Children and Youth Services Review

Transcription

1 Children and Youth Services Review 31 (2009) Contents lists available at ScienceDirect Children and Youth Services Review journal homepage: Lighthouse independent living program: Characteristics of youth served and their outcomes at discharge Mark J. Kroner a,, Alvin S. Mares b a Lighthouse Training Institute, 1501 Madison Rd., Cincinnati, OH 45206, United States b The Ohio State University, College of Social Work, United States article info abstract Article history: Received 18 August 2008 Received in revised form 16 October 2008 Accepted 21 October 2008 Available online 28 October 2008 Keywords: Foster care Independent living Treatment outcomes This study examined the outcomes of 455 young people who entered the Lighthouse Independent Living Program during the period On average, clients were admitted shortly before their 18th birthdays, and remained in the program for just under 10 months. At discharge, 60% had completed high school/ged program, 31% were employed, and 33% were independently housed. However, there were significant differences in outcomes across subgroups. Clients who presented with four or more clinical risk factors were less likely to have completed high school/ged program, less likely to be employed, and less likely to be independently housed than those who presented with fewer risk factors. Those staying in the program for less than 6 months were more likely to complete high school, but less likely to be employed and to be independently housed than those remaining in the program longer. Clients entering the program at ages years showed significantly better outcomes than younger clients. Female clients were more likely to be living independently at discharge, while no other gender or racial/ethnic group differences in outcomes were found. These descriptive data may provide useful benchmark data for independent living program planning, development, administration, and policy-making purposes Elsevier Ltd. All rights reserved. 1. Introduction 1.1. Challenges facing youth aging-out of foster care Youth emancipating from foster care face a number of challenges, including completing high school, coping with mental illness and substance abuse, attaining health insurance, finding employment and earning a living wage, securing housing, and completing school (Cook, Fleishman, & Grimes, 1991; Courtney, Piliavin, Grogan-Kaylor, & Nesmith, 2001; Courtney & Hughes-Huering, 2005; Dworsky, 2005; Festinger, 1983; Pecora, Kessler et al., 2005; Roman & Wolfe, 1997; Pecora, Williams et al., 2006). For example, a recent study of former foster youth from Wisconsin, Iowa and Illinois who were 21 years of age found that: 23% had dropped-out of high school; 18% had experienced homelessness since leaving care; only half were employed, with median annual earnings of less than $5500; 51% had health insurance, 70% of whom were covered by Medicaid; 71% of the young women had ever been pregnant, with 62% of those having had multiple pregnancies, half of the young men having ever impregnated a female, and more than half of the young women and nearly one-third of the young men having at Corresponding author. Tel.: ; fax: address: mkroner@lys.org (M.J. Kroner). least one child; and, 77% of the men and 54% of the women having ever been arrested (Courtney et al., 2007) Emergence of independent living The independent living (IL) field emerged officially in 1986 with passage of the Title IV-E Foster Care Independent-Living Initiative. This Initiative provided $70 million to states for the development of IL services for youth aging out of the child welfare services (Mech, 1988). The Initiative came out of a groundswell of expressed concerns in the field resulting from studies that were showing high rates of homelessness, public assistance, and incarceration among child welfare youth (Cook, 1988). Some communities had already begun to formally address this issue in previous years, often as pilot programs with no official licensing or standards in place (Mayne, 1988). Interest in housing assistance and support for emancipating foster youth increased and independent living programs started to experiment with different housing models, such as scattered-site or supervised apartments (Kroner, 1988; Brickman, Dey, & Cuthbert, 1991; DeWoody, Ceja, & Sylvestrer, 1993). Most of the housing-based independent living programs (ILPs) were supported by local funds as the Initiative did not allow funds to be used for direct housing costs such as ongoing rental payments. Moreover, the effectiveness of these housing models had not been formally evaluated (Barth, 1990; Waldinger & Furman, 1994) /$ see front matter 2008 Elsevier Ltd. All rights reserved. doi: /j.childyouth

2 564 M.J. Kroner, A.S. Mares / Children and Youth Services Review 31 (2009) The John Chafee Foster Care Independence Act (Public Law , 1999), which was passed by Congress in 1999, doubled the funding given to states up to $140-million and allowed for 30% of these funds to be used for housing. Program models emerged that reflected the fiscal and cost-of-living realities of individual states and communities (Kroner, 2001; White & Rog, 2004). The field became more sophisticated as federal support, communication between programs, focus on special populations and housing models increased and improved (Mech, 2003) Brief description of Lighthouse Youth Services, Inc. Lighthouse Youth Services is a multiservice agency providing social services to children, youth and families in south eastern Ohio. Operating since 1969, Lighthouse is nationally recognized as an innovator in services for families in crisis, for homeless youth and young adults, for youth learning to become self-sufficient, and in foster care for abused or neglected children ( In 2006, Lighthouse had six service divisions, a budget of $16- million, and a staff of 300. The revenues the agency received came from purchase of service contracts, local, state and federal grants, foundations and private donations Overview of the Lighthouse Independent Living Program Guiding principles The evolution of the Lighthouse Independent Living Program over the past 25 years, and experience working with a wide range of emancipating youth, has established the following guiding principles or beliefs: First, foster youth preparing for emancipation need time to adjust to the real world, to make decisions on their own, within limits, and to make mistakes while still under the support of caring adults. The analogy of driver's education to independent living applies; namely, that effective independent living training requires some type of transitional living bridge period where emancipating youth are able to live more on their own, out in the community, with agency/program supports to learn by doing, just as effective driver's training requires youth to spend some time behind the wheel, driving out on the roads and practicing maneuverability in parking lots, with a supportive adult occupying the passenger's seat. Next, while the average 10-month bridge period currently provided to 18 year-old clients is preferred to receiving no such transitional living experience, it is too brief to meet the numerous challenges facing emancipating youth. Consider the considerable level and growing number of years' support and assistance provided by parents to young adults from intact families (Arnett, 2000) young adults with much less troubled childhoods than emancipating foster youth who have experienced years of childhood abuse and neglect, multiple foster home placements and changes in schools, separation from siblings, etc. For example, twothirds of 19 year-olds from intact families live with their parents (Current Population Survey (CPS), 2005), and young adults receive an average of $2200 per year from their parents between the ages of 18 and 34 to supplement wages, pay for college tuition, help with housing costs, etc. (Schoeni & Ross, 2005). Finally, risk is part of change (McMillen, 1999). Very few 17 or 18 year-olds are ready to live on their own, be they foster youth or youth from intact families. When set out to do so, they are all but certain to make mistakes, both big and small, and in doing so (hopefully) learn as they go. So housing-based independent living programs must be designed to accommodate the full range of mistakes which their clients will make, despite the best efforts of program staff to minimize such mistakes, and to protect clients from the potential harm or consequences resulting from these mistakes Funding The Independent Living Program currently receives $65 a day for a youth in a scattered-site apartments, and $85 a day for other living arrangements such as a shared-home, supervised apartment or outof-county placement. The program accepts pregnant and parenting youth and receives an additional $10 a day for a child in the mother's custody. Per diem revenues cover about 85% of program expenses. The ILP billed Medicaid for mental health services provided for many of its eligible youth. The program also brings in revue from smaller grants and private donations Staffing The ILP has a full-time director, a full-time assistant director and clinical supervisor, six licensed Social Workers (with a BSW or MSW), a full-time mover who transported furniture to/from a central storage facility into and out of clients' apartments, a full-time housing specialist/case-aides and resident managers for the shared homes and supervised apartments. Each social worker typically carries a caseload of 10 12, depending on caseload complexity. All Social Workers are licensed by the State of Ohio. The ILP often hires staff from other Lighthouse programs or students completing field placement requirements at the ILP while pursuing a Social Work degree. Staff are expected to conduct at least one face-to-face visit and one additional apartment visit per week with each client. High-risk youth are contacted and seen several times a week, or even daily, when necessary. Assigned social workers are responsible for overall casemanagement and for working with referring agencies to develop a treatment team and support network for each client. There is a Social Worker on-call 24/7/365 for emergencies or crisis counseling Clients served The program serves current dependent foster youth referred by the Hamilton County Children's Services, and delinquent youth referred by the Ohio Department of Youth Services. Most are referred shortly before reaching the age of majority, and when transitional housing with supervised independent living skills training is needed. The program also serves as a placement setting of last resort for harder to place youth with unsuccessful placement histories and those with involved with multiple systems of care (e.g., child welfare, juvenile justice, mental health). In most cases, clients officially emancipate from public care at the time of discharge from program Housing placement and support The scattered-site housing model is utilized as the primary living arrangement (Kroner, 1999). The program pays the security deposit, rent, utilities, and phone bills, and provides furniture and house wares for each client, with clients taking over some of their bills toward the end of their stay in the program. Clients receive $55 weekly, $10 of which is placed into savings accounts for their use once they leave the program. The remaining $45 covers food, transportation and personal care items. Clients may attain additional spending money through competitive employment in the community, outside of the program. The program mainly uses apartments rented from private landlords. Clients are placed throughout Cincinnati, Ohio, near a bus line. Attempts are made to place youth into apartments that they can likely afford after emancipating from public care. In addition, attempts are made to place youth in areas with which they are familiar, close to school, work and supportive adults. The youth are involved in choosing their housing placements. Lighthouse signs the lease and assumes responsibility for the youth's overall behavior. Clients are allowed to keep their apartments, furniture, supplies and security deposits if they are employed at discharge and have proven to the

3 M.J. Kroner, A.S. Mares / Children and Youth Services Review 31 (2009) landlord that they are responsible. Clients who do not have a stable source of income at discharge receive assistance in finding other living arrangements, including low-income/subsidized housing. The ILP also has several other living arrangement options, such as the Anna Louise Inn, a boarding home for women, two small semisupervised shared-homes with live-in staff and two supervised apartment buildings with staff who lived in on-site apartments. Occasionally, for youth with special needs, the program utilizes Host Homes, which are essentially boarding homes for one youth, for special situations. Lighthouse Youth Services, of which the ILP is a part, also runs a temporary shelter, which is available to ILP clients for short-term housing and emergencies. These additional options provide a menu of short-term respite and/or alternative living arrangements for those clients temporarily unable to live independently. Youth are sometimes removed from individual apartments if they fail to progress or continuously violate program or landlord rules. In such circumstances, youth are placed into more supervised settings and then given a chance to earn their way back to their own apartments after demonstrating positive behavioral changes. The program has learned that many youth do better the second or third time they are given a chance to live alone. While being allowed to make and learn from their mistakes, clients are occasionally discharged from the program for continued rules infractions or involvement in illegal activities. Some of these clients are able to return to the program under a Chafee-funded aftercare program Purpose The purpose of this study is to empirically describe the clients served by the Lighthouse IL Program, the services they received, and the outcomes they achieved at discharge. Included in the description is a comparison of client outcomes across various client sub-groups (i.e., length of stay, client age at admission, risk groups, gender, and racial/ethnic groups). These descriptive data may provide useful benchmark data for IL program planning, development, and administration, as well as for IL policy making. 2. Methods 2.1. Sample and data collection The sample consisted of all youth (N=455) admitted into the Lighthouse Independent Living Program during the six-year period Experienced Lighthouse staff compiled existing clientlevel administrative and clinical records from the agency's management information system and paper files, under the direction of the first author. De-identified data were then analyzed by the second author in accordance with procedures approved by the institutional review board at the second author's academic institution Measures Client characteristics at program entry Intake/admission measures included year of admission, referral source, prior living arrangement, and length of stay. Demographic characteristics included age at admission, gender, and race/ethnicity (defined dichotomously, based on minority status). Clinical characteristics, as assessed by Lighthouse staff within 60 days of admission, consisted of one measure of overall functioning, the Global Assessment of Functioning (GAF; Endicott, Spitzer, Fleiss, & Cohen, 1976), and a unique self-sufficiency rating score developed by the program and assigned by case managers which ranged from 0 to 130, with higher scores indicating a higher level of knowledge of everyday living skills. The GAF a single-item measure ranging from 0 (lowest level of functioning) to 100 (highest level of functioning) based upon overall level of functioning in social, work, and school life domains is one of the two most commonly used mental health measures of functioning and has been found to have interrater reliability coefficients for use with children and adolescents ranging from.54 to.92 overall, although with somewhat less reliability among traumaexposed youth (Blake, Cangelosi, Johnson-Brooks, & Belcher, 2007). All program social workers were trained in GAF assessment by the agency's clinical director, a licensed psychologist. Twenty two dichotomous clinical risk factors or barriers/challenges facing foster youth preparing for emancipation were developed by the first author, who served as director of the IL program for nearly 18 years. Two Lighthouse staff then checked all applicable risk factors for each subject, based upon a review of various records, including: (a) intake/admission records (i.e., the Referral Sheet, Social History Form, and Intake Screening Form; (b) treatment records (i.e., Diagnostic Assessment Form, Incident Report, and progress notes; and, (c) discharge records (i.e., Termination Summary Form). A summary of the operational definitions used during this chart abstraction process is provided in Table 1. The abstraction process required over 100 total person hours of effort, spread over a five-month period (January through May 2008), and divided between two abstractors. The first abstractor served as Administrative Assistant for the ILP for over 15 years; the second was a licensed clinical social worker (LSW) who had worked at Lighthouse for several years. Cases were split between these two abstractors, with one taking those clients admitted during the period , whose records were archived in paper form, and the taking clients admitted more recently (i.e., during the period ), whose records were accessible in electronic form. Individual risk factor measures were then classified into six, nonmutually exclusive risk factor groups, based upon face validity: 1) Mental health and substance abuse risks (i.e., having an ongoing mental health issue, being on psychotropic medication, attempting suicide during the past year, and being chemically dependent) 2) Teen parenting risks (i.e., being pregnant or having one or more child(ren)) 3) Delinquency risks (i.e., committing a felony offense or having two or more misdemeanors during past year, involvement in gang activities, involvement of family or friends in illegal activities, and being violent towards others or committing a sex offense during past several years) 4) Learning disability risks (i.e., being diagnosed with a development disability, having limited intellectual abilities, and being unable to read or write) 5) Social adjustment risks (i.e., having a chronic history of truancy or school problems, running away from a stable placement during past year, and having no known social supports) 6) Other risks (i.e., having chronic medical issues, a history of poor judgment, little or no work experience in the private sector, and avoiding responsibilities as much as possible) Service receipt while in the program A service use checklist is completed by IL case managers when clients are discharged from the program. The list contains 38 specific services which clients may have received at any time during their stay in the program. These specific services were classified into four major types: 1) basic and community, 2) direct treatment, 3) skills training, and 4) referral. Basic and community services included food, clothing and shelter assistance, provision of furniture and house wares for clients' apartments, and recreational and community service activities for the youth. Direct treatment services were those delivered by program staff, in contrast to other services which were provided by agency staff from other programs. Skills training included a broad range of life skills (e.g., education, career, activities of daily living) that were taught by program staff, mostly in clients' apartment and employment settings in the community.

4 566 M.J. Kroner, A.S. Mares / Children and Youth Services Review 31 (2009) Table 1 Chart abstraction coding methodology used in developing risk factor categories. Risk factor categories and items Motivation and health Has little or no work experience in the private sector Has history of poor judgment Avoids responsibilities as much as possible Has chronic medical issues Mental health and substance abuse Has chronic mental health issues Is on psychotropic medication Is chemically dependent Has made a suicide attempt in last year Socialization Has chronic history of truancy or school problems Has runaway from a stable placement in the last year Has no known social supports Delinquency Has been violent toward people in last several years Has friends/family members involved in illegal activities Has had more than two misdemeanors in last year Has committed a felony offense in last year Has committed a sex offense in last several years Has been involved in gang activities Teen parenting Has a child, has more than one child Is pregnant Learning disability Has limited intellectual abilities Has a diagnosed developmental disability Cannot read or write Operational definition and data source(s) Short periods of multiple jobs, as indicated on Termination Summary Form Indicated on Interview Screening Form, Diagnostic Assessment Form, and/or Termination Summary Form Overall record indicated that youth accepted little to no responsibility in most or all areas of life One or more illnesses requiring on-going medical care or affecting youth's functioning was noted in medical history or diagnosed during treatment (e.g., diabetes, asthma, high blood pressure) Any diagnosed mental health problem, as indicated on Referral Sheet or on Diagnostic Assessment Form Indicated on Referral Sheet or on Diagnostic Assessment Form Mentioned on the Diagnostic Assessment Form; OR residential treatment within last 3 years; OR progress notes indicated daily living affected by use of illicit drug(s) Indicated on Interview Screening Form, Diagnostic Assessment Form, or Incident Report Clearly identified on Social History and/or Interview Screening Form(s) Indicated in Social History and/or Interview Screening Form(s) at intake; OR progress notes while in program Minimal contact with family indicated throughout records Any indication of fighting or domestic violence, either at home or at school, indicated in records Indicated in Social History; OR mention of recent family involvement in progress notes Impression gathered through review of Social History Indicated in Social History or Interview Screening; OR through Incident Report; OR referred by Ohio Department of Youth Services Indicated on Referral Sheet, Social History, or Interview Screening Clearly stated in Social History Indicated on Referral Sheet, Interview Screening, or Termination Summary (Same as above) Low IQ listed, very low grade level for age, or Individual Education Plan, as indicated on Referral Sheet, Interview Screening, or in progress notes Clearly indicated in Diagnostic Assessment; including: Autism, Asberger, and pervasive developmental disabilities Clearly indicated within the clinical records Data sources: Referral Sheet, Social History, and Intake Screening Form (upon entry into the program); Diagnostic Assessment Form, Incident Report, and Progress Notes (while in the program); and, Termination Summary Form (at discharge) Outcomes at discharge from the program Three dichotomous measures included 1) whether the youth had completed high school or the equivalent (i.e., received either a high school diploma or GED), 2) whether the youth was employed or had completed a vocational training program, and 3) whether the youth was living independently in his/her own place (i.e., renting an apartment or a private room in a house) at the time of discharge. It should be noted that Hamilton County generally emphasizes discharging youth from care as soon after their 18th birthday as possible. Unlike other states where youth are able to remain in care until the age of 21, such as Illinois, most youth in this study were therefore discharged before their 19th birthday Data analysis Independent samples t-tests, Chi-square tests, and analysis of variance (ANOVA) tests were used to examine possible differences in service use and outcomes between client sub-groups of interest. First, t-tests and chi-square tests were used to compare continuous and dichotomous service utilization measures, respectively, between clients without any risk factors and those with any risk factors (Table 4). Chi-square tests were also used to compare client outcome measures with two dichotomous client characteristic measures gender and race/ethnicity (i.e., minority vs. non-minority) (Table 6). ANOVA was used to compare client outcome measures and three four-level ordinal measures of client characteristics (sub-groups) of interest; namely, risk factor categories (Table 5) and length of stay and age at admission (Table 6). Tukey pair wise comparisons were included in these ANOVA analyses to compare outcomes between risk factor category, length of stay, and age of entry sub-groups. 3. Results 3.1. Client characteristics Intake/admission characteristics During the 6-year period examined, an average of 76 youth entered the program each year, an average of just over six new clients each month, on average. Most (82%) of these youth are referred by public agencies in Hamilton County, Ohio, primarily from the Department of Children's Services. Forty percent come to the program from foster care homes, 19% from group homes, and 41% from other living arrangements (Table 2). Once admitted, clients stayed in the program an average of 292 days, or 9.6 months, in the program. Twenty percent of clients stayed less than 3 months, nearly half (48%) stayed between 3 and 12 months, and another 28% stayed between one to two years in the program. Less than 5% of the clients remained in the program for longer than two years (Table 2) Demographic and clinical characteristics The mean age of admission into the program was 17.9 years, and ranged from 16 to 20 years of age. Eighty-seven percent of clients entered the program at 17 or 18 years of age. More than half of the clients served were female (56%) and 70% belonged to a racial or ethnic minority, mostly (64%) African-American (Table 2). Clients' overall levels of functioning as measured by the Global Assessment of

5 M.J. Kroner, A.S. Mares / Children and Youth Services Review 31 (2009) Table 2 Description of clients (N=455). Percentage (%) or mean Intake/admission characteristics Year of admission Average no. admissions per yr. 76 Referral source Hamilton County Other source Prior living arrangement Foster care Group home Other arrangement Length of stay b3 mos mos mos mos mos Mean no. days Demographic characteristics Age at admission 16 years years years years years Mean age (years) Gender (female) Racial/ethnic minority African-American Frequency (N) or standard deviation (SD) Clinical characteristics GAF at intake (0 100) (n=392) Self-sufficiency rating at intake (0 130) (n=375) Risk factor categories Motivation and health (any) 56% 254 Has little or no work experience in 42% 192 the private sector Has history of poor judgment 35% 159 Avoids responsibilities as much 17% 78 as possible Has chronic medical issues 13% 57 Mental health and substance abuse (any) 49% 221 Has chronic mental health issues 47% 213 Is on psychotropic medication 17% 78 Is chemically dependent 10% 46 Has made a suicide attempt in last year 5% 24 Socialization (any) 47% 213 Has chronic history of truancy or school 41% 186 problems Has runaway from a stable placement 17% 76 in the last year Has no known social supports 11% 48 Delinquency (any) 41% 186 Has been violent toward people in 27% 123 last several years Has friends/family members involved 11% 52 in illegal activities Has had more than two misdemeanors 10% 46 in last year Has committed a felony offense 8% 37 in last year Has committed a sex offense in last 6% 26 several years Has been involved in gang activities 1% 3 Teen parenting 18% 84 Has a child, has more than one child 16% 71 Is pregnant 5% 21 (continued on next page) Table 2 (continued) Functioning (GAF) scale averaged 61, with 60 being the cut-point between the young adults having some difficulty (61 70) and moderate difficulty (51 60) in social, occupational, and/or school functioning (DSM-IV-TR, 2000). Self-sufficiency ratings at intake averaged 77, or squarely in the mid-range of the 130-point rating scale. Two-thirds of clients presented with one or more risk factors, 10% facing identified learning disability issues to 56% facing motivational and health care issues (Table 2) Receipt of services Overall, clients received an average of 6.8 individual services. Over half (56%) of clients received mental health, substance abuse, educational and/or vocation service from other providers and agencies, two-thirds (64%) received life skills training from program staff, three-fourths (77%) received direct treatment services, and 87% received basic services, most notably basic support in the form of food, clothing, and shelter. Nearly 40% of clients received all four types of services while in the program (Table 3). Basic services utilization rates were greater among clients having one or more risk factors than those without any risk factors (94% vs. 86%, pb.01). Surprisingly, no significant differences were found in either skills training or referral to services provided by other agencies between any risk factor and no risk factor client groups. Although any risk factor clients received a greater number of direct treatment services (2.7 vs. 2.0, pb.01), the difference was not statistically significant (83% vs. 75%, p =.07). Any risk factor clients were, however, more likely to receive multiple types of services (3.1 vs. 2.8, pb.05), than those without any risk factors (Table 4). Subsequent analyses comparing the number of types of services received among clients represented with zero, one to two, three, and four to six types of risk factors indicated that the above differences were mostly attributable to clients without any risk factors receiving fewer direct treatment services than clients having four or more types of risk factors (2.0 vs. 3.0, pb.05), and fewer types of services overall (2.8 vs. 3.1, pb.05) (data not shown). Thus, clients with one, two, or three types of risk factors appeared to receive comparable types of services during their stay in the program Outcomes at discharge Percentage (%) or mean Clinical characteristics Risk factor categories Learning disability 10% 47 Has limited intellectual abilities 7% 30 Has a diagnosed developmental disability 4% 20 Cannot read or write 1% 5 No. risk factor categories None 33% 151 One 5% 23 Two 11% 49 Three 18% 84 Four 24% 108 Five 9% 40 Six 0% 0 Frequency (N) or standard deviation (SD) Overall. At the time of discharge, 60% of clients had completed high school or obtained their GED, 31% were employed or had completed a vocational training program, and one-third (33%) were living independently, either by themselves or with a friend, in their own apartment, room, or house (Table 5) Relationship between risk factors and outcomes. Surprisingly, clients with one or two types of risk factors had better outcomes in the areas of employment and independent housing (54 55%) than clients

6 568 M.J. Kroner, A.S. Mares / Children and Youth Services Review 31 (2009) Table 3 Types of independent living services provided to clients (N=433 discharged clients). %/Mean N/SD Rank Types of services Basic services (any) 92% 397 Basic support (food, clothing) 82% Furniture provision 68% Recreational activities 32% Community service activities 5% No. other svcs Direct treatment (any) 81% 349 Group counseling 18% Respite services 13% Tutoring 12% Medication administration/monitoring 11% Sexuality/STD prevention 11% Psychiatric services 11% Psychiatric liaison 10% Family planning 10% Family counseling 7% Lighthouse Community School 5% Substance abuse 4% Sex offender 4% Child care 3% Sexual abuse prevention 2% Other counseling/intervention 54% Other educational/employment 45% Other health care 29% No. treatment svcs Skills training (any) 67% 292 Employment skills 54% Diagnostic assessment/testing 24% Parenting skills 16% Vocational training 16% GED preparation/testing 8% Violence prevention 5% No. skills training svcs Referral services provided by another program (any) 59% 256 Outside mental health 20% GED 19% Optical 15% Respite 7% Substance abuse information 6% Legal 6% Early intervention 5% Child care 5% Substance abuse 3% Parent mental health 1% 6 38 Other 38% No. referral svcs No. types of services received (0 4) One Two Three Four Total no. services received (0 38) without any risk factors (31 35%) (Table 5). Clients with one or two risk factor types also did better than those with four or five risk factor types on all three outcomes, along with clients with three risk factors for employment. The only significantly better outcome found among clients without any risk factors was in comparison to those with four or five types of risk factors in the area of employment (35% vs. 19%, pb.05). It is possible that clients with one or two risk factors outperformed other clients because they were dysfunctional enough to draw the attention of program staff and resources, while being healthy enough, and possibly motivated enough, to take advantage of their likely final chance to get their lives together before being discharged from the child welfare system. While the only statistically significant difference in service utilization between risk factor category groups was a tendency for those without any risks to receive fewer types of services than clients with four or more risk factor categories, the Table 4 Types of independent living services provided to select client risk factor groups. Type of service No risk factors Any risk factors %/Mean N/SD %/Mean N/SD Chi-sqr. t p Any basic No. basic svcs Direct treatment ns (.07) No. treatment svcs Skills training ns No. skills svcs ns Referral ns No. referral svcs ns No. svc. types (0 4) pb.05; pb.01; ns pn.05. rudimentary nature of the dichotomous service utilization measures at any time while in the program precludes ruling-out (or in) the possibility that clients with one or two risk factor categories may have received a greater level of services while in care. Regardless, even for this most successful group, nearly one-fourth had not completed high school and nearly half were unemployed and did not have an affordable place to live on their own at the time of leaving the program. Many youth enter the ILP one to two years educationally behind their peer group, and are not able to stay in care long enough to graduate. It is the program's hope that Ohio will eventually be able to keep cases open until the age of 21, providing these youth with more time to complete high school. Among risk factor groups, teen parents (mothers) appeared less likely to complete high school (45%) and more likely to be independently housed (46%), yet comparably employed (24%) compared with other risk factor category clients (Table 5). Although high in absolute terms, the 55% drop-out rate among teen parents is less than the 70% rate reported by the Robin Hood Foundation (1996), perhaps due to their benefiting from a longer average length of stay in the program (401 days vs. 266 days for non-parents, pb.001) and, ironically, a lower IL program drop-out rate (20% vs. 40% for nonparents, pb.01). Again, extending foster care to the age of 21 would likely provide teen mothers with a better chance to complete high school graduation or GED completion requirements. Outcomes for the remaining risk factor groups appeared to be surprisingly similar, with high school completion rates ranging from 52 59%, employment rates from 23 28%, and independent housing rates from 26 29% (Table 5) Other factors. Four other factors believed to possibly influence client outcomes included program length of stay, age of admission, gender, and race/ethnicity. Clients remaining in the program for Table 5 Comparison of client outcomes by risk factor category groups. Risk factors Client outcomes Completed h.s. Employed Independ. housed % N % N % N All clients No. risk factor categories Zero categories (A) One to two categories (B) Three categories (C) Four or more categories (D) Tukey pairwise comparisons BND AND; BNA, B NA,D C,D Risk factor categories Motivation and health Mental health and substance abuse Socialization Delinquency Teen parenting Learning disability

7 M.J. Kroner, A.S. Mares / Children and Youth Services Review 31 (2009) Table 6 Comparison of client outcomes by length of stay in program, age of entry, gender, and race/ethnicity. Length of stay in program A B C D ANOVA Pairwise comparisons b6 months months months 18+ months Mean SD Mean SD Mean SD Mean SD df f p Completed high school ANB,D Employed A bb D; BbD Living independently AbB D Age at admission into program A B C D ANOVA Pairwise comparisons years years years years Mean SD Mean SD Mean SD Mean SD df f p Completed high school A CbD Employed BbD Living independently B,CbD Gender Race/ethnicity Male Female Chi-sqr. p Minority Caucasian Chi-sqr. p % N % N % N % N Completed high school ns ns Employed b0.1 ns ns Living independently ns pb.05; pb.01; pb.001;ns pn=.05. longer than 6 months were more likely to be employed and more likely to be independently housed, but less likely to have completed high school, compared with those who left within 6 months of admission (Table 6). This may be explained by the fact that some clients enter the program for a limited time with the expressed purpose of finishing their final year of high school. As one might expect, clients who entered the program at ages 19 or 20 generally had better outcomes than younger clients who entered at ages Older clients were more likely to have completed high school, more likely to be living independently than 17 and 18 year-old clients, and were more likely to be employed than 17 year-old clients (Table 6). No significant differences in outcomes were found among younger age groups (i.e., 16, 17, and 18 year-olds). The only significant difference in outcomes found among gender or racial/ethnic groups was that female clients were more likely to be living independently at discharge than male clients (39% vs. 25%, pb.01) (Table 6). 4. Discussion This study is among the first to empirically describe a housingbased independent living program, the characteristics of emancipating foster youth clients served in the program, and primary treatment outcomes for these youth upon leaving the program. In doing so, all three outcome evaluation domains of program structural characteristics, program process characteristics, and case outcomes (Courtney, 1993) have been described, to the extent possible using existing administrative and clinical data routinely collected by program and agency staff over a six-year period. These data are believed to fairly accurately describe the program characteristics, and the socio-demographic characteristics and clinical risk factors of the clients served by the program. Primary treatment outcomes at the time of discharge are also believed to be fairly accurately documented through these data, but are limited in scope (e.g., examine only three, dichotomously measured outcomes pertaining to educational attainment, employment, and housing), and fail to address longer-term outcomes post-discharge. Descriptive data on the treatment process are the least specific and are perhaps more suspect to unknown validity than program structure or outcomes data examined. Thus, this study adds to the existing scholarly literature which empirically describes specific independent living program models of practice. Scannapieco, Schagrin, and Scannapieco (1995) described the Baltimore County Department of Social Services Independent Living Program, and the 44 youth served during the five-year period , and found that 50% had completed high school, 52% were employed, and 36% were living independently at the time of discharge (which averaged 19 years of age). Mallon (1998) described the Life Skills Program developed by Green Chimneys, a non-profit child serving agency contracting with the New York City Administration for Children's Services, and the 46 youth served during the six-year period , and found that 74% had completed high school, 79% were employed, and 61% were living independently (including sharing an apartment with another) at the time of discharge (which averaged 21 years of age). Most recently, a report by the Administration on Children & Families (ACF, 2008) described the Community College Foundation's Life Skills Training (LST) Program, operating out of Los Angeles County, California. Among the 222 youth randomly assigned to the LST treatment group, 60% had completed high school and 45% were employed at age 19. Among these, and other statewide studies (e.g., Lindsey & Ahmed's, 1999 study of the Independent Living Program in various counties in North Carolina), and regional/multiprogram studies (e.g., Georgiades's, 2005 study of Florida's Miami- Dade and Monroe Counties IL Programs), the current study is the first description of a larger-scale housing program reported on in the published child welfare literature. The client outcomes data reported in this study, as in previous studies, may be viewed from a strengths-based or deficits-based perspective, while providing comparative or benchmark data for future studies and policy-making and program development. High school completion, employment, and independent housing rates for the overall sample at 19 years of age (on average) at discharge of 60%, 31%, and 33%, respectively, demonstrate both accomplishment and room for improvement. Moreover, the variability observed in these overall rates by risk factor groups, age at admission, and length of stay among clients in this particular housing model of independent living suggests the need for additional model-specific structure, treatment, and outcomes data collection and reporting, and also additional research to better understand outcomes variability between client

8 570 M.J. Kroner, A.S. Mares / Children and Youth Services Review 31 (2009) sub-groups and program models. The three studies mentioned above and this study are somewhat like four very different types of fruit, precluding the possibility of comparing apples to apples client outcomes across these studies. While rigorous evaluations of specific independent living models of practice are most certainly needed to determine the efficacy and effectiveness of any given model, the many challenges and high costs of conducting such evaluations suggest the need to make do with smaller-scale, empirically descriptive studies, such as this, which could be conducted relatively easily and at low cost for most any established independent living program. Additional mid-level, quasi-experimental and/or longer-term outcomes evaluation studies for specific practice models are also clearly needed to fill the gap between small-scale descriptive studies and rigorous randomized clinical trial studies. The dirth of published empirical data on even the most basic descriptive level for specific models of independent living practice addressing all three evaluation domains program structure, treatment process, and client outcomes suggests both great need and opportunity to advance the independent living field within child welfare. While providing potentially useful descriptive data to researchers, policy makers, funders, program developers and managers, etc., the findings should be considered in light of three major limitations. First, the validity and reliability of the measures used are largely unknown, including risk factor classification data, GAF and self-sufficiency measures, and even, to a lesser extent, outcomes data. These data were compiled retrospectively from various administrative and clinical records, over a six-year period, and are believed to be accurate based primarily on the considerable experience of program staff, relatively advanced management information system, and longevity of the Lighthouse Independent Living Program. Next, the validity and utility of the service utilization data are particularly questionable. While the dichotomous measurement of various types of independent living services is the approach recommended by the Children's Bureau in the recently issued National Youth in Transition Database (NYTD) federal regulations (ACF, 2008), ideally service utilization measures would include some level or frequency data to distinguish a youth who received a single session of educational tutoring, vs. the youth who received 30 tutorial sessions, for example. Third, the descriptive and bivariate client sub-group comparisons presented here are unadjusted i.e., not statistically adjusted for potentially confounding differences in client characteristics or use of services. While beyond the scope of this descriptive study, a more detailed multivariate examination of client characteristics associated with outcomes at discharge is planned for a later date. Finally, the external validity, or generalizability, of these findings is limited, given the uniqueness of the Lighthouse model, the close working relationship of the agency with the various public systems of care operating in Hamilton County, Ohio, the highly developed housing continuum of care available to independent living clients, etc. While early efforts are currently underway to export the Lighthouse model to other locales, the extent to which the model can be feasibly implemented outside of Hamilton County is not yet known. The findings of this study raise numerous questions for communities considering developing independent living housing programs for their emancipating foster youth. One such question is how long should the average length of stay be? The fact that counties and the State of Ohio, like many other States, are under financial pressure to discharge youth soon after their 18th birthdays due to budget limitations, while at the same being reluctant to accept the risk and liability of placing minors in public care into their own apartments out in the community, explains Hamilton County's compromise position of typically placing youth into the Lighthouse Program just a few months before their 18th birthday, providing them with about 10 months of hands-on skills training while living in their own apartments, and then discharging them before their 19th birthday. While longer lengths of stay appear to be related to improved outcomes in the areas of employment and independent housing, extending average lengths of stay would likely result in smaller numbers of youth served each year, assuming no increases in public funding for independent living at the county or state levels. Such increases may be more likely, though, in light of the recent passage of The Fostering Connections to Success and Increasing Adoptions Act (H.R. 6893, 2008), which will expand federal Title IV-E reimbursements to states providing out-of-home placements during the ages years bridge period. A second key question is which foster youth sub-group(s) should be placed into independent living housing programs? The risk factors data reported herein suggest that Hamilton County places youth with many different types of needs into the Lighthouse ILP, including youth with no demonstrable special risk factors to those having five different types of risk factors. Many of the more troubled youth entered the program with no previous histories of success in school or at work. They were placed (and accepted) into the program as an option of last resort for the county and other referring agencies to provide these more challenged youth with some practical skills for a few months until they were discharged from public care. Decreasing the admission of the highest risk youth (i.e., those with four or more clinical risk factor categories) would most likely produce better program outcomes overall. However, Lighthouse has felt that providing services to last chance youth is more important than improving overall client outcomes. After all, it is the high-risk youth who are most in need of services and supports. And Hamilton County has been willing and able to fund such efforts, for many of these high-risk youth, over the years. Yet other counties or states developing housing programs may decide to be more selective in targeting certain client sub-groups, such as teen parents, delinquent youth, or youth with mental health problems. A third question is how to involve the public mental health system in meeting the mental health needs of emancipating foster youth? With nearly half of the Lighthouse clients having documented mental health issues, it seems likely that an important piece of the emancipation puzzle for those youth is the assumption of case management, medication management, and other basic services typically provided by the adult community mental health system after youth are discharged from the child welfare system. For example, Lighthouse is currently involved in a pilot project funded by the local county's mental health system, drawing from the experiences of others around the country who are serving this population (Davis & Vander Stoep, 1997; Clark & Davis, 2000) that is showing some promise in this area. The program connects youth in custody with an adult mental health system case-manager who takes over the case when the youth turn 18 and custody is terminated. Housing is provided throughout the transition process. Other counties or states may have other ways of addressing this issue, but somehow the mental health needs of emancipating youth with mental health problems ought to be assessed and addressed during the bridge period to independence. Finally, the findings of this study raise four important questions which may be examined in future research. First, what are the causes of low high school completion rates and high independent housing rates among teen parents (moms)? Dropping-out of high school portends a difficult future and life for teen parents and their children, alike (Robin Hood Foundation, 1996). Extending the length of stay for teen mothers may relieve some of the pressure to find a paying job and allow them to focus on acquiring parenting skills and completing high school. Next, to what extent did clients experience multiple moves or placements within Lighthouse's continuum of housing? Providing youth with a second or third chance in a different living arrangement may result in better outcomes. Having a better understanding of which client sub-groups spent the majority of their time in the program in various levels of care/various types of housing settings

9 M.J. Kroner, A.S. Mares / Children and Youth Services Review 31 (2009) may help to inform the question of which types of clients can be reasonably served by other independent living housing programs, which may have fewer, comparable, or greater alternative housing options available than does Lighthouse. Third, why did nearly two out of five clients drop-out of the program early (i.e., before accomplishing the individualized goals developed with their independent living program case manager)? The addition or extension of aftercare services may provide greater opportunity to serve those who drop-out of the program, only to return at a later date for assistance, as many do. Finally, is scattered-site housing the most effective housing model for year-old youth preparing for independence, given the current 10-month average length of stay in the program? There is limited research that looks at the outcomes for youth exiting foster care from different living arrangements (Mech, 1994), but additional research is needed in this area Conclusion The Lighthouse Independent Living Program, which started out in 1981 as a pilot project that many did not expect to succeed, is now an established part of the Hamilton County Children's Services system of care for youth aging out of foster care. The county supports the ILP as it has seen many youth do well in the program and knows that its youth have to leave care often before they are developmentally ready. The descriptive findings presented in this paper are offered to help inform and encourage the development of housing programs for emancipating foster youth in other counties in Ohio, and possibly other locales in other states, by including a fairly detailed description of the structure, treatment, and outcomes domains of the program over the past six years of operation. Acknowledgements This work was supported by a Research Enhancement Grant by the College of Social Work at The Ohio State University. The authors wish to acknowledge Joyce Cooper, Merry Paul, Christi Watson, and Jean Sepate for data collection assistance and Amy Dworsky, Bob Mecum, Thad Parker, and Bonita Campbell for insights in interpreting data findings. Finally, we wish to acknowledge the staff and clients of the Lighthouse Independent Living Program whose daily work and effort have made the collection and reporting of these data possible. References Administration for Children and Families (2008). 45 CFR Part 1356 Chafee National Youth in Transition Database; Final Rule; 2/26/2008. Federal Register. Administration for Children and Families (ACF), U.S. Department of Health and Human Services (2008). Evaluation of the Life Skills Training Program: Los Angeles County. July Washington, D.C. Downloaded on 9/30/08 at: programs/opre/abuse_neglect/chafee/reports/eval_lst/eval_lst.p Arnett, J. J. (2000). Emerging adulthood. American Psychologist, 55(5), Barth, R. P. (1990). On their own: The experiences of youth after foster care. Child and Adolescent Social Work, 7(5), Blake, K., Cangelosi, S., Johnson-Brooks, S., & Belcher, H. M. E. (2007). Reliability of the GAF and CGAS with children exposed to trauma. Child Abuse & Neglect, 31, Brickman, A. S., Dey, S., & Cuthbert, P. (1991). A supervised independent-living orientation program for adolescents. Child Welfare, 70(1), Clark, H. B., & Davis, M. (Eds.). (2000). Transition to adulthood: A resource for assisting young people with emotional or behavioral difficulties. Baltimore: Brookes Publishing Company. Cook, R. (1988). Trends and needs in programming for independent living programs. Child Welfare, 67(6), Cook, R., Fleishman, E., & Grimes, V. (1991). A national evaluation of Title IV-E Foster Care Independent Living Programs for Youth: Phase 2. Final Report, Vol. 1 Westat, Rockville, MD. Courtney, M. E. (1993). Standardized outcome evaluation of child welfare services outof-home care: Problems and possibilities. Children and Youth Services Review, 15, Courtney, M. E., Dworsky, A., Cusick, G. R., Havlicek, J., Perez, A. & Keller, T. (2007). Midwest evaluation of the adult functioning of former foster youth: Outcomes at age 21. Chapin Hall Center for Children, University of Chicago. Courtney, M. E., & Hughes-Huering, D. (2005). The transition to adulthood for youth 'aging out' of the foster care system. In W. Osgood, C. Flanagan, E. M. Foster, & G. Ruth (Eds.), On your own without a net: The transition to adulthood for vulnerable populations (pp ). Chicago, IL.: University of Chicago Press. Courtney, M. E., Piliavin, I., Grogan-Kaylor, A., & Nesmith, A. (2001). Foster youth in transitions to adulthood: A longitudinal view of youth leaving care. Child Welfare, 80, Current Population Survey (CPS) (2005). As cited by: Rumbaut, R.G. & Komaie, G. (2007). Young adults in the United States: A mid-decade profile. Network on Transitions to Adulthood Research Network, University of Pennsylvania. Davis, M., & Vander Stoep, A. (1997). The transition to adulthood for youth who have serious emotional disturbance: Developmental transition and young adult outcomes. The Journal of Mental Health Administration, 24(4), DeWoody, M., Ceja, K., & Sylvestrer, M. (1993). Independent living services for youth in out-of-home care. Washington, D.C.: Child welfare League of America. DSM-IV-TR (2000). Diagnostic and statistical manual of mental disorders. Arlington, VA: American Psychiatric Publishing, Inc. Dworsky, A. (2005). The economic self-sufficiency of Wisconsin's former foster youth. Children and Youth Services Review, 27, Endicott, J., Spitzer, R. L., Fleiss, J. L., & Cohen, J. (1976). The global assessment scale. A procedure for measuring overall severity of psychiatric disturbance. Archives of General Psychiatry, 33(6), Festinger, T. (1983). No one ever asked us: A postcript to foster care. New York: Columbia University Press. Georgiades, S. D. (2005). Emancipated young adults' perspectives on independent living programs. Families in Society, 86(4), House Resolution (H.R.) 6893 (2008). Fostering Connections to Success and Increasing Adoptions Act. Accessed on 10/10/08 at: Kroner, M. J. (1988). Living arrangement options for young people preparing for independent living. Child Welfare, 67(6), Kroner, M. J. (1999). Housing options for independent living programs. Annapolis Junction, MD: CWLA Press. Kroner, M. J. (2001). Moving in: Ten successful independent and transitional living programs. Eugene, OR: Northwest Media. Lindsey, E. W., & Ahmed, F. U. (1999). The North Carolina independent living program: A comparison of outcomes for participants and nonparticipants. Children and Youth Services Review, 21(5), Mallon, G. P. (1998). After care, then where? Outcomes of an independent living program. Child Welfare, 77(1), Mayne, K. G. (1988). Pathways to adulthood: Creating successful programs to prepare teens for independence. National Resource Center for Youth Services, University of Oklahoma. McMillen, C. (1999). Better for it: How people benefit from adversity. Social Work, 44 (5), Mech, E. V. (1988). Editor. Independent-living services for at-risk adolescents. Child Welfare League of America: Washington. Mech, E. V. (1994). Foster youths in transition: Research perspectives on preparation for independent living. Child Welfare, 73(5), Mech, E. V. (2003). Uncertain futures: Foster youth in transition to adulthood. Washington: Child Welfare League of America. Pecora, P., Kessler, R., Williams, J., O'Brien, K., Downs, C., English, D., et al. (2005). Improving family foster care: Findings from the Northeast Foster Care Alumni Study Seattle, WA: Casey Family Programs. Pecora, P. J., Williams, J., Kessler, R. C., Hiripi, E., O'Brien, K., Emerson, J., et al. (2006). Assessing the educational achievements of adults who were formerly placed in family foster care. Child and Family Social Work, 11, Public Law (1999). Title I, Subtitle A, 101(b), 113 Stat Robin Hood Foundation (1996). Kids having kids: A special report on the costs of adolescent childbearing. In R. A. Maynard (Ed.), Robin Hood Foundation. NY: New York. Roman, N. P., & Wolfe, P. B. (1997). The relationship between foster care and homelessness. Public Welfare, 55(1), 4 9. Scannapieco, M., Schagrin, J. L., & Scannapieco, T. (1995). Independent living programs: Do they make a difference? Child and Adolescent Social Work Journal, 12, Schoeni, R., & Ross, K. (2005). Material assistance received from families during the transition to adulthood. In R. A., F. F., & R. (Eds.), On the frontier of adulthood: Theory, Research, and Public Policy (pp ). Chicago: University of Chicago Press. Waldinger, G., & Furman, W. M. (1994). Two models of preparing foster youths for emancipation. Special issue: Preparing foster youth for adulthood. Children and Youth Services Review, 16(3 4), White, R., & Rog, D. (2004). Introduction. Special issue on housing and homelessness. Child Welfare, 83(5).

10 Children and Youth Services Review 33 (2011) Contents lists available at ScienceDirect Children and Youth Services Review journal homepage: Living arrangements and level of care among clients discharged from a scattered-site housing-based independent living program Mark J. Kroner a,, Alvin S. Mares b a Lighthouse Youth Services, Inc., Lighthouse Training Institute, United States b The Ohio State University, College of Social Work, United States article info abstract Article history: Received 18 May 2010 Received in revised form 25 October 2010 Accepted 2 November 2010 Available online 9 November 2010 Keywords: Living arrangements Level of care Housing Independent living There is little recent research in the field of adolescent independent living that looks at the specific living arrangements of youth, who at the point of discharge from the child welfare system, had experienced living independently and had access to a choice of housing options. Administrative data and reviews of client records were used in this study to look at the choice of housing option and the change in level of care of 367 young adults who emancipated from the Lighthouse Youth Services Independent Living Program in Cincinnati, Ohio, during the five-year period Given a range of housing options at the time of discharge, over half (55%) chose an independent living arrangement, including 41% who lived in their own place, either alone (28%) or with a roommate (13%). Only 21% decided to live in someone else's home, including just 7% with one or both birth parents, 10% with some other relative, and 4% with a non-relative. The remaining 24% of youth were discharged from independent living to a more restrictive living arrangement (a. k.a., higher level of care, including a residential treatment program, group home, foster care, or supervised independent living program) (11%) or whose whereabouts were unknown (13%). The outcomes of this study suggest that, when presented with a choice of housing options, most of this county's emancipating foster youth would prefer to live on their own, rather than to return to live with their families of origin. The study also suggests that many youth who participate in a scattered-site housing-based independent living program can succeed in leaving care with affordable housing in place and avoid immediate homelessness Elsevier Ltd. All rights reserved. 1. Introduction In spite of the ongoing efforts of child welfare systems to pursue family reunification, adoption or other family/relative connections, there has always been a significant number of youth aging out of foster care annually who are not connected to stable family supports (Freundlich & Avery, 2005). The number has been ranging from 20,000 25,000 since the original Federal Independent Living legislation was passed over 20 years ago (U.S. General Accounting Office, 1999). While the total number of youth in foster care has decreased over the past decade, the number of youth aging out or emancipating from care each year has increased (Mares, 2010), and equaled 29,471 during fiscal year 2009 (U.S. Department of Health and Human Services, 2010). In 1986, Title IV-E of the Social Security Act was amended to establish a new federal initiative, named the federal Independent Living Initiative (ILI) (42 U.S.C. Sec. 677), to help foster youth ages 16 or older to live independently after reaching adulthood by enabling states to develop life skills, academic achievement, and vocational Corresponding author. Lighthouse Training Institute, 401 East McMillan St., Cincinnati, OH 45206, United States. Tel.: ; fax: address: mkroner@lys.org (M.J. Kroner). training programs to avoid homelessness, dependence on public assistance, and institutionalization after emancipating from care. Those new funds were not, however, allowed to be used for room and board (Hardin, 1987; Mech, 1988). An evaluation of ILI-funded programs conducted by Westat (1991) involving 810 youth in eight states found that skills training in budgeting, credit, consumer skills, education, and employment lead to positive outcomes in those areas. Moreover, a study conducted by Harding and Luft (1993) found that youth who participated in their state's Preparation for Adult Living program demonstrated overall greater housing stability, moving significantly fewer times than non-participants. However, some questioned the effectiveness of life skills training alone, and began advocating for the development of housing-based independent living programs to provide youth with an opportunity to apply life skills learned in a classroom setting in a supervised independent living setting, such as an apartment of their own, prior to emancipating from care (Brickman, Dey, & Cuthbert, 1991; DeWoody, Ceja, & Sylvestrer, 1993; Kroner, 1988). In 1999, Congress responded to independent living research findings and child welfare advocates by amending Title IV-E of the Social Security Act and redesigning the federal ILI by enacting the Foster Care Independence Act of 1999 (FCIA; H.R. 3443/P.L ) and the John H. Chafee Foster Care Independence Program (CFCIP) /$ see front matter 2010 Elsevier Ltd. All rights reserved. doi: /j.childyouth

11 406 M.J. Kroner, A.S. Mares / Children and Youth Services Review 33 (2011) This Act and Program provide states with more funding and greater flexibility in carrying out programs designed to help children make the transition from foster care to self-sufficiency, including for the first time authorization for states to use up to 30% of their allotted federal independent living funds to provide housing assistance for current and emancipated foster youth (Allen & Bissell, 2004). In 2009 the federal government provided $140 million in CFCIP grant funds to states, In 2007, all 52 states received CFCIP grants, which averaged $3.9 million per state, and ranged from a minimum of $500,000 awarded to less populated and smaller states (e.g., Alaska, Delaware and six other states) to $12 million for New York and $21 million for California (NRCYD, 2010). More recently, the Fostering Connections to Success and Increasing Adoptions Act of 2008 (H.R. 6893/P.L ) was enacted to improve outcomes for children in foster care, in part by authorizing states to claim IV-E reimbursement for foster care maintenance payments made on behalf of children up to 21 years of age who are in a supervised independent setting (Geen, 2009). 2. Review of independent living housing studies literature A review of the independent living housing studies literature was conducted by the second author. The universe of studies examined included those cited in five independent living literature review articles published since 2000 following the enactment of the federal Independent Living Initiative in 1986 and the Foster Care Indepedence Act in 1999 including: Loman and Siegel (2000), Collins (2001), Montgomery, Donkoha, and Underhilla (2006), Naccarato and DeLorenzo (2008), and Barth, Greeson, Zlotnik, and Chintapalli (2009). These five literature review articles cited a total of 13 studies in which housing-related outcomes were reported. A summary of those studies are presented in Table 1. In five of those 13 studies, outcomes from specific independent living programs (ILP) were reported. Those studies included data from two ILPs operated by non-profit agencies: The Green Chimneys ILP in New York City from (Mallon, 1998), and Lighthouse Youth Services ILP in Cincinnati from (Kroner & Mares, 2009). One state-wide administered ILP in North Carolina examined clients served during the period (Lindsey & Ahmed, 1999) was studied. Two state-monitored, county-administered ILPs serving clients in Baltimore County, Maryland, during the period (Scannapieco, Schagrin, & Scannapieco, 1995), and District 11 (Miami-Dade and Monroe Counties) in Florida during the period (Georgiades, 2005) were also examined. In contrast to the three public agency-administered programs which compared housing and other outcomes between ILP clients and non-ilp comparison group youth, both non-profit agency-based studies examined ILP client outcomes alone (Table 1). The remaining eight studies examined emancipated foster youth who may or may not have received independent living services, and were thus classified as ILP unspecified youth studies. Two of those studies have been the most influential within the field of independent Table 1 Summary of independent living arrangement rates by age reported in previous studies of emancipated foster youth. Authors Time period Geographic area Sample size Female White Black Independent living arrangement rates by mean age reported Comments ILP youth Georgiades (2005) Miami 49 78% 10% 60% 82% Includes own/rent independent housing Kroner and Mares (2009) Cincinatti % 30% 64% 33% Includes living in own place, either alone or with friends Lindsey and Ahmed (1999) NC 44 63% 38% 60% 68% Includes living alone, with partner, or with friends Mallon (1998) NYC 46 0% 4% 67% 68% 77% Includes own apartment (15 21%), shared apartment (46 51%) and furnished room (5 7%) Scannapieco et al. (1995) Baltimore 44 53% 68% NR 36% Includes living on own Non-ILP youth Georgiades (2005) Miami 18 78% 10% 60% 55% Includes own/rent independent housing Lindsey and Ahmed (1999) NC 32 48% 43% 57% 41% Includes living alone, with partner, or with friends Scannapieco et al. (1995) Baltimore 46 53% 68% NR 4% Includes living on own ILP unspecified youth Cook et al. (1991) AZ, CA, IL, MO, PA, NY and TN % 61% NR 49% Includes self (13%), self and child (8%) and sign other and child (28%) Courtney, Dworsky, Ruth, Keller, IA, IL and WI 321 a 51% 45% 41% 29% Includes own place Havlicek, and Bost (2005) Courtney et al. (2007) IA, IL and WI % 33% 56% 44% Includes own place Daining and DePanfilis (2007) Baltimore % 47% 61% 37% Includes living on their own ; plus 17% living with partner Fowler and Toro (2006) Detroit % 24% 75% 16% 42% Includes own apartment, house or dorm Lorentzen et al. (2008) 2008 CA % NR NR 13% Includes own apartment McCoy et al. (2008) MO 404 NR NR NR 24% 47% Undefined more independent ; 18 y/o=mean of 21% left care and 26% in-care groups McMillen and Tucker (1999) MO % 77% 22% 22% Includes college dorm and military Reilly (2003) Southern NV % 46% 30% 40% Includes living with partner (29%) Total N 3,712 b 57% 40% 54% 952 1, , Total N for age group Mean % 30% 57% 47% 37% Weighted mean % for age group NR=Not reported. a Total N=603, less those still in care (N=282), equals N=321 emancipated youth. b Excludes 19 year-old youth from Courtney et al. study to avoid double-counting.

12 M.J. Kroner, A.S. Mares / Children and Youth Services Review 33 (2011) living; namely, the national evaluation conducted by Westat conducted five years after the enactment of the federal Independent Living Initiative (Cook, Fleishman, & Grimes, 1991), and the ongoing Midwest Outcomes Study begun a decade later and four years after the enactment of the Foster Care Independence Act (Courtney & Dworsky, 2005; Courtney et al., 2007). These two studies are the most commonly cited independent living outcome studies conducted within the past three decades, and provide the most compelling documentation of the many challenges facing emancipated foster youth and need for independent living services in general, including housing assistance (Table 1). The six remaining studies reporting housing outcomes among emancipated foster youth include four state-based studies and two city-based studies. Two studies were conducted in Missouri the first in the early 1990's (McMillen & Tucker, 1999) and the second a decade later in the early 2000's (McCoy, McMillen, & Spitznagel, 2008). A third study involved a sample of emancipated youth in Southern Nevada conducted in the late 1990's (Reilly (2003). The fourth and largest statewide study is on-going and examines the Transitional Housing Program (THP-Plus), the largest housing-based supervised independent living program currently in operation in the United States (Lorentzen, Lemley, Kimberlin, & Byrnes, 2008). Finally, two city-specific studies were conducted in the early 2000's, including one in Baltimore, Maryland, and the other in Detroit, Michigan (Daining & DePanfilis, 2007; Fowler & Toro, 2006) (Table 1). Aggregated independent living arrangement rates data summarized in Table 1 provide potentially useful benchmark data for the study reported in this paper and future studies of housing-based independent living programs. Among the 3,700 emancipated youth studied across the 13 studies published over the past three decades, independent living arrangement rate estimates are as follows: 22%, 30%, 57%, 47%, and 37% for 18, 19, 20, 21 and 22 year-old youth, respectively (Table 1). Note that these estimates may be biased due to 1) potential differences between independent living program specific populations versus non-specified or general population-based samples, 2) unknown variations in housing outcomes across time periods and geographic regions examined, 3) differences in definitions of independent living status used (i.e., Comments column of Table 1) across these 13 studies, and 4) possible publication bias due to including only published housing outcomes studies. Regarding possible publication bias, there are several innovative housing-based interventions assisting emancipating foster youth currently in operation, including: First Place for Youth in San Francisco ( Urban Peak in Denver (Leeuvwen, 2004; html), HUD's Family Unification Program voucher initiative ( and the Transitional Housing Placement Plus (THP-Plus) Program in California ( about.html). Hopefully these programs will contribute to the scholarly housing outcomes literature in the future. 3. Level of care and living arrangements among housing-based independent living programs The term level of care is typically used in the fields of medicine (Phibbs, Bronstein, Buxton, & Phibbs, 1996), long-term care (Foley & Schneider, 1980), and mental health (Srebnik, Uehara, & Smukler, 1998). Within the field of child welfare, it is most often applied to youth requiring residential treatment for serious emotional or behavioral problems (Bates, English, & Kouidou-Giles, 1997). It has also been used to describe placement patterns for foster youth (Usher, Randolph, & Gogan, 1999). Level of care definitions vary both across and within fields. For example, within the field of mental health, Geller (1993) identified seven different level of care approaches for defining residential levels of care developed during the 1980s. One of those approaches was developed by the American Psychiatric Association in 1982 and identified seven levels of care, ranging from most to least restrictive, including: nursing facility, group home, personal care home, foster home, natural family placement, satellite housing, and independent living. While suggesting that level of restrictiveness is defined along two dimensions (i.e., a comparison of the similarity to the children's home environments and the amount of supervision and monitoring children receive) and noting that placing agencies are legally obligated to place youth into the least restrictive setting upon entering foster care and to continuously assess the appropriateness of the placement setting based on level of restrictiveness, Usher et al. (1999) failed to offer a similar residential level of care continuum. Instead, they identified eleven placement settings, including: adoptive home, children's residential center, detention facility, agency foster home, group home, hospital, unrelated person (unlicensed), other care provider, private foster home, relative, and Youth Services foster home. Similarly, the federal government has yet to provide a level of care continuum for foster youth or independent living, instead identifying in its annual adoption and foster care statistical reporting system (a.k. a., AFCARS) a list of seven living arrangements for youth currently in care, which include: pre-adoptive home, foster family home (relative), foster family home (non-relative), group home, institution, supervised independent living, runaway, and trial home visit (U.S. Department of Health and Human Services, 2010). Neither has the concept of level of care been applied to housingbased independent living programs. The study reported here is one of only five housing-based ILP studies published since Those four studies examined the living arrangements at or following discharge of 183 clients served in four different cities/geographic regions (Table 2). Each used a different definition of independent housing. None reported housing arrangements at the time of entering independent living (Table 1, last column). Neither did any attempt to classify specific living arrangements into levels of care. Thus, this is to our knowledge the first study to examine level of care at both admission and discharge to an housing-based independent living program among a large sample (n=367) of clients served in what may be the longest-standing, most well-developed, and most extensively described program in the country (Kroner, 1988, 1999, 2001; Kroner & Mares, 2009). Thus, this study seeks to contribute to the scholarly literature on housing-focused interventions for older adolescent and young adult foster youth by 1) describing the specific living arrangements and levels of care among clients served over a five-year period established by one well-established housing-based independent living program, 2) describing changes in the level of care from admission to discharge, and 3) comparing client characteristics across levels of care at discharge. In doing so, the major question of Where do emancipating foster youth live, following their custodial placement in a scatteredsite, housing-based independent living program? is addressed and the concept of level of care is applied to housing-based independent living programs. 4. Description of lighthouse youth services and the lighthouse independent living program Lighthouse Youth Services is a private non-profit organization established in 1969 in Hamilton County, Ohio. The agency has seven divisions which served over 4000 children, youth, and families in 2008, employed a staff of over 370, and maintained an annual operating budget of $18 million. The seven service divisions include: Homeless and Runaway Youth, Early Childhood Service, Juvenile Corrections, Foster Care and Independent Living, Home Based Service, Community Based Residential Treatment, and Education. The Lighthouse Independent Living Program (ILP) began in 1981 in order to help youth leaving the child welfare and juvenile justice systems, who were unable to return home, to make the transition to

13 408 M.J. Kroner, A.S. Mares / Children and Youth Services Review 33 (2011) Table 2 Comparison of living arrangement level of care at discharge or post-discharge among housing-based ILP clients. Kroner & Mares (2009) Mallon (1998) Scannapieco et al. (1995) Lindsey & Ahmed (1999) Georgiades (2005) Lighthouse Green Chimneys Baltimore County North Carolina Florida District 11 (n=367 a ) (n=46) (n=44) (n=44) (n=49) Age 19 Age 20 Age 19 Age 19 Age 20 % n % n % n % n % n Independent Independent living by self School dormitory 1 3 Independent with friend Subsidized/other supported housing Military Enlistment Own Apartment 15 7 Shared Apartment Living on own Living independently Own/rent independent housing With others NR NR 16 8 Home of family friend 4 13 Home of relative Biological father 1 3 Biological mother 6 22 Home of both biological parents 0 1 With Family Member With Friends Relatives/Foster Parents /Friends 16 8 Supervised/system NR NR 2 1 Jail Juvenile detention center 1 4 Group setting emergency shelter (in. YCC) 1 2 Group home 3 11 Supervised ind/trans. living (inc. ILP,TLP) 3 12 Unknown NR NR NR 0 0 Unknown NR not reported. Note: living arrangement and level of care data at intake were only reported for Lighthouse ILP, and thus are not presented in this ILP comparison table. a N=367 for those clients with complete living arrangement data at discharge; N=455 for clients with less detailed independent living rate data available at discharge. self-sufficiency. Youth often come from the agency's foster or group homes or a correctional program Lighthouse operates in rural Ohio. The ILP has served over 2400 youth since then and has served as a model program for many new programs around the country. Most youth referred to the program presented with a range of risk factors. The average Global Assessment of Functioning Scale (GAF) was 60.9 for all referred youth indicating moderate difficulty in social, occupational or school functioning (e.g., few friends, conflicts with peers or co-workers). Youth scored an average of 76.7 on a 130 point selfsufficiency skills scale created by the program. Nearly half (47%) had chronic mental health issues, 41% had a history of chronic history of truancy and school problems, 41% had a history of delinquency, 42% had little or no work experience and 27% had been violent toward people in the last several years (Kroner & Mares, 2009). Lighthouse decided years ago that it would accept high risk youth into the ILP, knowing that many had never shown much success in previous placements. The agency felt that high risk youth needed at least a chance to learn to live independently with hope that new challenges directly related to the youth's survival would serve as a source of motivation for learning. It is important to note that most foster youth in Ohio are discharged prior to their 19th birthdays. To meet the needs of these clients the Independent Living Program (ILP) program provides the following services: basic, direct treatment, independent living life skills training, and referral. Basic services include shelter, food, clothing, transportation, and education. All youth are assisted with finding an appropriate living arrangement. The ILP rents apartments from private landlords in the county in neighborhoods that are affordable and close to the client's school, job, and social supports. The program also operates two shared-homes, one for males and one for females, which have 3 4 beds and a live-in resident manager. The program pays the security deposit and furnishes the apartments with necessary supplies and a telephone. If the client does well and has a job at termination, s/he can keep the apartment and all of the furnishings and assume responsibility for the lease. The ILP also provides a weekly allowance of $55 for basic support, $10 of which is saved in an aftercare fund. This money is used for food, personal items, cleaning supplies etc. The agency also covers utility, phone, and rental payments until the last few months in the program, when bills are taken over by the client, if possible. The ILP assists clients with work clothing, minor school fees, and miscellaneous expenses. Most clients are expected to work a part-time job and purchase any items beyond the basic necessities. Direct treatment includes case management and clinical treatment provided by ILP and Lighthouse staff. Each youth is assigned to a licensed social worker with a caseload of 8 14 clients. Other program staff members also assist with client problems as they arise. Clients are usually seen or contacted several times during each week including regular phone contact. Vulnerable or new clients are asked to call in daily. The program staff maintain a 24 h on-call system. Specific needs are addressed depending on the client's situation. For example, 18% of all clients received some type of group counseling, 12% had tutors, 11% were monitored for medication compliance, The ILP staff provide 24- hour crisis management, which can involve hospital runs, resolving client/tenant problems, apartment maintenance issues, confronting client friends/family who are causing problems at the apartment, and so forth. This activity is time consuming but is a critical part of the process of learning responsible behavior. Independent living life skills training is provided, in collaboration with the local county child welfare system to provide 13, four hour

14 M.J. Kroner, A.S. Mares / Children and Youth Services Review 33 (2011) classes on self-sufficiency, hopefully completed prior to the youth's placement in an IL arrangement. Topics include an assessment of current level of functioning, money management, time-management/ planning ahead, use of community resources, apartment management, nutrition/food preparation use of public transportation, social skills, employment skills/ finding and holding a job, problem solving and decision making, self-care, and building a support network. Sixtyseven percent (67%) of all youth received some form of formal skills training while in the program. A portion of referred youth have already completed the county's life skills training program and focus more on education and employment activities than life skills. Some youth (16%) received parenting classes, 54% were assisted with employment skills development, and 24% received diagnostic assessment services (Kroner & Mares, 2009). Referral for relevant educational, vocational, therapeutic, medical, dental, and other needed resources and services provided by other service agencies is arranged by ILP staff as-needed. Everyone works toward the goal of the youth gaining maximum self-sufficiency given the time available and the developmental capabilities of the youth. Fifty-nine percent (59%) of all youth were connected with outside community resources such as mental health (20%), GED (19%), Child care (5%), and legal help (6%) (Kroner & Mares, 2009). 5. Methods 5.1. Sample The subjects in this study include all clients who were both admitted into and discharged from the Lighthouse Independent Living Program during the fiscal year period. Of the total 455 discharged ILP clients, discharge living arrangement data were available for 367 clients (81%). Missing data was attributed to the unfinished work of a small group of staff during the first year of the data period. Except for clients for whom discharge living arrangement data were available being more likely to have mental health problems than clients with missing data (64% vs. 48%, χ 2 =7.9, pb.01), no significant differences were found on demographic characteristics, clinical risk group, or length of stay measures examined (Table 3). Thus, subsequent analyses of data for the data available clients (N=367) appear to be justified, given the similar demographic, clinical risk, and length of stay profiles of these two client groups Data collection Experienced Lighthouse staff members compiled existing clientlevel administrative and clinical records from the agency's management information system and paper files, under the direction of the first author. De-identified data were then analyzed by the second author in accordance with procedures approved by the Institutional Review Board at the second author's academic institution Measures Client characteristics Demographic characteristics included age at admission, gender, and race/ethnicity (defined dichotomously, based on minority status). Clinical risk group classifications were made based upon an exploratory principal components factor analysis of 22 dichotomous clinical risk items conducted by the second author for this study. The 22 dichotomous clinical risk factors or barriers/challenges facing foster youth preparing for emancipation were developed by the first author, who served as director of the IL program for over 21 years. Two Lighthouse staff then checked all applicable risk factors for each subject, based upon a review of various records, including: (a) intake/ admission records (i.e., the Referral Sheet, Social History Form, and Intake Screening Form; (b) treatment records (i.e., Diagnostic Assessment Form, Incident Report, and progress notes; and, (c) discharge records (i.e., Termination Summary Form). A summary of the operational definitions used during this chart abstraction process has been reported previously (Kroner & Mares, 2009). The abstraction process required over 100 total person hours of effort, spread over a five-month period (January through May 2008), and divided between two abstractors. The first abstractor served as Administrative Assistant for the ILP for over 15 years; the second was a licensed clinical social worker (LSW) who had worked at Lighthouse for several years. Cases were split between these two abstractors, with one taking those clients admitted during the period , whose records were archived in paper form, and the taking clients admitted more recently (i.e., during the period ), whose records were accessible in electronic form. The results of the exploratory factor analysis generated the four non-mutually exclusive clinical risk groups: 1) mental health problems, 2) delinquency problems, 3) cognitive impairments, and 4) teenage parents. Table 3 Comparison of client characteristics among those with and without discharge living arrangement data (N=455). With discharge living arrangement data Without discharge living arrangement data Independent samples t-test or Chi-sqr test (N=367) (N=88) %/Mean N/SD %/Mean N/SD t or χ 2 df p Demographic characteristics Age ns Race White 31% % ns Black 62% % ns Other 7% 24 2% ns Minority 69% % ns Gender (female) 55% % ns Clinical risk groups (non-mutually exclusive) Mental health problem 64% % ** Delinquency problem 9% 33 6% ns Cognitive impairment 10% 37 8% ns Teen parent 18% 67 19% ns Program characteristics Length of stay (days) ns **pb.01, ns pn=.05.

15 410 M.J. Kroner, A.S. Mares / Children and Youth Services Review 33 (2011) Three additional characteristics included length of stay in the program (in days); and two outcome measures: completing high school (including GED), and working (either part or full-time) at the time of discharge from the program (both dichotomous measures) Living arrangements and levels of care Agency and program staff documented clients' living arrangements both at entry into the program (i.e., where clients were living at the time of referral to ILP) and at the time of discharge (i.e., immediately after leaving the program). A checklist of 22 specific living arrangements, (from which and to which a youth could be referred) was used to document admission and discharge living arrangements, including unknown for those clients whose living arrangement was unable to be determined by agency or program staff Living arrangements A brief description of the 22 living arrangements follows: 1. Independent, living by self: this is when a youth lives in his/her own apartment, usually with the lease in his/her name. (Some of these youth continue to receive financial support and case-management through the Transitional Youth Program, a collaborative program involving several agencies, for youth with serious mental health issues). 2. Independent, living with friend: this is when a youth lives with a friend in the friend's current residence. (ex. when a female moves in with a boyfriend). 3. Subsidized/other supported housing: the youth (often a teen mom) lives in housing that is based on income such as Section 8 or other types of subsidized housing. 4. School dormitory: the youth lives in a university or collegeoperated residence hall. 5. Military/Job Corps: although it has occurred, very few youth leave to these entities. 6. Home of relative (excluding biological parent(s): a youth lives with someone related to the youth. 7. Biological mother: the youth lives in biological mother's residence. 8. Biological father: the youth lives in biological father's residence. 9. Home of both biological parents: the youth lives with both bioparents. 10. Home of family friend: a youth lives with adults who know the youth through his/her family agree to allow the youth to move in with them. (Some of these youth return to the ILP when things do not work out in these settings after discharge). This can also include Host homes, non-licensed homes where the adult(s) who owns the home agrees to provide a room in their house for an agreed upon time. At times a youth chooses to stay at this site at discharge. 11. Jail: this means that a youth is incarcerated in a state correctional facility due to either a new crime or re-incarcerated due to parole rules violations. (Some youth who commit offenses while in the ILP return to the ILP after incarceration). 12. Juvenile detention center: the youth is in a juvenile detention center. Often these youth are sent to group or foster homes and at times can be re-referred later to an ILP. 13. Inpatient psychiatric hospital/secure residential facility: these youth presented mental health issues significant enough to require either emergency or long-term hospitalization. These Youth who enter these facilities while in the ILP can be re-referred/returned to the ILP again if it appears it is the most appropriate setting. 14. Drug/alcohol rehabilitation center: these youth are identified as having a drug or alcohol problem requiring treatment. Youth who leave an ILP setting for treatment can be re-referred to the ILP if it appears it is the most appropriate setting. 15. Residential treatment center (RTC): this is often an intensive, inpatient program for youth with serious behavioral issues who were not able to maintain stability in an independent living setting. Some youth sent to RTCs return to the ILP after treatment. 16. Emergency youth shelter: this is a temporary shelter for youth, run by Lighthouse, where youth can stay for up to three weeks. At times, youth who are violating program rules are placed here temporarily and then given another chance in another ILP site with a strict behavioral contract in place. 17. Group home: this is most often a group home of youth with 24 h coverage. Lighthouse operates one of these for females and one for males. Some youth are returned to the group home they came from if they fail to make an attempt at responsible behavior. Adult group settings can be used for some youth who leave the child welfare system but still need ongoing supervision due to a disability or mental health issue. 18. Therapeutic foster care: this is a licensed foster home with foster parents who have received extra training to address youth behavioral or mental health issues. 19. Foster care: this is a traditional foster home. At times, pregnant females are placed back in a foster home in order to have more supervision and monitoring during a pregnancy with complications. The mom can return to an ILP setting after the birth and the stabilization of the child. 20. Supervised ILP: this is another agency's ILP which is smaller than a group home but has 24-hour awake coverage. 21. Homeless: a youth leaves the program with no place to go. At times s/he can enter an adult homeless shelter. This is a 24 h shelter for homeless adults age 18 and over. Lighthouse has a strong working relationship with the local adult homeless shelter provider system through the Continuum of Care process. 21. Unknown: the youth leaves the program without notifying ILP or county staff. Many of these youth move in with friends and/or family of unknown stability and some return asking for help and housing within weeks Levels of care In the absence of either preexisting literature defining levels of care among housing-based independent living programs or existing administrative data quantifying the level of restrictiveness of each living arrangement, the first author classified the 22 living arrangements into four levels of care, ranging from lowest (most independent, least restrictive, most stable) to highest (least independent, most restrictive, least stable). This classification was based on face validity and nearly 20 years of clinical and administrative experience locally in Cincinnati, Ohio, and in presenting at national independent living conferences and consulting with independent living programs across the country. The independent level of care was defined as including living arrangement numbers 1 5. The living with others level of care included living arrangement numbers The supervised/system level of care included living arrangement numbers Finally, the unknown level of care included living arrangement number Data analysis Descriptive statistics were used to describe client living arrangements and levels of care. Independent samples t-tests and Chi-square procedures were used to compare client characteristics between those for whom discharge living arrangement data were available and those for whom such data were not available. Finally, analysis of variance (ANOVA) was used to compare client characteristics between clients who were discharged to various levels of care and those who had entered the program from a supervised/system level of care setting.

16 M.J. Kroner, A.S. Mares / Children and Youth Services Review 33 (2011) Table 4 Admission and discharge living arrangements among clients with discharge information (N=367). Living arrangements by level of care At admission At discharge % N % N Independent 1% 2 55% 202 Independent living by self 1% 2 28% 101 Independent with friend 0 13% 49 Subsidized/other supported housing 0 13% 47 School dormitory 0 1% 3 Military enlistment/job Corps 0 1% 2 With others 14% 50 21% 76 Home of relative (excl. biological parent) 7% 25 10% 37 Biological mother 2% 9 6% 22 Biological father 1% 3 1% 3 Home of both biological parents 0 0% 1 Home of family friend 4% 13 4% 13 Supervised/system 86% % 41 Jail 2% 8 3% 12 Juvenile detention center 2% 8 1% 4 Inpatient Psych Hosp./ Secure Resident Facility 1% 4 0 Drug/alcohol rehab center 2% 8 0 Residential treatment center 8% 29 0 Emergency youth shelter 2% 7 1% 2 Group home 18% 65 3% 11 Therapeutic foster care 5% 20 0 Foster care 43% Supervised ind/trans. living (incl. ILP,TLP) 2% 6 3% 12 Homeless 1% 3 0 Unknown 0% 1 13% Results 6.1. Living arrangements and changes in levels of care at admission and discharge Most clients entered the program from either a supervised/system setting (n=314; 85.6%) or from living with others (n=50; 13.6%). Thus, fewer than 1% (n=3) entered the program from independent or unknown levels of care settings. Over 70% of clients entered the program from foster care (43%), a group home (18%), or a residential treatment or drug/alcohol rehabilitation center (10%) (Table 4, admission column). While most clients were admitted from a supervised/system level of care setting, only one in ten clients (11.2%, n=41) were discharged to the same level of care. Threefourths of ILP clients were discharged to an independent level of care setting (55%, n=202) or to a living with others level of care setting (21%, n=75) (Table 5). The most common living arrangements postdischarge were living independently, either with self (28%, n=101) or with a friend (13%, n=49), living in subsidized/other supported housing (13%, n=47), and living with a relative, either with a biological parent (7%, n=26) or with another relative (10%, n=37) (Table 4, discharge column). Among the majority of ILP clients who were admitted from a supervised/system level of care (86%, n=314), nearly three-fourths were discharged to a lower level of care; namely, 54% (n=169) to an independent setting and 19% (n=61) to a living with others setting. Twelve percent (n=37) were discharged to the same supervised/ system level of care. The remaining 13% (n=48) were discharged to an unknown level of care/living arrangement setting. Nearly all of the clients (n=364/367, 99%) entered the program from either a living with others (13.6%) or supervised/system (85.6%) level of care. Those entering from living with others (i.e., from the home of a relative or friend of the family) were more likely to be discharged to highest, independent level of care than those entering the program from living in a supervised or system level of care (i.e., from an institution, group home, foster home, or from homelessness) (60% vs. 54%, respectively, χ2=8.7, pb.05) (data not shown). Those who entered from living with others were also less likely to be discharged to a supervised/system level of care (8% vs. 12%) or to an unknown level of care (2% vs. 15%) than clients who entered the program from a supervised/system level of care Characteristics of clients discharged to various levels of care A comparison of client characteristics across the four levels of care at discharge found significant differences primarily between those discharged to an independent level of care setting versus some other level of care. Clients discharged to an independent setting tended to be older, non-caucasian, and female, than other clients. They were less likely to have mental health and delinquency problems, and were more likely to be teenage parents, than other clients. They also remained in the program for longer and were more likely to be working at the time of discharge than other clients (Table 6). 7. Discussion 7.1. Major findings In this study, just over half (55%) of the 19 year-old clients, on average, had attained a living arrangement classified at the independent level of care. This rate is between the 36% rate reported by Scannapieco et al. (1995) in Baltimore, Maryland, and the 68% rate reported by Lindsey & Ahmed (1999) in North Carolina, for emancipated foster youth of the same average age. The 55% rate is somewhat lower than the 685 rate reported by Mallon (1998) for youth in New York City, and 82% reported by Georgiades (2005) in Miami, Florida, for emancipated youth that were an average of 20 years of age. As noted earlier, though, it is difficult to interpret or draw conclusions based upon this comparison of rates due to differences in ways of defining independent living, time period differences, and geographic/regional differences. Harder still is to compare the rates of living with others after leaving independent living across these five programs, since only two of the other four programs reported any such information, and both of those programs were from emancipated youth who averaged 20 years of age. Neverthe-less, the 21% rate observed in this study was between the 16% reported by Georgiades (2005) and 26% reported by Mallon (1998). No meaningful comparisons can be drawn between this study and the Table 5 Changes in living arrangement level of care from admission to discharge (N=367). Level of care at discharge Independent With others Supervised/system Unknown Total Level of care at admission Independent 2 (0.5%) 0 (0%) 0 (0%) 0 (0%) 2 (0.5%) With others 30 (8.2%) 15 (4.1%) 4 (1.1%) 1 (0.3%) 50 (13.6%) Supervised/system 169 (46.0%) 61 (16.6%) 37 (10.1%) 47 (12.8%) 314 (85.6%) Unknown 1 (0.3%) 0 (0%) 0 (0%) 0 (0%) 1 (0.3%) Total 202 (55.0%) 75 (20.7%) 41 (11.2%) 48 (13.1%) 367 (100.0%)

17 412 M.J. Kroner, A.S. Mares / Children and Youth Services Review 33 (2011) Table 6 Comparison of client characteristics by level of care at discharge among those admitted from a supervised setting (N=314). A B C D ANOVA Tukey pairwise Independent With others Supervised/system Unknown comparisons (N=169) (N=61) (N=37) (N=47) %/Mean %/Mean %/Mean %/Mean df f p Demographic characteristics Age A NC,D Race White A bd Black ns Other ns Minority A ND Gender (female) ANB,C Clinical risk groups (non-mutually exclusive) Mental health problem AbC Delinquency problem A,BbC; AbD Cognitive impairment ns Teen parent ANB,C Program characteristics Length of stay (days) ANB D; BND Client outcomes at discharge Completed high school ns Working or completed voc training ANB D ns pn=.05. p b.05. pb.01. p b.001. other four housing-based ILP studies for either the supervised/system or unknown levels of care at discharge, nor for any of the four levels of care at the time of admission into the program. This underscores the first major finding of this study; namely, the paucity of comparative or benchmark data for levels of care and living arrangement outcomes among housing-based independent living programs. Focusing on the examination in this study of housing outcomes data from the Lighthouse Independent Living Program, it should be noted that out of 367 discharges in this study, only seven percent (25 youth) went directly to the home of biological mothers or fathers. Historically, youth who make this choice at times return to the program at a later date, asking for housing assistance. This does not mean that they youth had no contact with family, but given a choice in maintaining their own living arrangement and moving back home immediately at discharge, the majority chose to live separately. These findings also indicate that a housing-based IL program can also be successful in moving emancipating youth from more restrictive to less-restrictive and, in most cases, less expensive, living arrangements chosen by clients at the time of leaving the program. This study found that only 11% of youth were discharged to a higher level of care, which has implications for cost savings for county and state child and adult welfare systems. Another significant observation is that no youth were discharged to homelessness by the local child welfare system during this 5 year period. Perhaps some of the 48 youth who fell in the unknown category were homeless at discharge. But this was due to personal choice or rejection of the numerous options available that could have prevented their homelessness. The fact that only 16 of the 367 youth in this sample left the ILP to either an adult jail or juvenile detention facility challenges the assumption that youth allowed to live alone while still in the custody of the state will cause continuous major problems or come into harm. Similarly, with 64% of youth being discharged from the program with an identified mental health problem, it is noteworthy that none left the program to enter an inpatient psychiatric unit. Those who might have had the need for inpatient treatment were allowed to return to their previous living arrangement, often with more focused supports in place. Also worth mentioning is the decrease in expense that comes with each decreased level of care. Only 11% of the youth in this study left the program or system via an increased level of care. Per diem rates in Hamilton County (Cincinnati) are currently approximately $275 for residential treatment, $170 for group homes, $ for therapeutic foster care homes, $85 for shared homes, and $65 for scattered-site apartments. If a youth proves able to live in a less-supervised setting, the savings to communities can be considerable. If less-supervised living arrangements such as scattered-site apartments, shared-homes, and host homes could become reimbursable through Title IV-E, as is the case for foster and group homes currently, the savings to child welfare agencies could be significant. Obviously, this type of housing model is much more possible in the mid-size cities where demand for housing is not as intense and rents are typically below $550 $600 a month. Finally, the finding that clients with mental health problems were more likely to remain in this housing-based independent living program may be interpreted as an indicator of program effectiveness by retaining a hard-to-reach population. Another implication of this finding is that more intensive and expensive housing-based independent living services may not be wanted or needed among healthier emancipating foster youth without mental health problems who may do quite well given the natural supports of their environments. Those dropping-out of the program early, around 18 years of age on average, grow over the five or six years to follow to become what Courtney, Hook, and Lee (2010) have described as accelerated adults the 36% of emancipated foster youth at ages 23 and 24 who are the most successful sub-group of emancipated foster youth. Two-thirds (63%) of accelerated adults are female, Most (84%) are stably housed in their own place. Nearly all (98%) have completed high school, including over half (54%) who have attended some college and one-in-ten (12%) who have completed an Associate's Degree or higher. Three-fourths (75%) are currently employed. Slightly over half (52%) have children, among whom 89% have retained physical custody of their children. And only one-in-seven (14%) have been convicted of committing a crime since leaving foster care. While some (17%) of these accelerated adults report experiencing PTSD symptoms, these symptoms are not serious enough to seriously impair social, vocational or educational

18 M.J. Kroner, A.S. Mares / Children and Youth Services Review 33 (2011) functioning. Thus, more formal, intensive, and expensive housingbased independent living program services may be more helpful and needed among the other three sub-groups of emancipating foster youth identified by Courtney and colleagues; namely, struggling parents (25%), troubled and troubling young adults (18%) including those with more serious behavioral and emotional problems, and the more generalized sub-group of emerging adults (21%) (i.e., those emancipated youth most closely resembling the non-foster youth, general population of year-olds in today's society) (Arnett, 2006) Limitations First, the validity of client clinical risk factor measures and living arrangements and levels of care at intake and discharge is unknown. While three Master's level and licensed social worker professionals with years (if not decades) of experience delivering independent living services to emancipating foster youth at Lighthouse Youth Services were responsible for the assessment of client risks, documentation of client living arrangements, and development of the levels of care presented herein, the validity of these measures remains unknown. Next, the findings are subject to selection bias due to clients with mental health problems being more likely to remain in the program than clients without mental health problems who were more likely to drop-out of the program and to be missing data on living arrangements at discharge. However, this particular limitation may indicate that the findings reported here may be viewed as conservative estimates of program effects, in that had the less severe cases who may be more likely housed remained in the sample, then the percentage of failed cases might have been lower. Third, outcomes data are censored, that is no information was available on these youth after discharge. Given that living arrangements and levels of care can, and often do, change a lot in a fairly short period of time after being terminated from state custody, the same may be true after youth were discharged from this housing-based independent living program. Thus, the stability of living arrangements and levels of care findings for the clients reported herein during the months and years following discharge from the Lighthouse Independent Living Program are unknown. However, the program's ability to use aftercare funds to stabilize youth for years post-discharge increases the odds for housing stability. Finally, the generalizability of these findings to other housing-based independent living programs and to emancipating foster youth in other communities is limited. For example, median housing costs among rental properties in 2009 ranged from $552 in West Virginia to $1293 in Hawaii, with the cost in Ohio being $670, the 13th lowest housing costs state in the nation among renters (American Community Survey, 2009). Thus, independent housing rates among emancipating foster youth in Ohio may be higher than in other states where housing is less affordable. Another factor limiting the generalizability of these findings is the unique nature of the Lighthouse Independent Living Program. Clients in this program are connected to numerous other Lighthouse operated services such as aftercare, in-home therapy, a wide range of various types of housing, along with a high-level of collaboration and cooperation between the local child welfare system, juvenile justice system, community mental health system, and homeless services system, which all together help to stabilize youth in the ILP. Few independent living programs currently in operation are able to provide such a wide range of supportive services to emancipating foster youth Implications for policymakers, child welfare system administrators, and independent living program directors Policy makers The evolution of the federal Independent Living Initiative in 1986 to the Foster Care Independence Act in 1999 to the Fostering Connections to Success and Increasing Adoptions Act in 2008 has increased both federal funding and flexibility to states to encourage the development of housing-based interventions for emancipating foster youth. This trend of increased funding and flexibility for housing over the past 25 years should be continued, given the preliminary promising housing outcomes reported here, which are generally congruent with those previously reported in the scholarly literature. However, more rigorous and routine process and outcomes data collection requirements should be imposed upon states as a condition of receiving these funds so that additional comparison or benchmark data may be generated and made available to state and local governments and philanthropic sector funders interested in investing in evidence-supported housing-focused interventions. The Foster Care Independence Act (FCIA) requirements for formal evaluations of Chafee-funded independent living programs and services and the National Youth in Transition Database (NYTD) data collection requirements of states receiving FCIA funds are useful examples of reasonable outcomes data reporting strings attached to federal funding strings which should be strengthened and perhaps further tightened to promote the responsible use of limited federal resources. Given that the federal government is the primary funder of independent living services, and given the paucity of published housing-focused intervention outcomes data (Tables 1 and 2), policy makers are advised to make reasonable efforts to encourage the generation of potentially useful comparative benchmark data, and to target federal funding of evidence-supported housing-focused independent living interventions Child welfare system administrators and magistrates The descriptive levels of care housing outcome data reported in this study offer several implications for administrators of state and local child welfare systems. First, the finding that over half (55%) of the 19-year old emancipating foster youth were discharged from the program to an independent level of care (i.e., living on their own) (Tables 3 and 4) provides a useful benchmark, especially for administrators in systems in which state custody typically is terminated around a foster youth's 18th birthday. In such systems, administrators and magistrates are advised to consider promoting and supporting the development and use of short-term, housingbased independent living interventions such as the one described herein where foster youth typically are admitted into a scattered-site apartment of their own around their 18th birthday or graduation from high school and then provided with an average of 10 months of independent living case management and life skills training in their own apartment setting before being finally discharged or emancipating from the child welfare system (Kroner & Mares, 2009). While most youth would likely benefit, those with mental health problems, delinquency problems, and those who are single parents may be especially helped well-served by such a step-down or transitional independent living housing-based intervention. Using such an approach may make it easier for administrators and magistrates to balance the competing goals of protecting foster youth safety and well-being (i.e., minimizing risks) at earlier ages when they are more vulnerable and preparing older emancipating foster youth to live independently (i.e., promoting independent decision-making and monitored the inherent risks thereof). The 18th year of life or year following graduation from high school in 18-year old custody states (e.g., Ohio) or the 21st year of life in 21-year old custody states (e.g., California) may be ideal years for emancipating foster youth to gain from independent living housing-based interventions Directors and developers of housing-based independent living programs The profiles of emancipating foster youth clients by level of care at discharge (Table 6) offer implications for directors and developers of housing-based independent living programs. First, admitting clients

19 414 M.J. Kroner, A.S. Mares / Children and Youth Services Review 33 (2011) at 18 years of age or older and providing a full year of support, on average, is recommended to assist clients attain independent living by the time of leaving the program. Next, female clients and teen parents respond more favorably to this type of intervention than male clients and clients with mental health problems and delinquency problems who may need a more intensive or specialized level of support. Third, employment assistance and vocational training appears to be an important element of support provided to make independent living possible when clients exit the program. Finally, with programs who accept high-risk youth with little previous history of responsibility, one-in-four clients may be expected to fail or bomb-out of scattered-site housing by the end of their stay in the program, despite the best efforts of program staff. While disappointing, the findings of this study suggest that program directors and developers may expect half of clients to succeed, and a fourth of clients to fail, with the remaining one in five (or so) to fall somewhere between success and failure by leaving the program to live with others, including family or friends, at the time of leaving the program Conclusion The Fostering Connections to Success and Increasing Adoptions Act of 2008 proposes to expand federal Title IV-E reimbursements to states providing out-of-home placements during the ages years bridge period. Should states choose to approve this expansion, the need for new housing and placement options will increase significantly. Fostering Connections provides states an opportunity not just to provide older youth help for a longer period of time, but also to reconsider how they are serving older youth, both in terms of their preparation for adulthood and their achievement of permanency. Our data indicates that creating a transition system of care that opens up living arrangements based on a youth's personal choice and level of maturity and gives discharged youth chances to return for additional housing support resembles the reality of youth from normal families who are often in transition well into their late 20s (Arnett, 2000) may yield positive outcomes for emancipating foster youth. Acknowledgements This work was supported by a Research Enhancement Grant by the College of Social Work at The Ohio State University. The authors wish to acknowledge Joyce Cooper, Merry Paul, Christi Watson, and Jean Sepate for data collection assistance and Bob Mecum, for insights in interpreting data findings. We would also like to acknowledge Jennifer Pokempner of the Juvenile Law Center and Richard Hooks- Wayman, Senior Youth Policy Analyst of the National Alliance to End Homelessness for their invaluable comments and suggestions. The staff and clients of the Lighthouse Independent Living Program are also acknowledged for their daily work and effort have made the collection and reporting of these data possible. Finally, we appreciate the extensive comments and suggestions offered by one of the anonymous reviewers of our initial manuscript. References Allen, M., & Bissell, M. (2004). Safety and stability for foster children: The policy context. The Future of Children, 14(1), American Community Survey. (2009). U.S. Bureau of the Census. Arnett, J. J. (2000). Emerging adulthood. American Psychologist, 55(5), Arnett, J. J. (2006). Emerging adulthood: The winding road from the late teens through the twenties. New York: Oxford University Press. Barth, R. P., Greeson, J. K. P., Zlotnik, S. R., & Chintapalli, L. K. (2009). Evidence-based practice for youth in supervised out-of-home care: A framework for development, definition, and evaluation. Journal of Evidence-Based Social Work, 6(2), Bates, B. C., English, D., & Kouidou-Giles, S. (1997). Residential treatment and its alternatives: A review of the literature. Child & Youth Care Forum, 26(1), Brickman, A. S., Dey, S., & Cuthbert, P. (1991). A supervised independent living orientation program for adolescents. Child Welfare, 70(1)(January February), Collins, M. E. (2001). Transition to adulthood for vulnerable youths: A review of research and implications for policy. The Social Service Review, 75(2), Cook, R., Fleishman, E., & Grimes, V. (1991). A national evaluation of Title IV-E foster care independent living programs for youth, Phase 2. (Final Report for Contract No ). Rockville, MD: Westat, Inc. Courtney, M. E., & Dworsky, A. (2005). Midwest evaluation of the adult functioning of former foster youth: Outcomes at 19. Chicago, IL: Chapin Hall Center for Children, University of Chicago. Courtney, M. E., Dworsky, A., Ruth, G., Keller, T., Havlicek, J., & Bost, N. (2005). Midwest evaluation of the adult functioning of former foster youth: Outcomes at 19. Chicago: Chapin Hall Center for Children at the University of Chicago. Courtney, M. E., Dworsky, A., Cusick, G. R., Havlicek, J., Perez, A., & Keller, T. (2007). Midwest evaluation of the adult functioning of former foster youth: Outcomes at age 21. Chicago, IL: Chapin Hall Center for Children, University of Chicago. Courtney, M. E., Hook, J. L., & Lee, J. S. (2010). Distinct subgroups of former foster youth during the transition to adulthood: Implications for policy and practice. Chicago, IL: Chapin Hall Center for Children, University of Chicago. Daining, C., & DePanfilis, D. (2007). Resilience of youth in transition from out-of-home care to adulthood. Children and Youth Services Review, 29(9), DeWoody, M., Ceja, K., & Sylvestrer, M. (1993). Independent living services for youth in out-of-home care. Washington, D.C.: Child welfare League of America. Foley, W. J., & Schneider, D. P. (1980). A comparison of the level of care predictions of six long-term care patient assessment systems. American Journal of Public Health, 70(1), Fowler, P. J., & Toro, P. A. (2006). Youth aging out of foster care in southeast Michigan: A follow-up study. Final report. : Research Group on Homelessness and Poverty. Department of Psychology. Wayne State University. Freundlich, M., & Avery, R. J. (2005). Planning for permanency for youth in congregate care. Children and Youth Services Review, 27(2), Geen, R. (2009). The Fostering Connections to Success and Increasing Adoptions Act: Implementation issues and a look ahead at additional child welfare reforms. Washington, DC: Child Trends. Geller, J. L. (1993). The linear continuum of transitional residences: Debunking the myth. American Journal of Psychiatry, 150(7), Georgiades, S. (2005). A multi-outcome evaluation of an independent living program. Child and Adolescent Social Work Journal, 22(5), Hardin, M. (1987). Legal issues related to the federal Independent Living Initiative. Washington, DC: American Bar Association. Harding, J. T., & Luft, J. L. (1993). Outcome evaluation of the PAL (Preparation for Adult Living) program. Abstract presented at the University of Illinois invitational research conference on preparing foster youth for adult living. In E. V., & J. R. (Eds.), Preparing foster youth for adult living: Proceedings of an invitational research conference (pp ). Washington, DC: Child Welfare League of America. Kroner, M. J. (1988). Living arrangement options for young people preparing for independent living. Child Welfare, 67(6), Kroner, M. J. (1999). Housing options for independent living programs. Annapolis Junction, MD: CWLA Press. Kroner, M. J. (2001). Moving in: Ten successful independent and transitional living programs. Eugene, OR: Northwest Media. Kroner, M. J., & Mares, A. S. (2009). Lighthouse independent living program: Characteristics of youth served and their outcomes at discharge. Children and Youth Services Review, 31(5), Leeuvwen, J. V. (2004). Reaching the hard to reach: Innovative housing for homeless youth through strategic partnerships. Child Welfare, 83(5), Lindsey, E. W., & Ahmed, F. U. (1999). The North Carolina independent living program: A comparison of outcomes for participants and non-participants. Children and Youth Services Review, 21(5), Loman, L. A., & Siegel, G. L. (2000). A review of literature on independent living of youths in foster and residential care. St. Louis, MO: Institute of Applied Research. Lorentzen, B., Lemley, A., Kimberlin, S., & Byrnes, M. (2008). Outcomes for former foster youth in California's THP-plus program: Are youth in THP-plus faring better? San Francisco: John Burton Foundation. Mallon, G. P. (1998). After care, then where? Outcomes of an independent living program. Child Welfare, 77(1), Mares, A. S. (2010). Personalcommunication based on unpublishedanalysisof AFCARS data from available at: McCoy, H., McMillen, J. C., & Spitznagel, E. L. (2008). Older youth leaving the foster care system: Who, what, when, where, and why? Children and Youth Services Review, 30 (7), McMillen, J. C., & Tucker, J. (1999). The status of older adolescents at exit from out-ofhome care. Child Welfare, 78(3), Mech, E. V. (1988). Editor. Independent-living services for at-risk adolescents. Washington, DC: Child Welfare League of America. Montgomery, P., Donkoha, C., & Underhilla, K. (2006). Independent living programs for young people leaving the care system: The state of the evidence. Children and Youth Services Review, 28(12), Naccarato, T., & DeLorenzo, E. (2008). Transitional youth services: Practice implications from a systematic review. Child and Adolescent Social Work Journal, 25(4), National Resource Center for Youth Development (NRCYD) (2010). John H. Chafee Foster Care Independence Program, Allotments for Accessed at:. nrcyd.ou.edu/programs/pdfs/fy07_cfcip_alloc.pdf Phibbs, C. S., Bronstein, J. M., Buxton, E., & Phibbs, R. H. (1996). The effects of patient volume and level of care at the hospital of birth on neonatal mortality. Journal of the American Medical Association, 276(13),

20 M.J. Kroner, A.S. Mares / Children and Youth Services Review 33 (2011) Reilly, T. (2003). Transition from care: Status and outcomes of youth who age out of foster care. Child Welfare, 82(6), Scannapieco, M., Schagrin, J. L., & Scannapieco, T. (1995). Independent living programs: Do they make a difference? Child and Adolescent Social Work Journal, 12(5), Srebnik, D., Uehara, E., & Smukler, M. (1998). Field test of a tool for level-of-care decisions in community mental health systems. Psychiatric Services, 49(1), U.S. Department of Health and Human Services, Administration for Children and Families, Administration on Children, Youth and Families, Children's Bureau. (2010). AFCARS Report PreliminaryFY2009EstimatesasofJuly2010 (17). Available at:. acf.hhs.gov/programs/cb/stats_research/afcars/tar/report17.htm U.S. General Accounting Office (1999). Foster care: Effectiveness of independent living services unknown. Washington, DC: Author. Usher, C. L., Randolph, K. A., & Gogan, H. C. (1999). Placement patterns in foster care. The Social Service Review, 73(1), Westat, Inc. (1991). A national evaluation of title IV E foster care independent living programs for youth. (Phase 2 Final report, Vol. 2).