 1995 scale that supplied three of the new items.

The researchers then checked how well the new scale and the five most-cited older scales matched real-world indicators of close family ties and treatment of strangers, drawn from census data, field experiments, and other international datasets. Responsibilism scores lined up strongly with these indicators. The older scales did not, and on average they pointed slightly in the wrong direction: countries scoring higher on some of them actually had weaker clan traditions and fewer people living with extended family.

Countries scoring higher on responsibilism tended to score higher on census-based measures of family life, such as household size and the ratio of marriages to divorces, and on an index of extended-family and clan traditions. People in these countries were more likely to agree that a friend has the right to expect them to lie in court on the friend’s behalf. They also donated more blood to family members but less to strangers, and returned fewer lost wallets in field experiments.

United Nations diplomats from more responsibilistic countries also ran up more unpaid New York City parking tickets while protected by diplomatic immunity. That behavior’s costs fall on strangers and the public system rather than on family or friends. Nearly all of these links held after accounting for national wealth; the exception was blood donations to family.

This emphasis on binding interpersonal obligations aligns with an earlier study of relationship scenarios in China and the United States, which found that Chinese participants were less likely than Americans to feel and directly express gratitude toward close others, a pattern its authors linked to collectivist norms that treat helping a close friend or relative as an expected duty rather than a favor. That earlier work examined reactions to specific relational situations in two countries rather than applying a broad cross-cultural scale, but both studies suggest that East Asian relationship norms prioritize fulfilling duties over expressing warm sentiments.

Beyond family life, more responsibilistic countries tended to have more perceived government corruption, which the authors suggest fits with putting personal connections ahead of impersonal rules. They also tended to score lower on measures of democracy, innovation, and economic freedom, and had fewer deaths from COVID-19 and from drug use disorders. These links held after accounting for national wealth. Links with holistic thinking (a style that focuses on context and relationships) and with less migrant-friendly policies did not survive once wealth and the number of statistical tests were taken into account.

In exploratory analyses, responsibilism tended to be higher in places with a history of more infectious disease and in places where more farmland is devoted to rice. Earlier research by Talhelm and colleagues suggests that rice farming required close coordination among families and neighbors. Wealthier countries, and those with a history of livestock herding, a more independent way of making a living, tended to score lower.

Because the scale asks people about their own choices rather than asking them to rate their country, it can also detect differences between people in the same country. In the United States, Southerners scored slightly higher than Northerners, and Americans with Asian or Latin American heritage scored higher than European Americans. Students at the University of Illinois Chicago, which enrolls many first-generation college students, also scored higher than students at highly selective universities.

Across the full dataset, people who placed themselves higher on a social class ladder scored slightly lower on responsibilism, while people who described themselves as more socially conservative scored somewhat higher. The older scales, on average, showed almost no link with either social class or politics.

“I think the responsibilism-vs-warm-fuzzy distinction is a useful way to see the world, even if you never leave your hometown,” Talhelm noted. “You can see the differences in other places too, like between working class and middle class, as well as liberal politics (warm fuzzy) and conservative politics (responsibilistic).”

In the authors’ reading, warm fuzzy collectivism and responsibilism represent two contrasting ways of approaching human relationships rather than a rejection of social ties altogether. They suggest that in societies with fewer binding obligations to family and close circles, people have more freedom to choose relationships based on personal enjoyment and to leave unsatisfying friendships, which may explain why respondents in individualistic cultures agree more with warm fuzzy statements.

“One pushback I get is along the lines of ‘What about socialism? That’s the real collectivism!'” Talhelm said. “Socialism (at least as an ideal) is the opposite of responsibilism. Responsibilism is about prioritizing close relationships over distant relationships. Socialism is all about big groups, ‘the many,’ the collective.”

“Some people want to say that’s the real collectivism, and that’s OK with me,” Talhelm continued. “As long as people understand that responsibilism is different, and responsibilism is a more accurate description of cultures that people typically describe as collectivistic, then we’re on the same page.”

As with all research, there are some caveats to consider when interpreting these findings. The links with real-world behavior compare national averages drawn from other datasets, so they describe countries rather than individuals; the diplomats and wallet finders were not the study’s participants. Because these cross-cultural comparisons and links to societal indicators such as corruption, innovation, and public health are observational, they also cannot establish that responsibilism directly causes those societal outcomes. In addition, the researchers relied primarily on university students in the 100 cultures to keep the samples demographically similar across countries, meaning the scores reflect a younger, more educated segment of each society rather than nationally representative populations.

There are also boundaries to what the new scale measures. The authors designed the 14-item questionnaire specifically to capture duties in close relationships versus generalized warmth toward strangers, which leaves out other potential facets of collectivism, such as how much people adjust their behavior to fit changing social contexts. The study also does not settle ongoing debates over whether individualism and collectivism represent opposite ends of a single spectrum or separate dimensions that can coexist in the same person or culture. Future studies can test whether adding other behavioral dimensions improves the scale and can develop items simultaneously in languages with features like grammatical gender, going beyond the English and Chinese drafts used here.

“We have a fun follow-up right now showing people across cultures different messages to try to encourage people to donate money to climate change,” Talhelm said. “The idea is that we can make these messages more impactful if we fit them to the culture–either warm fuzzy messages or duty-bound responsibilistic messages. So far, we’re finding that people in cultures like the U.S. and Australia are more willing to donate if they see a warm fuzzy message that emphasizes caring about all people around the world.”

“In responsibilistic cultures, that message does not help,” he concluded. “The bottom line is that we can be more effective at communicating if we understand some basics of culture.”

The study, “Defining collectivism as duties in relationships rather than generalized warmth improves predictive validity in 100 cultures,” was authored by Thomas Talhelm, Liuqing Wei, Rui Sun, Danila Medvedev, Álvaro San Martin, May Helmy, Adil Samekin, Alejandra Zaragoza Scherman, Alexander S. English, and The Responsibilism Collaboration Team.