The Society of Light and Lighting’s ‘Illuminating the Future’ conference in March considered the links between darkness, crime and safety, and lighting, with Sheffield University’s Dr Jim Uttley among those unpicking the complexities of this area. Here is an abridged version of his presentation.
In any discussion about the relationship between darkness, lighting and crime there is often an inherent assumption that lighting can influence perceptions of safety and actual crime as well.
For example, BS 5489-1:2020 states that lighting has ‘the potential of helping to reduce crime and the fear of crime’. As an academic, my job is to question this assumption that lighting can reduce crime and the fear of crime.
For millennia, darkness and night have been associated with the fear and anxiety of ‘what is out there?’, whether that could have been an animal trying to eat us or a criminal trying to rob us, or worse. Humans have an inherent fear of the dark; darkness produces feelings of anxiousness, anxiety and a lack of safety.
When we start to look at the evidence, it strongly supports a relationship between darkness, lighting and perceptions of safety; how safe we feel. There have been lots of studies which have demonstrated that when it’s dark we feel a lot less safe. But, many studies have also demonstrated that lighting can offset the reduced perception of safety we feel after dark.
RATING SAFETY AT STREET LOCATIONS
For example, a recent study in Slovenia by Wei et al (2024) used what we call a day and dark approach to look at whether people feel less safe at night. People were taken to 12 different locations and asked to give ratings of how safe they felt both in daytime and at night.
For all the locations, the day/dark difference was positive. In other words, feelings of safety were higher in the daytime compared with after dark. And that, by and large, is what we’d expect; that people feel less safe when it’s dark.
One interesting element of this study was that the 12 locations had a range of lighting conditions, with a range of minimum illuminance. When you increased the minimum illuminance, the difference between feelings of safety in daylight and darkness started to reduce.
This study is just one example of many studies that use subjective evaluations to demonstrate perceptions of safety are lower after dark but that lighting can make a difference.
People have this fear of crime, a fear of being victimised. That’s why they feel less safe; that’s why they’re less likely to walk at night. But is this fair? Is it accurate to assume that you are going to be less safe because of darkness? Does darkness actually increase the risk of crime? And if this is the case can lighting help with that? Can it offset that increased risk?
Again, lots of research has been done on these questions. However, unlike the evidence about the impact of light on fear of crime, evidence is very mixed about the impact that light has on actual crime.
SHEFFIELD UNIVERSITY RESEARCH
This is where research we’re carrying out at the University of Sheffield comes in. To understand the effect lighting has on crime, you first have to understand whether darkness does in fact influence crime or not, whether the change from daylight to darkness does change crime rates.
If darkness doesn’t change the risk of crime taking place, then it doesn’t make sense to think that lighting will make any difference – street lighting provides a much smaller change in light level compared with the change between daylight and darkness.
Furthermore, if we can measure what effect darkness has on crime we can use that as a measure to evaluate street lighting and whether different lighting strategies change the risk of crime taking place.
To that end, this is a research area where we’re working on with South Yorkshire Police. I ought to caution here that this research is at a very preliminary stage and our results are currently under academic review.
We’re using what is called a ‘case and control’ approach. We select certain hours that transition between darkness to daylight and back to darkness during the year, and compare the counts of crimes in the daylight and dark periods of these hours. This controls for the influence time of day has on risk of crime being committed. However, seasonal changes, such as changes in weather conditions, will also influence crime risk. We therefore count crimes at hours of the day that remain in the same light condition throughout the year, and use these to control for seasonal effects.
We looked at crimes taking place over a 10-year period, so a fairly large sample of data. From these counts of crimes in our case and control hours, we created an ‘odds ratio’. This is a measure of the effect that darkness has on crime. An odds ratio greater than 1 indicates a greater risk of crime after dark compared with daylight. We calculated this for a number of different crime categories and then for crime overall, adding confidence levels as a measure of uncertainty – we tended to be less certain about the effect of darkness on the risk of crime for those crime categories that had fewer occurrences.
We found that certain crime categories and total crime overall did show an increase in risk when it was dark. There does, on these findings, seem to be an increased risk of crime taking place when it is dark.
The next question, of course, is can lighting can offset this? Can lighting reduce the negative effect darkness has on risk of crime?
Again, I need to emphasise that our results here are preliminary. We’re still working through the data and our final findings could be slightly different to those discussed here.
We looked at the locations of all street lighting in Sheffield, using lighting data from the local authority. We then plotted this data on a map and categorised every single road in our study area based on how many street lights it had.
We put each road into one of four categories: whether no lighting was present at all or whether there was a low, medium or high density of lighting. From here we calculated the odds ratio to measure the effect darkness was having on crime – with, again, a score higher than 1 suggesting an increase in risk after dark – calculating this for each light density category, and for different crime categories as well as crime overall.
The results so far – and I stress our findings are preliminary – have been nuanced. For bicycle theft, for example, as you increase the density of lighting, so add more lighting, the risk of crime after dark starts to reduce, the odds ratio starts to reduce. That might be the effect you would anticipate – that more lighting equals less crime.
But for burglary, the risk of crime after dark actually seems to be slightly higher if you have a low level of lighting compared with no lighting at all. It is only when you get to quite high densities of lighting does the risk of burglary after dark start to reduce.
We found a similar pattern of effect for robbery. A low density of lighting almost seems to be slightly worse than having no lighting at all. Equally, with criminal damage, it is only when you have get really high densities of lighting that it seems to bring down the risk of this taking place.
CONCLUSIONS AND NEXT STEPS
So, what is the key message from all this? It is that the question of whether lighting can reduce crime, and how we use lighting to reduce crime, is not simple to answer. There are nuances; it is dependent on the crime type that you’re looking at, which, in turn, will influence the answer as to whether lighting has an effect or not.
To summarise, we do have strong evidence that darkness and lighting influence perceptions of safety, the fear of crime. But previous evidence about whether lighting and actual crime have a relationship is quite mixed. Our research seems to suggest that, yes, darkness does increase the risk of crime overall, and specifically for certain crime types
However, the influence that lighting has on crime seems to vary depending on the crime type. Our initial results even suggest that, for some types of crime, an absence of lighting helps to reduce the risk of crimes taking place after dark.
Finally, what are our next steps? The initial results I have highlighted were just for Sheffield. We are now extending this study to three other local authorities in South Yorkshire. We are also extending it to local authorities outside South Yorkshire, through the LiveLabs2 project.
We have also so far been using a very crude measure of lighting: the density of lighting – so the number of lamps. We’ve not taken account of more refined metrics, such as illuminance levels or uniformity.
So that will be a further next step, to use more refined measures of lighting. We also want this research to have impact, not just to sit in academic journals.
To that end, we are working with stakeholders – local police and local authorities – to co-produce recommendations for how we might use this research to influence actual policing and the use of street lighting to reduce crime.
Dr Jim Uttley BSc, PGDip, PhD is lecturer in architectural science at the University of Sheffield
Image: Sheffield at night (Shutterstock)
This is an abridged version of the article that appears in the May edition of Lighting Journal. To read the full article, simply click on the page-turner to your right.
