Americans’ Declining Trust in Each Other and Reasons Behind It – !

Americans trust each other less than they did a few decades ago. The share of adults who said “most people can be trusted” declined from 46% in 1972 to 34% in 2018, according to the General Social Survey.
In a 2023-24 Pew Research Center poll, an identical 34% said most people can be trusted.
Who says most people can be trusted?
Social trust seems to be rooted partly in personal experience. People learn to trust others based on how they themselves have been treated. But scarring events that reduce trust – like losing a job or experiencing discrimination – may happen to people in some groups more often than others, leading to differences in trust across society.
Racial and ethnic differences persist even after we take income, education, partisanship, age and other factors into account. Previous research shows that collective as well as individual experiences with racial discrimination can have a lasting impact on levels of trust.
Why does trust matter?
On the one hand, common sense tells us that people sometimes can be too trusting – falling prey to scams, for example. And distrust can be a rational response to a life full of hardship.
On the other hand, trust is the oil that lubricates the frictions of daily life. Trust makes it easier for people to work together to solve problems. It is beneficial for the economy because it’s related to confidence that other people will respect contracts, repay loans and behave honestly. And higher trust is associated with better-functioning democratic institutions.
In short, overall levels of social trust seem to go hand in hand with many features of a healthy society, creating a “virtuous circle.” (Though, as with any circular set of attitudes and behaviors, it’s difficult to say what is cause and what is effect.)
A Pew Research Center survey from March 2025 offers a few examples. Americans who express trust in others are more likely to say they would help their friends and neighbors in various ways.
Trust also plays a role in how people stay abreast of what’s happening around them. Americans who trust others are more likely than those who don’t to trust the information they get from friends, family and acquaintances. This kind of word-of-mouth news is now the most common source of local news for Americans.
Also, trust in other people is associated with trust in institutions and experts. For example, people with lower levels of social trust tend to express less confidence in national news organizations, local news outlets, public school principals, the police and the federal government. While few U.S. adults trust the government in Washington much at all these days, Americans who say most people can be trusted are twice as likely to say the federal government can be trusted most of the time (22% vs. 11%). However, there is no consensus on whether declines in institutional trust lead to declines in social trust, or vice versa.
Trust may offer some personal benefits, too. People who say most people can be trusted report greater life satisfaction than those who say you can’t be too careful. They tend to report being happier, more satisfied with their health and more likely to describe their family life as excellent or very good.
Where is trust highest?
Some parts of the United States stand out for higher – or lower – levels of trust.
Take, for example, New Hampshire, Oregon and Utah. In these states, close to half of adults say “most people can be trusted.”
But the share of people in some other states – like Alabama, Arkansas, Louisiana, Mississippi and West Virginia – who say “most people can be trusted” is about half of that in the highest trusting states. These “lower trusting” states have fewer college graduates and lower household incomes, both of which are associated with lower levels of trust.
Within a state, trust also can vary dramatically from place to place. Take California: 35% of its residents (about the national average) say most people can be trusted, but San Franciscans are nearly twice as likely as people in Riverside to feel so trustful (46% vs. 24%).
Notably, the San Francisco metropolitan area, which encompasses both Oakland and Berkeley, is one of the most highly educated places in the U.S., with more than half of adults 25 and older holding at least a bachelor’s degree.
In the Riverside metro area, by contrast, 25% are college graduates. The two metro areas also differ in average income levels and racial composition in ways that are related to trust.
While the level of trust within a community may be predicted largely by the wealth, education and other characteristics of the people who live there, that’s not the entire story. Even after controlling for demographic differences, neighborhoods seem to matter. For example, people who live in an area where more people are college educated tend to have higher levels of trust, regardless of their own level of education. (For more about the impact of neighborhoods, refer to the “Geography” section below.)
People who say that drug addiction, poverty, crime and availability of jobs are major problems in their community are – perhaps not surprisingly – more likely to say that most people can’t be trusted than those who don’t face those problems. And Americans who feel only “somewhat safe” in their neighborhood at night are considerably more likely to say most people can’t be trusted, on average, than those who feel very or extremely safe.
As numerous studies have noted, demographic and social diversity also can present barriers to trust. A March 2025 survey found that more than two-thirds of people who say their neighbors are similar to them racially, politically and educationally also say they trust most or all of their neighbors. By contrast, most people who see their neighbors as very different say they trust only some or none of them.
Still, causality is hard to pin down. Do differences between people increase mistrust? Or do mistrustful people tend to perceive more differences between themselves and others? (For more about the complicated relationship between neighborhoods and trust, read “How connected do Americans feel to their neighbors?”)
Why might trust be declining?
While it’s difficult to know for sure what is causing the long-term decrease in trust, we can learn a bit by querying the same people at different points over time. In two surveys roughly five years apart, we asked Americans whether most people can be trusted – first in March 2020 (starting about a week after the World Health Organization declared COVID-19 a pandemic), and then in February 2025 (less than a month after President Donald Trump’s inauguration). More than 4,000 people responded to the question in both years.
Over this tumultuous five-year period, 25% of people surveyed changed their minds about whether most people can be trusted. While slightly more gained trust than lost it (14% vs. 11%), financial well-being is tied to losing trust after controlling for other demographic characteristics. People who said they didn’t live comfortably were more likely to lose trust than those who did live comfortably. (More details of this longitudinal analysis are in the Methodology.)
Looking at the longer term, declining trust may also be related to other significant shifts we’ve documented in decades of Center surveys. While we cannot say that any of these shifts caused the decrease in trust, these major changes in American society may have something to do with the drop in trust.
Political polarization
Republicans and Democrats are now further apart – in both political values and demographic characteristics – than they were a few decades ago. On both sides, views of the opposing party and its supporters have grown more negative, which may contribute to the decline in social trust.
Political polarization may result in people believing they share less in common with their fellow citizens and reinforce an “us versus them” outlook.
While this polarization is a nationwide phenomenon, it is likely reinforced by the fact that many Americans live in places with little opportunity to interact meaningfully with someone from the opposite party.
Internet and technology
The share of Americans who use the internet has risen sharply over the past quarter century. Most now use the internet daily – including 41% who say they use it almost constantly.
How this affects social networks and social ties has long been a matter of debate, but some of the ills laid at the feet of the internet and screens include distracting from face-to-face interactions, increasing isolation and loneliness, and causing or exacerbating mental health issues, particularly among teens and young adults. (On the other side of the coin, technology and social media use also can facilitate offline engagement or even broaden social networks, and the work tying the internet to societal ills has many critics.)
News and information
Fewer Americans have a shared media diet today than a few decades ago. Over the past 10 years or so, daily newspaper circulation has fallen precipitously, audiences for local TV news have decreased, and fewer Americans regularly watch nightly network TV news.
Meanwhile, social media news consumption has risen dramatically. With these changes, Americans are having trouble telling the difference between what’s true and what’s not true. Many are also concerned about made-up news and information, and 54% say this has a big impact on Americans’ confidence in each other.
Diversity
Scholars have argued that higher levels of ethnic diversity are related to lower levels of social trust, and the U.S. population has been growing more racially and ethnically diverse. From 2000 to 2018, at least 109 counties in 22 states – from California and Nevada to Kansas and North Carolina – went from majority White to majority non-White.
Americans now see more racial diversity around them: 50% say that all or most of their neighbors are the same race or ethnicity as they are, down from 55% in 2018. And, particularly among White Americans, the sense that one lives in a neighborhood with a lot of people who share the same race and ethnicity is related to higher levels of social trust.
Inequality
Income inequality has risen in recent decades. The richest families in the U.S. have experienced greater gains in wealth than other families in recent decades. And the share of total U.S. household income held by the middle class has fallen almost without fail in each decade since 1970.
Americans broadly see the system as unfair and favoring powerful special interests – and this discontent may contribute to a broader sense that others are selfish or profiting at one’s expense, possibly leading to a decline in trust.
Religion
There has been a long-term decline in the percentages of Americans who identify with a religion and say religion is important in their lives. Some scholars, including political scientist Robert Putnam, have argued that the decline of organized religion is part of a broader shift of Americans toward living increasingly isolated lives (“bowling alone”), with fewer opportunities to create interpersonal connections and build trust between different kinds of people.
How and where time is spent
About a quarter of Americans know most or all of their neighbors – down from three-in-ten in 2018. Time-use studies show that Americans now spend more time alone and less time socializing than in the past. They are also spending more time at home. They even walk more quickly and socialize less in public than a few decades ago.
While some of these trends were fueled by COVID-19, they have persisted since the pandemic abated, suggesting a more fundamental shift in how and where people spend time. For example, more Americans with “teleworkable” jobs report working from home most or some of the time today than in 2020, though the share who work from home all of the time has shrunk since the height of the pandemic.
Collectively, these societal shifts may have driven down social trust by increasing many people’s feelings of isolation and undercutting the sense that “we are all in this together.”
Social trust and generational change
Whatever the underlying reasons, each new birth cohort of individuals reaching adulthood (each “new generation”) has been raised in a less trusting climate, creating a striking generational pattern: Each recent birth cohort is less trusting than the previous one. And as older, more trusting cohorts have gradually died and been succeeded by younger ones, the population’s average level of trust has been dropping.
For instance, adults born in the 1990s (people ages 26 to 35 today) express lower levels of trust than those born before the 1990s. Likewise, those born in the 1970s (people ages 46 to 55 today) express less trust than those born in earlier decades.
These differences between birth cohorts consistently show up in surveys, even as overall levels of trust vary somewhat from survey to survey.
Is the U.S. unique?
America is not unique – either in how much trust we have in one another, or in the broad patterns that connect who is more trusting with who is better off in other ways.
In international surveys in 2020, Americans’ levels of trust were about average among the 14 high-income countries we surveyed: lower than in places like Denmark, the Netherlands and Sweden but higher than in Spain, France and Italy. And in most of the high-income countries, older people, people with higher incomes, and people with more years of formal education tend to be more trusting – just as we find in the United States.
One big difference, though, is that while levels of trust have dropped and then remained stable in the U.S., there is evidence they may be rising in some other high-income countries.
According to the World Values Survey, trust has increased around 10 points or more over the past two decades in Canada, Germany, the Netherlands, Singapore and the UK.
Using a slightly different measure – a 10-point scale where 0 means “you can’t be too careful” and 10 means “most people can be trusted” – the European Social Survey also finds that many European publics are becoming more trusting.
Where are we headed?
Fewer than half of Americans (44%) now say they can trust all or most of the people in their neighborhood, down from 52% a decade ago. And only around a third say most people can be trusted – also down around 10 points over the past 10 years.
But in more recent years, the decline in trust has largely abated. Levels of social trust have been largely flat – or even slightly up – since 2018, regardless of the way we ask the question. (For several different ways to measure trust, refer to the Methodology.)
So, what does this mean? Americans are emerging from a traumatizing global pandemic that they say did more to drive the country and their communities apart than to bring them together. People see more rudeness in public now, and some suggest the pandemic turned Americans into hyperindividualists.
Yet despite – or perhaps in part because of – this disorienting event, Americans may be interacting in new ways. COVID-19 may even have some positive consequences for how people relate to one another.
Our 2023-24 Religious Landscape Study shows Americans are no longer leaving organized religion at rates seen in the past. Other studies find that national rates of volunteerism have bounced back to pre-pandemic rates – or higher. And a worldwide Gallup study of happiness in 2024 reported that a “global surge in benevolent acts,” especially the helping of strangers, began during the pandemic in 2020 and has continued in subsequent years.
Explore more about trust among U.S. adults by clicking on the menu below.
Geography
Levels of social trust do not vary much by whether people live in an urban, suburban or rural county. Based on a government system for classifying counties by population size and the presence or absence of a city in the metro area, trust levels vary from 31% in the two least urban groups of counties to 35% and 37% in the two most urban groups.
Region, state and metro area
Trust varies across census regions in the United States, with the lowest overall rates in the South (where 30% say most people can be trusted) and the highest (38%) in the Midwest. But within each region, trust levels vary from state to state. In the West, for example, Oregon has one of the most trusting populations (46%), while New Mexico has one of the least (26%).
And trust can vary dramatically even among metro areas within the same state. For example, California boasts one of the country’s most trusting metro areas (among the 34 for which estimates can be made) as well as one of the least: San Francisco (46%) and Riverside (24%).
Some of these differences simply reflect the kinds of people who live in each geographic area. For example, in the South, where the majority of the Black population lives, trust levels are lower partly because Black adults tend to have lower rates of trust. To put it another way, trust varies across regions, states and metro areas in part because of who lives in each place and how they differ by race, income, education and other demographic traits.
Education level of neighborhoods
Even after accounting for some of these “compositional” differences, the characteristics of one’s neighborhood still may play a small role in people’s levels of trust. For instance, people living in locations with a highly educated population tend to be more trusting.
The survey’s respondents were sorted into five roughly equal sized groups (or “quintiles”) based on the share of people in their neighborhoods who have college degrees. In neighborhoods where few people (an average of 8%) have a college degree, 24% of adults surveyed say most people can be trusted. But in neighborhoods where many people are college graduates (an average of 68%), twice as many (49%) express trust in others. (In this analysis, we define “neighborhoods” as census block groups, which typically contain 600 to 3,000 individuals. We use the terms neighborhoods, places and locations somewhat interchangeably.)
Even focusing on people at the same educational level – such as college graduates – those living in areas where many other people are college grads tend to be more trusting. When we rank places by the share of residents who have a college degree, in the top one-fifth of those places, 54% of college graduates are trusting. In the bottom fifth, 37% of college grads say most people can be trusted.
The same pattern appears if we look only at people with a high school diploma. In places with the lowest share of people with college degrees, 21% of high school graduates are trusting, while in places with the highest share of college graduates, 28% of high school grads are trusting.
At the same time, people vary in numerous ways that are relevant to social trust – even if they have similar levels of education. For example, Black and White college graduates differ in trust by about 25 percentage points: 53% of White college grads are trusting, compared with 28% of Black college grads.
To isolate how characteristics of a neighborhood or community may relate to trust, we built a statistical model to compare people who are as similar as possible on a wide range of characteristics known to affect levels of trust – for example, gender, age, race and ethnicity, education and more (refer to the Methodology for more detail). We also included a measure of the share of college graduates in one’s neighborhood.
Results indicate that an average person who lives in a neighborhood with a high ratio of college graduates is 7 points more likely to think most people can be trusted than that same person would be if they lived in a neighborhood with a very low ratio of college graduates. Once we take into account the share of college graduates in a person’s neighborhood, that neighborhood’s affluence – measured by the share of people who have no health insurance, live in an owner-occupied dwelling or receive some sort of public assistance – does not seem to play a role in trust levels.
Still, other neighborhood factors may be related to people’s levels of social trust, including the partisan composition of one’s neighborhood or even its ethnic diversity. Previous work in Denmark has found that ethnic diversity at the neighborhood level negatively affects trust.
Education
Americans with higher levels of education are more likely to say most people can be trusted. Around half of U.S. adults with a master’s degree or higher say that, generally, most people can be trusted (52%), compared with 44% of those with a bachelor’s degree and around a third or fewer who have either an associate degree or some college but not a degree. Around a quarter of those with a high school diploma or less formal schooling say most people can be trusted.
The pattern of people with higher levels of education being more trusting of others is consistent regardless of gender or political affiliation.
Income and employment status
The likelihood of believing that most people can be trusted rises steadily with income. Around one-in-five Americans whose annual household income is under $30,000 say most people can be trusted. Generally speaking, the share of people who trust others rises with each step up the income ladder. Half of those with household incomes of $150,000 or higher say that most people can be trusted.
Employment
Retirees are more likely to view others as trustworthy than are people who are working either full time or part time for pay. And trust levels are particularly low among U.S. adults who are not currently working or who are unable to work due to disability.
While retirees tend to be older and more affluent – two factors associated with trust – they remain more trusting even after accounting for a variety of demographic differences, including age, income, education, nativity, race and ethnicity, and sex. However, demographic differences do help to explain the relatively low levels of trust expressed by people who are not currently working.
Age
Older adults generally are more trusting than younger adults. Among Americans ages 65 and older, 44% say most people can be trusted, compared with 26% of those ages 18 to 29. Among adults ages 30 to 49 and 50 to 64, roughly one-third say the same.
Moreover, older people are more trusting than younger ones who share the same education level, partisanship and gender. For example, 37% of college graduates under 30 say most people can be trusted, while 59% of college grads ages 65 and older say the same.
However, these patterns are not seen across all racial and ethnic groups. Although older White, Hispanic and Asian Americans tend to be more trusting than younger people with the same race or ethnicity, levels of trust among Black Americans generally do not vary much by age.
Gender, marital status and sexual orientation
Men are slightly more likely than women to say most people can be trusted (37% vs. 32%). While the gender difference is fairly small, it is persistent: Men tend to express a little more trust in other people than women do, regardless of how we ask the question.
And men of various racial and ethnic groups, education levels and ages tend to be more trusting than women in the same groups. For example, White men are more trusting than White women; men under 30 are more trusting than women under 30; and men with a high school diploma or less schooling are more likely than women with the same education to say most people can be trusted (27% vs. 21%).
Marital status
Trust is also related to personal relationships. On average, married people express higher levels of trust (40% say most people can be trusted) than adults who live with a partner (29%), are divorced (29%), are separated (25%) or have never been married (25%). But Americans who have been widowed are about as trusting as those who are married (38% and 40%, respectively).
Still, gender differences tend to persist even among people who are married. For example, married men are more likely than married women to trust others. The same is true for men and women who have never been married, as well as men and women living with a partner.
Sexual orientation
People who identify as gay or lesbian (43%) are more likely than those who identify as straight (35%), bisexual (31%) or something else (29%) to say most people can be trusted.
Some of the gap in trust between gay or lesbian Americans and straight Americans has to do with party affiliation: Most gay and lesbian adults identify as Democrats or lean Democratic, and Democrats tend to be somewhat more trusting than Republicans.
However, party affiliation is not the whole story. Gay and lesbian Democrats are slightly more trusting than straight Democrats (45% vs. 40%). Among Republicans, 35% of gay and lesbian adults say most people can be trusted, while 31% of straight Republicans say the same. (Party affiliation in this case includes both those who identify with and those who lean toward each party.)
Although straight men are a bit more likely than straight women to say they trust others (37% vs. 32%), there is no significant difference in the levels of trust expressed by gay men (43%) and lesbian women (41%).
Race and ethnicity
Trust varies greatly by race and ethnicity. About equal shares of White Americans (40%) and Asian Americans (38%) say most people can be trusted. However, Black adults (21%) and Hispanic adults (23%) express trust at about half that rate, on average.
Racial and ethnic differences in saying that most people can be trusted persist even after statistically accounting for differences in education, income, gender and party affiliation, among other factors.
For example, White college graduates are more likely than Asian, Hispanic or Black college graduates to say most people can be trusted.
And White respondents with household incomes of $100,000 or more (51%) are also significantly more likely to say most people can be trusted than Asian (44%), Hispanic (32%) or Black households (28%) at similar income levels.
Party affiliation and political ideology
Democrats are somewhat more trusting than Republicans. Overall, among those who identify as Democrats or lean toward the Democratic Party, 39% say that most people can be trusted. Among Republicans and those who lean toward the GOP, 31% do so.
The gap between the parties on trust is especially large among non-Hispanic White adults. Half of White Democrats say most people can be trusted, compared with about one-third (34%) of White Republicans.
There is also a 10 percentage point gap in trust between Asian American Democrats and Republicans (41% vs. 31%). But the differences are much smaller between Black Democrats and Black Republicans, or between Hispanic Democrats and Hispanic Republicans.
Self-described liberal Democrats express more trust (with 47% saying most people can be trusted) than Democrats who describe themselves as ideologically moderate (35%) or conservative (24%).
On the other hand, among Republicans, conservatives (33%) are a little more trusting than those who describe themselves as moderate (28%) or liberal (25%).
Religious affiliation and practice
While just over a third of U.S. adults believe most people can be trusted, around half of Jews, Hindus and members of the Church of Jesus Christ of Latter-day Saints – widely known as Mormons – tend to trust others. Atheists and agnostics also stand out as particularly trusting groups (45% each).
On the other end of the spectrum, people in historically Black Protestant denominations are among the least likely to say that most people can be trusted (19%), followed by evangelical Protestants (30%) and those who say their current religion is “nothing in particular” (30%).
People within religious groups vary in ways beyond their beliefs. For example, Jews and Hindus are the two U.S. religious groups analyzed here with the highest shares of college graduates, and Americans with higher levels of education tend to be more trusting. Similarly, the lower levels of trust among members of historically Black Protestant churches is related to the fact that Black Americans tend to express lower levels of trust than White Americans. Still, the different trust levels between religious groups largely persist even after accounting for education, race and other demographic factors.
Religious practice
People who attend religious services at higher rates tend to be more trusting than those who attend less. Again, these differences remain even after accounting for demographic factors. This greater trust may be because of the social nature of participating in a church, synagogue, mosque, temple or other religious community, as academics like Robert Putnam have theorized.
But while attending religious services is related to higher levels of trust, the same is not true for praying or considering religion to be important in one’s life. If anything, less religious Americans tend to be more trusting by these measures.
For example, U.S. adults who say they seldom or never pray and those who say religion is not too or not at all important to them show higher levels of trust than their more religious counterparts.
Notably, though, differences in trust by prayer and importance of religion are largely explained by other demographic factors, such as political partisanship, age, income and education.
About this essay
Results presented in this data essay come from the Religious Landscape Study (RLS), the American Trends Panel and other sources. More information on each can be found in the Methodology. The RLS was made possible by The Pew Charitable Trusts, which received support from the Lilly Endowment Inc., Templeton Religion Trust, The Arthur Vining David Foundations and the M.J. Murdock Charitable Trust. Pew Research Center is a subsidiary of The Pew Charitable Trusts, its primary funder.
This data essay is a collaborative effort based on the input and analysis of a number of individuals and experts at Pew Research Center.