Appendix: Narrative Description of Studies Summarized in Table 1

1. A Declining Inner-City Neighborhood in Hartford, Conn.

One of the first reported projects to use street closures in the United States was undertaken in Asylum Hill, a declining inner-city neighborhood in Hartford, Conn.13 In an attempt to deal with burglary, mugging, and purse-snatching, four streets were closed using large planters, some streets were made one-way, and entrances to several streets were narrowed. At the same time, neighborhood policing was introduced, as well as a scheme to encourage the development of community groups and residents' organizations. Subsequent evaluations compared crime in the area with that in an adjacent control area. Victim surveys showed that crime dropped immediately following the street closures, but this result did not last for long. There was little evidence that the other changes had any effect on crime, though they did reduce fear of crime and improve community cohesion.14

Research design: Adequate. This is a small but careful case study using sound crime measures.

2. A Transitional Neighborhood in Dayton, Ohio

Another widely reported project involved a 10-square-block Dayton, Ohio, neighborhood known as Five Oaks. During the late 1980s and early 1990s, this once stable, middle-income neighborhood rapidly changed into a working-class area with increasing poverty and neighborhood decay. This was accompanied by an increase in crime problems, including drug houses, gunshots, prostitution, and speeding traffic. To regain control, a neighborhood stabilization project was implemented. A major component of the project was a traffic management scheme in which 11 streets from the surrounding areas were closed to traffic, as well as 24 streets within the grid. Twenty-six alleys were also closed so that the gates could not be circumvented, creating several sub-neighborhoods. Brick columns with metal gates served as barriers, and the remaining entrances to the area were identified with brick columns bearing a logo and the name Five Oaks.

One year after these changes were implemented, overall crime had dropped by 25 percent, with an even larger decline of 40 percent in violent crime. Resident surveys showed a reduction in the perceived seriousness of crime, including drug-related offenses, prostitution, gang problems, burglary, and violence.15

An active residents' association was extensively involved in planning the project, and took responsibility for it. A high level of media attention may have promoted images of a cohesive neighborhood and deterred potential offenders.

Research design: Strong. This is a large study, using sound crime measures, with street closings as a major component.

3. A Crime-Ridden, Inner-City Neighborhood in Los Angeles

In 1990, the city of Los Angeles and the police department decided to implement "Operation Cul-de-Sac," a community-based policing program to restore order to crime-ridden, inner-city neighborhoods. Because of the problems resulting from the Rodney King beating, the program never progressed beyond its trial in Newton, an area in south central Los Angeles. Newton covers approximately one square mile, with 5,000 residents in some 500 dwellings. In 1990, over half the households were below the poverty line. In 1988, the community was 95 percent African American, and by 1990, 60 percent of its residents were Hispanics—most of them illegal immigrants. Newton had one of the highest recorded levels of serious crime in the city and was plagued by drug activity, gang activity, and drive-by shootings.16

Fourteen iron gates were placed on streets to mark Newton's outer boundary. Barriers were installed to impede drive-by shootings and drive-up drug purchase. Patrols (foot, bicycle, and horseback) were stepped up to suppress these crimes and to improve police-community relations. Officers also joined in cleanup efforts with community groups and the high school. A survey of 350 residents taken in both the first and the last month of the program found that their ratings of police officers' politeness and helpfulness improved by over 33 percent.

The barriers brought about an immediate reduction in serious crimes, including drive-by shootings and homicides.† For example, in 1989, the year before Operation Cul-de-Sac, seven homicides were committed in the area. In the two years after the barriers were installed, only one homicide was recorded. There was no evidence that homicides had been displaced to another neighborhood. When the barriers were removed (in the aftermath of the Rodney King beating), homicides returned to their previous level.17

† The traffic barriers prevented cars from entering the street, or required those that did enter to leave the same way. The latter increased the risks for shooters, because those whom they shot at would have their weapons ready when the car returned.

Research design: Strong. This is a large study, with street closings as a major component of the intervention. Displacement/diffusion was assessed. The study's major strength is its assessment of the effect of reopening the streets.

4. A Public-Housing Project in Hartford, Conn.

In response to a drive-by shooting that wounded four adolescents in a large public-housing project in Hartford, the housing authority erected a barrier across the street at the site of the shooting. Violent crimes on the street decreased by 33 percent (from nine to six) during the 15 months after it was barricaded, compared with the 15-month period before. On adjoining streets and blocks, violent crime also decreased in similar proportion, indicating that no displacement occurred. The barrier had no effect on drug-related crimes on the street.†18

† The authors' claim that increased drug activity on nearby streets was due to displacement therefore seems unlikely.

Research design: Weak. A carefully designed study, but only one barrier was installed, and the reduction in the number of violent crimes (from nine to six) could have been due to chance.

5. An Inner-City, Drug-Dealing Neighborhood in Charlotte, N.C.

Belmont is a deprived inner-city neighborhood in Charlotte, well known locally for being an easy place to buy drugs on the street. The streets are laid out in a grid, and the neighborhood is easily reached from several nearby highways. Five drug-related homicides and more than 100 aggravated assaults in a nine-month period in 1998 to 1999 led to the establishment of a problem-oriented policing project in the northeastern part of the neighborhood. Analysis revealed that 60 percent of those arrested for buying or selling drugs in the area were not Belmont residents. It also revealed that distinct travel routes for drug trafficking fed vehicles from nearby highways into the area. The police decided to block two of the busiest routes by installing concrete barriers at the end of two streets.

Beautified street barriers in Charlotte, N.C., helped control neighborhood violent crime and drug problems.

A 12-month before-and-after comparison of reported crime data showed that after the barriers were installed, violent offenses decreased by 54 percent (from 59 to 27) in the northeastern part of Belmont, and arrests fell by 42 percent. The largest drops were on the two barricaded streets. There was no evidence that violence had been displaced elsewhere in Belmont (in fact, violent offenses for Belmont as a whole dropped by 12 percent, from 236 to 206), though there was some evidence that drug activity had been displaced.

Encouraged by these results, the police sought to install more barriers in Belmont, but the community opposed this on the grounds that the barriers were ugly and were not a substitute for proper policing. Even after "beautification" of the two existing concrete barriers (which were replaced by posts and chains in a mulched garden), community objections to installing more barriers persisted, and the police withdrew the plan.†

† Matt White, crime analyst for the Charlotte-Mecklenburg Police Department, evaluated this project, with advice from Herman Goldstein and the author of this guide. The report has not been published, though the Charlotte-Observer has published an article about the initiative (Markoe 2001).

Research design: Adequate. A careful analysis was done, but only two barriers were installed.

6. A Run-Down, Inner-City Neighborhood With Long-Established Street Prostitution, in London

Finsbury Park, a run-down North London neighborhood, was known for years as an area to solicit prostitutes. Residents, disheartened by police failure to control the problem, petitioned the local authority to reduce vehicle access to the area, in hopes of deterring men from cruising for prostitutes. As a result, seven streets were closed in 1985. This was preceded by an intensive police crackdown that involved a range of interventions directed toward prostitutes and their clients, pimps, and local landlords who rented short-term accommodation.

As judged by official crime statistics, resident surveys, traffic counts, and interviews with prostitutes, this combined approach was successful. It increased residents' sense of security, reduced the traffic volume, reduced serious crimes by about 50 percent, and improved the relationship between the police, the public, and the local authority. Finally, it did not displace the problems to adjacent communities. This seemed due to the prostitutes' lack of deep commitment to their profession. Few were addicted or controlled by pimps. In fact, the most common reasons they gave for being prostitutes were that they could earn more money from that than from other types of work, they enjoyed the independence, and they enjoyed meeting a variety of men. Many of them came to Finsbury Park from outlying areas on cheap "away day" rail tickets. Together with other women, they rented rooms in one of the many local boarding houses or residential hotels, or they conducted business in clients' cars. When not working as prostitutes, many of them worked as barmaids, go-go dancers, or shop assistants.

The prostitutes' relatively light commitment to their work, and the availability of alternative ways to make money, might help explain why the researchers could find little evidence of their displacement to nearby areas in London. Of 253 women arrested for prostitution in 1984 (the year before the street closures), only 65 were still involved in prostitution in North London as of 1991. Another 50 had convictions in other parts of the country, but for the remaining 138 women, there was no record of their having been involved in prostitution after Finsbury Park was "closed down."19

Research design: Strong. Multiple before-and-after measures were used. A careful attempt was made to measure displacement.

7. A Middle-Class, Residential Area With a Recent Problem of Street Prostitution, in London

A similar project in Streatham, an inner-city suburb in South London, also reduced street prostitution and related problems, but overall, it was not as successful as the Finsbury Park project. Again, the impetus for the project grew from local residents who sought to create a partnership with the police and the local authority to develop a traffic management scheme, introduced in December 1989.20 Several streets were closed, and "no entry" signs were installed.

The traffic management scheme achieved many of its goals. Traffic was reduced, especially late at night, and cruising for prostitutes declined by 60 percent. Furthermore, burglary, assault, and street robbery decreased. Residents' fear of crime decreased, and there was also evidence of improved dialogue with the police and increased community cohesion. However, there was substantial "benign" displacement of the problem to the nearby park and main commercial street—"benign" because prostitution there was considered less offensive than in the residential area. The reason given for the greater amount of displacement in Streatham was that the prostitutes there were much more committed to prostitution than those in Finsbury Park.

Research design: Strong. Multiple before-and-after measures were used. Some attempt was made to measure displacement.

8. Downtown Prostitution Strolls in Vancouver, British Columbia

At one time or another between 1970 and 1989, downtown Vancouver had numerous prostitution strolls. Pressure from local residents and businesses generated numerous initiatives to "get tough" with the prostitutes. These included a series of police crackdowns, a "shame the johns" campaign, civil injunctions forbidding prostitutes from entering certain areas, and the installation of a series of "traffic diverters" to prevent cars from cruising in the strolls.

An evaluation of these initiatives concluded that, in every case, the prostitutes adapted to the changes. They moved to new strolls in the downtown area or changed their way of doing business.21 With regard to the traffic diverters, the evaluation reported that shortly after these were installed, a local newspaper published a photo of a woman sitting astride one of them, waiting for a customer. Other prostitutes were reported as saying that the diverters "were 'good for business' because they slowed traffic down nicely." The evaluation proposed that the traffic diverters and other measures to prevent prostitution had failed in Vancouver (while appearing to have worked in London) because more of the Vancouver women might have been supporting heroin habits or had fewer opportunities to engage in off-street prostitution.

Research design: Weak. An interesting and persuasive case is made for adaptation and displacement as the result of street closings (and other measures), but very limited use is made of data.

9. A Citywide Program of Street Closures in St. Louis

In January 1984, the St. Louis Metropolitan Police Department initiated "Operation Safestreet," a multifaceted program with five components: (1) "Project Porch Light," in which people were asked to keep porch lights on from dusk until dawn; (2) "Project Home Security," which target-hardened homes, (3) traditional "Neighborhood Watch"; (4) "Operation Safestreet Newsletter," which regularly informed residents of the current crime situation; and (5) "Project Quiet Street," a traffic management program using street closures and diversions.

This program was phased in for the entire city over four years. Project Quiet Streetgenerated considerable public debate, two lawsuits, and one unsuccessful recall election of an alderman. Negative public reaction grew from the failure to involve citizens at the planning stage—residents began to be involved only after the program began. Consequently, after four years, only two out of nine targeted neighborhoods had permanent barriers in place.

Results were studied in only one of those neighborhoods, but a comparison was made with a nearby "control" area that did not have street closures. It was found that crime rates fluctuated randomly, with no real decrease attributable to the street closures. However, a review of five years of data following introduction of the barriers showed lower rates of increases in burglary where streets were modified. 22

Evaluation design: Adequate. The study is distinguished by an unusually long follow-up, but only one neighborhood was studied.

10. A Citywide "Alley-Gating" Program in Liverpool, England

Liverpool is an older city in the United Kingdom. Much of the city's housing consists of row houses, which can be accessed from lanes running behind them. These lanes have contributed to high burglary rates in many parts of the city, and for a number of years, the city has pursued an intensive program of "alley-gating." This involves installing robust, lockable gates to block alleys and thus restrict burglars' access to the rear of houses. Gate keys are available only to residents of the houses secured by the gates.23

Alley gates, installed extensively in Liverpool, England, have proven a cost-effective method of reducing residential burglaries.

A recent evaluation covered a total of 3,178 alley gates, protecting 106 blocks of housing.24 The gates protected distinct blocks of adjacent housing, typically containing around 360 houses. It was found that burglary decreased by approximately 37 percent in the gated areas, and that burglary declined in direct proportion to the number of gates installed over time. Moreover, there was a large reduction in burglaries where offenders gained access via the rear of the property. There was a small increase in the proportion of burglaries where offenders gained access through the front or side of the property, indicating possible displacement, but the changes observed were unrelated to the timing and intensity of implementation. Finally, burglaries declined in nearby areas not within the boundaries of the alley-gating scheme, suggesting there had been a diffusion of benefits to unprotected houses.

A simple cost-benefit analysis indicated that once the gates had been in place for a year or more, they became cost-beneficial, with a return of around $1.86 for every dollar spent.

Research design: Strong. In fact, the combination of the large number of alley gates covered in the evaluation, the effort made to examine displacement/diffusion, and the cost-benefit analysis undertaken make this by far the strongest study reviewed here.

11. An Affluent Suburban City in Florida

Miami Shores was once a quiet suburban community near Miami. Following major growth in Miami- Dade County, commuter traffic increased, and soon after, crime also increased substantially. In 1986, city officials decided to close 67 streets as part of a citywide strategy to curb traffic, speeding, and crime problems—primarily property crime. The referendum on the street closures passed with a 58 percent majority vote, despite much negative publicity generated by a small but vocal minority. Implementation started in July 1988 and ended in March 1991. In August 1992, a second phase of 28 street closures was proposed, but only eight were approved in the referendum.

A before-and-after examination of crime rates found that Miami Shores showed small declines for burglary, larceny, and auto theft. Rates were unchanged for robbery and aggravated assault. In contrast, Miami showed significant increases for all of the above crimes, and Miami- Dade County showed a general upward trend across crime categories. The evaluators attributed the generally favorable results in Miami Shores to the barriers.25

Research design: Weak. This was a large study, but the evaluation did not explore alternative explanations for the unchanged crime rates in Miami Shores compared with the rest of Miami- Dade County. Nor did it examine possible displacement.