The confidence pill: The causal effect of exposure to police social media

Social media has become an important channel for police forces to communicate directly with the public. Police use these platforms to:

• report enforcement activity
• highlight engagement with local communities
• request assistance from the public
• communicate information about crime

However, there is limited causal evidence on how exposure to this content affects public attitudes towards the police or the wellbeing of those viewing it.

This project examines whether:

• police social media communication affects public confidence in the police
• these effects depend on the type of content being communicated

It also considers whether increasing the visibility of crime and enforcement activity through social media may have unintended effects on viewers’ wellbeing.

We conducted a large-scale online randomised experiment with participants in the UK. Participants were randomly exposed either to real Instagram content previously posted by English police forces or to neutral social media content. Police posts were divided into three categories:

• enforcement outcomes
• community engagement
• requests for public assistance

This design allows us to identify the causal effect of exposure to police social media and to compare the effects of different communication strategies.

Objectives

The project’s main objectives are to:

• establish whether police social media affects confidence in the police
• examine whether it changes perceptions of police effectiveness, crime concerns, reporting intentions and support for policing
• assess its immediate effects on subjective wellbeing
• determine whether different types of police content produce different outcomes

The randomised controlled trial (RCT) was designed to test whether:

• exposure to police social media causally affects public confidence in the police and other attitudes towards policing
• effects differ according to the type of police content shown

The main hypotheses were that exposure to police social media would affect: 

• confidence in the police
• perceptions of police effectiveness and crime
• willingness to cooperate with and support the police
• participants’ subjective wellbeing

The experiment additionally tested whether these effects differed across enforcement-focused posts, community-engagement posts and requests for public assistance.

The null hypotheses were that exposure to police social media would have no effect on these outcomes relative to the control condition, and that there would be no differences in effects across the three types of police content.

• UK (participants).
• Social media content from police forces in England.

Approximately 6,300 participants.

Participants were adult in the UK, recruited to take part in an online survey experiment. Eligibility required participants to meet the study’s recruitment criteria and provide informed consent to participate. Participants were randomly allocated to one of four experimental conditions.

Participants assigned to a treatment condition were exposed to real Instagram posts previously published by police forces in England.

The intervention had three treatment arms representing common forms of police social media communication:

• enforcement – posts communicating police enforcement activities and outcomes
• community – posts presenting police engagement with local communities
• public assistance – posts in which police requested information or assistance from members of the public

Random assignment to these conditions allows the effects of exposure to police social media to be separated from differences between people who ordinarily choose whether or not to follow or engage with police accounts.

Randomised comparing treatment to control. Participants assigned to the control group were exposed to neutral social media content rather than police social media content. The control condition therefore provides the counterfactual against which the effects of exposure to police content are estimated.

The trial measured a range of outcomes concerning perceptions of policing, intended behaviour and subjective wellbeing.

The principal outcomes included:

• confidence in the police
• concern about crime
• perceptions of police effectiveness (including perceived crime clearance rates)
• willingness to report crime
• preferences concerning police funding
• willingness to make a real monetary donation

The experiment also measured participants’ subjective wellbeing, including happiness following exposure to the experimental content.

Outcomes were collected through the online survey following random assignment and exposure to the experimental material. Where applicable, survey questions used rating scales or elicited participants’ stated beliefs and behavioural intentions. The donation outcome provided a behavioural measure.

The experiment shows that police social media communication can affect both public attitudes towards the police and the wellbeing of those exposed to it.

Exposure to police social media increased confidence in the police. However, the effects depended importantly on the type of content participants viewed. In particular, posts emphasising community engagement performed better than enforcement-focused communication in generating favourable responses towards the police. The findings therefore indicate that the content of police communication matters, rather than social media exposure having a uniform effect irrespective of what police forces communicate.

At the same time, exposure to police social media had an unintended cost. Participants exposed to police content reported lower short-term happiness. This suggests a potential trade-off in police communication: content that increases the visibility of policing and improves confidence in the institution may simultaneously make viewers feel worse in the short term.

Across the broader set of outcomes, the results do not suggest that police social media uniformly changes all attitudes or intended behaviours. Instead, the clearest effects concern confidence in the police and immediate subjective wellbeing, with important differences according to the content of the communication.

Overall, the RCT provides causal evidence that police social media is not simply an informational tool. The way police forces present their activities online can influence how the public evaluates the police and can also affect viewers’ emotional responses. The results suggest that police forces should consider both the institutional benefits and potential wellbeing consequences of different social media communication strategies.

Additional resources

The full research reportis available to download.