(source, the murder of Kitty Genovese is the archetypical although contested example of the bystander effect)
The bystander effect can explain the persistence of certain types or instances of rights violations. If many people witness a person in distress, then it’s the less likely that any one person will help. “I could help, but I’m sure someone will”. Numerous experiments have proven the effect. Virtually all of them find that the presence of others inhibits helping, often by a large margin. The probability of help is indeed inversely related to the number of bystanders, although not necessarily one-on-one. More precisely, the effect occurs when bystanders are strangers; when bystanders are friends help is usually forthcoming.
What are the reasons for this effect? Hard to tell, but social influence may be one: bystanders monitor the reactions of other people in an emergency situation to see if others think that it is necessary to intervene. If everyone first looks at the others, then you have a vicious circle of influence. Since everyone is doing exactly the same thing – i.e. nothing – they all conclude from the inaction of others that help is not needed. Diffusion of responsibility may be another reason: when a lot of people are present, they all assume that others carry more responsibility to intervene, because others may be seen as closer or stronger or first on the spot (this is also the thinking behind the firing squad or the Japanese procedure for capital punishment). The fear of being harmed or of offering unwanted assistance may also explain the effect.
Increasing urbanization and improved knowledge of everyday events (by way of better information systems such as the internet) can make the bystander effect more common, and can therefore make it more difficult to stop rights violations.