Broken windows theory: why small signs of neglect change how people behave
People read a place before they read its rules.
In 1969, a Stanford psychologist named Philip Zimbardo parked a car on a street in the Bronx, New York. He removed its number plates and left the bonnet up, as if abandoned. What happened next became the founding story of the broken windows theory.
Within ten minutes, the first visitors arrived: a father, a mother and their young son. They took the radiator and the battery. Within 24 hours, almost everything of value had gone. Then the vandalism began. Windows were smashed and the seats ripped open.
Zimbardo left a similar car in Palo Alto, California, a well-off university town. Nobody touched it for more than a week. So Zimbardo took a sledgehammer to it himself. Within a few hours, passers-by had joined in, and the car was turned upside down and destroyed.
Thirteen years later, two American scholars built a theory of policing on those two cars.
The broken windows theory holds that visible signs of neglect, such as a broken window, graffiti or litter, signal that nobody is in charge of a place. That signal invites more rule-breaking, from small disorder to serious crime. The theory was proposed by James Q. Wilson and George Kelling in 1982.
Where did the broken windows theory come from
Zimbardo was not studying broken windows. His 1969 paper was about anonymity: how people behave when they feel nobody can see who they are. The link to disorder came later, in March 1982, when The Atlantic magazine published an essay called “Broken Windows: The police and neighborhood safety“. Its authors were James Q. Wilson, a Harvard political scientist, and George Kelling, a criminologist.
The essay began with a puzzle from New Jersey. In the mid-1970s, the state paid cities to take police officers out of patrol cars and put them back on foot. A careful evaluation, carried out chiefly in Newark and published by the Police Foundation, found that foot patrol had not reduced crime. Yet residents in foot-patrol areas felt safer, thought better of the police and took fewer precautions at home. Kelling had helped run that evaluation. He had also spent hours walking the beat with Newark officers, watching how they kept informal order on a busy street near a train station.
Wilson and Kelling argued that the residents were not fooled. What frightens people in public, they wrote, is often disorder rather than crime: the drunk sprawled on the pavement, the rowdy group at the bus stop. Disorder left alone tells everyone that nobody cares, and that is when serious crime moves in. Their image was the window. Leave one broken and unrepaired, and the rest soon go too, because breaking more costs nothing.
Zimbardo’s cars fitted that picture. In the Bronx, where abandoned cars and theft were common, the damage began almost at once. In Palo Alto, it took one visible act of vandalism to tell passers-by that the car was fair game.
The original essay was more cautious than the policy it inspired. Wilson and Kelling asked how anyone could ensure that police did not become “the agents of neighborhood bigotry”, and admitted they had no wholly satisfactory answer. On how a police chief should deploy officers, they wrote that nobody knew for certain. The essay was about foot patrol, fear and helping neighbourhoods enforce their own standards. It did not call for mass arrests.
Broken windows theory examples, from New York to Bengaluru
New York City took the idea furthest. William Bratton, who ran the city’s transit police from 1990 to 1992, went after fare-dodgers on the subway. After Rudy Giuliani became mayor in 1994 and made Bratton police commissioner, the approach spread to the streets as a crackdown on minor offences. Crime in New York fell steeply through the 1990s, and broken windows got much of the credit.
The later years were darker. Stops and searches of young men multiplied, and in August 2013 a federal court ruled that New York’s stop-and-frisk policy was unconstitutional because of the way it singled out young Black and Hispanic men. Kelling himself was not an advocate of stop-and-frisk, but it grew in broken windows’ name.
Bengaluru offers a gentler version. In November 2010, an anonymous volunteer group calling itself The Ugly Indian cleared a filthy corner of Church Street, outside the offices of Wipro and The Times of India. They removed the garbage, repaired the footpath, painted the walls and added plants to discourage people from spitting and urinating there. Citizen Matters reported in December 2013 that the spot was still clean 38 months later. The group has since repeated these “spot fixes” across the city, with no arrests and no fines.
The most careful test came from the Netherlands. In 2008, Kees Keizer, Siegwart Lindenberg and Linda Steg of the University of Groningen tied paper flyers to the handlebars of bicycles parked in an alley with no dustbin. When the alley walls were clean, 33% of cyclists dropped the flyer on the ground. When the walls were covered in graffiti, 69% did. In another of their six experiments, they left an envelope sticking out of a postbox with a five-euro note visible inside. At a clean postbox, 13% of passers-by took it. At a graffiti-covered one, 27% did.
Why would graffiti make someone more likely to steal? Keizer and his colleagues argued that people constantly read two signals: what they are supposed to do, and what others around them actually do. When the second signal says rules are being broken, the first one weakens. The effect crossed categories. Graffiti, which the person had not added, made them more willing to litter or take money.
You meet this at a smaller scale all the time. The office kitchen with one unwashed mug in the sink soon has five. The shared drive where one folder is named “final_final_v3” soon has no naming convention at all. In both cases, the first lapse left alone tells everyone else what is normal.
Does broken windows theory actually work
It depends on which version you mean, and the difference matters.
The theory has taken serious hits. In 2006, Bernard Harcourt and Jens Ludwig argued in the University of Chicago Law Review that New York’s crime drop did not prove the theory, since crime also fell sharply in many cities that never adopted broken windows policing. They also looked at a US government programme that moved poor families to tidier, more orderly neighbourhoods. The families who moved did not commit fewer crimes.
In 2019, Daniel O’Brien, Chelsea Farrell and Brandon Welsh reviewed the research on whether disorder makes individual people more aggressive. They concluded that the apparent link was largely a product of how the studies were designed.
The most useful evidence comes from Anthony Braga, Brandon Welsh and Cory Schnell, who pooled studies of police efforts against disorder in 2015 and updated the work in 2024 with 59 tests. Tackling disorder did reduce crime modestly. But the gains came from community and problem-solving work that changed conditions at particular places. Aggressive order-maintenance, meaning arrests and crackdowns on individuals for minor offences, produced no significant reduction in crime.
Kelling watched his idea travel. In a 2016 interview with NPR, he said he had read with dismay about new police chiefs promising to “implement broken windows tomorrow”, because it needed training, guidelines and oversight. He even wondered aloud whether it was time to back away from the metaphor.
The fair conclusion is narrow. Fixing the window works. Arresting everyone who walks past the window does not.
How to use the broken windows theory at work and at home
Fix the first small thing fast, whether it is the unwashed mug or the missed deadline nobody mentioned. The Groningen experiments showed that one visible lapse spreads to rules that have nothing to do with it.
Fix things before you punish people. Braga’s reviews found that changing the conditions of a place worked, while crackdowns on individuals did not. At work, that means a clear folder structure before a stern email about file names.
Make the expected standard visible and easy. The Ugly Indian’s Church Street fix relied on paint and potted plants to discourage spitting, which made the right behaviour the natural one.
Test your fix the way The Ugly Indian tests theirs. The group’s stated standard is that a spot fix should hold on a public street for at least 90 days with no supervision. If yours needs someone standing guard, you have fixed the symptom, not the signal.
A broken window tells people what a place will tolerate. So does a repaired one.
Daily Reflection
Look for one small thing left broken in your office or home today. Fix it before anyone else copies it.
