The Relationship Between Neighborhood Criminal Behavior and Oxford Houses

Abstract

The present study investigated crime rates in areas surrounding 42 Oxford Houses and 42 control houses in a large city in the Northwestern United States. A city-run Global Information Systems’ (GIS) website was used to gather crime data including assault, arson, burglary, larceny, robbery, sexual assault, homicide, and vehicle theft over a calendar year. Findings indicated that there were no significant differences between the crime rates around Oxford Houses and the control houses. These results suggest that well-managed and governed recovery homes pose minimal risks to neighbors in terms of criminal behavior.

Introduction

One form of aftercare treatment that has been gaining popularity is the Oxford House model (Jason, Ferrari, Davis & Olson, 2006). Oxford Houses were started in 1975, and have grown to include over 1,250 Houses in the United States, Canada, and Australia. They are recovery homes that offer a sober living environment for individuals recovering from alcohol and/or substance abuse. The three main requirements for living in an Oxford House are to remain abstinent from substance use, fulfill financial obligations to the House (e.g. rent), and to contribute to the House through chores and other agreed-upon activities (Oxford House Manual, 2006). Oxford Houses are run democratically with members of each House (typically seven to twelve individuals) holding elected positions and making decisions on a majority-rule basis. Oxford Houses differ from traditional recovery homes by using exclusively mutual accountability for sobriety, financial responsibilities, and the enforcement of house rules (Polcin, in press). Traditional recovery environments, such as therapeutic communities, have professional staff that provide services and typically have limitations on length-of-stay. These are the key differentiating features of Oxford Houses, which do not limit duration of residence and do not include professional staff (Oxford House Manual).

The residents of an Oxford House are accountable for their own financial responsibilities and deviating from these requirements, along with disruptive behaviors or resuming the use of drugs and/or alcohol are all grounds for dismissal from the House (Jason, Olson, Ferrari & Lo Sasso, 2006). Oxford Houses are self-supporting entities, as each member is responsible for his or her portion of the House rent and expenses, and no outside funding is used to sustain the House. With regard to the neighborhoods that Oxford Houses are located in, they tend to be in middle-class areas with low drug-trafficking and criminal activity (Jason, Davis, Ferrari & Bishop, 2001; Ferrari, Jason, Blake, Davis & Olson, 2006).

There have been several recent evaluations of the Oxford House model. For example, Jason, Davis, Ferrari and Anderson (2007) found that only 18.5% of the participants who left Oxford House reported any substance use over the course of a year. Furthermore, Jason et al. (2006) reported findings from another study that recruited 150 individuals who completed treatment at alcohol and drug abuse facilities in the Chicago metropolitan area. Half of the participants were randomly assigned to live in an Oxford House while the other half received community-based aftercare services (Usual Care). At a 24-month follow-up, positive outcomes were evident in terms of substance use (31.3% of participants assigned to the Oxford House condition reported substance use at 24 months compared to 64.8% of Usual Care participants), employment (76.1% of Oxford House participants versus 48.6% of Usual Care participants reported being employed at the 24-month assessment) and days engaged in illegal activities during the 30 days prior to the final assessment (M=0.9 for Oxford House and M=1.8 for Usual Care participants).

Jason, Roberts and Olson (2005) found that the vast majority of neighbors that lived next to an Oxford House had positive views of these recovery homes. Despite the positive attitudes of neighbors toward Oxford House, some communities have expressed concern about the introduction of recovery homes to their neighborhoods. Lawsuits have been filed around the country in efforts to reduce the ability of these houses to enter middle-class neighborhoods (Jason et al., 2008). Some local governments have expressed a ‘not in my back yard’ (NIMBY) attitude toward recovery homes in general. The National Law Center (1997) polled 89 supportive housing programs and found that 41% had experienced the NIMBY opposition from neighbors and governments. The most prevailing fears of recovery homes being brought into neighborhoods are declining property values and increased crime.

Homes that have individuals in recovery for substance abuse may face more neighborhood resistance because of potential past criminal activity in the prior lifestyles of the residents, and also due to stigma that is associated with substance abuse recovery. There is little available evidence to determine if neighborhood residents’ concerns regarding crime are legitimate. The present study was conducted to examine crime rates in urban neighborhoods with Oxford Houses. We hypothesized that there would not be a significant difference in crime rates for neighborhoods with Oxford Houses versus neighborhoods without Oxford Houses.

Method

Procedure

The Oxford House website was used to locate the addresses of Oxford Houses in Portland, Oregon. This study utilized a yoked-control design involving 42 Oxford Houses and 42 control houses. Each Oxford House was plotted into the geospatial mapping program Google Earth, and then a control house was randomly selected by moving a quarter mile to the west of the Oxford House. We chose a quarter of a mile as a distance, as we believed it would provide enough distance to allow for separate areas of crime statistics while still remaining in the same general neighborhood.

A global information systems’ (GIS) website controlled by the City of Portland (www.cgis.ci.portland.or.us), was used to gather crime data including assault, arson, burglary, larceny, robbery, sexual assault, homicide, and vehicle theft. We selected the time period of 1 January 2005 to 31 December 2005. Only houses within Portland were used because crime data was not available for neighborhoods outside of city limits. A two block radius was mapped around both the experimental and control houses. Only crimes that occurred within this two block radius were included in the data collection. A two block radius was used because it was thought that this was where most of the criminal impact would occur if members of these houses were involved in criminal activity.

Results

An independent samples t-test was used when comparing the 42 Oxford Houses with the 42 control houses, which were not group homes. We analyzed the following data: assault, arson, burglary, larceny, robbery, homicide, and vehicle theft. Sexual assault was not included in the analysis due to the fact that no sexual assaults occurred in the two block radius of any of the 42 Oxford Houses or control houses. The results showed that there was no significant difference in the amount of any of the tested crimes that occurred in the area of the Oxford Houses versus the control houses. Table 1 reports the means and standard deviations for each type of crime, as well as the statistical results.

Table 1.

Types of crime comparing areas with Oxford Houses (OH) versus controls

Crime	OH Mean (SD)	Control Mean (SD)	t	p
Assault	0.50 (0.97)	0.31 (0.60)	1.08	.28*
Arson	0.07 (0.34)	0.10 (0.30)	−.34	.73
Burglary	1.07 (0.87)	1.10 (1.16)	−.11	.92
Larceny	2.07 (1.37)	2.64 (1.81)	−1.64	.11*
Robbery	0.24 (0.49)	0.20 (0.51)	.44	.66
Sexual	0.00 (0.00)	0.00 (0.00)	--	--
Homicide	0.02 (0.15)	0.00 (0.00)	1.00	.32*
Vehicle Theft	0.90 (0.96)	0.81 (1.13)	.42	.68
Total Crime	4.83 (3.00)	5.22 (3.16)	−.57	.57

^*Due to a violation of Levene’s test, equal variances were not assumed.

Discussion

With all of Portland’s Oxford Houses included in the analysis, there was no difference in the amount of crime committed around Oxford Houses and control houses. This suggests that the presence of a recovery home, more specifically an Oxford House, is not associated with higher crime. These findings suggest that well-managed self-governed recovery homes, such as Oxford Houses, pose a minimum risk to neighbors in terms of criminal behavior.

One of the potential reasons that low crime levels were found is that Oxford House members are actively involved in their communities. Jason, Schober and Olson (2008) found that Oxford House residents reported spending around 10.6 hours per month on neighborhood involvement. Their study found that large percentages of Oxford House residents were involved in activities such as community support groups, participating in anti-drug campaigns, and working with community organizations, among many other activities (Jason, Schober & Olson, 2008). Because of this high level involvement in anti-drug campaigns and involvement in the community, it is possible that these activities could potentially facilitate reductions in crime rates. These findings suggest that not only do residents help themselves stay abstinent by living in the Oxford Houses (Jason et al., 2006), but important contributions are made by Oxford House members to their neighborhoods and communities.

Of course, these results may not transfer to all other types of recovery homes. Because Oxford Houses discipline their members if they break the rules and will dismiss House members if they engage in illegal activity or begin using substances again, they may have lower crime rates as a result of these contingencies. Many recovery homes tend to house larger numbers of individuals and some even allow for substance use, and these practices could very well result in different findings.

There were some limitations in this study. For example, because we relied on publicly available data, we were unable to evaluate some potentially important variables that could have explained outcomes. These might have included specific neighborhood involvement among Oxford House residents, other institutions attracting substance abusers to the neighborhood, or other neighborhood characteristics not quantified by the GIS program. As the data were collected in 2005, we also were unable to use features that were not in existence, or examine crime data that was not available that may have been relevant (e.g. prostitution, DUIs, substance-related crimes).

Future studies in this area might involve longitudinal data collection in order to examine crime rates within an area before the establishment of Oxford Houses and then compare them to the crime rates after the Houses had been introduced to the blocks or communities. In future studies, it would also be useful to compare the crime rates over longer periods of time. As advances in GIS technology provide more information about neighborhood characteristics and crime data, investigators can examine more variables that may influence crime rates. Finally, other researchers might use these methods to evaluate crime rates and other recovery settings in neighborhoods. Even with the limitations in the current study, findings do suggest that Oxford Houses do not increase crime levels, and this outcome has important policy implications for the acceptance of Oxford Houses within communities.