Do 'watching eyes' influence antisocial behavior? A systematic review and meta-analysis

Eye cues have been shown to stimulate rapid, reflexive, unconscious processing and in many experimental settings to cue increased prosocial and decreased antisocial behaviour. Eye cues are being widely applied in public policy to reduce crime and antisocial behaviour. Recently, failed replication attempts and two metaanalyses examining the eye cue effect on generosity have raised doubts regarding earlier findings. Much of the wider evidence on eye cues has still not been systematically reviewed, notably that which is most relevant to its practical application: the effect of eye cues on antisocial behaviour. Given the evidence of humans' heightened sensitivity to threat and negative information, we hypothesized that the watching eyes effect would be more consistent in studies examining antisocial behaviour. In our meta-analysis of 15 experiments from 13 research papers we report a reduction in the risk of antisocial behaviour of 35% when eye cues are present. By contrast, systematic reviews have suggested CCTV cameras reduce crime by only 16%. We conclude that there is sufficient evidence of a watching eyes effect on antisocial behaviour to justify their use in the very low-cost and potentially high-impact real-world interventions that are proliferating in public policy, particularly in the UK. Public significance statement: Our meta-analysis of 15 experiments involving 2035 participants shows that photographs and/or stylized images of eyes reduced antisocial behaviour by 35%. Our findings support public policy initiatives employing pictures of ‘watching eyes’ to reduce crime. Furthermore, in an age when we are watched more than at any time in modern history – both online and on the street – our findings highlight an urgent need to fully understand the effect that perceived surveillance, feeling watched, has on our decisions and action