Broken windows thesis springs a leak

The broken windows theory posits that minor misdemeanors, like littering or graffiti spraying, stimulate more serious anti-social behavior. LMU sociologists now argue that the idea is flawed and does not justify the adoption of hardline policies.

The origins of the broken-windows theory go back to New York City in the 1980s, when it was first suggested that signs of disorder or physical degradation in a neighborhood, such as broken windows in unoccupied or disused buildings, graffiti spraying or other acts of vandalism, were correlated with an increase in the incidence of destructive behavior – including overtly criminal activities. The basic idea was that signs of disarray in the environment signaled that public order had broken down, and that violators of social norms therefore had no sanctions to fear. The theory was subsequently invoked to justify the introduction of a zero-tolerance policy in some US cities, which meant that perpetrators of minor infringements of public order were unrelentingly prosecuted and punished. However, at the time, really convincing evidence for the validity of the theory was lacking. This explains why a study published by Dutch sociologists in the leading American journal Science garnered a great deal of attention. Its findings were based on relatively simple field experiments, and appeared to provide compelling empirical support for the theory.

Sociologists Dr. Marc Keuschnigg (LMU) and Dr. Tobias Wolbring (now a Junior Professor at Mannheim University) have now used a similar experimental paradigm to investigate the broken windows effect in Munich. However, their results suggest that, in some important respects, the theory is in need of modification. The results of their study recently appeared in the journal "Rationality and Society".

Fliers and litter

For their first experiment, Keuschnigg and Wolbring chose two student dormitories in Munich. According to the tenants themselves, the two buildings differed with respect to the degree of social cohesion prevailing among their inhabitants. This difference is presumably related to the fact that, in one case, new residents were allocated by the landlord, while the occupants of the other were able to select from among the applicants for vacancies. Importantly, none of the tenants were informed that they had been chosen to take part in a field experiment. One morning, the research team attached fliers bearing a meaningless text to the handlebars of all bicycles parked in front of the residence hall. How many of these leaflets would be ripped off and simply thrown away – a harmless form of norm violation – instead of being disposed of in the nearest litter bin? In the second run, the researchers also left refuse strewn around the parking area. Here, the trash signal indeed stimulated a much larger proportion of bike-owners to leave the fliers lying around. Interestingly, the increase was significantly higher in the case of the dormitory with the higher degree of social cohesion. But this result actually corresponds to what the two sociologists had expected on the basis of their theory: "Such a stimulus is more potent if it is perceived as defying the conventional expectation," Keuschnigg explains.

A second experiment conducted by Keuschnigg und Wolbring confirmed the results of the first. The norm infringement at the center of this experiment involved pedestrians ignoring a red traffic light. Here too, seeing someone else disregard the red light acted as a stimulus for others to do the same. The important variable in this experiment was the location of the traffic light. The two sociologists made use of official data to rank different districts in Munich according to their assessed level of neighborliness and social control, two measures of what sociologists call 'social capital.' Once again, they found that violation of the norm had a more powerful effect in the area with the higher level of social cohesion.

The power of suggestion

"These findings, however, clearly contradict the notion of the downward spiral, which is the basis of the broken windows theory," Keuschnigg points out, and these results certainly do not provide an argument in favor of a zero-tolerance policy. For what is the point of cleaning up in a disadvantaged area when a zero-tolerance approach would be more effective in better-off parts of the city?

In the third experiment, the point at issue was an actual offense rather than a misdemeanor. The researchers placed stamped-addressed panel envelopes, each visibly containing a banknote, in front of public mailboxes, and asked how many passers-by would deposit the envelope in the mailbox, and how many would simply pocket the envelope and its contents. The treatment in this case was a badly battered bicycle, parked by the mailbox. The results again indicated that the bicycle acted as a broken-window signal, because a higher proportion of pedestrians took the envelope with them as compared to the control scenario without the bicycle. In other words, the norm violation represented by the damaged bike encouraged the infringement of a different norm.

Weighing up the cost

However, subsequent experiments demonstrated that the signal's stimulus effect vanishes altogether when the envelope contained 100 euros instead of only 5 euros. "When something of real value is at stake, people are no longer susceptible to the suggestive power of weak environmental cues," says Keuschnigg. This finding suggests that typical broken-window signals have an impact only on relatively innocuous norm violations – and do not actually encourage outright criminality. "At all events, our results provide no support for policies of hardline response in cases of minor misdemeanors," says Keuschnigg.