Lane Kenworthy, The Good Society
May 2023

Humans are social creatures. We need connection to other people. Isolation and loneliness can be uncomfortable, even painful, and they can have damaging effects on mental and physical health.

How common is social disconnection? Has it increased? How big a problem is this?

Skip to:

• Consequences of loneliness and isolation
• 150 years of predicting declining social connections
• Isolation
• Loneliness
• Teenagers
• What to do
• Summary

CONSEQUENCES OF LONELINESS AND ISOLATION

Research comparing across individuals finds that isolation and loneliness are linked with less frequent experience of positive emotions, with more frequent experience of negative emotions, with anxiety, with depression, and with lower levels of happiness and life satisfaction.¹ They may also worsen physical health and shorten life. A 2015 meta-analysis of 70 studies concludes that “Substantial evidence now indicates that individuals lacking social connections (both objective and subjective social isolation) are at risk for premature mortality. The risk associated with social isolation and loneliness is comparable with well-established risk factors for mortality, including … physical activity, obesity, substance abuse, responsible sexual behavior, mental health, injury and violence, environmental quality, immunization, and access to health care.”²

What can we learn by comparing across nations? Figure 1 includes three charts. On the horizontal axis in each chart is the share of adults who don’t live alone — that is, who live with one or more other persons. In the first chart, the vertical axis has a measure of happiness. In the second chart it is suicides. In the third it is life expectancy. Life satisfaction tends to be lower, not higher, in countries where fewer people live alone. On the other hand, the suicide rate tends to be lower in such countries, although the difference is modest. There is no apparent association between living alone and life expectancy.

Figure 2 has three analogous charts, but with the horizontal axes showing the share of adults who say they get together with friends several times a month or more. Here we see a positive association with happiness. There again is a negative association with suicides, and this association is fairly strong. These first two charts suggest a beneficial impact of social connections. Once again, however, there is no apparent relationship with life expectancy.

In figure 3, the horizontal axes have a third indicator of social connections: the share of adults who say yes in response to the question “If you were in trouble, do you have relatives or friends you can count on to help you whenever you need them, or not?” Here the first chart suggests a beneficial effect while the other two don’t.

These cross-country patterns suggest that isolation may be bad for happiness and suicides. Two of the three measures of social connectedness are associated with these outcomes in the predicted way. There is no indication of an effect on life expectancy.

150 YEARS OF PREDICTING DECLINING SOCIAL CONNECTIONS

For at least 150 years, analysts and pundits have worried that modernity would reduce social connections and ties. Better transportation, greater access to college, and the concentration of jobs in cities increases the incentive for people to move away from family and childhood friends. Cities are crowded, which can make it more difficult to develop lasting bonds. Technological advance provides more sources of distraction and access to individualized entertainment, reducing the time available for friends or family. As more women move into paid work, they have less time to socialize. And as government public insurance programs cushion more of the risks and hardships we face, voluntary organizations, a key source of interaction and community, may diminish in prevalence.

On the other hand, most of these developments could potentially have the opposite effect. Geographic mobility might increase the number of friends we have. Living in a more populous area, such as a city or suburb, boosts the number of potential friends. New technologies — telephones, email, text messaging, social media — enhance our ability to communicate with family and friends. For women, work is a potential source of contacts and friendships. And by reducing deprivation and improving capabilities among the less advantaged, government services and transfers might decrease isolation.

In the late 1800s and early 1900s, sociologists Ferdinand Tonnies, Max Weber, and Emile Durkheim described the shift from the “gemeinschaft” society of small villages, which emphasizes personal relationships and family, to the individualistic, atomistic “gesellschaft” society common in large cities. In 1950, David Riesman’s The Lonely Crowd portrayed an America in which adults are less and less connected with one another, whether by traditions or traditional institutions. In the 1970s, Vance Packard’s A Nation of Strangers warned that personal connections of all sorts were weakening.

In recent decades pessimistic assessments in this vein have proliferated. In Bowling Alone, published in 2000, Robert Putnam examined a host of indicators of social connections. The evidence, he concluded, suggests that “across a very wide range of activities, the last several decades have witnessed a striking diminution of regular contacts with our friends and neighbors. We spend less time in conversation over meals, we exchange visits less often, we engage less often in leisure activities that encourage casual social interaction, we spend more time watching (admittedly, some of it in the presence of others) and less time doing. We know our neighbors less well, and we see old friends less often.”³

Over the ensuing decades, David Brooks, an op-ed writer for the New York Times, penned dozens of columns suggesting that “social fragmentation and social isolation are the fundamental problems afflicting America today.”⁴

In 2016, Yuval Levin posited, in his book The Fractured Republic, that “As the national government grows more centralized, and takes over the work otherwise performed by mediating institutions — from families and communities to local governments and charities — individuals become increasingly atomized.” The result, according to Levin, is that “We have set loose a scourge of loneliness and isolation.”⁵

In 2017, Jean Twenge examined patterns of loneliness and isolation among teenagers since the advent of the smartphone a decade earlier. Her conclusion: “The twin rise of the smartphone and social media has caused an earthquake of a magnitude we’ve not seen in a very long time, if ever. There is compelling evidence that the devices we’ve placed in young people’s hands are having profound effects on their lives — and making them seriously unhappy.”⁶

That same year, Vivek Murthy, two-time US Surgeon General, suggested that “Loneliness is a growing health epidemic. We live in the most technologically connected age in the history of civilization, yet rates of loneliness have doubled since the 1980s.” He reiterated that conclusion in a 2023 report.⁷

A major exception to these pessimistic assessments comes from Claude Fischer. In his 2011 book Still Connected, Fischer undertook the most comprehensive and detailed examination of available data on social connections and ties. He concluded as follows⁸:

“The question that this book has posed is whether and how Americans’ relationships with family and friends changed between 1970 and 2010. The short answer, based on a canvass of published research and available survey data, is: not much. Some of the ways in which Americans engaged with people in their immediate circles changed, but the intimacy and support of close family and friendship ties stayed about the same. Few Americans were socially isolated, and the percentage of those who were did not increase. The number of family and friends with whom people reported being close stayed about the same. Americans got together with one another in set-aside home activities like dinner parties less often, but they communicated with one another electronically more often. Americans expected to get about as much help from family and friends as they had earlier. And American feelings about their social relationships stayed about the same or became more upbeat.”

What does an updated look at the evidence tell us about social connections, isolation, and loneliness in the United States?

ISOLATION

Let’s begin with the question of how much time Americans spend with family. Figure 4 shows that the share of families who frequently eat dinner together declined in the 1980s but has been fairly constant since then.

Robert Putnam reported in 2000 that “Beyond mealtime, virtually all forms of family togetherness became less common over the last quarter of the twentieth century. Between 1976 and 1997, according to Roper polls of families with children aged eight to seventeen, vacationing together fell from 53 percent to 38 percent, watching TV together from 54 percent to 41 percent, attending religious services together from 38 percent to 31 percent, and ‘just sitting and talking’ together from 53 percent to 43 percent.”⁹

On the other hand, the amount of time parents spend with their children — playing, reading, doing homework, driving to and from school and extracurricular activities — has increased since 2000.¹⁰ And as figure 5 shows, there has been no decline since the mid-1970s in the share of Americans, around 55-60%, who say they spend social evenings several times a month or more with relatives.

In Still Connected, Claude Fischer concludes that “This pattern of results suggests that Americans were no less committed to their immediate families and other relatives in the 2000s than before….”¹¹

What about time with friends and neighbors? In Bowling Alone, Robert Putnam reported declines in the 1970s, 1980s, and 1990s in playing cards, participating in league bowling, going out to bars and nightclubs, participating in youth sports (boys only), and engaging in informal socializing (visiting with friends, attending parties, informal conversation, and so on).¹²

The General Social Survey periodically asks Americans how often they spend a social evening out with neighbors. As figure 6 shows, the share who say they do so several times a month or more dropped in the late 1970s and through the 1980s. Since, then, however, there has been no change. The survey asks the same question about friends who live outside the neighborhood. As we see in figure 7, here there has been no decline at all. Data from DDB Needham surveys are consistent with this picture.¹³

Figure 8 shows the share of Americans who say that yesterday they called a friend or relative just to talk. The line is flat, suggesting no decline. We don’t have good data on contact with family and friends via texting or social media, but given that neither existed prior to the late 2000s, there is good reason to suspect they have increased.

Has there been an increase in the share of Americans who don’t have any friends?¹⁴ Since 1976, Gallup has occasionally asked “Not counting your relatives, about how many close friends would you say you have?” And since the mid-1980s the General Social Survey has occasionally asked about “close friends.” The trend for average number of friends has been flat. The share saying “zero” has been less than 5%, and it too hasn’t increased.¹⁵

In recent decades the share of adults who live alone has risen steadily, in the United States and in a number of other countries. This appears to be largely a function of affluence. As people have more disposable income, more of them choose to live by themselves.¹⁶ (Some of it is due to rising life expectancy, which means more elderly persons live alone for lengthy periods after a spouse or partner dies.) Eric Klinenberg, who has studied this phenomenon in the US in detail, concludes that those who live alone aren’t condemned to be isolated or feel lonely: “On the contrary, the evidence suggests that people who live alone compensate by becoming more socially active than those who live with others.”¹⁷ And some of the countries in which the largest share of adults live alone are among those in which loneliness is lowest and perceived social support is highest.¹⁸

In his review of the data in 2011, Claude Fischer concluded that there has been no decline in American adults’ overall contact with family and friends.¹⁹ That conclusion seems likely to still hold.

Since the mid-2000s the Gallup World Poll has regularly asked “If you were in trouble, do you have relatives or friends you can count on to help you whenever you need them, or not?” More than 90% of Americans answer yes to this question. And even though the data don’t go very far back in time, the fact that the level is so high means there can’t have been much decline.

LONELINESS

Isolation and loneliness aren’t strongly correlated across individuals. Some isolated people are lonely, but others aren’t. Some non-isolated persons aren’t lonely, while some are.²⁰ (The one notable exception is people without a romantic partner/spouse, who tend to feel lonelier than those who do have one.) So even though isolation doesn’t seem to have increased to any appreciable degree, it’s possible that Americans have gotten lonelier.

Loneliness is “the subjective feeling that you’re lacking the social connections you need.”²¹ Social scientists typically measure loneliness via answers to questions such as

• How often do you feel that you lack companionship?
• How often do you feel that there are people you can turn to?
• How often do you feel that there are people who really understand you?
• Do you have someone with whom you’re comfortable discussing important matters?
• How satisfied are you with your friendships?
• Would you like to spend more time with friends?

Reviewing the available data for the years 1970 to 2010, Claude Fischer finds that fewer than 10% of American adults reported being lonely according to these types of questions, and there was no change during this period.²²

What do we find when people are asked more directly about being lonely? One question, asked only a handful of times over the past half century, is “During the past few weeks, did you ever feel very lonely or remote from other people?” Figure 10 shows the share responding yes to this question from the early 1960s to the early 2000s. The data suggest no noteworthy change over that period. In 2018, a different survey found 11% saying they feel lonely always or often; 22% saying they feel lonely, isolated, left out, or lack companionship always or often; and 34% saying they feel lonely always, often, or sometimes. In 2023 another survey found 17% saying they experienced loneliness “a lot of the day yesterday.”

These data suggest that the prevalence of loneliness among American adults may be somewhere between 10% and 30% (rather than Fischer’s estimate of 10%).²³

The data in figure 10 give us little reason to conclude that there’s been an increase in loneliness. A 2018 assessment by a congressional committee concludes similarly: “The discussion of loneliness has suggested to media consumers and policymakers that it is an epidemic — that loneliness has increased substantially in recent years and is a pressing problem in need of urgent attention. These claims, however, are based on a flawed interpretation of the research literature. In fact, there is little evidence that loneliness has increased.”²⁴

TEENAGERS

The story is different for American teenagers.

Smartphones and social media have shifted the type of contact teens have with one another, though they likely haven’t decreased the amount of contact. As figure 11 shows, we see sharp declines over the past decade in the share of 12th graders (high school seniors) who ever go out on dates, who attend parties once a month or more, and, most revealingly, who say they get together with friends every day or nearly every day. While these forms of face-to-face contact have decreased, they’ve been replaced by contact via texting and social media. As Jean Twenge puts it, “The timing of the recent, severe drop in going out and getting together with friends … occurred right when smartphones became popular and social media use really took off. Time spent with friends in person has been replaced by time spent with friends (and virtual friends) online.”²⁵ Texting and social media were nonexistent prior to 2007. As figure 12 shows, both are now heavily used by 12th graders.

Loneliness appears to have increased among American teenagers, as we see in figure 13. This began around 2013.

Jean Twenge, Jonathan Haidt, and colleagues note that there also has been a significant rise among teens in unhappiness, depression, self-harm, and suicide.²⁶ In a 2017 book, Twenge suggested that smartphones and social media increased perceived loneliness among teens, and that the rise in loneliness caused the worsening of mental well-being. She also concluded that the harm has been greater among girls than among boys, because girls “use social media more often, giving them additional opportunities to feel excluded and lonely when they see their friends or classmates getting together without them.”²⁷

In more recent writings, these scholars have described the trend of increased loneliness as one component of declining well-being among American teenagers, rather than as a cause of the other components.²⁸ This conclusion seems more consistent with the fact that self-reported loneliness among teens appears to have begun increasing at exactly the same time, around 2013, as unhappiness, depression, self-harm, and suicide — not a few years before. This coincident timing suggests there may have been another driving force, and that loneliness is as much or more a consequence of deteriorating happiness and mental well-being as a cause of it.

What might that alternative cause be? There are at least four candidates. The first is pressure to succeed in school. Rigorous and intense education is very likely helpful for an economy oriented around complex problem solving. But it may also increase anxiety. In countries with higher average test scores and higher student assessments of academic competition, anxiety among teens tends to be greater. And while a country’s income level is positively correlated with life satisfaction among adults, for adolescents the correlation is negative.²⁹

The second is the perception among American teenagers that things have gotten worse — in life, in the United States, and in the world. Negative headlines in popular news media outlets have become more common.³⁰ This, coupled with increased dissemination and amplification of such headlines via social media, may have contributed to growing perception among US teens that certain types of speech aren’t just objectionable but also harmful, that you or someone you love might be a victim of a school shooting or other mass shooting, that police are a danger to Blacks, that women’s progress in education and in the workplace is blocked by discrimination, that “every time I try to get ahead, something or somebody stops me,” that climate change is leading inexorably to species extinction, and that “it is hard to have hope for the world.”³¹

The third is comparison with idealized lives and bodies. This has been around at least since the rich, beautiful, and famous have adorned magazine covers in grocery store checkout lanes. But social media puts these images in our face all day every day.

Fourth, American teens have been getting less sleep. This too seems to be largely a function of smartphones and social media. Sleep deprivation contributes to depression.³²

This interpretation, whereby the rise in loneliness is an element or a consequence of worsening subjective well-being among US teens rather than its cause, also is more consistent with two things we observe about suicide trends. Although girls seem more likely than boys to feel lonely as a result of social media, there was a larger increase in teen suicides among boys than among girls, as we see in figure 14. And while social media use and loneliness among teens has increased in other rich democratic countries too, in most of them teenage suicides were flat or decreasing in the 2010s, as figure 15 shows.³³ (The appendix has a separate graph for each country.)

Finally, it’s worth noting the possibility that smartphones and social media have themselves played little or no causal role in these developments. Two recent meta-analyses find no compelling evidence that teenagers’ use of digital technology or social media significantly worsens their subjective well-being.³⁴ And the most rigorous empirical study to date concludes that if there is a harmful effect, it is a very small one.³⁵

Indeed, whether or not loneliness itself has increased, and if so how much, remains in question. The most comprehensive analysis of trends in loneliness among young Americans — a meta-analysis of studies of persons age 17 to 30 covering the years 1976 to 2019 — finds evidence of a real but very small rise in loneliness.³⁶

WHAT TO DO

The evidence gives us little reason to conclude that isolation or loneliness have increased significantly in the United States, at least among adults. But that doesn’t mean they aren’t a problem. As noted earlier, they’re bad for mental and physical well-being. What, if anything, can we do address this problem?

One approach focuses on limiting smartphone and social media access for preteens and teenagers: encourage parents to limit smartphone access to children who are 13 or older, require that social media platforms implement strict screening and monitoring techniques to prevent preteens from getting access, and prohibit smartphone use in K-12 schools.³⁷

Relatedly, we might consider offering a required class students would take around age 13 on how to use social media safely and effectively. We spend enormous quantities of time on social media. It’s a core feature of modern life. It is, arguably, just as important as other topics for which we offer targeted courses, such as civics, personal finance, and sex education.

Another common suggestion is to change the culture, to encourage more people to find ways to have more in-person contact with family, friends, and others. That’s a reasonable hope, but intentionally shifting culture is notoriously difficult. Moreover, shifting it in this particular direction would be swimming against the past half century’s trend toward greater individualism and personal freedom.³⁸

We can create structures that make it easier for people to have regular contact with others. One example is a program of universal (voluntary or mandatory) national service for one year following high school. Another is facilitation (perhaps via subsidies) of communal living arrangements, similar to college dormitories.³⁹

Perhaps most important, we should acknowledge that no matter what steps we take some people will end up isolated and/or lonely, and we should ensure that they receive appropriate mental and physical health services.

SUMMARY

Since the onset of industrialization, societal observers have predicted a rise in isolation and loneliness, and in recent decades this view seems to have become increasingly prominent. However, in his thorough examination of the data in 2011, Claude Fischer concluded that over the past half century “Americans’ relationships with family and friends were — perhaps surprisingly to some commentators — robust and lasting.”⁴⁰ With the possible exception of the recent impact of smartphones and social media on teen loneliness, that assessment continues to hold.⁴¹

Even so, isolation and loneliness affect a nontrivial share of Americans, with harmful consequences for happiness, mental well-being, and perhaps physical health.

APPENDIX

The appendix has additional data.