Review question
Does improving street lighting in a residential area reduce crime, and if so, through what mechanism?
The question is deliberately narrow. "Does lighting work" cannot be answered. "Does improving lighting in residential areas reduce recorded crime, and is the reduction confined to hours of darkness" can be, because it specifies the intervention, the setting, the outcome measure and a test that distinguishes between explanations.
Why the mechanism matters
If better lighting reduces crime by making offenders more visible, the reduction should appear at night and not during the day. If it reduces crime by signalling investment and increasing the use of public space, daytime crime may fall as well. The two explanations imply different things about where else the intervention might work, so a review that reports only the headline effect is less useful than one that examines the pattern.
Search and screening
A real review records what was searched and what was excluded, so that someone else can check it or repeat it. The counts below are synthetic and illustrate the format.
| Stage | Records | Reason for exclusion |
|---|---|---|
| Identified through database searching | 214 | |
| Duplicates removed | -47 | Same study in multiple databases |
| Title and abstract screened | 167 | |
| Excluded at title and abstract | -121 | Not an evaluation, or no crime outcome measured |
| Full text assessed | 46 | |
| Excluded at full text | -33 | No comparison area, or lighting changed alongside other measures |
| Included in the review | 13 |
The largest exclusion category at full text is the important one. Where lighting was improved at the same time as patrols increased or vacant buildings were cleared, the evaluation cannot separate the effect of lighting from everything else that changed. Studies of that kind say nothing about lighting specifically, however encouraging their results look.
What the evidence indicates
Systematic reviews of street lighting interventions have found that improved lighting is associated with reductions in recorded crime in the areas where it is installed, relative to comparison areas. The reviews have been conducted across multiple studies in several countries and are among the more established bodies of evidence in situational crime prevention.
The more interesting finding concerns timing. Where reviews have been able to separate daytime from night-time crime, reductions have not been confined to the hours of darkness. That pattern is difficult to explain through improved visibility alone, and has been taken to support an interpretation in which lighting improvements signal investment in an area, increase its use, and strengthen informal surveillance.
Effects have varied considerably between studies. Context appears to matter: the condition of the lighting before the intervention, what else is happening in the area, and how crime is recorded locally all affect the result.
Confidence in these conclusions
| Statement | Confidence | Basis |
|---|---|---|
| Improved lighting is associated with reduced recorded crime | Reasonable | Multiple studies, several systematic reviews, consistent direction |
| The effect is not confined to hours of darkness | Moderate | Found where studies allow the comparison, but not all do |
| The mechanism is increased area use rather than visibility | Tentative | An inference from the timing pattern, not directly tested |
| Results transfer to a given setting | Low | Effects vary by context; most studies are from other countries |
Limits of this review
- Most of the underlying evaluation literature comes from high-income countries. Whether the findings transfer to a different setting is an open question, not an established one.
- Recorded crime is the outcome in most studies. If better lighting changes reporting behaviour as well as offending, recorded crime moves for two reasons at once.
- Displacement is rarely measured well. A reduction in the treated area is not the same as a reduction overall.
- Published evaluations may over-represent interventions that worked.
What this means for a programme decision
The evidence supports lighting improvement as a reasonable component of a crime prevention approach, particularly where existing lighting is poor. It does not support treating lighting as a solution on its own, and it does not allow a specific reduction to be forecast for a particular area. Anyone proposing lighting improvements should plan to measure the outcome locally rather than assume the effect found elsewhere.
Sources
Abbreviated for this sample. A real deliverable carries full citations with access dates for every source, and the search strings used.
- Welsh, B. C. and Farrington, D. P. Systematic review of the effects of improved street lighting on crime, published through the Campbell Collaboration and in related Home Office research.
- College of Policing, Crime Reduction Toolkit, entry on street lighting, which summarises the review evidence for a practitioner audience.