
Do Broken Windows
Encourage Criminality?
BROKEN-WINDOWS, CRIME, DISORDER
The broken windows theory is a kind of criminology theory, which thinks that such behaviors as breaking windows, even littering, will have a negative impact on the whole community environment, and even cause crime. The theory has had a significant impact on police policy in the United States. In 1990, the New York City Police Department introduced this theory into police enforcement. Then New York City Police Commissioner Bratton introduced his Broken Windows-based Quality of Life Initiative between 1994 and 1996. New York's felony rate dropped by nearly 40 percent within those two years, and the homicide rate dropped by half. In the past, the New York City Police Department and the community attributed the decline in crime to broken windows policing, which believes that stopping, questioning, and frisking the perpetrators of minor crimes such as disturbing public order will help to reduce the occurrence of felonies. However, in recent years, many skeptical voices have emerged. Critics argue that broken windows policing may not be directly responsible for the drop-in crime rates. There is no evidence that disorderly conduct leads to crime.
Therefore, the goal of this research would be to verify whether the broken windows theory actually causes the crime rate to increase by using New York City crime incident data and disorderly incident data to further discuss whether broken window policing is effective or not. This project aims to improve the New York police policy-making and arrangement by justifying the significant level of the broken windows theory in New York City. By analyzing the regionality differences in New York City, seasonality change, and various kinds of crime and disorder behavior, this research set the purpose which is to dig deeper and find how the theory reflected differently on different regions and changed with time.
PROBLEM
From previous studies, it is known that the broken window theory has been applied to policy arrangement a long time ago. Nevertheless, there are some analyses showing that there is no significant amount of evidence to support the broken windows theory. So we are facing the question of whether broken window policing really works or is a waste of manpower and resources.
In this research, the goal is to explore how the broken windows theory reflected differently in different regions in New York City and how the theory changed over time. This project hopes to verify whether the relationship between crime and disorder reflected by broken windows theory is true or not and put forward some suggestions on improving police measures through our research results to improve the police's efficiency and allocate resources better to address specific criminal activities.
HYPOTHESIS
According to the broken windows theory, higher levels of disorder cause higher levels of crime.
Higher levels of police activity cause lower levels of disorder and crime.
The demographic census data might impact the disorder and crime data. Lower levels of education, income, and house price cause higher levels of disorder and crimes
METHOD
In this study, we mainly use four types of data: major crime, disorder, police behavior, and census data to investigate our hypotheses.
we compute the disorder density of each unit in terms of population in different zip codes in Equation (1).
In addition, Equation (2) computes the crime density of each unit in terms of population in different zip codes.
Based on Equations (1) and (2), we compute the distribution of disorder density and crime density of each unit in terms of population.
Then, based on the disorder dataset and crime dataset, we utilized the python method “pandas.corr()” to compute the correlation of different disorder types and different crime categories.

RESULT
The spatial distribution of correlation between Trash and Property Crime

The spatial distribution of correlation between Uncivil Use of Space and Property Crime

The spatial distribution of correlation between Trash and Violent Crime

The spatial distribution of correlation between Uncivil Use of Space and Violent Crime

There is definitely a relationship between disorder behavior and crime number, but we still cannot say that the broken window theory and broken window law enforcement are true and effective.
SUGGESTION
In order to further study the effectiveness of broken windows enforcement in the future, we put forward the following policy suggestion:
NYPD should make the specific data of quality-of-life summonses and misdemeanor arrests public, including time, place, type and other information.
Project Sponsor
Maurizio Porfiri
Project Advisor
Alain Boldini
Project Mentor
Emma Schechter
Project Mentor
Roni Barak Ventura
Project Mentor
OUR TEAM
Catherine Liu
Master's Student
NYU CUSP
Mengyuan Guo
Master's Student
NYU CUSP