The Connection Paradox
A new consumer survey finds people who use a lot of social media and video games are more lonely — but say that they feel more connected
Key Findings
- Forty-six (46%) of US adults are lonely. Loneliness is most prevalent among people aged 18–29 (55%), people earning less than $50,000 per year (52%), people of Hispanic origin (51%) and women (50%).
- Heavier users of both social media and video games are more likely to be lonely than lighter users.
- Paradoxically, heavier users of both social media and video games rated their influence on social connection more positively than lighter users.
- Also, paradoxically, heavier users of social media are more likely to report that it has made them more connected to friends, family and neighbors than lighter users.
- People with higher incomes are less lonely and rate both social media and video games more positively on their influence on social connection.
Introduction
Do social media and video games make people less social? Or could they catalyze more socialization and less isolation?
These questions inform the foundations of what we do at Building H. Our focus is on the “product environment” — the commercial products and services that collectively weave the fabric of everyday life — and how it shapes people’s health. For the Building H Index, we assess how popular products and services influence the health behaviors of their users. In order to understand the ways that products like Facebook, TikTok and PlayStation affect social connection, we needed to dig deeper.
With support from Einhorn Collaborative, Building H contracted with the research firm Verasight to conduct a survey of US consumers about their use of certain types of commercial products and services and their health habits. Respondents were asked about their use of and experiences with popular social media platforms, video and computer games, automobiles and bicycles, along with questions about their eating habits, levels of physical activity, sleep, the amount of time they spent outdoors and, in particular, their social connection and social experience. The online survey was completed by a demographically representative sample of over 3,100 US adults and was conducted between August 22 and September 5, 2023.
This report focuses primarily on one behavior — social engagement — and use of the social media platforms Facebook, Instagram, Snapchat, TikTok and YouTube and gaming platforms by Nintendo, PlayStation, Xbox, Steam and Meta Quest. We worked with the Foundation for Social Connection and members of their Scientific Advisory Council, specifically Carla Perissinotto and Matthew Smith, to craft survey questions that would probe Americans’ social experiences generally and with these products/services specifically.
Findings
Loneliness and Social Connection
The US has been described as being in the midst of an “epidemic of loneliness and isolation,” which is serious enough to merit an advisory by US Surgeon General Vivek Murthy. Our surveys corroborate current trends on loneliness and isolation.
- Forty-six (46) percent of survey respondents were lonely, using the UCLA Three-Item Loneliness scale. Loneliness varied somewhat by gender, with women more likely to be lonely (50%) than men (42%) and by race, with Whites, Blacks and Hispanics at 44%, 45% and 51% respectively.
- Loneliness varied considerably by age and income, with both younger people and people with lower incomes reporting higher levels of loneliness.
In addition to the loneliness scale, we asked all respondents two questions about their relationships and their social activity:
- Sixty-one (61) percent of respondents indicated that they were satisfied with the kinds of relationships they have with family and friends, versus somewhat dissatisfied (31%) and very dissatisfied (8%).
- Twenty-five (25) percent of respondents indicated that they had not spent time with someone who didn’t live with them — meaning a visit to, a visit from, or going out together — in the past week. Fifty-three (53) percent reported having two or more visits.
Social Media Use
Social media use in the US continues to be widespread and robust. Overall, 68% of US adults report spending an hour or more per day on at least one of the five platforms — Facebook, Instagram, Snapchat, TikTok and YouTube — we asked about. Thirty-eight (38) percent of respondents indicated using at least one social media platform “almost constantly.”
Use of the different platforms broke down as follows:
We asked the usage question two different ways — asking overall frequency (e.g. daily, several times a day) and then, for people who use a given platform at least daily, approximately how much time they spend on the platform per day.
We also looked to see if there were demographic patterns across social media use. The most notable differences were:
- Women (41%) were somewhat more likely than men (35%) to use at least one social media platform “almost constantly.”
- Black people (55%) were more likely to report “almost constant” use on one or more platforms than White people (33%) or people of Hispanic origin (45%).
- Younger users tended to use the platforms more intensely, with 48% of those aged 30–49 using at least one platform “almost constantly,” as compared with people between the ages of 50 and 64 (32%) and people 65 and over (24%).
Platform preferences also varied considerably by age, with Facebook users skewing older and the other platforms better represented in the 18–29 and 30–49 age groups.
Overall social media use did not vary substantially by income categories.
Use of time limits
Most social media platforms offer users opportunities to set limits on use, such as overall time per day, time per session, or blocking certain times of the day. Our survey suggests that these limits have had a relatively small impact. Two-thirds of all social media users have not set limits on any of the platforms and no more than 20% have set limits on any particular platform. Of those who have used these capabilities, 38% of them override them “frequently” (20%) or “always or almost always” (18%).
Game Playing
We asked survey respondents about their experiences playing several of the major game platforms: PlayStation, Nintendo, Xbox, Steam, and two virtual reality (VR) platforms: Meta (formerly Oculus) Quest and PlayStation VR.
Game play is less widespread than social media use, although, among daily users, it appears to capture comparable time.
- Fifty-three (53) percent of respondents play games on at least one of the major platforms and 33% spend an hour or more per day on them.
- Very few respondents (no more than 6%) reported playing games “almost constantly” on any of the platforms, although 25%-34% reported playing two hours or more per day on at least one of the platforms.
Time spent playing video games varied considerably by demographic factors.
- Fifty-eight (58) percent of men play on one of the major platforms, versus 49% of women.
- More people of Hispanic origin (68%) and Black people (64%) report playing on at least one major platform than Whites (47%).
- There is a sharp gradation by age, with people aged 18–29 and 30–49 playing substantially more than older respondents.
- There is also a gradation by income: people with higher incomes are more likely to play video games and more likely to play them for two or more hours per day than people in lower income brackets.
Social Connection and Social Media
Social media platforms are, by design, about supporting social interactions, but there is plenty of controversy about the effects that social media have on social connection. To examine this relationship, we looked at a general measure of loneliness; several measures of social connection; and a longer scale that probed users’ perceptions of how social media platforms are fulfilling different social functions and providing social experiences.
We defined several groups of social media users:
- light social media users, meaning that they didn’t use any of the five platforms more than 30 minutes per day
- moderate social media users, who didn’t use any of the platforms more than one hour per day
- heavy Facebook users, meaning those who reported using Facebook two hours or more per day
- heavy Instagram users, who reported using Instagram two hours or more per day
- heavy Snapchat users, who reported using Snapchat two hours or more per day
- heavy TikTok users, who reported using TikTok two hours or more per day
- heavy YouTube users, who reported using YouTube two hours or more per day
- heavy social media users, meaning those who reported using at least one of the five platforms for two hours or more per day
First, we looked at the association between respondents’ scores on the Three-Item Loneliness scale and their social media use and found a correlation between greater use of social media and increased likelihood of loneliness.
Second, we asked about satisfaction with relationships — about how satisfied people were with their relationships with family and friends — and, for a marker of social activity, how many times they had social visits with people outside their household in the past week.
Linking social media experience to social relationships, we asked respondents whether their use of social media had made them more connected to family, friends, and neighbors or others in their community. The results suggest that heavier users of social media do feel that it has helped them become more connected — especially to friends and family.
We saw two notable demographic variations to this question. Younger people were more likely to report that social media use made them feel more connected to neighbors and others in their community, but also less likely — in the case of ages 18–29 — to report that it resulted in more connection to family.
There also appeared to be an income divide over whether social media had made people feel more connected — with those in the $100,000 — $150,000 income group reporting better connections to friends and family than any other income group.
We then asked respondents to consider the social media platform they use most often and rate how much social media had influenced their relationships and social interactions, for better or for worse, using a version of the Retrospective Assessment for Connection Impact (RACI) scale developed by Matthew Smith and adapted for this project. The trend from these responses is clear — heavier users of social media found social media to have a greater positive impact on their social relationships and interactions than moderate or light users of social media.
Looking at the averages across the 13 different aspects of social connection in the above scale, light social media users reported the least positive influence (0.26). Moderate social media users rated the influence an average of 0.46 and heavy users came in at 0.97.
On average women (0.79) rated social media’s influence higher than men (0.70). Black people had the highest assessment (0.96), as compared to 0.83 for people of Hispanic origin and 0.70 for Whites.
We also found considerable differences by age and income groups. People aged 30–49 averaged ratings of 1.04, as compared with 18–29 (0.88), 50–64 (0.57) and 65+ (0.37). People earning between $100,000 and $150,000 had the highest ratings (0.99), followed by people making $150,000 or more (0.97), as compared with people in the $50,000-$100,000 group (0.74) and people making less than $50,000 (0.56).
Games and Social Connection
In looking at social connection and game playing, we first examined social practices associated with game playing: whether people played single or multiplayer games, whether they played with others in person or online, and whether they used text chatting while playing.
A majority of game players report playing single player games “almost always” or “mostly,” but there is a considerable amount of play in multiplayer games.
According to the survey, playing with friends takes place in person or online in roughly equal measures.
Text chatting with other players while gaming is also relatively common.
All of these practices had some demographic variation:
- Women were more likely than men to “always or almost always” play single player games (40% to 32%) and less likely to “always or almost always” or “frequently” text with friends while playing (25% to 36%).
- Age is also a factor, with younger generations more likely to play multiplayer games and text with each other while playing.
As with social media, we also looked at the question of how playing video games relates to social connection by looking at a general measure of loneliness; other measures of social connection; and the longer scale, from Smith, on users’ perceptions of how their game play fulfills different social functions and provides social experiences.
We defined several groups of game players:
- non-gamers, who reported that they did not use any of the six platforms — PlayStation, Nintendo, Xbox, Steam, PlayStation VR or Meta Quest
- light users, meaning that they reported playing no more than two hours per week
- moderate users, who reported playing between two and five hours per week
- heavy users, who reported playing more than five hours per week
- heavy PlayStation users, who reported playing games on their PlayStation for at least one hour per day
- heavy Nintendo users, who reported playing games on their Nintendo system for at least one hour per day
- heavy Xbox users, who reported playing games on their Xbox for at least one hour per day
- heavy Steam users, who reported playing games on Steam for at least one hour per day
- heavy PlayStation VR users, who reported playing games on their PlayStation VR for at least one hour per day
- heavy Quest users, who reported playing games on their Meta (or Oculus) Quest for at least one hour per day
We found a correlation between higher levels of overall game play and increased likelihood of loneliness.
We found only minor differences in associations between loneliness and the different game platforms used.
In terms of relationship satisfaction, there weren’t significant differences across levels of play, but more time spent gaming was associated with higher levels of in-person social activity, as measured by the question on the number of visits with people outside of their household in the past week.
Very similar to the reports from social media users, most game players report that using games has been a net positive for many aspects of social connection and that more usage is associated with more positive assessments of the influence on social connection.
Averaging the scores across these 13 different elements of the scale, the results were: light users 0.07, moderate users 0.81 and heavy users 0.96.
We did see some demographic variations in these scores:
- The average score for men was 0.80, versus 0.50 for women.
- The average score for Black people was 0.79, versus 0.71 for people of Hispanic origin and 0.63 for White people.
- People aged 30–49 (0.94) and 18–29 (0.80) gave the highest ratings, contrasting sharply with those aged 50–64 (0.23) and 65+ (-0.28).
- The highest ratings came from people making $150,000 or more (1.10) and $100,000 to $150,000 (0.96), as compared with people making $50,000 to $100,000 (0.54) and people making less than $50,000 (0.50).
Some of these differences might be due to different levels of usage among the different groups. For example, “heavy” users are more likely to be men. The gap between scores for men and women is narrowed considerably (1.00 to 0.90) if one looks only at heavy users.
Discussion
The survey results put additional contours around the loneliness epidemic. The epidemic is hitting hardest among young Americans and among those with lower incomes — but it is also clearly widespread: no demographic group we examined had less than one-third of its members assessed as lonely.
The usage numbers for both social media and video games show the extent to which they have become — and remained — deeply embedded in everyday life. Virtually all Americans use at least one of the major social media platforms we examined and more than a third use one or more of the platforms “almost constantly.” Video game use has a different pattern — roughly half the country doesn’t use any of the major game platforms we referenced, but among those who do, roughly two-thirds spent an hour a day or more playing on them.
The high usage rates for both social media and video games and the high rates of loneliness leads to an apparent paradox: higher usage of both is associated with higher rates of loneliness, but heavy users reported higher levels of social activity and they rate the utility of social media and games higher than lighter users. Heavier users of both had more social visits and there was a clear trend among usage and ratings on how the products had influenced many aspects of social connection — the heavier users consistently rated them higher. Heavy social media users were also more likely to report that social media has made them feel better connected to friends, family and neighbors than lighter users.
It’s important to be clear that this survey does not establish causation — heavy users of games and social media might be spending time on them because they are lonely; they might be lonely because they are spending some much time on them; or their loneliness could be unrelated to their engagement in social media and video games.
The apparent paradox could have something to do with the subjective nature of loneliness — it is a perception rather than an objective measurement of one’s social connection — and also the way it was calculated for the survey. We used the Three-Item Loneliness Scale that is based on questions about companionship, feeling left out and feelings of isolation. Just as it is possible to feel lonely at a party, it is conceivable that heavier users of social media and games are more socially engaged, yet have higher expectations or greater needs for companionship and inclusion.
The higher ratings for the influences that both social media and games have in social connection associated with more usage suggests people are spending more time on each at least in part because they find that it has value to them. That implication seems obvious, but it also seems to suggest more agency on the part of users. It suggests that heavier use of both social media and games could be rational and purposeful. Of course social connection is not the only value one gets from social media and from video games and might not be a driver of time spent on either.
The positive ratings for social media could also reflect the extent to which these services, whose use is so widespread, have become indispensable for social connection. Given the ebbing of in-person interaction resulting from remote work and the increasing availability of services (e.g. shopping, entertainment) that can be performed from home and the fact that so many people can be reached via social media, the utility of these platforms has been elevated.
Several of the demographic variations, such as younger generations using social media and playing games more than older generations and finding more value in them, are unsurprising. Others, such as the gender gap in rating the utility of games for social connection, are troubling and bear further research to probe on the underlying reasons. A consistent trend is the income divide — with people of higher incomes being less lonely, believing that social media has made them more connected, and gaining more social connection utility from both social media and video games than people at lower income levels. This divide could be a reflection of how the services and products are designed and targeted and, if so, could also represent an opportunity for companies to adapt them to be more valuable to a wider range of people. More research on this issue is certainly warranted.
As noted earlier, this report represents our analysis of the influences of social media and video games on one behavior — social connection. Subsequent reports will look at how these products and services affect other behaviors, including eating habits, physical activity, sleep and time spent outdoors, that affect people’s physical and mental health.
Acknowledgments
Building H’s Alec McMorris and Evan Cook helped with analysis of the survey data. The team at Verasight — Amelia Goranson, Jake Rothschild and Ben Leff, executed the survey and prepared results. Abigail Barth, Carla Perissinotto and Matthew Smith from the Foundation for Social Connection and their Scientific Advisory Council collaborated on the development of the survey questions. Building H co-founder Thomas Goetz and core collaborator Sara Singer contributed to the conceptualization and design of the surveys.
Building H is a project of the Public Health Institute.
Additional Reports:
Additional reports in this series include:
