By Philip Ball

The idea that neglected environments encourage crime and antisocial behavior has been around since the 1980s. Now, a study shows that messy surroundings also make people more likely to stereotype others.

Diederik Stapel and Siegwart Lindenberg, social scientists at Tilburg University in the Netherlands, asked subjects in messy or orderly everyday environments (a street and a railway station) to complete questionnaires that probed their judgments about certain social groups. They found small but significant and systematic differences in the responses: there was more stereotyping in the disorderly areas than the clean ones.

The researchers suggest that local authorities could therefore counteract social discrimination by diagnosing and removing signs of disorder and decay in public environments. They report their findings today in Science.

David Schneider, a psychologist at Rice University in Houston, Texas, and a specialist in stereotyping, calls the study "an excellent piece of work that not only speaks to a possibly important environmental cause, but also supports a major potential theoretical explanation for some forms of prejudice."

Zero tolerance

Social scientists and criminologists have long suspected that environment has an influence on behavior. The "broken windows" hypothesis, developed by sociologists James Wilson and George Kelling, supposes that people are most likely to commit criminal and antisocial acts when they see evidence that others have already done so--for example, when they are in public places that show signs of decay and neglect.

This idea motivated the New York subway system's famous zero-tolerance policy on graffiti in the late 1980s (for which Kelling acted as a consultant), which is credited with improving the safety of the network. Lindenberg and his coworkers tested the hypothesis in 2008 with a study in Dutch urban settings; their findings suggested that surroundings do have an influence on people's readiness to act unlawfully or antisocially.

But could evidence of social decay, even at the mild level of littering, also affect our unconscious discriminatory attitudes towards other people? To test that possibility, Stapel and Lindenberg devised a variety of disorderly environments in which to test people's mind-sets.

In one experiment, passers-by in the busy Utrecht railway station were asked to sit in a row of chairs and answer a questionnaire for the reward of a chocolate bar or an apple. The researchers took advantage of a cleaners' strike that had left the station dirty and litter-strewn to create their messy environment; they returned to do the same testing once the strike was over and the station was clean.

In the questionnaires, participants were asked to rate how much certain social groups--Muslims, homosexuals and Dutch people--conformed to qualities that formed part of positive and negative stereotypes, as well as qualities unrelated to stereotypes. For example, the "positive" stereotypes for homosexuals were (creative, sweet), the 'negative' were (strange, feminine) and the neutral terms were (impatient, intelligent).

As well as probing these responses, the experiment examined unconscious negative responses to race. All the subjects were white, but when they were asked to sit down, one chair at the end of the row was already occupied by a black or white Dutch person. In the messy station, people sat on average further from the black person than the white one, whereas in the clean station there was no statistical difference.

To eliminate effects caused by differences in the environments' cleanliness while preserving the disorder, the researchers ran a second experiment. Subjects were approached on a street in an affluent Dutch city; in one case, the street was orderly, but in the other the same street had been made more disorderly by the removal of a few paving slabs and the addition of a badly parked car and an 'abandoned' bicycle. Again, disorder boosted stereotyping.

Imposing order

Stapel and Lindenberg say that stereotyping may be an attempt to mentally compensate for mess: "a way to cope with chaos, a mental cleaning device" that partitions other people neatly into predefined categories.

In support of that idea, they showed participants in a lab pictures of disorderly and orderly situations, such as a bookcase with either disheveled or regularly stacked books, before asking them to complete both the stereotyping survey and another one that probed their perceived need for structure by asking them to rate the truthfulness of statements such as, "I do not like situations that are uncertain". Both stereotyping and the need for structure were higher in people viewing the disorderly pictures.

Robert Sampson, a sociologist at Harvard University in Cambridge, Mass., says that the study is "clever and well done," but is cautious about how to interpret the results. "Disorder is not necessarily chaotic," he says, "and is subject to different social meanings in ongoing or non-manipulated environments. There are considerable subjective variations within the same residential environment on how disorder is rated--the social context matters."

He adds, "Once we get out of the lab or temporarily induced settings and consider the everyday contexts in which people live and interact, we cannot simply assume that interventions to clean up disorder will have invariant effects."

Schneider agrees that the implications of the work for public policy are not yet clear. "One question we'd need to answer is how long these kinds of effects last," he says. "There is a possibility that people may quickly adapt to disorder. So I would be very wary of concluding that people who live in unclean and disordered areas are more prejudiced because of that."

Stapel acknowledges this. "People who constantly live in disorder get used to it and will not show the effects we find," he says. "Disorder in our definition is something that is unexpected."