Most people probably don’t give much thought to the subway. You ride it every day to get into and back from work, and to move around the city to visit friends, grab a drink, or go shopping. Subways are clean, affordable transportation used by millions in many big cities around the world.
And because they’re so affordable and reliable, they’re used by criminals, too. How much are they used by criminals and exactly what kind of impact they may have on the crime rate is, however, largely an open question. Could subways help explain why the crime rate is higher in urban areas?
So researchers in Washington, D.C. decided to find out.
It’s understood that anything available for the public good — like subways — can just as easily be used to undermine the public, too. Researchers occasionally have the opportunity to study things like this due to natural events or, in this case, scheduled maintenance. Subways need to be maintained, and in Washington, D.C.’s case, they occasionally need to close an entire station in order to conduct such maintenance.
Previous research on crime generally demonstrates that most crime is local. Criminals rarely travel (or don’t travel very far) in order to commit crime. Therefore, access to cheap public transportation may be helpful to criminals looking to expand their enterprise.
The researchers (Phillips & Sandler, 2015) used geo-coded data1 on reported crimes made available by the Washington Metro Police from 2011 to 2013 for eight types of crime: assault, sexual assault, robbery, arson, burglary, stolen auto, theft, and theft from auto. (They didn’t include homicides in their analysis because the data did not include the time of day.) The researchers combined this data with ridership data from the D.C.’s subway authority (the WMATA), as well as station closure data. (The study coincided with a large-scale maintenance program undertaken on the D.C. subway system, that included the complete temporary closure of stations on all four subway lines.)
After crunching the numbers and conducting their analysis, what did the researchers find? “We find strong evidence that closing down stations reduces crime at other connected stations but no evidence of changes at stations on other lines.”
Closing one station leads to a five percent reduction in crime at other stations on the same line, which amounts to a two percent decrease in crime across the entire rail network. This matches patterns in access to transit as ridership falls at stations on the same line but not stations on other lines.
They also found that when you close a subway station, it appears to entirely geographically shift the distribution of crimes throughout the system. That is, crime fell where few previous offenders lived, suggesting criminals weren’t getting to these neighborhoods as often. Crime also fell when the station that closed was in a neighborhood where many previous offenders lived.
Crime reductions also follow the structure of the rail network, falling more in stations that become disconnected from a larger part of their train lines.
Finally, we rule out the possibility that crime falls near stations simply because it is displaced to locations further from stations. If anything, crime falls in areas further from train stations.
Most of the drop in crime appears to be related to automobiles: “A decrease in theft from automobiles accounts for the majority of the drop in crime […]”
Our results provide support for the theory that perpetrators respond sensitively to transportation costs. A relatively modest change in public transit leads to noticeable changes in perpetrator behavior and a spatial redistribution of crime toward the residences of those committing crimes.
Closures only affect crime on the same line, suggesting that the affected perpetrators are unlikely to change trains to commit a crime. Closures matter far more for crime within 1/4 mile of a station than for crime further afield, suggesting that the affected perpetrators do not walk large distances to commit a crime.
The study goes against some of the previous research in this area, suggesting we need more studies — or a replication of this one — before we can draw policy-making conclusions (and I’m not sure what they would even be).
This study shows what is perhaps common sense that, yes, criminals take the subway too.
Phillips, D.C. & Sandler, D. (2015). Does public transit spread crime? Evidence from temporary rail station closures. Regional Science and Urban Economics, 52, 13-26.
- One of the innovations of this study is the use of geo-coded data, something not readily available to most researchers two decades ago. [↩]