Crime Trends and Micro Places

This blog is a part of our series, "Perspectives in Crime" where we explore leading academic studies that touch on crime data.

An increasing body of research has demonstrated that significant percentages of crimes occur at a limited number of addresses or street segments. These micro places are smaller than police precincts, neighborhoods, or census tracts, the more common spatial units used by criminologists, economists, and data scientists to analyze community life.

In recent years, researchers have deployed novel computing methods to survey robberies, showing that in many American cities reported robberies and other crimes are distributed unevenly, with most incidents occurring at a small percentage of micro places.

A paper by Anthony Braga, David Hureau and Andrew Papachristos, “The Relevance of Micro Places to Citywide Robbery Trends: A Longitudinal Analysis of Robbery Incidents at Street Corners and Block Faces in Boston” published in the Journal of Research in Crime and Delinquency, surveys 29 years of Boston robbery data, from 1980 to 2008, to determine what percentage of Boston robberies occurred in these micro areas.

Violent crime and location, location, location

Robbery, a violent crime ranging in definition from street muggings to bank robberies, produces resounding waves of negative impact manifesting in traumatized individuals, fearful communities, and emboldened criminals. According to the FBI’s Uniform Crime Reporting (UCR) data, approximately 87% of robberies occurring in 2007 took place in cities. Robbers make seemingly rational decisions in selecting targets and locations for their crime that revolve around minimizing resistance, securing as much property as possible, and escaping. Location figures heavily into these decisions, with robbers attracted to places where targets are “vulnerable, accessible, and profitable.” This means locations that offer successful robbery opportunities are likely to experience multiple incidents.

The researcher’s main goal during this study was to analyze to what degree crime patterns in micro places influenced city wide crime trends in Boston during the study period. A 1989 study found that 5% of city addresses were responsible for over 50% of civilian calls for police services in Minneapolis (Sherman, 1989). Even within high crime neighborhoods, analysis revealed that hotspots were responsible for the majority share of area crime, leaving other areas relatively crime free. A 14-year study of crime in Seattle found that about 50% of crime was concentrated in 4.5% of street segments and found the consistency of those hotspots relatively stable year to year (Weisburd et al. 2004). Another study by this paper’s authors found that 5% of street segments and intersections hosted 74% of gun assaults in Boston during the same time span, 1980-2008 (Braga et al. 2010)

Researchers used computerized records from the Boston Police Department of reported robberies during the study period, revealing a total of 142,213 robbery incidents. Records also show robbery incidents decreased dramatically during the study period, falling 68% from 1980 to 2008. The steepest period of decline occurred during the 1990s, falling by about two thirds from 1992 to 1998, mirroring national level trends in robbery incidents during the 1990s.

Street segments, intersections, and addresses

For this study focused on micro places, researchers chose two geographical units on which to focus, street segments and intersections. Street Segments, sometimes also called block faces, are defined as “the two block faces on both sides of a street between two intersections.” Intersections are where two or more streets intersect. Crime reporting often uses intersections as a proxy for location, and this measurement captures crimes reported using intersections.

Researchers geocoded incident data where location was a given street address or intersection. This captured 95.1% of reported robberies during the study period. Locations like parks could not be geocoded to a specific street segment or intersection and were excluded from analysis.

Street robberies and commercial robberies

Street robberies were the most common robbery type during the study period, with approximately 82% of recorded robberies taking place on streets. Commercial robberies, such as robberies of banks, liquor stores, or gas stations, comprised about 10% of recorded robbery incidents.

Aggregating intersections and street segments as street units, researchers found that 47.5% of street segments experienced at least one robbery over the 29-year study period. About 12% of Boston street segments experienced one robbery during the study period. Consistent with previous research, 8.1% of street units generated over 60% of all robbery incidents. The most active street unit experienced 638 robberies. The top 52 street units, some 0.18% of all street segments, generated 10,886 robbery incidents.

Figure 1 shows the share of robbery incidents among Boston street units, broken up into 4 quartiles: street units with a low, medium, or high frequency of robbery incidents, and street units that only experienced once incident during the study period. In all types or robbery, streets with a high frequency of robbery incidents hosted the majority share of all robbery incidents.

Street-level crime trajectory

While these conclusions supported the researcher expectation of crime being distributed among a relatively small number of streets, these findings did not describe whether individual street units yielded high robbery incidents from year to year. Researchers employed growth curve regression models to determine the trajectory of crime at street units from year to year.

Running numerous models analyzing street segments, intersections, and a combination of the two, researchers found that street units with higher robbery incidents continue to have higher levels of relative robbery through time.

Figure 2 shows that roughly 2% of street units produced half of all robberies during the study period, even during the significant crime decline of the 1990s.

Models accounting for level of traffic (vehicular and pedestrian) found that major thoroughfares tend to experience less robbery incidents, likely because high passive guardianship by other motorists and pedestrians, meaning a robbery would be more likely to be witnessed or interrupted; whereas long segments of secondary streets, with less traffic and situational advantages such as poor lighting, were likely to experience more robbery incidents over time.

Crime concentrated

The paper’s findings support the notion that a small percentage of micro places consistently are responsible for most robbery behavior in Boston during the 29-year study period. Roughly 8% of street segments and intersections in Boston are responsible for more than 60% of street robbery incidents. Incredibly, just over 1% of street segments and intersections were responsible for nearly 50% of commercial robbery incidents. Because crime concentrates in limited areas, law enforcement, security, and business professionals can and should make targeted decisions regarding the allocation of security management resources. Learn how the Pinkerton Crime Index can help.