Research & Evaluation
High School Graduation and Child Welfare: A Description of the Education Status of Older Minnesota Adolescents in the Academic Year After Substantiated Child Maltreatment Findings
Title IV-E Curriculum Module
Module Elements
- Summary of findings (Below)
- PowerPoint PDF
- Selected references for additional readings
- Resource list for more information
- Potential guest speakers
- Discussion Questions
Module Summary
Introduction
The adult outcomes of children who have contact with the child welfare system are important to communities, child advocates, and policymakers. Young adults with poor educational outcomes, untreated mental health problems or other personal and social challenges may not reach their full potential in terms of earnings, physical and mental health, aspirations, or become effective parents. Adults who cannot contribute positively, frequently generate costs to the larger community in terms of welfare utilization and crime. The economic growth of any country depends upon a workforce that is well-educated and productive (National Science Foundation, 1996).
Education is key to economic and personal well-being. This report provides information on the educational outcomes of older adolescents who have had contact with the child welfare system in Minnesota, over 30 months, prior to the 2002-2003 academic year. Very few of these youth were in out-of-home placement, and it is likely that many of them no longer had active county case workers when their educational status was reviewed.
The report also provides descriptive data on the study group that show similarities between them and national educational outcome trends observed for long-term foster care youth – a different, but related population. It suggests additional research questions and prompts consideration of policy and practice implications for the adult outcomes of Minnesota youth at risk for school failure. The scope of the examination captures older youth for whom there was a substantiated (i.e. confirmed after investigation) maltreatment findings in Minnesota during January 2001 through June 2003 (with no out-of-home placement at the time of maltreatment) as well as those few who were later placed out of home as a result of the maltreatment. This paper is the first in a series of Minn-LInK special topic reports related to child well-being in Minnesota.
The report is structured to focus first on the graduation outcomes of 12th grade adolescents with recent child protection system involvement, followed by an overview of the educational status of students in the entire group at all ages and grade levels (including, but not limited to 12th graders). Related analyses that describe the original child protection group and information on adolescents whose education records could not be matched are included in Appendices A and C respectively.
Background
Literature on child welfare youth most often focuses on those who have been placed long-term out of their homes due to abuse and neglect. Many foster care youth - those who have spent some or a large portion of their childhoods in relative and non-relative homes – constitute a population that faces significant challenges to becoming successful adults. Foster care youth arguably represent the most at-risk of all children who have contact with the child welfare system. Although this study examines a broader population of adolescents having contact with the child welfare system in Minnesota, many children cycle in and out of placement and therefore the most current findings related to adult outcomes of foster care youth likely apply to a portion of the child welfare population in general and at any point in time.
Older adolescents constitute a bit more challenging group to study because their contact with the child protection system becomes more tenuous and inconsistent as they approach age 18. Some county agency staff indicated that they do not always investigate alleged maltreatment of older adolescents - or, if an older adolescent has contact with any public system, it is more often the corrections area. County agencies also acknowledge that staff cuts have created some limitations on what can and cannot be investigated with older youth. Youth without placement often do not have active case plans in place, which also limits the services they receive. For these reasons and probably many others, much of the literature on older adolescents in public child welfare systems focuses on those in long-term placement.
This study is different than most in that it casts a broader net to include older adolescents who have had contact with child protection (in this case, via a substantiated maltreatment finding) – acknowledging that many other adolescents are no longer even involved in child protection by this age and are, therefore, not represented here.
Social and Emotional Well-being and Disability Most studies of foster care youth show that they have more developmental problems (including disabilities) than other children. These social and emotional challenges can continue into adulthood: over two-thirds of young adults discharged from foster care were found to have emotional disturbances (38%) or were using illegal drugs (40%) (Wertheimer, 2002). A study of foster care children preparing to leave state care found that rates of special education receipt among sample subjects were as high as 47% (Courtney, et al, 2004) with similarly disproportionate rates for other studies (ranging from 19-36%) (MacArthur Foundation, 2005). Up to 38% of former foster care youth report being emotionally disturbed (Wertheimer, 2002).
Youth with child protection contacts and histories of placements are at risk for homelessness. Homeless youth surveyed in Minnesota reported high rates of suicidal thoughts (34%), attempts (67%), or a significant mental health problem (42%) (Owen, 2003). The Midwest Evaluation of the Adult Functioning of Former Foster Youth found that foster care youth had disproportionately higher rates of serious fighting, theft, running away and causing injury than the national average (Courtney, 2004). A 2004 study of need and use of mental health services among child welfare populations showed that only 25% of youth in need of mental health services received them (Burns, et al).
Contact with the Justice System Juveniles with a history of foster care placement have high rates of crime and incarceration during their adolescence and after exiting care. There continues to be an important overlap between youth experiencing maltreatment, foster care, homelessness, and criminal activity. Of homeless youth surveyed, 10% reported that they had traded sex for shelter, clothing, or food (Owen, 2003). One-quarter to over 1/3 of adolescents exiting foster care reported that they had done something illegal for money (Barth, 1990; Casey Family Programs, 2001) and other studies indicate that the range of contact with the justice system can be anywhere from 18-45% of a given sample of former foster care children (Wertheimer, 2002; Casey Family Programs, 2001; Courtney, 2001). The interconnectedness of juvenile justice involvement, foster care involvement, and emotional well-being is also evident by 20% of juveniles arrested having a serious mental health disorder (Mac Arthur Foundation) and 50-75% of incarcerated youth having a diagnosable mental health disorder (National League of Cities, 2005).
Education and Work The multiple effects of the challenges presented by social and emotional issues, disability, poverty, and contact with the justice system are manifested in the educational outcomes of former foster care youth. This is a critical issue because education is universally considered the foundation of economic and personal well-being in adulthood. In addition to having disproportionate rates of special education receipt, foster care students have high rates of tutoring, when available (35%), and low rates of high school completion. For those students still in foster care, academic failure becomes a risk factor for further maltreatment and ongoing academic problems (Evans, 2004). While trying to finish school, older foster care youth experience a host of challenges.
A comparison study of foster care and non-foster care high school students showed that foster care youth experienced more discipline problems, higher rates of school changes, lower (foster) parent school involvement, and for those who graduated from high school, less financial aid for post-secondary education from their biological family. Foster care students were more likely than non-foster care students to be tracked into “general” post-secondary planning (e.g. vocational and community college or employment) than “college prep,” even after controlling for test scores and grades. When student groups had similar post-secondary plans, there was a lag in former foster care students being able to keep up with those plans at the same rate as non-foster care students post graduation. Foster care students who were due to graduate were much more likely than non-foster care students to say that they would not be disappointed to not graduate with their classmates on time. Among drop-outs in the sample, 25% reported that not liking the teachers was a reason why they left school (Whiting Blome, 1997).
Follow-up studies show that former foster care youth graduate in lower numbers than non-foster care populations with graduation rates that range from 45% - 50% (Barth, 1990; Wertheimer, 2002; Casey Family Programs, 2001). By age 18, adolescents, who have had disrupted school attendance due to moves and foster care placements, poor academic performance because of multiple stressors, also abruptly lose access to the supportive services that could ease the transition to adulthood.
Challenges faced by adolescents in the child welfare system must also be viewed in the broader context of changes in how all youth now transition to adulthood in American society. A growing body of literature suggests that for most young people in the United States, traditional “adulthood” no longer actually begins at ages 18 or 21 (i.e., the legal voting or drinking age in most states). In fact, with many young adults remaining at home with parents during their college years or when first securing employment, it seems that youth without the assets of supportive families or economic resources begin adulthood even further behind their peers (Arnett, 2001).
Although older adolescents having contact with the child welfare system face significant challenges as they enter adulthood, this paper is intended to describe the secondary educational outcomes that result from these challenges, exploring potential new research and policy and practice innovations that could improve them.
Methods
Minn-LInK
The Minn-LInK project at the Center for Advanced Studies in Child Welfare at the University of Minnesota School of Social Work relies on secondary administrative data obtained from statewide public programs. Minn-LInK provides a unique collaborative, university-based research environment with the express purpose of studying child and family well-being in Minnesota. The administrative data sets used in this descriptive analysis originate in the Minnesota Department of Human Services (utilizing the Social Services Information System, or SSIS), which oversees the state child protection system in Minnesota and student public school education records from the Minnesota Department of Education. All data use has been within the personal privacy guidelines set by strict legal agreements between these agencies and the University of Minnesota.
The Data
Human service programs collect data for multiple purposes: program administration, compliance with federal and state reporting, fiscal management, and local outcome measures. Policy and practice research has rarely been the focus of either automated system development or data collection. While these realities do not prohibit the successful design, implementation, and completion of research, it does present researchers with unique challenges related to study design and time-frames for study group selection that do not occur when collecting and working with primary data. Instances in which data system conditions drove the structure of this study have been noted in this report.
In this study, the entire universe of adolescents experiencing substantiated (confirmed) maltreatment or neglect for a given time period and meeting age criteria were selected. There is no comparison group, and the ability to match the child protection records to education records was a function of unknown factors. For these reasons, statistical tests were not run on the data.
Graduation Rates
The graduation percentage rates for this study were calculated by creating a ratio of the number of students in 12th grade who graduated over all 12th grade students. This is different from the methodology used in some other studies. In fact, over the past five years there has been a great deal of interest in and reconsideration of the ways in which school districts across the nation calculate and report their high school graduation rates largely driven by the passage of the No Child Left Behind (NCLB) education act.
The national research group, the Urban Institute, and others, recommend a multi-year method that takes into account the status of 9th graders by 12th grade, following those students over the entire time period. The Urban Institute method, called the Cumulative Promotion Index or CPI, is similar to that used by schools to report for NCLB (Swanson, 2004). In 2001, the Manhattan Institute, using National Center for Education Statistics data, used a method that creates a ratio of the number of public high school diplomas awarded by an estimate of the number of students who would have received diplomas had the graduation rate in that school been 100 percent. This method also requires multiple years of data, as it follows students who begin in eighth grade and adjusts by student population changes over six years (Greene, 2001). Because the child welfare adolescent group consisted of students of different age and grade ranges, many years of education data would have been required to calculate graduation rates in this manner.
Studies of specific populations of students – such as adolescents in long-term foster care – frequently rely upon student report for information about graduation or are limited by the number of years of administrative data they have available for study (Courtney, et al, 2004; Burley, et al, 2001). This study was also limited in the number of years of data available for study, and while the method used for calculating graduation did not involve multiple years, it is intuitive, and the same method was used to calculate the graduation rates for all 12th grade Minnesota graduates as for the child welfare adolescent Graduate group. Further, while the method relies upon a single year of data, it did produce an overall graduation rate for all 12th grade graduates that is similar (within four to five percentage points) to most other sources reporting on Minnesota graduation rates (in this report, the overall graduation rate for 12th graders in June 2003 was 74%). What is most important to keep in mind, is that this report is about the gap between the graduate rates – however it may be calculated – between all Minnesota 12th graders and 12th graders who have had recent contact with the child protection system. For more detailed information on the calculation of graduation rates, see Appendix D.
Match for Graduation Status
The child welfare and education data systems in Minnesota do not share a unique identifier by which individual child records can be easily matched. Recordmatching must be accomplished according to algorithms that combine first and last name and birthdates and occasionally, social security numbers. The education records for some adolescents simply cannot be matched because of insufficient or inaccurate information in either or both files (or because adolescents are not enrolled in the public education system).
Prior to matching to education records, a group of child protection records were selected. This group was comprised of all adolescents who experienced substantiated maltreatment or neglect over the period of January 2001 and June 2003, and who were projected to be near graduation age (at least 18) by June 2003. (For more information on the characteristics of this group, see Appendix A.)
In addition to the inevitable inability to match some records due to inaccuracies and lack of data, the oldest members of the adolescent sample were less likely than younger members to appear in the education record. In particular, 115 students were projected to turn 20 during the 2002-2003 school year and have a lower likelihood of being enrolled in the K-12 school system.
Findings
Teens with recent child protection involvement had much lower high school graduation rates than all 12th graders in 2003. Of the 387 seniors in the study group, 47% successfully graduated from high school. Minnesota’s high school graduation rate is anywhere from 78% - 92%, depending upon the calculation method used.
Graduation rates were different depending upon
- • Race
- • Geography (a metro or non-metro county)
- • Gifted and talented status
When comparing the characteristics of 2003 child welfare graduates and non-graduates to all Minnesota 12th grade graduates, child welfare 12th graders were more likely to
- • have lower incomes (indicated by meal program eligibility
- • participate in special education
- • be a child of color (non-Caucasian)
Overall Progress
The full group of child welfare teens (N=501) spanned multiple grades and not all students would be expected to graduate by June 2003. When examining the status for students in all grade levels:
- • 70% were making educational progress.
- • 21% had experienced set-backs
- • 9% had an unknown status
Recommendations
Practice
Have a clear idea of current practice and potential enhancements and examine them. Minnesota has over 300 school districts and 87 counties. This study examines the condition of graduation and educational outcomes for a group of adolescents, but does not examine current or innovative practices. County staff observe that they are doing the best that they can with current resources and wonder if other jurisdictions are having better results with other models of delivery. In particular, outcomes may be different in schools where county social work staff are housed in the schools. Schools are the most logical (and sometimes only) point of contact for adolescents who are in danger of school failure but may be without an ongoing relationship with county social services staff.
Consider options for proactive, rather than reactive relationships between county, school-based social work staff, and juvenile justice. In many counties, current practice – and funding realities – result in these three support systems not interacting until there is a crisis. This can be aggravated by how cases are assigned to county social workers (making it challenging for school social workers to locate their county counterparts) or by the physical locations of social work staff (on site at schools versus elsewhere in county government buildings).
Consider a different role for Guardian ad Litem in the process of assuring older child welfare adolescents who are at risk of not graduating have the necessary services for success. Likewise, examine the need for another academic support role such as an education surrogate. Case workers in county agencies have responsibility for connecting youth with needed services particularly when the child cannot stay with their biological family after maltreatment occurs. In this study, the average number of placements was 1.4 – usually a short-term placement followed by another of longer duration. It may be unrealistic for county social workers to expect a short-term caregiver to take a great deal of responsibility for the school attendance, homework, test performance, or school engagement of the youth. If out-of-home placement case plan outcome expectations do not include attention to academics, who is giving this attention? Is it appropriate to expect social workers to take the lead in making sure foster care youth are engaged in their schools and working up to their full academic potential? How much time should pass between a placement start date and a regular “check-in” on academic status? Are some counties already doing so, and at what points in time? Does this indicate a need for an educational surrogate or advocate?
A recent national Child and Family Services Review (CFSR) summary notes that for Minnesota, areas of needed improvement related to educational outcome were: 1) there was too much reliance on foster parents to see to children’s educational needs; 2) there were multiple school changes related to placement changes; and 3) the inclusion of school records in case records was the exception rather than the rule (Minnesota Department of Human Services, 2003). Although educational needs are recognized as an important performance measure for the state, there is no statutory or rule-based requirement that county agency staff track academic outcomes for child welfare youth, and monitoring varies by county agency across the state.
When students are not placed out of their homes – a condition that applied to the majority of this group - are steps taken to be sure that biological families take an active role in the educational outcomes of their children? Given that this study found considerable challenges to fully matching a group of child welfare adolescents using administrative data (likely due in part to the mobility of these youth and their families), it would seem that other options for monitoring academic progress should be considered. For those adolescents whose situations do not warrant out of home placement, what county-school system strategies can be employed to assure ongoing educational support in the absence of an active social services case plan? What positive and respectful supports can be put in place for parents to keep their adolescents in school and progressing?
Counties may consider closer monitoring of child welfare adolescents who are receiving special education. This may prevent some portion of the low graduation rates observed for special education recipients.
When a youth is involved in the child welfare system, there are multiple workers and entities involved in the services they receive. Some research implies that there is sometimes a bias at work in schools in which, regardless of their academic performance, foster care youth are not encouraged to set their post-secondary sights any higher than vocational education or immediate employment (Whiting Blome, 1997). Minnesota schools must be vigilant to potential biases about child welfare youth and assure objectivity when making post-secondary recommendations. Failure to do so sends a message to these youth that the educational system does not have faith in their ability to graduate, earn college degrees, and reach their fullest potential. Programs, teachers, and advisors supporting the educational outcomes of child welfare youth would be well-advised to monitor their working relationships with youth and be aware of any potential limitations they unintentionally convey that may be based on anything other than academic performance and ability.
Further Research
Further examine the educational outcome differences observed between metro and non-metro counties in Minnesota. Explore some of the reasons why non-metro counties have higher rates of educational progress as well as higher rates of adolescents for whom educational status is unknown. Explore differences in practice, resources, or philosophies of agencies serving these geographical areas of the state. Surely some child welfare youth possessing all of the risk factors repeatedly associated with educational failure have “made it” and graduated. Some of them will also go onto college and earn degrees that will assure them a much higher standard of living than had they gone straight to employment after high school, or dropped out of high school altogether.
Qualitative methods could be used to complete detailed, case-level analyses of the services received, and educational outcomes of all child welfare adolescents (not just those in placement), thereby controlling for as many independent variables as possible (e.g. race, income, special education receipt, etc.) Detailed case studies of these successes and those with less successful outcomes may reveal what specific factors make key differences. Case studies should take into account the specific policy and practice differences in place that affected the outcomes of students. Giving attention to the outcomes of students who were expelled could also lend insight into long-term outcomes. Some of those factors may vary by other attributes and assets the youth possesses. These studies can advise the ways in which specific service delivery practice can change and where additional investments of time and money should be made or existing investments shifted.
Quantitative methods employed in these evaluations could statistically measure the relationships and the related effect sizes of particular interventions – something that could not be done in this descriptive study. Careful randomized group construction could remove the effects of the various factors shown here as potentially confounding (for example, race and geography) in order to isolate the effects of specific interventions. On future evaluations of interventions and ongoing tracking of graduation rates, the results of this study should provide a baseline measure for future studies of educational outcomes.
Study what proportions of graduates and non-graduates are later found to have entered the higher education system. Examine the post-secondary status of graduates and non-graduates, exploring what types of programs they enter and complete can give an indication of future earnings potential.
Examine wages one, two, or more years after June 2003 for graduates and non-graduates. Study whether there are differences between the average wages earned, hours worked, and industries of employment for these students to get a better sense of the degree to which completing high school (or not) influences their later ability to earn a livable wage.
Examine public assistance program (welfare) use of graduates and non-graduates for time periods after June 2003. “Use” could include public assistance program receipt as well as applications for programs that were denied (indicating that the individual applied for assistance, but may have been denied due to wage level or other reasons).
Study whether these adolescents appear in vital statistics birth records as new parents, during their high school year or years immediately thereafter, to determine whether they are facing additional new responsibilities as young adults leaving the child welfare system.
Explore how the educational outcomes of this group compare to others for whom we know we provide ongoing supportive services - such as those in the adoption system, those in long-term placement, those who are state wards, etc. In addition to racial disparities, one of the differences between 12th grade child welfare youth who do not graduate on time, or at all, or experience great challenges compared to those who do graduate, and all other 12th grade graduates in Minnesota are their high rates of special education receipt. Further research is needed into some of the adult outcomes of these youth, to consider the challenges they face and the difficulties they have getting through high school in comparison with their peers.
Explore the reasons why records from some racial and geographic groups are more “match-able” than others. Are the child welfare records in better condition for matching (i.e. names more likely to be correctly spelled, social security numbers more likely to be present, etc.) if a child has been in placement? What practices and conditions might be in place in metro counties that are lacking in non-metro counties that dramatically improve the match-ability of their child welfare records? Maximizing match rates are critical to building the very best populations from which to generalize findings.
Policy
Experiments with funneling more services through schools could require funding shifts or increased funding that might need legislative or policy change to support. School-based initiatives would need to be sensitive to current pressures related to No Child Left Behind and recent public school funding changes. True partnerships targeted at shared outcomes would be necessary, along with sufficient financial supports. All new service models should be rigorously evaluated.
Revisit the Structure of the Family Services Collaboratives. The Collaboratives showed promise, but were under funded. Additional modifications might be needed to improve service delivery outcomes. For instance, some parents served by the Collaboratives are quite challenged by service choice. A more family-specific approach might help in general, and improve educational outcomes of children in particular.
Consider expansion of out-of-home placement related services to serve young adults after age 18. (see CASCW’s online conference proceedings from “Leave No Adolescent Behind.”) Evaluations of these investments have shown positive education and earnings results. This model represents added expense for Minnesota as it is currently structured, but may save tax dollars in other systems.
Explore the cost-benefit work done by others and consider doing it here in Minnesota. Use the findings of institutions such as the Washington State Institute for Public Policy (WSIPP) that has analyzed the cost-benefit ratios of an array of prevention and early intervention programs for children and youth. Their analyses have shown remarkable savings to taxpayers for a number of programs aimed at improving child outcomes ultimately saving money for all taxpayers.