Why Good People Do Nothing: The Psychology of Moral Inaction

In 1964, a young woman was attacked outside her apartment building in New York. The incident became one of the most cited cases in the history of social psychology — not primarily because of what happened, but because of the newspaper story that followed: that 38 witnesses had watched from their windows and done nothing. The number and the story were later shown to be largely fabricated by the press. But the question the case sparked was real: why do groups of people sometimes fail to act when individuals are in obvious need?

Two psychologists, John Darley and Bibb Latané, decided to find out. The experiments they designed over the next decade produced results that are still deeply uncomfortable: the presence of other people does not make us more likely to help. In many circumstances, it makes us significantly less likely to help. They called it the bystander effect.

The experiment that changed how we think about moral action

In Darley and Latané's most famous study, participants sat alone or in groups while they believed a fellow participant began to have an epileptic seizure — communicated via an intercom. When participants were alone, 85 percent sought help within six minutes. When they believed four other people were also listening, that number dropped to 31 percent. The emergency was identical. The only variable was the number of observers.

What was happening was not callousness. Most participants who did not act were visibly distressed — they fidgeted, looked anxious, seemed genuinely upset. They were not indifferent to what they were hearing. They were caught in a social and cognitive trap that paralysed their response.

Diffusion of responsibility

The first trap is diffusion of responsibility. When many people witness an emergency, each individual feels personally less responsible for resolving it. If you are the only one who sees someone collapse, the moral weight falls entirely on you. If twenty people see it, your share of the responsibility feels like one twentieth — even though the person still needs help just as urgently. The group doesn't pool its capacity; it dilutes it.

This is not a character flaw unique to cowards. It is a feature of how human beings process shared moral situations. We evolved in small groups where responsibility was naturally clear. We did not evolve good intuitions for mass-observer scenarios. The same person who would act decisively alone may freeze completely when surrounded by others who are also watching.

Pluralistic ignorance

The second trap is pluralistic ignorance — a phenomenon where every individual in a group privately doubts the prevailing norm, but assumes everyone else endorses it. Each person looks at the others for signals about how to interpret an ambiguous situation. If nobody is acting, everyone assumes there must be a reason — maybe it is not really an emergency, maybe help is already on the way, maybe the right response is to wait.

The silence of the crowd is read as evidence that silence is appropriate. Everyone suppresses their individual concern and mirrors the inaction around them — while privately feeling uneasy. The result is a group that collectively produces a response (none) that no individual actually endorses.

Moral disengagement — how we quiet the alarm

Psychologist Albert Bandura identified a different but related mechanism: moral disengagement. Rather than being blocked from acting, people actively construct reasons why they are not obligated to act. Bandura identified several mechanisms through which this happens: moral justification (convincing yourself the action would cause greater harm), euphemistic labelling (calling inaction 'not getting involved' rather than 'abandoning someone in need'), diffusion of agency across a group, and dehumanising or depersonalising the victim.

These are not conscious strategies of self-deception. They operate quickly and automatically. By the time you have talked yourself out of intervening, you may genuinely believe the intervention was unnecessary or harmful. The alarm was real; the disengagement reasoning came after it, manufacturing justification for an inaction that was already attractive for other reasons.

The capable bystander problem

There is a philosophical tradition that holds we are more culpable for harms we cause than for harms we merely allow. The distinction between commission and omission is built into many legal systems and many people's moral intuitions. But the bystander research complicates this framework in an important way: being capable of preventing a harm and choosing not to — while being aware of the harm and the capacity — starts to look much more like a moral choice than a passive default.

A bystander who watches someone drown while holding a life ring has not merely failed to help. They have made a series of real-time decisions: not to throw the ring, not to call for help, not to shout. Inaction, in circumstances where action was available and obvious, is itself an act. The question is whether our moral intuitions are well-calibrated to treat it that way.

What SplitVote data shows about inaction

On SplitVote, dilemmas that involve clear action — pulling a lever, making an intervention, overriding a system — tend to produce closer splits than their omission equivalents. Scenarios where the question is whether to do something active to cause a harm consistently produce more resistance than scenarios where the same harm results from choosing not to act. The commission-omission asymmetry that philosophers have debated for decades appears clearly in how real people vote.

What is less clear from voting data is whether people are aware of this asymmetry in their own reasoning. Many voters who would decline to actively cause a harm would also decline to prevent an equivalent harm — and often feel the two positions are entirely consistent. The moral discomfort shows up differently depending on which direction the action faces.

What actually changes bystander behaviour

Darley and Latané also studied interventions that reduce the bystander effect. The most reliable is specificity: being directly addressed by name, or being clearly designated as the person responsible, dramatically increases the probability of action. "Someone call an ambulance" produces less response than "You, in the blue jacket — call an ambulance now." Responsibility that is diffuse stays diffuse; responsibility that is made explicit lands differently.

A second intervention is prior awareness of the effect itself. People who have been told about the bystander effect and diffusion of responsibility before entering a scenario are meaningfully more likely to act against the crowd signal. This is one of the few cases in social psychology where knowing the bias genuinely seems to reduce it — which is one reason the research continues to be worth discussing.

References to bystander effect research, diffusion of responsibility, pluralistic ignorance, and moral disengagement are for contextual and educational purposes. Original research described is attributed to Darley, Latané, and Bandura. The Kitty Genovese case summary reflects revised historical accounts that challenged early press reports. SplitVote is an entertainment and reflection platform, not a scientific study. Voting data described represents community patterns, not controlled experimental results.