Methods

Data from Hartford Police Department violent and drug crime reports for 15 months before and after the erection of the barrier were analyzed. Violent crimes include murder, aggravated assault, rape, and robbery; drug crimes include possession and sale of illegal drugs. Although a drive-by shooting was the impetus for erecting the barrier, these shootings are infrequent (although highly publicized) and were not separately coded as such by the police until the end of the study period. Only one police crime report is written per incident, regardless of the numbers of complaints made or police officers responding.

Violent and drug crime incidents in the housing project were plotted by street address, before and after erection of the barrier, using a computerized geographic information system (Arc Info, Version 3.4D software, ESRI). The housing project consists of two story garden apartments in two neighborhoods separated by a small river, functionally forming two projects. The residents of the project are 68% Hispanic and 27% African-American (compared with the City of Hartford which is 32% Hispanic and 40% African-American). One in four Hartford families earns less than the federal poverty standard.⁵

Four geographic areas within the housing project were defined: the “intervention street” containing the barrier, the “adjoining streets” containing the intervention street and those streets abutting it, the “same section of the project” containing all streets in the project lying east of the river including the barrier street, and the “opposite section of the project” with all streets in the project lying west of the river and not including the barrier street. Violent and drug crime incidents were plotted by street address using a computerized geographic information system (as above). The number of drug and violent crime incidents occurring before and after erection of the barrier were tabulated, along with per cent change, for the four geographic areas and for the City of Hartford as a whole.

Results

The maps in fig 1 show the distribution of violent crimes, 15 months before and after erection of the barrier respectively, in and around the area with the barrier and in the separate part of the housing project. Figure 2 contains similar maps for drug crimes. After erection of the barrier, violent crime incidence decreased 33% on the intervention street, but there was no change in drug crimes. On adjoining streets, violent crimes decreased 50% but drug crimes increased 100%. Overall, the section of the housing project containing the barrier had a 30% decrease in violent crimes and 109% increase in drug crimes. In comparison, the other section of the housing project experienced a 26% decrease in violent crimes, and a 414% increase in drug crimes. The decrease in violent crime in the four geographic areas studied is generally consistent with the trend observed in the city as a whole (15% decrease). For drug crimes, the city wide data showed a 25% increase; the intervention street showed no change while sharp increases were seen in the adjoining streets, same section, and opposite sections of the housing project (table 1). This suggests that the presence of the barrier displaced drug crimes previously occurring on the intervention street to other areas of the housing project.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Distribution of violent crimes before and after erection of the barrier.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Distribution of drug crimes before and after erection of the barrier.

Discussion

We evaluated the effect of a physical barrier in reducing violent injury and drug related crime. Environmental modification using physical barriers to separate potential victims of injury from injury agents is an accepted method of unintentional injury control.⁶ Examples include highway median barriers, swimming pool fences, and window guards. Accepted environmental measures to prevent intentional injury include metal detectors in schools and better lighting in high crime areas.^7,8 Physical barriers to prevent violent street injury have not been evaluated.

Environmental change at the community level has been used as a crime deterrent with variable success. A crime prevention program altering vehicular and pedestrian traffic, implemented in a Hartford neighborhood in 1976, produced only short term decreases in the neighborhood's crime rates.⁹ Barriers are becoming popular as a new strategy to control violence and crime. Several cities erected or considered erecting similar barriers^10–13 or gates across public roads.¹⁴ One city erecting multiple barriers in a neighborhood beset by high rates of violent and drug related crime found they seemed to reduce crime rates, but the barriers' effect on business and quality of life in the neighborhood met with mixed reviews.¹⁵ Our analysis of a barrier found it may have reduced violent crime but appeared to displace drug crimes to surrounding streets and the other section of the housing project. Despite the reason for erecting the barrier, the numbers of violent crimes on the intervention street were too low to embrace street barriers as a method of preventing street violence without further study of other barriers in different settings.

The barrier's limited success may be in part due to the geography of the housing project itself. Focused drug enforcement operations are more successful in relatively isolated neighborhoods with physically distinct boundaries and few exits.¹⁶ Limiting exits limits drug dealer's ability to move their activities and makes neighborhoods easier for police to patrol. This particular housing project, because it has limited access roads, distinct boundaries, and is without major thoroughfares, is more amenable to focused crime control interventions.

There are several limitations to this study. First, violent and drug related crime reports are imperfect measures of the barrier's effect. Drug arrests may reflect actual drug activity, or merely police and/or resident's interest in halting this activity. The relative severity of violent crime is a better measure of the barrier's effect because it is more likely to be consistently reported by residents and responded to by police. Second, the barrier may cause decreased police patrols on the intervention street (and consequently fewer crime reports) due to the reluctance of policemen in a patrol car to enter a street that requires their turning around to exit. Third, we could not measure the barrier's effect on resident's sense of wellbeing or their sense of isolation in their neighborhood. Finally, confounding events, such as other targeted police activities or changes in street gang activities in the area might also have influenced the effect of the barrier. To our knowledge, there were no police activities targeting this area nor changes in gang activities during the study period.

Implications for prevention

This study adds to our knowledge about preventing urban violence through environmental alteration. A physical barrier erected across a thoroughfare in a public housing project to prevent street crime decreased violent crime on the intervention street and surrounding streets to a moderate extent, but appeared to displace drug related crimes to surrounding areas. Its effect was enhanced by the geographic characteristics of the neighborhood. Such barriers, whether erected individually or as multiple barriers in a neighborhood, should be evaluated in other settings before their use becomes more widespread.