What the data actually shows
The strongest framework here is Baumeister and Leary's 1995 review, 'The Need to Belong,' which gathered a wide body of evidence suggesting people are powerfully motivated to form and maintain lasting, positive relationships, react strongly to threats to their belonging, and suffer measurable effects when it is lacking. The authors treat belonging as a fundamental need on a par with other basic motivations, not a mere social nicety.
Evolutionary accounts add a plausible 'why.' For most of human history, group membership was closely tied to survival — access to food, protection, cooperation, and care all flowed through the group, and exclusion was genuinely dangerous. On this account, the discomfort of feeling left out is an old alarm system that once tracked real risk, which helps explain why social rejection registers so sharply even when nothing material is at stake.
Tajfel and Turner's social identity theory documents the flip side. Classic 'minimal group' experiments found that people will favour their own group over another even when the groups are formed on trivial, arbitrary grounds and members gain nothing personally. This suggests the pull toward in-group loyalty — and the tendency to draw a line against out-groups — is easy to trigger and partly independent of any real conflict of interest.
Why this feels different from how it actually is
The need to belong can feel like a personal failing because modern culture prizes independence and self-sufficiency, so wanting to be part of something can read as weakness. The research suggests the opposite: the drive is close to universal, and the people who seem most self-contained are usually embedded in groups they simply do not have to think about.
It also feels different because belonging is most visible by its absence. When you are securely part of a group, the need is quietly met and largely invisible; it announces itself loudly only when you are excluded, new, or between groups. That asymmetry can make the desire feel like a problem to be fixed rather than a standing feature of being human.
And the in-group pull can feel uncomfortable to admit. Because social identity attaches our self-worth to our groups, criticism of a group we belong to can feel like a personal attack, and we can find ourselves favouring 'our side' more than we would endorse on reflection. Recognising that this is a built-in tendency, not a private flaw, makes it easier to notice rather than deny.
The people who seem most self-contained are usually embedded in groups they simply do not have to think about.
What the research says to do about it
If belonging is a genuine need, the practical implication is to treat building and maintaining group ties as basic maintenance rather than a luxury — on the same footing as sleep or movement. The research on belonging and on isolation broadly suggests that stable, recurring membership in groups where you feel accepted is more protective than a large number of loose or one-off contacts.
Because belonging is met through being known and accepted over time, recurring, low-stakes contact tends to matter more than intensity. Groups built around a shared activity or routine — a team, a class, a regular gathering — give the repeated, structured contact that turns acquaintances into a sense of 'we,' which is harder to manufacture in occasional one-on-one meetings.
On the in-group front, the most consistent lesson from the social identity literature is that contact and shared, cooperative goals across group lines can soften out-group bias. You do not have to abandon your own groups to reduce the reflexive 'us versus them'; working alongside people from other groups toward a common aim tends to do more than simply telling yourself to be fair.
What the research says does not help
Trying to talk yourself out of needing groups rarely works, because the drive is not a bad habit but a basic motivation. Treating the desire for belonging as something to suppress tends to leave the need unmet rather than resolved, and the discomfort of exclusion does not respond well to being argued away.
Substituting a large quantity of shallow or purely online affiliations for genuine acceptance is an unreliable fix. Counts of contacts, followers, or memberships are weak proxies for the felt sense of being known and accepted that the belonging research points to. The number is not the thing.
And leaning harder into in-group loyalty as a cure for feeling adrift can backfire. The same process that gives belonging also sharpens out-group bias, so intensifying 'us versus them' may deliver a quick hit of identity while narrowing your world. Strong belonging and rigid hostility to outsiders are not the same thing, and the second is not required for the first.
Strong belonging and rigid hostility to outsiders are not the same thing, and the second is not required for the first.
What this looks like in real life
Favouring 'us' over nothing at all
In the classic minimal-group experiments, people were sorted into groups on arbitrary, meaningless grounds and gained nothing personally — yet they still allocated more to their own group than to the other. That a bias toward 'us' appears under such bare conditions is a strong hint that group identity is a default setting, not a response to real competition. It's a useful caution when group divisions feel like they must reflect deep differences.
Working toward a shared aim
You don't have to abandon your own groups to reduce reflexive out-group bias. The social identity research suggests that contact and shared, cooperative goals across group lines do more than simply telling yourself to be fair. Working alongside people from other groups toward a common aim tends to loosen the 'us versus them' pull while leaving your own belonging intact.
Real numbers in context
There is no single statistic that captures a need, but the supporting data is broad. Baumeister and Leary's 1995 review remains one of the most cited papers in social psychology precisely because the pattern it describes — strong motivation to belong, strong reactions to its threat — recurs across so many studies. The need shows up less as a number than as a consistent shape across decades of research.
The minimal group experiments are the striking quantitative thread: people reliably allocate more to their own group than to another even when the groups are assigned at random and they gain nothing themselves. That a bias toward 'us' appears under such bare conditions is a strong hint that group identity is a default setting rather than a response to real competition — a useful caution when group divisions feel like they must reflect deep differences.