This post was written by Keith Dawson for UBM Tech’s community Web site All LED Lighting, sponsored by Philips Lumileds. It is archived here because the All LED Lighting site has gone dark. This material is Copyright 2013-2015 by UBM Americas.


Street Lights & Crime: It's Fractal

Do street lights deter crime? It depends. But we don't know on what.

See if you can isolate the causative factors behind the following true statements.

  1. Increased street lighting in Harrisburg; Indianapolis; New Orleans; and Portland, Ore., did not coincide with a drop in crime rates.
  2. Increased street lighting in Atlanta; Fort Worth; Kansas City, Mo.; and Milwaukee did correlate with a drop in crime.
  3. Fort Worth saw a decrease in all types of crime, but in Kansas City only violent crime abated.
  4. When street lights in Brantwood, Essex, UK were turned off to save money, burglaries, vandalism, and muggings spiked. The lights were turned back on.
  5. When street lights were turned off in Warwikshire, UK, to save money, violent crime and antisocial behavior fell by 20%. But a teenager was killed by a taxi, and investigation found that darkness was a factor.

(1., 2., and 3. come out of a 2007 systematic review of US and UK studies on light and crime, reported in The Atlantic's CityLab. 4. and 5. are based on anecdote and press accounts, reported in Lux Review.)

This has been the pattern since the 1980s when social scientists began in earnest to investigate the relationship of lighting to crime rates and safety: Results are all over the map.

Yet we feel safer when there is more light. The feeling is universal and ancient. But in today's complex society, that feeling of security might not translate into actual safety.

A theory of causation
The following is what I take away from a study of the conundrum of light and crime: there are factors we aren't considering, perhaps even factors we can't quantify.

Suppose the outcome of an experiment with more light, or less light, depends on the details of the local criminal population? Suppose even one criminal, bolder or smarter than the rest, uses increased lighting to improve his casing skills and to improve his perception of the chances of getting caught. A one-man crime spree could skew the statistics.

In most areas, police know the makeup of the local criminal population, and could probably characterize it if asked. The problem is, that population is nowhere static.

Suppose it depends on the population makeup of the locality under consideration? An older population might be more vulnerable. An area with two antagonistic ethnic groups could see more, or less, violence with less, or more, street lighting.

Suppose it depends on factors even less measurable, such as cultural history or the average of personal expectations of well-being in the population?

It would be excellent to be able to say, and know with a good certainty: "If we brighten the lights in this neighborhood of Billings, Mont., the crime statistics should improve." But we're a long way from that level of knowledge, and I'm not sure whether it will ever be possible to get there.

The question may be not just fractally complex, but unknowably so.

— Keith Dawson Circle me on Google+ Follow me on Twitter Visit my LinkedIn page, Editor-in-Chief, All LED Lighting