Skip to content

New research challenges ‘broken windows theory’ of crime prediction

Photo of Dan O'Brien teaching his Big Data for Cities course

In a recently published study, Northeastern University assistant professor Daniel T. O’Brien offered a new model for advancing the study of neighborhood dynamics. He leveraged Big Data to shed new light on what factors predict crime in urban neighborhoods, finding that private conflict—not public disorder—is a strong indicator.

The tra­di­tional “broken win­dows theory” goes that acts of public dis­order in neighborhoods—such as graf­fiti, litter, and aban­doned homes—can encourage future crime there. But now research led by North­eastern Uni­ver­sity assis­tant pro­fessor Daniel T. O’Brien has lever­aged Big Data to shed new light on the fac­tors that pre­dict crime in urban neighborhoods.

The researchers found that, in fact, pri­vate con­flict may be a stronger pre­dictor of crime in a community.

Our research sug­gests that the ‘broken win­dows model’ doesn’t effec­tively cap­ture the ori­gins of crime in a neigh­bor­hood,” O’Brien said. “What’s hap­pening is that vio­lent crime is bub­bling out from the social dynamics of the com­mu­nity, out from these pri­vate con­flicts that already exist, and then is esca­lating and spilling into public spaces.”

O’Brien holds joint appoint­ments in the School of Public Policy and Urban Affairs and the School of Crim­i­nology and Crim­inal Jus­tice, and his research uses Big Data—most often in the form of large admin­is­tra­tive data sets gen­er­ated by city government—in con­junc­tion with tra­di­tional method­olo­gies to explore the behav­ioral and social dynamics of urban neigh­bor­hoods. This work is emblem­atic of his class­room teaching, as stu­dents in his Big Data for Cities course are immersed in the emerging field of “urban informatics.”

O’Brien is also the research director of the Boston Area Research Ini­tia­tive, which under­takes and sup­ports cutting-​​edge urban research at the inter­sec­tion of social sci­ence and public policy.

The find­ings, O’Brien said, show how Big Data can be used to advance the study of neigh­bor­hood dynamics and to assess and mit­i­gate crime. His col­lab­o­ra­tors were Har­vard Uni­ver­sity pro­fes­sors Robert J. Sampson and Christo­pher Win­ship, who are director and co-​​director, respec­tively, of the Boston Area Research Initiative.

In a study with Sampson and Win­ship, O’Brien devel­oped an eco­metric method­ology that trans­lated more than 300,000 non-​​emergency calls to 311 in the city of Boston during 2011 and 2012 into mea­sures of phys­ical dis­order, such as graf­fiti and the accu­mu­la­tion of litter. The pur­pose of this method­ology, which was pub­lished this summer in the journal Soci­o­log­ical Method­ology, was to create a detailed set of met­rics for a city to use in assessing neigh­bor­hood dynamics sev­eral times a year.

Then, O’Brien and Sampson put the broken win­dows theory to the test by applying this method­ology to more than 1 mil­lion 311 and 911 calls in that same time period. The goal was to not only mea­sure social dis­order and crime, but to examine how it changed over time. The results, pub­lished this summer in the Journal of Research in Crime and Delin­quency, pointed to a “social esca­la­tion model” in which future dis­order and crime emerge not from public cues but from pri­vate dis­order within the community.

Using the 911 and 311 data the researchers devel­oped six measures—public social dis­order, public vio­lence not involving guns; domestic vio­lence and other pri­vate con­flicts; gun vio­lence, and pri­vate neglect in neigh­bor­hoods, and public den­i­gra­tion in neigh­bor­hoods. Upon exam­ining the con­nec­tions between these six fac­tors, here’s what they found:

•    Pri­vate con­flict was the strongest leading indi­cator of crime in the model, pre­dicting increases in social dis­order, public vio­lence, guns, and even phys­ical dis­order in pri­vately owned spaces.
•    Phys­ical and social forms of public dis­order were weakly pre­dic­tive of future vio­lence and dis­order, if at all; public den­i­gra­tion had no pre­dic­tive power, and the link from public social dis­order to later public vio­lence was half the mag­ni­tude of the reverse pathway from vio­lence to social disorder.

The researchers noted that the study did not demon­strate causality; in other words, why pri­vate con­flict was such a stronger indi­cator of crime than pri­vate neglect and public denigration.

O’Brien and his col­leagues the­o­rized that a stressful social ecology like that reflected in pri­vate con­flict among res­i­dents can drive behavior that leads to mul­tiple con­se­quences for a neigh­bor­hood as a whole. For example, friend­ship dis­putes or domestic vio­lence can spill out into the public space, and those inci­dents are likely to increase in severity over time—a pro­gres­sion that until now has been largely invis­ible to researchers because tra­di­tional forms of assess­ment such as neigh­bor­hood sur­veys or obser­va­tions only cap­ture what’s vis­ible in the public space and not what’s behind closed doors.

With these data sets, the method­ology becomes a tool for reli­ably and con­tin­u­ously tracking and ana­lyzing the con­di­tions of the city,” O’Brien said.

From the city to the class­room
O’Brien said the type of large admin­is­tra­tive data sets that proved to be cru­cial in these studies will be the focus of his Big Data for Cities course this fall. In this course, under­grad­uate and grad­uate stu­dents learn how to manage and ana­lyze large data sets—such as restau­rant inspec­tion vio­la­tion records and the tax assessor’s data­base for the city of Boston—by pur­suing group research projects focused on spe­cific city resources and services.

The course, which launched in fall 2014, dove­tails with the university’s new M.S. in Urban Infor­matics, which cou­ples com­pre­hen­sive data ana­lytics skills with an under­standing of the big ques­tions cities face in the 21st century.

What’s spe­cial about these Big Data is that we’re all still trying to figure out what to do with them, in terms of trans­lating them into soci­etal impact and improving munic­ipal ser­vices,” he said. “The stu­dents [last fall] were excited to take on this chal­lenge, and it’s been inspiring to see them draw out their ideas and conclusions.”

-By Greg St. Martin

More Stories

Photo of the Capitol Building at night

High stakes for politics, SCOTUS in 2018

Photo of the crashed truck that was used in the October 31st attack in Manhattan.

Weaponizing Language: How the meaning of “allahu akbar” has been distorted

Northeastern logo

Why I love studying Spanish