have a greater opportunity todevelop a sense of competencetoward maintaining abstinence.For more information, see theOxford House Web page: http://www.oxfordhouse.org.

Similar to those who partici-pate in Alcoholics Anonymous,members of an Oxford House re-ceive abstinence support frompeers; however, unlike AlcoholicsAnonymous, there is no single,set course for recovery that allmembers must follow. In fact,residents of Oxford House arefree to decide personallywhether to seek psychological orsubstance abuse treatment byprofessionals or a 12-step organi-zation. In short, Oxford House of-fers residents the freedom to de-cide whether to seek and choosewhich (if any) treatment they de-sire while receiving constant sup-port and guidance within an ab-stinent communal setting. Theinvolvement of the individual inthe course of treatment may en-

courage him or her to learn howto cope effectively and independ-ently with stressful situations thatpromote substance abuse. Thissense of competence and self-efficacy may reduce the risk ofrelapse when the person returnsto former high-risk situations.

To be admitted into an OxfordHouse, applicants fill out an ap-plication form and are inter-viewed by existing residents.Eighty percent or more of thecurrent house members mustvote in favor of the applicant’sadmission to the house for theapplication to be accepted. Thisis an example of how each Ox-ford House operates democrati-cally with majority rule. Every 6months, the residents elect offi-cers (e.g., president, secretary)who facilitate the handling of Ox-ford House clerical responsibili-ties (e.g., convene weekly meet-ings, collect rent). Oxford Housemembers maintain financial inde-pendence with each resident pay-ing rent and doing chores. Devia-tions from financial responsibilitiesto the Oxford House; behaviorsthat are disruptive, antisocial, orboth; or resumption of drug use,alcohol use, or both, result ineviction. The houses are rented,multibedroom dwellings forsame-sex occupants. At present,70% of the facilities house malesand 30% house females; 55% ofoccupants are White, 35% areAfrican American, 5% are His-panic, and 5% are other.

Oxford Houses are democratic, mutual help–oriented recoveryhomes for individuals with substance abuse histories. There aremore than 1200 of these houses in the United States, and eachhome is operated independently by its residents, without helpfrom professional staff. In a recent experiment, 150 individualsin Illinois were randomly assigned to either an Oxford House orusual-care condition (i.e., outpatient treatment or self-help groups)after substance abuse treatment discharge. At the 24-month fol-low-up, those in the Oxford House condition compared with theusual-care condition had significantly lower substance use, sig-nificantly higher monthly income, and significantly lower incar-ceration rates. (Am J Public Health. 2006;96:XXX–XXX. doi:10.2105/AJPH.2005.070839)

Communal Housing Settings Enhance Substance Abuse Recovery| Leonard A. Jason, PhD, Bradley D. Olson, PhD, Joseph R. Ferrari, PhD, and Anthony T. Lo Sasso, PhD

FIELD ACTION REPORT

SUBSTANCE ABUSE RECIDIVISMafter treatment is high for bothmen and women.1 Under mod-ern managed care, private andpublic sector inpatient substanceabuse facilities have reducedtheir services dramatically. Thus,there is a need to develop, evalu-ate, and expand lower-cost, resi-dential, nonmedical, community-based care options for substanceabuse patients.

Oxford House, founded in1975 by Paul Molloy, illustrates acommunity-based approach to-ward substance abuse absti-nence. The Oxford House model,described in the box on page 3,offers a community where resi-dents live without professionaltreatment staff and length-of-stayrestrictions, unlike traditionalhospital care, where trained pro-fessionals are necessary, or thera-peutic communities, where resi-dents have a maximum lengthof stay. Because there is nomaximum stay, residents may

In 1988, Congress passed theAnti-Drug Abuse Act, which allo-cated federal funds to any state toestablish recovery homes such asOxford Houses. A group of recov-ering substance abusers, throughthe support of an established Ox-ford House, may request fromtheir state a $4000 interest-freeloan to begin a new OxfordHouse. Payments on the loan arereturned to that state’s revolvingloan start-up fund so that addi-tional homes can be rented. In thelate 1990s, states were no longerrequired to administer a state loanprogram, but many states contin-ued to offer these loans to OxfordHouses. This loan program incombination with the OxfordHouse grassroots movement wasinstrumental in helping to expandthe number of Oxford Houses tomore than 1200 in 48 states.

During this expansion, manycommunities attempted to limitthe presence of Oxford Houses intheir neighborhoods. In 1995, theUS Supreme Court heard the caseof City of Edmonds v Oxford House.The purpose of this case was todecide on the Constitutionality ofthe city of Edmonds’ law that stip-ulated that no more than 5 indi-viduals who were unrelated bio-logically or legally could live in ahouse together. Using this law, thecity tried to close down an OxfordHouse that consisted of 13 resi-dents. The Oxford House organi-zation argued that this law, andothers like it, was discriminatory

American Journal of Public Health, Oct 2006; Vol. 96, pp1727–1729

FIELD ACTION REPORT

Oxford House condition com-pared with the usual-care condi-tion had significantly lower sub-stance use (31.3% vs 64.8%),significantly higher monthly in-come ($989.40 vs $440.00),and significantly lower incarcera-tion rates (3% vs 9%).

The economic data gathered(i.e., productivity and incarceration),which represent only a fraction ofthe total economic impact of sub-stance abuse,8 shows that OxfordHouse participants, by month 24,earned roughly $550 more permonth than participants in theusual-care group. In a single year,the income difference for the entireOxford House sample correspondsto approximately $494000 in ad-ditional production. In 2002, thestate of Illinois spent an average of$23812 per year to incarcerateeach drug offender.9 The lowerrate of incarceration among OxfordHouse versus usual-care partici-pants at 24 months (3% vs 9%)corresponds to an annual saving ofroughly $119000 for Illinois. To-gether, the productivity and incar-ceration benefits yield an estimated$613000 in savings per year, oran average of $8173 per OxfordHouse member.

Perhaps the recovery-home ex-perience of communal living

with a vacancy to apply for resi-dency. Using this system, we weresuccessful in placing each individ-ual assigned to the Oxford Housecondition into an Oxford House.In contrast, those in the usual-carecondition went to a variety of set-tings after discharge from the sub-stance abuse treatment settings.

Overall, participants’ question-naire completion rates across the24-month assessment period werecomparable for the 2 conditions(at the last assessment, more than90% of participants had filled outthe questionnaires). In addition, toincrease the validity of self-reportabstinence data, a person in eachparticipant’s support network listedon the final follow-up assessmentwas required to confirm the partic-ipant’s self-reported abstinence at24 months.

DISCUSSION ANDEVALUATION

We found that in importantareas such as substance use,criminal activity, and employ-ment, participants assigned to theOxford House condition showedsignificantly greater positive out-comes than those assigned to theusual-care condition. At the 24-month follow-up, those in the

one of the early studies,4 the re-search team found that OxfordHouse residents typically werenever married (53%), young(early- to mid-30s), White (58%),and male (70%). In another study,individuals who had entered Illi-nois Oxford Houses were followedup for a 2-year period, and 62%of those interviewed either re-mained in the house or had lefton good terms.5 In a study of indi-viduals in Missouri who had beenliving in Oxford Houses, 69%were abstinent at a 6-month fol-low-up interview.6 However, nei-ther study included a control orcomparison group. Other studieson Oxford House from the DePaulUniversity research team are avail-able elsewhere.7

In our study, 150 individualswere randomly assigned to eitheran Oxford House or usual after-care condition (i.e., outpatienttreatment or self-help groups)after they had received inpatienttreatment for substance abuse.We hypothesized that individualsassigned to the Oxford Housewould experience lower rates ofsubstance use and criminal activ-ity and higher rates of employ-ment compared with those receiv-ing the usual after-care services.

Of those approached to bepart of the study, only 4 refusedparticipation. Sociodemographicvariables did not differ signifi-cantly between participants as-signed to the Oxford House andthose assigned to usual-care serv-ices. Enrolled participants wereinterviewed every 6 months for a24-month period. All 75 individ-uals assigned to the OxfordHouse condition gained admis-sion to an Oxford House. Whenvacancies occurred, our researchteam was notified, and if an indi-vidual had not been voted intoone Oxford House, we broughtthat person to the Oxford House

KEY FINDINGS

• Given the high costs of substanceabuse disorders to society in gen-eral, and to the health care de-livery system in particular, the re-sults of this randomized test ofthe efficacy of a low-cost, self-help housing intervention com-pared with the usual services pro-vided after inpatient substanceabuse treatment have major pub-lic health implications.

• Because residents pay all ex-penses, these types of self-governed settings have importantpublic policy implications for sta-bilizing individuals with substanceabuse histories, especially in anera of cutbacks in funding for a va-riety of social service programs.

and that the city of Edmonds hadfailed to make reasonable accom-modations for the group home.

The Supreme Court decided infavor of Oxford House, citing theFair Housing Amendments Actof 1988 and upholding the 1993Washington State law stipulatingthat no Washington State citymay enact a zoning ordinancethat treats a residential structurehousing a collection of individu-als with disabilities different thanone housing individuals relatedgenetically or by law. Under theFair Housing Amendments Act,substance abuse was considereda disability, and therefore the cityof Edmonds was required tomake a reasonable accommoda-tion for Oxford House. This deci-sion had widespread implicationsfor the growth of Oxford Houseand other existing and futureprograms of this type.

Since 1992, a team of re-searchers at DePaul University, inChicago, has been studying theOxford House organization.1–3 In An Oxford House and its residents.

American Journal of Public Health, Oct 2006; Vol. 96, pp1727–1729

FIELD ACTION REPORT

helped reduce substance abuseproblems because residents wereable to develop a strong sense ofbonding with others who sharedcommon abstinence goals. To re-main in the Oxford House, mem-bers needed to pay weekly rent;this policy probably encouragedresidents to seek and maintainemployment. Reduced criminalactivity might have been influ-enced by the structure and loca-tion of the homes; houses wererented, multibedroom dwellingslocated in low-crime, residentialneighborhoods.

Our long-term collaborative re-lationship with Oxford House wasa significant factor in ensuring thesuccessful implementation of thisevaluation.10 The DePaul Univer-sity research team had workedwith Paul Molloy, the CEO of theOxford House organization, andthe Illinois Oxford House chapterfor 9 years prior to initiating thisstudy. Over this time period, Ox-ford House members and DePaulUniversity researchers collaboratedon pilot studies and developedhigh levels of mutual trust. By in-volving the Oxford House organi-zation in the planning of the re-search study, we gained a greaterappreciation of the culture andunique needs of the Oxford Housecommunity, and this collaborativeprocess led to the self-help mem-bers valuing and working coopera-tively with the research team. Thisprocess is well illustrated by Mol-loy’s decision to allow the re-searchers to employ randomizationin the experimental design, some-thing that is often difficult to ac-complish with self-help groups.The support from the OxfordHouse organization was critical inenabling the DePaul research teamto design a study with the method-ological rigor to secure fundingfrom the National Institute on Al-cohol Abuse and Alcoholism.

NEXT STEPS

After treatment for substanceabuse (e.g., hospital-based pro-grams or therapeutic communi-ties), many clients return to for-mer high-risk environments. Forboth men and women, returningto these settings without a sup-portive abstinence network in-creases the chance of relapse andrecidivism. The results of thisstudy suggest that Oxford Housemay be a promising type of recov-ery home for individuals attempt-ing to maintain abstinence. Manyexoffenders, individuals with psy-chiatric disorders, and others whoare homeless have substanceabuse problems, and similar typesof community-based housing set-tings might be effective with thesepopulations.11 The relationship be-tween outcomes and individualdifferences among residents inOxford Houses needs further ex-ploration. In addition, it is impor-tant to better understand whetherresidents need a minimum lengthof stay in Oxford Houses to ob-tain positive outcomes.

About the AuthorsLeonard A. Jason, Bradley D. Olson, andJoseph R. Ferrari are with DePaul Univer-sity, Chicago, Ill. Anthony T. Lo Sasso iswith the University of Illinois, Chicago.

Requests for reprints should be sent toLeonard A. Jason, PhD, Center for Com-munity Research, DePaul University, 990 W Fullerton Ave, Chicago, IL 60614(e-mail: ljason@depaul.edu).

ContributorsL.A. Jason supervised the overall studyimplementation, data analysis, and writ-ing of the article. B. Olson managed thedata collection and was responsible forthe data analysis. J.R. Ferrari con-tributed to study implementation andwriting of this article. A. Lo Sasso ana-lyzed the economic data and helpedwrite that portion of the article.

AcknowledgmentsThe authors received financial supportfrom the National Institute on AlcoholAbuse and Alcoholism (grant AA12218).

Human Participant ProtectionThis study was approved by the DePaulUniversity institutional review board.

References1. Jason LA, Pechota ME, BowdenBS, et al. Oxford House: community liv-ing is community healing. In: Lewis JA,ed. Addictions: Concepts and Strategiesfor Treatment. Gaithersburg, Md: AspenPublications; 1994:333–338.

2. Ferrari JR, Jason LA, Olson BD,Davis MI, Alvarez J. Sense of communityamong Oxford House residents recover-ing from substance abuse: making ahouse a home. In: Fischer A, ed. Psycho-logical Sense of Community. New York,NY: Kluger/Plenum Inc; 2002:109–122.

3. Olson BD, Curtis CE, Jason LA, etal. Physical and sexual trauma, psychiat-ric symptoms, and sense of communityamong women in recovery: toward anew model of shelter aftercare. J PrevIntervent Community. 2003;26:67–80.

4. Jason LA, Ferrari JR, Dvorchak PA,et al. The characteristics of alcoholics inself-help residential treatment settings: amulti-site study of Oxford House. Alcohol Treatment Q. 1997;15:53–63.

5. Bishop PD, Jason LA, Ferrari JR,

Cheng-Fang H. A survival analysis ofcommunal-living, self-help, addiction recovery participants. Am J CommunityPsychol. 1998;26:803–821.

6. Majer JM, Jason LA, Ferrari JR,North CS. Comorbidity among OxfordHouse residents: a preliminary outcomestudy. Addict Behav. 2002;27: 837–845.

7. Jason LA, Ferrari JR, Davis MI,Olson BD, eds. Creating Communities forAddiction Recovery: The Oxford HouseModel. Binghamton, NY: Haworth; 2006.

8. Department of Health and HumanServices. Economic Cost of Alcohol andDrug Abuse in the United States, 1992.Available at: http://www.nida.nih.gov/EconomicCosts/Index.html. AccessedApril 26, 2005.

9. Illinois Attorney General. Strategiesfor Fighting Meth. Available at: http://www.ag.state.il.us/methnet/fightmeth/courts.html. Accessed April 26, 2005.

10. Jason LA, Keys CB, Suarez-Bal-cazar Y, et al., eds. Participatory Com-munity Research: Theories and Methodsin Action. Washington, DC: AmericanPsychological Association; 2004.

11. Jason LA, Perdoux M. Havens: Stories of True Community Healing.Westport, CT: Praeger; 2004.

The Oxford House Model

• The house must be democratically self-run.• The house membership is responsible for all household

expenses. Each house is fully responsible for its ownexpenses and debts.

• An individual recovering from drug or alcohol addiction canlive in an Oxford House for as long as he or she does notdrink alcohol, does not use drugs, and pays an equalshare of the house expenses. The house must immedi-ately expel any member who uses alcohol or drugs.

• The average stay is a little more than a year, but many res-idents stay 3 or more years.

• There are men-only houses and women-only houses, butno houses for both groups to live together.

• Any recovering alcoholic or drug addict can apply to getinto any Oxford House by filling out an application andbeing interviewed by the existing members of the house.

• Any group of individuals recovering from alcohol or drug ad-diction can start a new Oxford House. All they need to dois to find a house to rent in the name of the group andapply to Oxford House Inc for a charter.

• Oxford Houses have 6 to 10 members. A house with fewerthan 6 individuals is difficult to maintain because of thesmall size of the group and the fact that any vacancycauses a greater disruption of the financial welfare of thehouse. A house must have 6 or more residents to be rec-ognized or chartered by Oxford House.

• There is no time limit on sobriety before coming into anOxford House. Generally, an individual comes into an OxfordHouse after a 28-day rehabilitation program or a 5- to 10-daydetoxification program.

American Journal of Public Health, Oct 2006; Vol. 96, pp1727–1729