Friday, November 21, 2008

Broken window effect size – surprisingly large. Mow the lawn?

I love to see experimental testing of human culture and behavior. Maybe the experiments aren’t as well funded or robust as well done clinical trials they’re still a big improvement over our unreliable “common sense”. On the other hand, if they get published in Science they’re probably pretty darned good.

We now have the results of several experiments that, together, suggest that “broken windows” and other features of a disordered environment affect the incidence of crime and antisocial behavior. This is not a surprise to teachers, parents or police, but the effect size is notable.

A doubling of anti-social behavior is enough to justify a lot of investment in the prevention of ‘social crimes’. I now have a business justification to mow my lawn more frequently …

The “broken windows” theory of crime is correct | Can the can | The Economist

A PLACE that is covered in graffiti and festooned with rubbish makes people feel uneasy. And with good reason, according to a group of researchers in the Netherlands. Kees Keizer and his colleagues at the University of Groningen deliberately created such settings as a part of a series of experiments designed to discover if signs of vandalism, litter and low-level lawbreaking could change the way people behave…

… The idea that observing disorder can have a psychological effect on people has been around for a while. In the late 1980s George Kelling, a former probation officer who now works at Rutgers University, initiated what became a vigorous campaign to remove graffiti from New York City’s subway system, which was followed by a reduction in petty crime…

… Dr Kelling’s theory takes its name from the observation that a few broken windows in an empty building quickly lead to more smashed panes, more vandalism and eventually to break-ins. The tendency for people to behave in a particular way can be strengthened or weakened depending on what they observe others to be doing. This does not necessarily mean that people will copy bad behaviour exactly, reaching for a spray can when they see graffiti. Rather, says Dr Keizer, it can foster the “violation” of other norms of behaviour. It was this effect that his experiments, which have just been published in Science, set out to test.

… When the alley contained graffiti, 69% of the riders littered compared with 33% when the walls were clean

… In the “order” condition (with four bicycles parked nearby, but not locked to the fence) 27% of people were prepared to trespass by stepping through the gap, whereas in the disorder condition (with the four bikes locked to the fence, in violation of the sign) 82% took the short cut.

… With no fireworks, 48% of people took the flyers with them when they collected their bikes. With fireworks, this fell to 20%.

.. In a condition of order, 13% of those passing took the envelope (instead of leaving it or pushing it into the box). But if the post box was covered in graffiti, 27% did…

The power, and the glory of a Science pub, comes from the experimental designs and the consistency across multiple settings. These experiments were done in the Netherlands, so of course effects may vary in different cultures.

Doubling. That’s big.

Local police should have no trouble justifying budgets to prevent ‘social crimes’ with these results. There are obvious implications for military and economic interventions in the world’s trouble spots as well.

I guess I need to rake my leaves too …

Update 11/22/08: Obvious follow-up study. Is the effect relative? So if you eliminated all the Graffiti, would rust become the new differentiator between order and disorder? If the effect is relative, then we'd always be chasing our proverbial tails. The return on broken window policing would be transient.

No comments: