The purpose of the study was to examine criminal behavior and criminal justice system involvement among youth making the transition from out-of-home care to independent adulthood. The larger purpose of the Midwest Study of the Adult Functioning of Former Foster Youth (Midwest Study) was to gather information about services provided to selected youth served in participating states and the adult self-sufficiency outcomes achieved by the youth. The research team considered the importance of experiences within the child welfare system on criminal behavior during the transition to adulthood. In addition, they examined whether social bonds predict criminal behavior and the risk for criminal justice involvement among former foster youth. The following research questions were addressed:
- Are offending patterns during the early transition to adulthood (ages 17-22) among youth formerly in out-of-home care different from those of the general population?
- Do experiences in out-of-home care, including number of placements, placement type, age at entry, and receipt of independent living services predict later criminal behavior or criminal justice involvement during the transition to adulthood among youth aging out of the child welfare system? Is the relationship between out-of-home care experiences and crime moderated by race?
- Above and beyond prior experiences with maltreatment and within the child welfare system, do the bonds to parents or caregivers, education, or employment that foster youth have as they approach the transition to adulthood predict later criminal behavior or criminal justice involvement?
The study collected data from two sources: (1) survey data from the Midwest Study of the Adult Functioning of Former Foster Youth (Midwest Study), and (2) official arrest data.
The Midwest Study was a longitudinal panel study that was part of a collaborative effort of the state public child welfare agencies in Illinois, Iowa, and Wisconsin, Chapin Hall at the University of Chicago, and the University of Washington. The participating states funded and/or operated the full range of services supported by the Chafee Foster Care Independence Program (e.g., life skills training, mentoring, tutoring, employment services, transitional housing), but these services varied considerably in availability between and within the states. In addition, the policy regimes differed across the three states in ways that affected the supports available to youth who aged out of care. For example, at the time of this study, youth in Illinois were legally permitted to remain in care until age 21, while youth in Iowa and Wisconsin remained in care until approximately age 18.
The Midwest Study survey data were collected directly from the youth in the sample every two years over three waves, between May 2002 and January 2007. A total of 732 respondents participated in at least one of the in-person interviews over the three waves. The interviews were approximately 90 minutes long and were made up almost entirely of fixed response questions. A portion of the survey was administered using Audio Computer Aided Self Interviewing (ACASI). Study participants listened to a recording of questions about criminal behavior through headphones and entered their responses into a computer.
During the first wave of interviews, conducted between May 2002 and March 2003, 732 interviews were completed. Youth were between the ages of 17 and 18 when they were first interviewed. The second interview took place in March through December 2004, between respondents' 19th and 20th birthdays. A total of 603 youth completed the Wave 2 survey. The third wave of interviews, conducted between March 2006 and January 2007, took place as soon as possible after respondents' 21st birthday. A total of 590 individuals from the original sample completed the Wave 3 survey. This data collection includes some variables that were directly measured from the original Midwest Study survey instrument and other variables that were computed or derived from variables in the original data for purposes of the current study.
To supplement the survey data, the research team accessed official arrest data from each state for this study. Youth were matched to arrest records based on identifying information such as name, date of birth, social security number, gender, and race. In each of the 3 state databases, researchers obtained data on all criminal arrests that occurred between the respondents' Wave 1 interview and August 31, 2007, a date by which all of the study participants were at least 21 years old. Official arrest information was obtained for 728 of the original 732 sample.
Youth from Illinois, Iowa, and Wisconsin who met the following basic criteria were eligible for inclusion in the original sample: (1) they reached the age of 17 years old while in out-of-home care, (2) they had been in care for at least one year prior to their 17th birthday, and (3) they were placed in out-of-home care for reasons of abuse or neglect rather than delinquency. The only exclusions from this population were based on random sampling of the eligible population in Illinois and the inability of youth to participate in the survey due to developmental disability, mental illness, or inability to participate in an interview in English.
In the states of Iowa and Wisconsin, the sample reflected the universe of youth in out-of-home care who fit these criteria. In Illinois, a random sample of youth was selected, who represented two-thirds of those youth in out-of-home care that fit the sampling criteria. A group of 767 youths who met the sample selection criteria was identified.
All youth from Illinois, Iowa, and Wisconsin between May 2002 and August 2007 who reached the age of 17 years old while in out-of-home care, were in care for at least one year prior to their 17th birthday, and were placed in out-of-home care for reasons of abuse or neglect rather than delinquency.
The Midwest Study of the Adult Functioning of Former Foster Youth (Midwest Study)
official criminal arrest records
administrative records data
The study contains a total of 85 variables including indicator variables, demographic and background variables, delinquency and crime variables, out-of-home care experiences variables, and social bonds variables.
Indicator variables include three variables that indicate whether the respondent participated in the Wave 1 survey, Wave 2 survey, and Wave 3 survey.
Demographic variables include gender, age at Wave 1, Hispanic ethnicity, race (White, Black, mixed or other race), and state of residence at Wave 1. Background risk factors include indicator variables for behavioral or mental health disorders, teen parenthood, maltreatment histories, and alcohol or drug dependence diagnoses.
Delinquency and crime variables measured at Wave 1, Wave 2, and Wave 3 indicate whether the offender deliberately damaged property, stole something worth less than $50, stole something worth more than $50, went into a house or building to steal something, sold drugs, hurt someone badly enough to need medical care, used or threatened to use a weapon to get something, participated in a group fight, pulled a knife or gun on someone, and shot or stabbed someone. Other variables include the number of total delinquency items at Wave 1, self-reported arrest at Wave 1, whether the offender had any violent offense or any nonviolent offense at each of the three waves, Wave 3 violent behavior scale, Wave 3 nonviolent behavior scale, whether the respondent was ever arrested at Wave 1 or Wave 2, whether the respondent was arrested since Wave 1 at Wave 2 interview, whether the respondent was arrested since Wave 2 at Wave 3 interview, whether the respondent has an official arrest record, time since Wave 1 interview to first arrest or end of observation period, and latent class analysis assignment.
The out-of-home care experiences variables contain measures that track an individual's experience within the child welfare system. These include age at the time of first placement, the total number of foster care placements, type of placements at the Wave 1 survey (foster care, kinship care, group care, independent living or other care), and sum of any independent living services at Wave 1.
Social bonds variables include measures of parent (both maternal and paternal) and foster caregiver attachment, and two variables indicating whether a respondent had no mother or father present in their life. Other variables include employment, educational aspirations, and school/college enrollment status. There are also two scale variables, one measuring perceived social support (based on the MOS Social Support Survey) and another measuring how likely an individual was to turn to the child welfare system for help.
Of the 767 youths who comprised the final sample, a total of 732 Wave 1 interviews were completed, yielding a response rate of 95.4 percent. At Wave 2, 603 of the 732 youth completed interviews, for a follow-up response rate of 82.4 percent. At Wave 3, 590 of the 732 youth completed interviews, for a response rate of 80.6 percent.
Of the 732 Wave 1 respondents, 730 (99.7 percent) reported on criminal behavior. Of the 603 Wave 2 respondents, 574 (95.2 percent) reported on criminal behavior. Of the 590 Wave 3 respondents, 504 (85.4 percent) reported on criminal behavior.
The study includes a violent crime scale, nonviolent crime scale, maternal closeness scale, paternal closeness scale, caregiver closeness scale, social support scale (based on the Medical Outcomes Study (MOS) Social Support Survey), and a scale of the likelihood of turning to the child welfare system for assistance after discharge.