What the data actually shows
Several motives appear to operate together. Reciprocity-based accounts note that helping can pay off over time through returned favours and reputation, which gives even self-interested individuals reasons to help. Norm-based accounts point to widely shared expectations — to help those in need, and to help those who have helped us — that guide behaviour without requiring conscious calculation in the moment.
Batson's empathy-altruism hypothesis goes further, arguing that feeling empathy for someone in need can produce a motivation to help that is genuinely aimed at their welfare rather than at relieving our own discomfort or earning a reward. Across a series of experiments, Batson and colleagues reported that people high in empathy still helped even when it would have been easy to walk away unnoticed — a pattern they argued is hard to explain by self-interest alone, though the debate over whether 'pure' altruism exists is not fully settled.
The situational side is just as well documented. The bystander effect, studied by Latané and Darley after the Kitty Genovese case, found that people are often less likely to help when others are present, partly through diffusion of responsibility. But later reviews have qualified this: in clearly dangerous emergencies, bystanders frequently do intervene, and analyses of real-world footage suggest help arrives in a large majority of public conflicts. The effect is real but conditional, not a blanket law that people ignore those in need.
Why this feels different from how it actually is
Cynicism about human nature is loud and memorable, so it can feel as though people rarely help. Dramatic stories of bystanders doing nothing travel widely and stick, while the far more common cases of ordinary, unremarkable help go unreported precisely because they are unremarkable. The available examples are skewed toward the worst cases.
It also feels different because we tend to assume helping must be secretly selfish. When someone helps, it is easy to reach for the hidden motive — reputation, guilt relief, reciprocity — and conclude that 'real' altruism is a myth. The empathy-altruism research complicates that reflex: while mixed motives are common, the evidence suggests genuine other-focused concern is also part of the mix, not merely a cover story.
And the bystander effect gets remembered as a verdict on people rather than a description of a situation. The finding is about how the presence of others changes any one person's behaviour, not proof that people are indifferent. Reframed as 'situations shape helping,' it points to something more useful and less bleak than 'no one helps.'
Across a whole scene, help usually arrives — even though the presence of others can lower any one person's odds of acting.
What the research says to do about it
If you want help to be more likely — for yourself or others — the research on the bystander effect points to reducing ambiguity and diffusion of responsibility. Making a need clearly visible, and directing a request to a specific person rather than a crowd, tends to make help much more likely than a general appeal, because it removes the assumption that someone else will handle it.
Because empathy is a strong driver, anything that makes the person in need feel like a specific individual rather than an anonymous member of a crowd tends to increase helping. The research on the 'identifiable victim' pattern suggests people respond more to a single, concrete person than to abstract or large-scale need — a reason that personal, specific appeals often outperform statistics.
And it is worth correcting your own baseline expectation. Since people generally underestimate how willing others are to help — and how positively help is received — assuming the worst can stop you from asking when asking would have worked. The data suggests the realistic expectation is that most people help in most ordinary situations, which is a more accurate starting point than cynicism.
What the research says does not help
Assuming people are fundamentally indifferent is both inaccurate and self-defeating. The research suggests helping is common and that bystander reluctance is situational rather than a fixed trait, so treating cynicism as realism tends to discourage the very requests and offers that usually succeed.
Relying on a crowd to help is a known failure mode. The bystander effect shows that a vague appeal to many people can leave each one assuming someone else will act. Broadcasting a need to everyone is less effective than directing it clearly to a specific person, even though it feels like the opposite should be true.
Insisting that all helping is secretly selfish, and therefore not worth crediting, is not well supported and is unhelpful in practice. The evidence points to mixed motives — reciprocity, norms, and genuine empathy together — so dismissing altruism wholesale both overstates the case and discounts a real and common part of why people help.
Most people help more readily than a cynical view of human nature would predict.
What this looks like in real life
Ask one named person, not the crowd
Directing a request to a specific individual — 'you, in the blue jacket, can you call an ambulance?' — tends to make help far more likely than a general appeal to everyone. It removes the diffusion of responsibility that lets each bystander assume someone else will handle it. Reducing ambiguity about what's needed works the same way: a clearly visible need is acted on more readily than a vague one.
A situation, not a verdict on people
The bystander effect gets remembered as proof that people don't care. But it describes how the presence of others changes any one person's behaviour, not that people are indifferent to those in need. Read as 'situations shape helping' — and set against footage showing bystanders intervene in most public conflicts — it points somewhere more useful and far less bleak than 'no one helps.'
Real numbers in context
The cynical mental image is largely wrong. Analyses of real-world surveillance footage of public conflicts have found that bystanders intervene in a large majority of cases — one widely reported cross-national study put it at roughly nine in ten incidents — which sits in tension with the bleak 'nobody helps' reading of the bystander effect. The presence of others can lower any one person's odds of acting, yet across a whole scene help usually arrives.
On motivation, the empathy-altruism experiments are about pattern rather than a single statistic: people high in empathy tended to help even when escape was easy and unnoticed, which Batson argued is hard to square with pure self-interest. The honest summary is that helping is common and multiply motivated — partly reciprocity, partly norms, partly genuine concern — and that most people help more readily than a cynical estimate would suggest.