That figure — drawn from the College of Policing’s systematic review of environmental crime prevention measures — sits at the centre of a debate with direct operational consequences for North East Lincolnshire. As the borough continues its LED conversion programme and responds to reports of faulty columns across residential estates, the question is not simply whether illumination matters. It is how confidently the wider evidence can be translated into local decisions.
For Grimsby, Cleethorpes and Immingham, the research offers a useful planning framework, but not a guaranteed local result. A systematic review can indicate the likely direction and scale of an intervention’s effect; it cannot establish that every replacement programme will produce the same reduction, or that a particular scheme will pay for itself within a defined period. Local layout, baseline crime, maintenance standards, reporting behaviour and the design of the intervention all matter.
What follows is therefore not an argument for lighting as a universal answer. It is an examination of what the evidence shows, where it remains uncertain, and what it implies for the borough’s preventative infrastructure.
The 21% Effect: What Systematic Reviews Reveal About Lighting and Crime
The most frequently cited figure in assessments of street lighting and crime comes from the Farrington and Welsh meta-analysis, first published in 2002, together with later reviews by the College of Policing and Welsh and colleagues. Across 13 controlled evaluations, areas receiving improved lighting showed a reduction in recorded crime of approximately 20% to 21% compared with matched control areas.
The importance of that result lies less in the number itself than in the design of the evidence behind it. The finding does not come from a single neighbourhood, a single lighting technology or one local authority programme. It is an average drawn from multiple evaluations in different settings. That makes it more useful than an isolated success story, while also making it less suitable as a direct forecast for one borough.
For a local authority allocating capital expenditure, the correct interpretation is cautious. The research indicates that lighting improvements can produce a measurable reduction in recorded crime under the conditions examined by the studies. It does not mean that a 21% reduction should automatically be applied to every estate, town-centre route or rural road. The local baseline, the condition of existing columns, the visibility of surrounding spaces and the way the scheme is implemented will influence the outcome.
A 21% reduction is a documented average across controlled evaluations, not a promise attached to every lighting upgrade.
That distinction matters when a project is assessed before installation. A local comparison needs to establish what is happening in the intervention area, what is happening in comparable areas without the upgrade, and whether both are being measured over a meaningful period. Otherwise, normal variation can be mistaken for an effect. Crime can rise or fall for reasons unrelated to lighting, including changes in policing, local development, seasonal activity, transport patterns or the availability of other services.
The mechanism also requires careful parsing. Improved lighting is not a single-cause variable. It may increase the perceived and actual risk of being observed in a public space, but it can also signal that an area is being maintained and monitored. Better visibility may affect pedestrian movement, informal supervision and the willingness of residents to use a space. Those effects can overlap, and the research does not reduce the result to one simple visibility calculation.
That is why a lighting audit should look beyond lamp output. It should consider whether a footpath is overlooked by homes, whether vegetation creates concealed approaches, whether a column illuminates the intended pavement rather than the carriageway, and whether darker gaps remain between upgraded points. A technically brighter road is not necessarily a safer public space if the light is poorly distributed or leaves key routes in shadow.
Beyond Night Vision: Why Improved Lighting Deters Crime During the Day
One of the more analytically useful findings in the evidence base is that reductions following lighting upgrades are not confined to nighttime hours. Several evaluations recorded decreases during daylight periods as well. That pattern complicates the narrow explanation that lighting works only by making offenders easier to identify after dark.
If the effect were purely about night-time visibility, a daytime reduction would be difficult to explain. The alternative account focuses on the way environmental improvements influence how a place is used and perceived. A well-maintained, adequately illuminated street can communicate that an area is watched, valued and subject to active stewardship. Residents may be more willing to walk through it, while potential offenders may face more informal observation and fewer opportunities to operate unnoticed.
This does not mean that a new LED fitting creates community cohesion by itself. Nor does it establish that every daytime reduction is caused by a change in public behaviour. It does mean that the intervention should not be evaluated solely as a technical response to darkness. Lighting forms part of the setting in which movement, supervision and confidence take place.
In practical terms, an LED upgrade on a residential cul-de-sac in the West Marsh is not simply a night-time intervention. Its effect may depend on what the surrounding environment communicates throughout the day: whether paths are usable, whether sightlines are open, whether public space appears maintained and whether people continue to pass through it.
For North East Lincolnshire Council, this points towards sequencing rather than isolated replacement. Lighting improvements may be more useful when considered alongside other environmental measures, such as vegetation management, clearer signage, repairs to surfaces and the removal of obstructions that create hidden areas. The available evidence supports thinking about the whole setting, but it does not justify claiming that bundled improvements will always produce a particular additional reduction. The benefit of integration should be tested locally rather than assumed.
Maintenance Is Part of the Intervention
The Council’s established reporting channels for defective street lighting provide a practical entry point for residents. A failed lamp, an intermittent fitting or a column that no longer illuminates the intended area is not merely a minor inconvenience. It reduces the performance of an asset that was installed to alter the conditions of a public space.
That does not mean every fault immediately reverses the crime-prevention effect of an entire scheme. The more defensible point is narrower: the findings from lighting evaluations relate to functioning interventions. If output is reduced, coverage is interrupted or important gaps remain unresolved, the local scheme may not resemble the intervention that was originally assessed.
Prompt reporting of broken street lights in Grimsby and elsewhere in the borough can therefore support both maintenance and evaluation. Residents often identify failures that are not obvious from a central inspection schedule, particularly where a fitting is flickering, partially obscured or illuminating unevenly. Reporting also creates a record of recurring problems, which can help distinguish a one-off fault from a wider issue involving equipment, positioning or maintenance practice.
A useful local audit should ask:
- Is the fitting working consistently, rather than only appearing operational during a daytime inspection?
- Does it illuminate the footway, crossing point or entrance that people actually use?
- Are there dark gaps between columns or around vegetation, walls and service entrances?
- Has the condition of the surrounding space changed since the lighting was installed?
- Are reported failures being resolved quickly enough to preserve the intended coverage?
- Can crime, disorder and resident-confidence data be compared before and after the change?
These questions do not turn a lighting scheme into a controlled experiment, but they prevent a simple installation count from being treated as evidence of success. The relevant asset is not a column on a register. It is a functioning public-space illumination system.
Economic Returns: What the New York Trial Can and Cannot Tell Us
The financial case for lighting has been examined directly in only a limited number of settings. One of the clearest examples in the supplied evidence is a randomised controlled trial conducted across New York City public-housing sites. That trial estimated that improved street lighting generated more than $750,000 per community annually in averted crime costs and produced a benefit-to-cost ratio of approximately four to one.
That is a significant result, but its scope must remain clear. The ratio belongs to the New York City trial. It is not a general ratio for street lighting, a standard return for LED conversion, or an established benchmark for North East Lincolnshire. The trial involved a particular population, housing environment, cost structure and pattern of implementation. Those conditions cannot simply be transferred to a UK borough without a separate local analysis.
The documented four-to-one return belongs to the New York City public-housing trial; it should not be presented as a universal payback rate for lighting schemes.
The same caution applies to the estimate of avoided crime costs. The value assigned to a prevented incident depends on what is included: policing and emergency response, criminal justice costs, victim services, property loss, healthcare, lost work and wider social impacts. Different systems price those consequences differently. A UK local authority cannot establish its own return merely by converting the dollar figure into pounds.
For North East Lincolnshire, a proper financial assessment would need to connect the cost of the local scheme with locally observed outcomes. That could include installation, electrical work, column replacement, design, inspections, repairs and ongoing energy or maintenance costs. On the benefits side, the authority would need to define which changes count as avoided costs and how confidently they can be attributed to the lighting intervention.
This is also why claims about payback should be kept separate from claims about public value. A scheme may improve residents’ confidence, make routes more usable and support a broader community-safety strategy even when the financial return cannot be demonstrated within a particular accounting period. Conversely, a scheme may show a favourable financial result in one setting without proving that the same return will appear in another.
The evidence supplied does not establish a UK or North East Lincolnshire payback period. Nor does it establish that lighting produces a higher return than most other infrastructure investments. Comparisons with transport, housing, public realm or other capital programmes would require a common valuation method and reliable evidence for each alternative. Without that work, cross-infrastructure rankings are not justified.
The 93% figure sometimes cited in discussions of lighting and women’s safety belongs to perception research rather than crime-cost analysis. It can help show why residents may value improvements even when crime incidence does not change in the same direction. But perceived safety, recorded crime and financial return are different outcomes. A serious business case should keep them visible as separate measures rather than combining them into one headline benefit.
The UK Context: Why British Studies Show Higher Success Rates
The Campbell Collaboration’s review of the international evidence identifies a substantial difference between the results of UK and US evaluations. Studies conducted in the United Kingdom recorded an average crime reduction of 38% following lighting improvements, compared with 7% in studies conducted in the United States.
That difference is important, but it should not be turned into a local target without qualification. The UK figure is an average from the studies included in that review. It does not mean that a North East Lincolnshire scheme should be expected to reduce crime by 38%, and it does not override the uncertainty attached to local conditions.
Several structural explanations may account for the divergence. Baseline lighting provision differs between jurisdictions, so the same upgrade can represent a different scale of change. Urban form also matters. More compact areas with clearer boundaries between treatment and control locations may make it easier to observe an effect than places where movement crosses several neighbourhoods. Differences in police recording practice, offence definitions and data quality can also influence the result.
There may be a selection effect as well. Schemes are not always introduced randomly. Authorities may choose locations with particular problems, visible disorder or strong community pressure. If an area improves after installation, it can be difficult to separate the contribution of lighting from the influence of other changes taking place at the same time.
For analysts working within North East Lincolnshire, the UK-specific result is relevant as context, not as a forecast. It suggests that British evaluations have sometimes found stronger effects than the international average. The wider 20% to 21% result may serve as a more conservative reference point, but even that should not be treated as a guaranteed local reduction. Expectations need to be accompanied by a monitoring design that can show whether a particular scheme is performing differently from comparable areas.
A sensible evaluation would combine several measures:
1. Recorded crime and disorder: track relevant categories before and after the change, while allowing for seasonal variation and wider trends.
2. Comparable locations: identify nearby areas that have not received the same upgrade, recognising that no comparison area is perfectly identical.
3. Night and day patterns: examine whether changes are concentrated after dark or appear across the full day.
4. Resident confidence: collect information about whether people feel safer and use the space differently.
5. Maintenance performance: record outages, repair times and recurring faults so that a failed asset is not counted as a functioning intervention.
6. Displacement and diffusion: check whether problems move to adjacent streets or whether improved use spreads beyond the upgraded location.
The point is not to burden every replacement with an elaborate research programme. It is to avoid making a strong financial or safety claim from a before-and-after comparison alone.
Nuances in Reporting: Understanding the Impact on Violent Crime Statistics
The evidence base is not uniformly positive across every crime category. The 2021 Welsh and colleagues’ systematic review, drawing on 21 studies, found that while street lighting improvements had a strong positive effect on property crime reduction, they were associated in some cases with a very small undesirable impact on reported violent-crime outcomes.
That result needs careful handling. A small increase in recorded violence after an upgrade does not automatically mean that the intervention caused more violence. Improved visibility can make incidents easier to identify. Greater confidence in the environment may also make victims and witnesses more willing to report harassment, assault or disorder. In that situation, the recorded total can rise because the reporting system is capturing more of what was already occurring.
The opposite interpretation is also possible in some settings: a change in lighting could affect how people use a space, who is present and when incidents occur. The available result does not allow every increase to be dismissed as better reporting. It requires a closer look at incident type, location, timing, repeat victimisation and the relationship between reports and other measures.
For local analysts, the operational warning is straightforward. Headline figures showing a small uptick in violent crime following a lighting upgrade should not be treated as conclusive evidence of failure, but neither should they be explained away without examination. The response should be to track reporting rates alongside severity, victim and witness information, and the distribution of incidents across the upgraded and comparison areas.
This is where coordinated preventative work becomes critical. Street lighting is a situational measure. It can alter visibility, guardianship and the use of public space, but it does not address the upstream drivers of violent behaviour, including substance misuse, domestic abuse dynamics or youth disengagement. It should not be asked to carry the whole burden of a community-safety strategy.
Where environmental improvements sit alongside diversionary activity, support services, appropriate enforcement and visible local engagement, the authority has more than one route for reducing harm. That layered approach may also make evaluation harder, because several interventions can influence the same area at once. Even so, the complexity is preferable to pretending that one infrastructure change can explain every movement in the crime data.
Projecting Forward: LED Conversion, Part-Night Dimming and the Local Horizon
The trajectory for North East Lincolnshire is shaped by the ongoing LED conversion programme and by wider questions about how lighting should be operated during low-footfall periods. Some local authority areas have adopted part-night policies under which lights are dimmed or switched off at specified times. The borough-level evidence supplied here does not provide a clean evaluation of how those arrangements affect antisocial behaviour or recorded crime.
That gap matters because an LED conversion is not one uniform intervention. The outcome depends on the fitting selected, its optical distribution, mounting height, spacing, controls, maintenance and operating schedule. A more efficient lamp can reduce energy use while still failing to illuminate the part of a route that residents need. Conversely, an apparently modest change can improve visibility if it corrects a badly positioned or unreliable fitting.
The broader evidence suggests that the documented benefit of improved lighting depends on the intervention being present and maintained. A failed lamp, a prolonged outage or a reduction in output changes the conditions under which the scheme is operating. It would be misleading to treat a dimmed or switched-off route as equivalent to a fully functioning upgrade when assessing the effect of lighting on night-time safety.
This does not settle the policy question in favour of continuous full illumination everywhere. Lighting also involves energy, environmental impact, nuisance, glare, dark-sky considerations and the needs of residents living close to columns. The point is that any operating policy should be evaluated against the function of the space, not just the nominal number of hours for which a fitting is switched on.
For residents, the practical implication is direct. Reporting a broken street light through the Council’s established channels is a way to protect the performance of the local network. It is especially useful to describe the location precisely, the nature of the fault and whether the problem affects a crossing, footpath, estate entrance or other route where people depend on consistent illumination.
For the authority, the equivalent task is to treat maintenance records as part of community-safety intelligence. Repeated failures in one area may reveal an equipment problem, a design weakness or a mismatch between the lighting plan and the way the space is used. Those records can support future community safety lighting audits and help distinguish an intervention that has not worked from one that has not been kept in working condition.
The forward projection should therefore be sober rather than celebratory. Street lighting is one of the better-evidenced situational crime-prevention measures available to a local authority. The research supports the view that improved lighting can reduce recorded crime, and the UK studies indicate that some British schemes have produced particularly strong results. But the evidence does not establish a universal return, a guaranteed local reduction or a standard payback period.
The New York City trial provides a documented example of a benefit-to-cost ratio of approximately four to one in a specific public-housing setting. It should remain a case study, not a template for North East Lincolnshire. Local financial returns, UK payback and comparisons with other infrastructure investments remain unverified unless they are supported by a separate local appraisal.
What lighting can provide, when correctly specified and properly maintained, is a stronger environmental baseline: a public space that is easier to navigate, easier to observe and potentially less attractive for certain forms of crime. Its value will depend less on the number of fittings installed than on whether the right places are illuminated, whether output is sustained, and whether the results are measured without overstating what the data can prove.