A Blog by

The spread of disorder – a repost in wake of London’s riot cleanup

Yesterday, I watched as hundreds of Londoners took to the streets in a heroic attempt to clean up the mess caused by rioters and looters the night before. Looking at pictures of large crowds getting off trains with cleaning equipment in hand and marching down streets with brooms held aloft, I’ve rarely felt so proud of my city.

The clean-up operation was a great move – a positive note in an otherwise depressing week and a chance for a beleagured capital to come together and reclaim its sense of community. But the act of cleaning away the preceding day’s damage was also important. To explain why, I’m reposting this piece from a few years back about a Dutch study which showed that signs of disorder only breed more disorder. To clarify, this is in no way an attempt to explain the psychology of the riots themselves; it simply suggests another reason why the clean-up operation was a smart move.


Imagine walking through a neighbourhood and seeing graffiti, litter, and shopping trolleys strewn about the place. Are these problems to be solved, or petty annoyances that can be ignored in the light of more serious offences? A new study suggests that the former is right – even the most trivial of transgressions can spread and spiral because their very presence stimulates more of the same behaviour. Through a series of stunning real-world experiments, Kees Keizer and colleagues from the University of Groningen have shown that disorder breeds more disorder. The mere presence of graffiti, for example, can double the number of people who litter and steal.

Their study provides strong support for the controversial Broken Windows Theory, which suggests that signs of petty crimes, like broken windows, serve as a trigger for yet more criminal behaviour. It follows that fixing small problems can prevent the build-up of bigger ones and the gradual decay of a neighbourhood. The idea was first proposed in a magazine article published in 1982, but soon became the basis of many a social policy.

It inspired Rudy Guiliani’s Quality of Life Campaign in New York, which focused attention on seemingly trivial fixes such as removing graffiti, clearing signs of vandalism and sweeping the streets. The campaign seemed to work, which motivated other cities to try the same tactics. But despite its popularity, the Broken Windows Theory still divides opinion, for it lacks the backing of hard evidence, it’s plagued by woolly definitions of “disorder” and critics have questioned its role in New York’s drop in crime. These are fairly hefty shortcomings for a concept that is so central to anti-crime measures and Keiser wanted to address them once and for all.

To do so, he took to the streets of Groningen and watched unknowing passers-by in real-life situations as they reacted to signs of disorder. The recurring question was this: would people exposed to inappropriate behaviour behave in a similar way themselves?

He began in an alleyway in a local shopping district, where bicycles are commonly parked and where a conspicuous red sign warned against graffiti. He attached a flyer from a fictional sportswear shop to the handlebars of parked bicycles and watched what people did as they returned to their rides. Under normal circumstances (picture on the left), most people took the flyer with them and just 33% littered by throwing it on the ground. But that all changed when Keiser covered the wall with graffiti (picture on the right). With this innocuous difference, the proportion of litterers doubled and 69% discarded their flyers on the street.


Keiser explains this behaviour in terms of “social norms” – the rules that separate appropriate behaviours from inappropriate ones. Problems arise when our view of what is common (in this case, graffiti) fails to mesh with our understanding of what society expects (as epitomised by the “No Graffiti” sign). Graffiti is frowned upon, but the covered walls send a message that it is common and therefore, more acceptable. Keizer calls this the Cialdini effect.

Fence.jpgTo see how far its influence would extend, he set up a temporary fence in front of a car park. He attached two signs to the fence, one banning people from locking their bicycles to it, and another saying that entry was forbidden and asking people to use a detour some distance away. When he placed four bicycles a metre away, just 27% of people disobeyed the detour sign and squeezed through the gap in the fence. But when the bikes were locked to the fence, in blatant disregard of the first sign, 82% of people ignored the detour sign too. With one rule broken, the other followed suit.

A third related study took place in a supermarket car park, where prominent stickers asked shoppers to return their carts to the main building. Keizer plastered various cars with the same flyer from the first study. If the garage was clear of carts, just 30% of shoppers littered with the flyer, but if four unreturned shopping carts were left lying about, 58% did so. Again, when people saw that one rule was broken, they felt less strongly about following another.


Together, these three experiments show that signs of disregarded rules can spread to affect commonly held behaviours (“don’t litter”) as well as specific requests from third parties (“don’t enter” or “return trolleys”).

Signs of disorder don’t even need to be seen to have such influences – they can be heard too. In the Netherlands, most people know that setting off fireworks in the weeks before New Year’s Eve is illegal and carries a small fine. Keiser found that he could trigger people to litter more frequently by giving them audible evidence that this law had been flouted. Again, he attached a flyer to bicycles parked near a train station. Under normal circumstances, 52% of cyclists littered but if they heard the sound of fireworks let off by Keiser at a nearby location, that figure grew to 80%.

Letterbox.jpgFor his final and most dramatic demonstration, Keiser showed that the mere presence of graffiti can even turn people into thieves. He wedged an envelope into the slot of a mailbox, with a 5 Euro note showing in the transparent window. If the mailbox and the ground around it were clean, just 13% of passers-by stole the envelope. If the mailbox was covered in graffiti, or if the ground around it was covered in litter, the proportion of thieves doubled to 27% and 25% respectively.

Keiser thinks that it’s unlikely that people inferred a reduced police presence by the presence of litter or graffiti – certainly, litter is generally tolerated by the police in Groningen. Instead, he thinks that one transgression was actually fostering another. This isn’t a simple case of imitation – littering doesn’t just beget littering. Keiser’s idea is that seeing the breakdown of one social norm makes it easier to ignore others, by weakening our general resolve to act appropriately and strengthening our temptations to act in our own self-interest.

All in all, the suite of experiments, all in a realistic setting, provide powerful evidence that the Broken Windows Theory is valid and all of Keiser’s results were statistically significant. Small, petty signs of disorder can indeed turn people away from the straight and narrow. His message to police and policy-makers is stark – it is worth spending time on small and seemingly trivial interventions, to prevent disorder from spreading and escalating.

Reference: Keizer, K., Lindenberg, S., & Steg, L. (2008). The Spreading of Disorder Science, 322 (5908), 1681-1685 DOI: 10.1126/science.1161405

12 thoughts on “The spread of disorder – a repost in wake of London’s riot cleanup

  1. Of course, it says that in the third paragraph of the article but don’t let the presence of actual words stop you from reflexively mashing your hands against the keyboard.

  2. @Adam: As a resident of Brooklyn, I can assure you Giuliani proved no such thing. He took credit for the city’s drop in crime despite the same thing happening in almost every major American city during the same period regardless of whether they implemented similar “quality of life” measures (most of which amounted to harassing non-whites and generally poor people. . . oh, kinda similar to what the police had been up to in London pre-current-riots).

  3. This is a very misleading article, as it suggests that Broken Windows had never been tested before Keizer’s experiment proved it. Actually it’s been tested over and over again, and almost always *dis*proved. Which isn’t all that surprising, because the original “BW” article is incredibly vague and points in several different directions – at one point it says that it doesn’t matter if reducing disorder makes the crime rate go up, as disorder is more important than crime. Keizer’s study is interesting, but “Broken Windows” is a long way from being proved – or even consistently defined.

  4. After the hockey riots in Vancouver last month, there was a mass of people going onto the street to clean up that mess too.

    btw, going to steal your response to adam and use it myself at some appropriate time.

  5. This is a very misleading article, as it suggests that Broken Windows had never been tested before Keizer’s experiment proved it.

    Point me to the bit which says this. I wrote, “… the Broken Windows Theory still divides opinion, for it lacks the backing of hard evidence…” This does not suggest what you have written.

    Keizer’s study is interesting, but “Broken Windows” is a long way from being proved – or even consistently defined.

    Again, where is the bit where I said that it had been proved? “Provides strong support”? “Provides powerful evidence”? In fact, the first time the word “proves” appears, it’s in comment #1.

    Based on this and #1, I have a new hypothesis that articles about disorder breed poor reading comprehension.

  6. Well done, Ed. Obviously, this is not meant to be an informative article requiring full comprehension and serious thought. It is no doubt an experiment to see how many people simply skim a few sentences, form an erroneous opinion and then shoot out a scathing comment. Your work is done here.

  7. I’m surprised no-one has mentioned Singapore – perhaps the world’s greatest experiment with the “Broken Windows” theory, and on the face of it a supportive example. But, in assessing the validity of the theory, do we need to factor out the threat of imminent punishment from strict criminal codes? It seems to me that, to be valid, the theory requires people to “self-police,” or consciously choose not to. If an omni-present, over-zealous police force is literally watching your every move, then we’re not dealing with self-policing any more.

    Thus, I’d be more inclined to look at other orderly societies, where the police presence is much less “in-your-face.” Instead of Singapore, or any number of oil-rich despotisms in the Gulf region, how about Switzerland as a prime example of order begetting more order? Having said that, it’s been twenty years since I was last there, and I gather there has been a deterioration in the country’s fabled neatness. Even I can recall seeing drug addicts shooting up openly in some Zurich parks.

Leave a Reply

Your email address will not be published. Required fields are marked *