Employment outcomes of young parents who age out of foster care

CYSR-03090; No of Pages 8 Children and Youth Services Review xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Children and Youth Services Review journal homepage: www.elsevier.com/locate/childyouth

Employment outcomes of young parents who age out of foster care Amy Dworsky ⁎, Elissa Gitlow Chapin Hall, University of Chicago, 1313 East 60th Street, Chicago, IL 60637, USA

a r t i c l e

i n f o

Article history: Received 31 August 2016 Accepted 28 September 2016 Available online xxxx Keywords: Former foster youth Adolescent parents Transition to adulthood

a b s t r a c t Despite the high rate of early parenthood among youths in foster care, relatively little is known about the employment outcomes of young parents after they age out of the child welfare system. This study analyzed administrative data for 1943 recently emancipated youths who were the parent of at least one child. Like prior studies of former foster youths, this study finds that only half of the parents were employed at any point during the first four quarters after exiting care, most of those who worked were not consistently employed, and the earnings of parents who were employed were very low. Several demographic and placement history characteristics were associated with an increase or decrease in the odds of being employed and/or with total earnings among the parents who worked. The implications of these findings for policy and practice are discussed. © 2016 Published by Elsevier Ltd.

1. Introduction Despite the high rate of early parenthood among youths in foster care (Dworsky & Courtney, 2010; Putnam-Horstein, Cederbaum, King, & Needell, 2013), and the fact that struggling parents have been identified as a distinct subgroup of former foster youths (Courtney, Hook, & Lee, 2010), relatively little is known about what happens to young parents after they age out of the child welfare system. To address this gap in our knowledge, this study used administrative data to examine the employment outcomes of former foster youths who became parents while in foster care. In addition to looking at whether these parents were employed and how much they earned during the first year post-emancipation, we investigated the relationship between those outcomes and an array of demographic and placement history characteristics. Before turning to a description of our data sources and methodology, we briefly review the literature on the employment outcomes of former foster youths, the effects of adolescent childbearing on employment, and the prevalence of early parenthood among youths in foster care. We also provide some information about the context in which the present study was conducted. 1.1. Employment and foster care Since it was created in 1999 by the Foster Care Independence Act, the John H. Chafee Foster Care Independence Program has been providing states with federal funds to help youths transitioning out of foster care become self-sufficient young adults. However, research on the employment outcomes of former foster youths suggests that far too many of these

⁎ Corresponding author. E-mail address: [email protected] (A. Dworsky).

young people are not able to support themselves. Some of these studies have used survey data collected directly from former foster youths (Courtney, Dworsky, Brown, Cary, Love, & Vorhies, 2011; Courtney, Piliavin, Grogan-Kaylor, & Nesmith, 2001; Hook & Courtney, 2011; Naccarato, Brophy, & Courtney, 2010). Other studies have linked the child welfare records of former foster youths to quarterly Unemployment Insurance (UI) wage records (Dworsky, 2005; Goerge et al., 2002; Macomber et al., 2008; Singer, 2006). Regardless of the type of data used, the overall pattern is the same: low rates of employment, unstable work histories, and earnings that are often less than the poverty threshold for a single adult. Two other consistent findings have emerged from this research. First, young people who age out of foster care are less likely to be employed, and earn substantially less when they work than young people in the general population who have not been in foster care (Courtney et al., 2010; Macomber et al., 2008), including those from low-income families (Macomber et al., 2008). Second, despite increasing over time, both the employment rates and earnings of young people who age out of care tend to remain low well into their mid-twenties (Courtney, Dworsky et al., 2011, Courtney, Zinn et al., 2011; Hook & Courtney, 2011; Macomber et al., 2008). A number of studies have gone beyond descriptive analyses to identify predictors of employment and/or earnings using multivariate techniques. Some of the identified predictors reflect patterns that are evident among young people in the general population (Baum, Ma, & Payea, 2013; Bird & Bryant, 2015; Bureau of Labor Statistics, 2015; Office of the President of the United States, 2015). For example, a number of studies have found that former foster youths who are African American are less likely to be working (Dworsky, 2005; Dworsky et al., 2010; Hook & Courtney, 2011) and earn less when they are employed (Dworsky, 2005; Dworsky et al., 2010; Goerge et al., 2002; Naccarato et al., 2010) than their white counterparts. Other studies have found that educational attainment is

http://dx.doi.org/10.1016/j.childyouth.2016.09.032 0190-7409/© 2016 Published by Elsevier Ltd.

Please cite this article as: Dworsky, A., & Gitlow, E., Employment outcomes of young parents who age out of foster care, Children and Youth Services Review (2016), http://dx.doi.org/10.1016/j.childyouth.2016.09.032

2

A. Dworsky, E. Gitlow / Children and Youth Services Review xxx (2016) xxx–xxx

positively associated with both employment (Hook & Courtney, 2011) and earnings (Hook & Courtney, 2011; Naccarato et al., 2010). Although both Goerge et al. (2002) and Macomber et al. (2008) failed to find a consistent relationship between foster care placement history characteristics and employment outcomes, other research suggests that the employment outcomes of former foster youths are related to their experiences while in foster care. For example, studies have found that exiting from congregate care rather than family foster care is associated with a lower likelihood of being employed and with lower earnings (Dworsky, 2005; Hook & Courtney, 2011); that placement instability is negatively related to earnings, but not to employment (Hook & Courtney, 2011); and that exiting foster care at an older age is associated with more favorable employment outcomes (Dworsky, 2005; Hook & Courtney, 2011; Macomber et al., 2008). There is also evidence that employment prior to age 18 is associated with an increase in the likelihood of being employed post-discharge (Goerge et al., 2002; Macomber et al., 2008).

Midwest Evaluation of the Adult Functioning of Former Foster Youth (Midwest Study) followed N 700 young people in Iowa, Wisconsin, and Illinois as they transitioned out of foster care and into adulthood. By age 19, 32% of the young women and 14% of the young men reported being the parent of at least one child, compared with 12% of 19-yearold women and 7% of 19-year-old men in the general population (Dworsky & Courtney, 2010). More recently, Putnam-Horstein et al. (2013) linked child protective services (CPS) records from the California Department of Social Services to birth records from the California Department of Public Health using probabilistic matching to estimate the percentage of 17-year-old female foster youths in Los Angeles County who gave birth during their teen years. Of the 6749 young women who were in foster care at age 17 between 2003 and 2007, 11.5% gave birth at least once before age 18 and 27.5% gave birth at least once before age 20.

1.2. Employment and early parenthood

Very little is known about the employment outcomes of youths who become parents while still in foster care. One study that did examine the effects of parenthood on the employment outcomes of former foster youths found that having a child increased the likelihood that young men would be employed, decreased the likelihood that young women would be employed, and had no effect on hourly wages among either young men or young women who worked (Hook & Courtney, 2011).3 Moreover, some of parents did not give birth to or father their first child until after they aged out of foster care. In the present study, we address three primary research questions:

The adverse consequences of teenage motherhood on economic wellbeing have been well-documented (An, Haveman, & Wolfe, 1993; Hofferth & Hayes, 1987). A number of studies suggest that adolescent childbearing has a negative effect on future employment and earnings, even after controlling for other factors that might explain the differences in labor market outcomes between young women who were teenage mothers and their peers who delayed becoming parents. For example, using propensity score matching with data from the National Longitudinal Study of Adolescent Health (Add Health), Lee (2010) found that teenage mothers were 15% less likely to be employed, 12% less likely to participate in work-related activities, and 10% less likely to be employed full-time than young women who postponed childbearing. However, there was no difference in weekly wages among full-time workers. Similarly, using data from the National Longitudinal Study of Adolescent Health (Add Health), Fletcher and Wolfe (2009) found that adolescent childbearing is associated with a substantial reduction in wages. One explanation for the poor labor market outcomes experienced by young women who were teenage mothers is that adolescent childbearing impedes the accumulation of human capital (Becker, 1993). Consistent with this theory, numerous studies have found that teenage mothers are less likely to complete high school and graduate from college than their peers with no children (Ashcraft & Lang, 2006; Levine & Painter, 2003; Perper, Peterson, & Manlove, 2010). However, some have argued that the seemingly adverse consequences of adolescent childbearing are actually an artifact of the disadvantaged backgrounds from which most teenage mothers come (Geronimus, 1991; Geronimus, Korenman, & Hillemeier, 1994).1 1.3. Early parenthood and foster care Despite a precipitous decline in the U.S. teenage birth rate among all racial and ethnic groups over the past two decades (Ventura, Hamilton, & Mathews, 2014), youths in foster care continue to be at very high risk of becoming young parents. Because the federal government does not require states to report the number of youths in foster care who are parents, and most states do not have a system to track this information, much of what we know about the high rate of early parenthood among this population comes from just a handful of studies.2 The

1 Although a number of approaches have been used to address this selection bias problem, each has limitations (see Lee, 2010, for a discussion). 2 The Preventing Sex Trafficking and Strengthening Families Act of 2014 (P.L. 113–183) will require the U.S. Department of Health and Human Services to include state-level data on the number of pregnant or parenting youths in foster care in its annual report to Congress on child welfare beginning in Fiscal Year 2016.

2. Present study

1. What are the employment outcomes of youths who become parents while they are in care during the first year after they emancipate? 2. Are there gender differences in the employment outcomes of youths who become parents while they are in care? 3. What demographic or placement history characteristics are associated with better or worse employment outcomes among recently emancipated parents? We address these questions by capitalizing on the existence of a unique service delivery network in Illinois that targets the state's pregnant and parenting foster youths. Although few state or local child welfare agencies systematically track young parents in foster care, Illinois has long been an exception in this regard due, in large part, to a lawsuit (Hill vs. Erickson) filed in 1988 against the state's public child welfare agency. That lawsuit alleged that the Department of Children and Family Services (DCFS) had violated the rights of the pregnant and parenting youths in its care by systemically failing to provide them and their children with adequate services or appropriate placements. The case was certified as a class action in1989 and settled with a consent decree in 1994. As part of that consent decree, DCFS agreed to provide pregnant and parenting youths with services and placements tailored to their needs. Ulich Children's Advantage Network (UCAN), a private child welfare agency, was contracted to develop a new case management system for pregnant and parenting youths, which led to the creation of the Teen Parenting Service Network (TPSN) in 1998. TPSN is charged with providing specialized services and placements to pregnant and parenting youths in DCFS care and monitoring the provision of those services and placements by other agencies that contract with TPSN or with DCFS. These services include parenting assessment, parenting education, safe sex and healthy sexuality education, family planning and subsequent pregnancy prevention, pregnancy options

3 The researchers speculated that a lack of affordable or reliable childcare may prevent young mothers from seeking or accepting employment, whereas child support obligations may have motivated young fathers, a majority of whom were noncustodial parents, to work.

Please cite this article as: Dworsky, A., & Gitlow, E., Employment outcomes of young parents who age out of foster care, Children and Youth Services Review (2016), http://dx.doi.org/10.1016/j.childyouth.2016.09.032

A. Dworsky, E. Gitlow / Children and Youth Services Review xxx (2016) xxx–xxx

counseling,4 prenatal care, developmental screenings and early education for children, children's medical/dental care, and outreach to nonward parents and partners. TPSN youths also receive a variety of home- and community-based services including individual, group, and family therapy and educational supports. TPSN does not provide pregnant or parenting youths with vocational training, job placement, or other employment-related services. However, the TPSN Educational Support team works to ensure that all of the youths it serves have an opportunity to participate in activities that promote employment and/or educational attainment. For more than a decade, only pregnant and parenting youths in Cook County (Chicago) and the surrounding collar counties (i.e., DuPage, Kane, Lake, McHenry, and Will) were included under the TPSN umbrella. However, as a result of a 2009 Supplement to the Hill Consent Decree, TPSN monitoring and service provision were expanded in January 2010 to include all pregnant and parenting youths in DCFS care regardless of the county in which they live. 3. Methods 3.1. Data sources The data for this study came from three sources. 3.1.1. Teen Parenting Service Network (TPSN) TPSN's Child and Family Profile database includes information about all the pregnant and parenting youths under the TPSN umbrella, as well as the children of those youths.5 Because each record included the TPSN client's DCFS case ID, it was possible to link the TPSN data to DCFS administrative data. 3.1.2. Illinois Department of Children and Family Services The Child and Youth Centered Information System (CYCIS) is a DCFS database that tracks the provision of child welfare services, including out-of-home care. For every out-of-home care placement a child experiences, CYCIS contains the entry and exit dates; the reason the child was placed in care (e.g., abuse, neglect, or dependency); the child's permanency goal (e.g., reunification, adoption); the type(s) of care in which the child was placed (e.g., foster home, relative home); the start and end dates of each placement; and the child's discharge outcome (e.g., reunification, adoption, subsidized guardianship). 3.1.3. Illinois Department of Employment Security (IDES) Unemployment Insurance (UI) is a federal-state program jointly financed through federal and state employer payroll taxes. Each state administers its own program, which is overseen by the U.S. Department of Labor. Because employer payroll taxes are based on the amount of wages paid to employees, employers are required to submit quarterly wage records for all employees covered under the state's Unemployment Insurance laws to the agency that administers their state's UI program. These data can be used to determine whether an individual was employed in a given quarter as well as their total wages for those quarters in which they were employed. 3.2. Sample UCAN provided the Principal Investigator with a file containing individual-level records for 1943 parents, 1591 mothers, and 352 fathers who emancipated from DCFS care between July 1, 2004, and June 30,

4

Options counseling refers to counseling to help a pregnant woman decide whether to carry the pregnancy to term, have an abortion, or place their child for adoption. 5 Prior to TPSN's statewide expansion, the database included records only for pregnant and parenting youths in Cook County and the collar counties. Beginning in 2010, it included records for pregnant and parenting youths throughout the state.

3

2013.6 We used their CYCIS IDs to link these records to the CYCIS data that provided information about their demographic and placement history characteristics (Table 1). This population of parents is predominantly African American. Eighty percent exited from DCFS care in Cook County, although Cook County currently accounts for only 40% of the state's child welfare caseload (Illinois Department of Children and Family Services, 2016). This is because TPSN covered only Cook County and the surrounding collar counties prior to its statewide expansion in 2010. Thirty-five percent of these parents were at least 19 years old when they gave birth to or fathered their first child, but 29% became parents for the first time before age 17. A third of these parents had given birth to or fathered more than one child by the time they exited DCFS care. The mean age at exit among these parents was 20.6 years old. This reflects the fact that Illinois was one of the few states in which youths could remain in foster care until their 21st birthday even before the Fostering Connections to Success Act of 2008 extended eligibility for Title IV-E foster care (for which states can claim partial federal reimbursement) to age 21 beginning in October 2010.7 These youths have experienced considerable instability since their initial entry into care. Almost three quarters experienced five or more placements before their 18th birthday (mean = 8) and just under half had been “on run” for at least 24 h two or more times (mean = 4). Their mean age at (most recent) entry into care was 10 years old. However, a quarter of these youths had entered care more than once. Although their placement histories were generally similar, the young mothers and the young fathers differed in two important respects. First, on average, mothers were almost a year younger than fathers when their first child was born. Second, fathers were almost twice as likely as mothers to have spent time in a juvenile detention or correctional facility while they were in care. 3.3. Outcome measures A file containing CYCIS IDs and Social Security numbers (SSNs) for the TPSN parents was provided to IDES. IDES matched the SSNs against quarterly wage records covering an 11-year period beginning with the first quarter of 2004 and ending with the fourth quarter of 2014, and provided the Principal Investigator with the results of that match. These data were used to measure three main outcomes: (1) whether youths were employed in the first four quarters after the quarter in which they exited care8; (2) the number of quarters in which youths were employed if they were employed in at least one of those four quarters; and (3) total earnings for those four quarters if youths were employed in at least one. Earnings were converted to 2014 constant dollars using the Consumer Price Index (CPI). 3.4. Analytic approach We examined the employment outcomes of the 1943 TPSN parents both overall and separately by gender using univariate and bivariate statistics. We also estimated two pairs of multivariate models to examine the relationship between two of our outcome measures and an array of demographic and placement history characteristics among the 1943 parents who emancipated from care between July 1, 2004, and June 30, 2013. The first was a pair of logistic regression models that predicted whether parents were employed in at least one quarter during the first four quarters after the quarter in which they exited care. The second was a pair of ordinary least squares (OLS) regression models that 6 Although the percentage of TPSN parents who are fathers more than tripled between FY 1999 when it was 5.4% and FY 2014 when it was 18.7%, some unknown number of fathers in DCFS care do not receive parenting services. 7 The state's plan to extend Title IV-E foster care was approved in October 2010. 8 Parents were counted as employed in a given quarter only if they had earned at least $100 during that quarter.

Please cite this article as: Dworsky, A., & Gitlow, E., Employment outcomes of young parents who age out of foster care, Children and Youth Services Review (2016), http://dx.doi.org/10.1016/j.childyouth.2016.09.032

4

A. Dworsky, E. Gitlow / Children and Youth Services Review xxx (2016) xxx–xxx

Table 1 Demographic and placement history characteristics of TPSN parents by gender.

Gender Female Male Race or ethnicity African American White Latino or Hispanic Other/Missing Age at first birth in years 14 or younger 15 16 17 18 19 20 Mean age at first birth Number of children at exit 1 2 3 or more Mean number of children at exit Exit cohort 2004–2007 2008–2013 DCFS Region Cook County Other Total number of spells One More than one Mean number of spells Age at most recent entry in years 0 to 5 6 to 11 12 to 14 15 to 17 Mean age at most recent entry Number of placements prior to age 18a 1 or 2 3 or 4 5 or 6 7 or 8 9 or 10 More than 10 Mean number of placements Number of times on run prior to age 18 Never ran away Ran away once Ran away more than once Mean number of times ran away Dually involved Yes No Age at exit in years 18 or 19 20 21 Mean age at exit

Table 2 Covariates.

Total

Females

Males

Covariate

N = 1943

n = 1591

n = 352

#

#

#

Gender Race/ethnicity

%

%

%

1591 81.9 1591 100.0 0 0.0 352 18.1 0 0.0 352 100.0 1676 137 115 14

86.3 7.1 5.9 0.8

1362 117 100 12

85.6 7.4 6.3 0.8

314 20 15 3

89.2 5.7 4.3 0.9

114 190 248 332 376 402 281 17.53

5.9 9.8 12.8 17.1 19.4 20.7 14.5

108 171 217 289 288 306 212 17.39

6.8 10.8 13.6 18.2 18.1 19.2 13.3

6 1.7 19 5.4 31 8.8 43 12.2 88 25.0 96 27.3 69 19.6 18.14

1292 66.5 1026 64.5 503 25.9 438 27.5 148 7.6 127 8.0 1.43 1.45

266 75.6 65 18.5 21 6.0 1.32

1555 80.0 1259 79.1 388 20.0 332 20.9

296 84.1 56 15.9

1467 75.5 1198 75.3 476 24.5 393 24.7 1.31 1.31

269 76.4 83 23.6 1.28

369 683 448 443 10.22

19.0 35.2 23.1 22.8

81 133 74 64 9.49

23.0 37.8 21.0 18.2

211 312 364 278 241 537 8.27

10.9 16.1 18.7 14.3 12.4 27.6

11.9 11.1 21.0 13.9 14.8 27.3

720 278 945 4.37

*

* *

*

288 450 374 379 10.38

18.1 34.6 23.5 23.8

169 273 290 229 189 441 8.27

10.6 17.2 18.2 14.4 11.9 27.7

42 39 74 49 52 96 8.29

37.1 593 14.3 221 48.6 777 4.48

37.3 13.9 48.8

127 36.1 57 16.2 168 47.7 3.89

568 29.2 355 22.3 1375 70.8 1236 77.7

213 60.5 139 39.5

11 0.6 749 38.6 1183 60.8 20.60

4 1.1 137 38.9 211 59.9 20.59

Description

Dichotomous variable coded 1 if the parent is female Dichotomous variable coded 1 if the parent is African American Age at first birth Continuous variable representing age in years when the parent's first child was born Number of children at Continuous variable representing the number of exit children the parent had given birth to or fathered prior to exiting care Exit cohort Dichotomous variable coded 1 if the parent exited care between 2004 and 2007 (prior to the recession) DCFS region Dichotomous variable coded 1 if the parent exited care in Cook County Number of runaway Continuous variable representing the total number of episodes per month times parent were absent from their placement without permission for at least one night prior to age 18 divided by the total number of months in care Number of placements Continuous variable representing the total number of per month placements the parent experienced prior to age 18 divided by the total number of months in care. Placements interrupted by non-placement events (e.g., runaway episodes, detentions/incarcerations, or hospitalizations) were not counted twice. Dually involved Dichotomous variable coded 1 if the parent was in detention or incarcerated while in care Number of months in Continuous variable representing the total number of care months the parent was in care across all out-of-home care spells Employed in year prior to Dichotomous variable coded 1 if the parent was exit employed in at least one quarter during the four quarters prior to exit

covariates listed in the table except employment during the year prior to exit. The second included all the covariates plus the measure of prior employment. *

4. Results 4.1. Univariate and bivariate results Table 3 shows the percentage of TPSN parents who were employed at any point during the first four quarters after the quarter in which they exited as well as the number of quarters in which they had any earnings if they were ever employed in any of those four quarters. Only half of the parents were ever employed during the first four quarters after the quarter in which they exited, and most of those who did work were not consistently employed. Parents who were ever *

7 0.5 612 38.5 972 61.1 20.60

predicted total earnings from employment during those four quarters among parents who were employed in at least one quarter. Because the earnings variable was positively skewed, we used a log transformation of earnings as the dependent variable. Table 2 lists the covariates included in the models.9 The first model in each pair included all the 9 The models did not include age at exit because virtually all the parents were at least 20 years old when they exited care.

Table 3 Employment during the first four quarters post-exit by gender.

Ever employed Number of quarters employed 1 2 3 4 Mean number of quarters employed

Total

Females

Males

N = 1943

n = 1591

n = 352

#

%

#

%

#

%

969

49.9

819

51.5

150

42.6

224 213 174 358 2.69

23.1 22.0 18.0 37.0

179 180 146 314 2.73

21.9 22.0 17.8 38.3

45 33 28 44 2.47

30.0 22.0 18.7 29.3

p

*

Please cite this article as: Dworsky, A., & Gitlow, E., Employment outcomes of young parents who age out of foster care, Children and Youth Services Review (2016), http://dx.doi.org/10.1016/j.childyouth.2016.09.032

A. Dworsky, E. Gitlow / Children and Youth Services Review xxx (2016) xxx–xxx Table 4 Earnings during the first four quarters post-exit by gender and number of quarters worked. N Total At least one quarter 1 quarter 2 quarters 3 quarters 4 quarters Females At least one quarter 1 quarter 2 quarters 3 quarters 4 quarters Males At least one quarter 1 quarter 2 quarters 3 quarters 4 quarters

Median

Mean

Table 6 Parameter estimates from OLS regression model predicting total earnings. Model I

Standard deviation

969 224 213 174 358

4584 685.5 2177 5132.5 13,118

7310 1228 2906 6429 14,164

7765 2214 2603 4961 7888

819 179 180 146 314

4879 624 2101 5114.5 13,048.5

7443 1234 2921 6547 13,990

7708 2424 2650 5150 7646

150 45 33 28 44

3072.5 839 2355 5401 14,075

6585 1203 2820 5813 15,403

8058 1041 2370 3847 9448

Intercept Gender Race or ethnicity Age at first birth in years Number of children at exit DCFS region Exit cohort Number of placements Number of runs Dually involved Months in care Employed in year prior to exit

employed had earnings in an average of 2.7 quarters. Although the mothers were more likely than the fathers to be employed in at least one quarter, there was no gender difference in the number of quarters worked among those who were employed. Table 4 shows the median and mean annual earnings of the TPSN parents who were employed at any point during the first four quarters after the quarter in which they emancipated. Parents who were ever employed had median and mean earnings over those four quarters of $4584 and $7310, respectively. The mothers had higher earnings, on average, than the fathers, but this difference was not statistically significant. Moreover, although there was a positive relationship between total earnings and the number of quarters with any employment, total earnings were still very low, even among parents who worked in all four quarters. The median and mean earnings for those who worked in all four quarters were only $13,118 and $14,164, respectively. 4.2. Multivariate results Table 5 presents the odds ratios from the pair of logistic regression models predicting employment. The odds ratios from Model I indicate

5

Model II

Beta

S.E.

p

Beta

S.E.

p

10.662 −0.034 −0.208 −0.081

0.698 0.141 0.137 0.032

b0.001 0.811 0.129 0.012 *

10.266 −0.025 −0.089 −0.082

0.612 0.122 0.118 0.028

b0.001 0.837 0.454 0.003 *

−0.206 0.085 0.016

* −0.199 0.074 0.007

*

−0.463 0.033 −0.300 −1.696 −0.661 −0.001

* −0.345 −0.038 −0.636 * −1.570 * −0.591 * −0.002 0.653

*

0.116 0.100 0.466 0.513 0.127 0.001

b0.001 0.743 0.520 0.001 b0.001 0.189

0.101 0.087 0.398 0.445 0.110 0.001 0.093

0.001 0.664 0.111 0.000 b0.001 0.036 b0.001

* * * *

that being African American,10 having more children at exit, exiting in Cook County, running away more frequently, and being involved in both the child welfare and juvenile justice systems were associated with a decrease in the odds of being employed, whereas exiting care prior to 2008 was associated with an increase in the odds of being employed. The odds ratios from Model II indicate that being employed prior to exit more than quadrupled the odds of post-exit employment. Although the odds ratios for the other covariates are similar to the odds ratios from Model I, three of the covariates that were statistically significant in the first model (i.e., being African American, having more children at exit, and number of runaway episodes) were no longer statistically significant once prior employment was added to the model. Table 6 presents the parameter estimates from the pair of OLS regression models predicting earnings. The coefficients from Model I indicate that that being younger at first birth, having more children at exit, exiting in Cook County, running away more frequently, being involved in both the child welfare and juvenile justice systems, and spending more months in care prior to age 18 were associated with lower earnings among parents who were employed. The coefficients from Model II indicate that being employed prior to exit was associated with higher post-exit earnings. All of the covariates that were statistically significant in the first model remained statistically significant even after prior employment was added.11

5. Limitations

Table 5 Parameter estimates from the logistic regression models predicting employment. Model I

Gender Race or ethnicity Age at first birth in years Number of children at exit DCFS region Exit cohort Number of placements Number of runaway episodes Dually involved Months in care Employed in year prior to exit

Model II

Odds Ratio

S.E.

p

Odds Ratio

1.061 0.657 0.972

0.133 0.655 0.141 0.003 0.031 0.346

1.011 * 0.762 0.990

0.141 0.939 0.150 0.071 0.032 0.758

0.799

0.079 0.004

* 0.875

0.083 0.109

0.623 1.483 0.933

0.120 b0.001 * 0.740 0.098 b0.001 * 1.414 0.493 0.888 0.657

0.128 0.019 0.104 0.001 0.521 0.420

0.266

0.418 0.002

0.437 0.120

0.517 1.000

0.114 b0.001 * 0.524 0.001 0.716 1.000 4.353

* 0.507

S.E.

p

+

* *

0.121 b0.001 * 0.001 0.970 0.101 b0.001 *

Before turning to a discussion of the study's major findings, it is important to acknowledge a number of the study's limitations. First, our exclusive reliance on administrative data restricted the range of predictors that we could include in our multivariate analysis to demographic and placement history characteristics that could be measured using the CYCIS data. By contrast, studies that have used survey data to identify predictors of employment outcomes among former foster youths have generally been able to include a much broader range of covariates such as measures of mental health or substance use disorders that could adversely affect employment. Particularly noticeable was the absence of data on educational attainment, one of the most consistent predictors of employment and earnings. Completing high school, attending college, and earning a college degree are each associated with an increased likelihood of being 10 The results were essentially unchanged when the analysis was limited to the 1591 mothers. However, the parameter estimate for being African American was only marginally significant when fathers were omitted from the analysis. 11 The results were essentially unchanged when the analysis was limited to the 1591 mothers. However, the parameter estimate for number of children was only marginally significant once employment prior to emancipation was added to the model.

Please cite this article as: Dworsky, A., & Gitlow, E., Employment outcomes of young parents who age out of foster care, Children and Youth Services Review (2016), http://dx.doi.org/10.1016/j.childyouth.2016.09.032

6

A. Dworsky, E. Gitlow / Children and Youth Services Review xxx (2016) xxx–xxx

employed and with higher wages (Bloom, 2010; Danziger & Ratner, 2010) in the general population. In fact, studies using survey data have found that educational attainment is positively associated with employment and earnings among former foster youths (Hook & Courtney, 2011; Naccarato et al., 2010). Unfortunately, the CYCIS data to which we had access do not include data on educational attainment, and the Illinois State Board of Education, the state education agency, will not provide researchers with individual-level data on youths in foster care due to concerns related to the Family Educational Rights and Privacy Act (FERPA). Second, among the potential advantages of using UI wage data to measure employment and earnings are their lower cost and greater accuracy than self-report data. However, using UI wage data does have some disadvantages. First, although over 99% of all U.S. workers and 89% of the civilian labor force is engaged in “covered” employment (where coverage is defined by the Federal Unemployment Tax Act), the earnings of individuals who are self-employed, certain agricultural labor and domestic service workers, military and federal civilian employees, and railroad workers are not captured in the UI data. Second, because the UI data do not indicate start or termination dates, it is not possible to determine if an individual was employed throughout an entire quarter or only some portion of it. Third, because total wages paid are reported but total hours worked are not, UI data cannot be used to calculate an hourly wage or to distinguish between full-time and parttime employment. Finally, because each state maintains its own UI database, the earnings of individuals who move to or work in another state cannot be observed (Feldbaum & Harmon, 2012; Hotz & Scholz, 2001; Kornfeld & Bloom, 1999). Third, an unknown number of fathers in DCFS care do not receive parenting services through TPSN. Some of these young men do not acknowledge being fathers. Others acknowledge being fathers but are not eligible for TPSN services because they are unwilling to share information about their child and the child's mother. Finally, fathers in DCFS care are generally not living with their children and many caseworkers are unaware that noncustodial parents are eligible for TPSN services. We do not know if or how the employment outcomes of the 352 fathers for whom we have data might differ from the employment outcomes of the fathers who were not served by TPSN.12 Fourth, although our multivariate analysis identified a number of demographic and placement history characteristics associated with more or less favorable employment outcomes, we cannot interpret these associations as evidence of causal relationships. Any of these associations could potentially be explained by unobserved or omitted variables for which our analysis did not control. Finally, our study focused exclusively on youths in a single state with a comprehensive service network for all pregnant or parenting youths in foster care. Although TPSN does not provide parenting youths with vocational training, job placement, or other employment-related services, it is possible that young parents experience even poorer employment outcomes in states without a wide array of specialized services and supports. Alternatively, some states may do a better job of preparing all youths in foster care, including those who are parents, for employment. 6. Discussion Despite a high rate of early parenthood among youths in foster care, relatively little is known about the employment outcomes of young parents after they age out. To address these gaps in our knowledge, we analyzed administrative data for 1943 youths who were the parent of at least one child when they aged out of care. The results of our analyses are consistent with the picture of generally poor employment outcomes 12 Because the percentage of TPSN parents who are fathers more than tripled between FY 1999 and FY 2014, the number of fathers in DCFS care for whom we are missing data probably declined over time.

among former foster youths. Only half of the parents were employed at any point during the first four quarters after the quarter in which they exited, and most of those who ever worked were not consistently employed. This certainly contributed to the very low earnings of parents who were ever employed. Their median and mean earnings were $4584 and $7310, respectively. However, the median and mean earnings of parents who worked in all four quarters were still very low ($13,118 and $14,164, respectively). This is considerably lower than the poverty threshold for a single parent with one child ($16,317) in 2014 (U.S. Census Bureau, 2015). Although mothers were more likely to have been employed in at least one quarter than fathers, there were no gender differences in either the number of quarters worked or total earnings among parents who were employed. Moreover, gender had no effect on the odds of being employed once we controlled for placement history and other demographic characteristics. Several other findings from our multivariate analysis are also worth noting. First, although parent age at birth of their first child had no effect on the odds of being employed, being younger at first birth was associated with lower earnings. Moreover, parents who had given birth to or fathered more children by the time they exited care were less likely to be employed and had lower earnings, all other things being equal. Second, consistent with a number of prior studies (Dworsky, 2005; Dworsky et al., 2010; Hook & Courtney, 2011), African American parents were less likely to be employed than their non-African American counterparts. However, unlike some of those prior studies (Dworsky et al., 2010; Goerge et al., 2002; Naccarato et al., 2010), we found no differences in earnings between African American and non-African American parents among those who were employed.13 Third, because frequent placement changes have the potential to impede the development of relationships with adults who could connect youths with employment and reduce opportunities for youths to acquire work experience, we expected placement instability would be associated with less favorable employment outcomes. However, we found no relationship between employment outcomes and the number of placement changes youths experienced prior to age 18. This null finding is not totally surprising; prior analyses of the relationship between employment outcomes and placement instability have yielded inconsistent results (Goerge et al., 2002; Hook & Courtney, 2011; Macomber et al., 2008). Fourth, we did find relationships between employment outcomes and two of the other placement history characteristics included in our models. Both running away more frequently while in foster care and being involved in both child welfare and juvenile justice systems were associated with a reduction in the odds of being employed and with lower earnings.14 It is not clear why running away more frequently, which often leads to placement changes, was consistently associated with poorer employment outcomes but the number of placement changes was not. One possibility is that frequently running away is a proxy for some unmeasured characteristic that is not associated with placement instability per se. However, the negative relationship between dual involvement and employment outcomes is consistent with the results of prior studies, which have found that having a history of delinquency or a felony conviction reduces young men's likelihood of being employed (Caspi, Entner Wright, Moffitt, & Silva, 1998; Pager, 2003; Raphael, 2007). Given that fathers were more likely than mothers to be involved in both the child welfare and juvenile justice systems, it makes sense that mothers were more likely to be employed than fathers. However, there was no gender difference in employment when

13 Hook and Courtney (2011) also found that being African American was associated with a reduction in the odds of being employed but not with lower earnings. 14 The parameter estimate for running away from care in the model predicting employment was in the expected direction but no longer statistically significant once prior employment was added to the model.

Please cite this article as: Dworsky, A., & Gitlow, E., Employment outcomes of young parents who age out of foster care, Children and Youth Services Review (2016), http://dx.doi.org/10.1016/j.childyouth.2016.09.032

A. Dworsky, E. Gitlow / Children and Youth Services Review xxx (2016) xxx–xxx

we estimated the multivariate model that controlled for placement history characteristics. Finally, parents who had worked during the year prior to exit were over four times more likely to be employed after exiting care and earned significantly more if they were employed than their peers who had not worked during that year even after controlling for differences in demographic or placement history. This finding is consistent with the results of earlier studies that found a positive relationship between employment prior to age 18 and employment post-discharge (Goerge et al., 2002; Macomber et al., 2008). 6.1. Policy and practice implications The results of this study have a number of policy and practice implications. First, like the results of prior studies, our findings suggest that far too many youths aging out of care are not receiving the education, training, and other services needed to prepare them for employment. Consistent with this conclusion, data from the National Youth in Transition Database (NYTD) indicate that more than half of the youths who received independent living services paid for or provided by the state agency that administers the Chafee Foster Care Independence Program in federal fiscal years 2011 to 2013 did not receive any career preparation or employment training services (U.S. Department of Health and Human Services, 2014). Better coordination between the child welfare system and other state or local agencies and not-for-profit organizations that can address the employment-related services needs of young parents as well as other youths in foster care could reduce the number who exit care without the knowledge or skills required for labor market success. These public agencies and not-for-profit organizations could also provide help securing employment so that youths in foster care, including young parents, have both work experience and jobs when they emancipate. Second, although the administrative data we analyzed provide no information about the reasons parents were not employed, another recent study suggests that a lack of childcare is often a contributing factor (Dworsky & Gitlow, 2015). Hence, although neither the Fostering Connections Act nor other federal child welfare policies address the unique challenges faced by parenting youths in foster care, it is imperative that child welfare agencies help connect young parents with childcare and other supports they need to balance the competing demands of work and parenting. And third, the consistent relationship we found between poor employment outcomes and a history of detention or incarceration suggests that child welfare and juvenile or criminal justice systems should collaborate more closely to address the employment needs of dually involved youths. It also draws attention to the importance of interventions aimed at preventing justice system involvement among youths, and especially young men, in foster care. 6.2. Future research There are a number of questions about the employment outcomes of youths who become parents while still in foster care that future research should address. First, our analysis was limited to employment and earnings during the first year after exiting. Subsequent studies should examine changes in employment outcomes over time, using growth curve models, for example. It would also be important to determine whether the difference in employment we observed between parents and non-parents persists beyond the first year. Second, future studies should also compare the employment outcomes of youths who became parents while in foster care to those of their peers who either delayed parenthood or did not become parents. In making those comparisons, it will be important to control for other differences between parenting foster youths and non-parenting foster youths that might contribute to differences in their employment outcomes.

7

Third, more than a decade and a half after the Chafee Foster Care Independence Program was created, there is relatively little evidence that the employment-related services youths receive lead to better employment outcomes (Edelstein & Lowenstein, 2014; Montgomery, Donkoh, & Underhill, 2006). Only one employment program targeting this population has been rigorously evaluated, and that evaluation found no positive impacts on employment or other key outcomes (Courtney, Zinn, Koralek, & Bess, 2011). Our findings clearly highlight the need for rigorous evaluations of programs aimed at improving the employment outcomes of youths in foster care, including programs that target young parents. Fourth, we had no information about circumstances such as school enrollment or incarceration that might explain why some of the youths in our study were not employed (or why the earnings of some who were employed were very low). Knowing more about the circumstances of these youths is important because they have very different implications for their future employment and earnings. Hence, future studies that use administrative data to examine the employment outcomes of former foster youths should link child welfare data not only to data from state Unemployment Insurance agencies but also to data from other sources. For example, data from the National Student Clearinghouse could be used to identify former foster youths who are not employed because they are enrolled in college. Likewise, criminal justice system data could be used to identify former foster youths who are not employed because they are incarcerated. Finally, the experiences of youths who become parents while in care in Illinois may be quite different from the experiences of youths who become parents while in care in other states due to the existence of the Teen Parenting Service Network. Thus, research on the employment outcomes of young parents who aged out of foster care in other states is needed to determine whether these findings generalize beyond Illinois. 7. Conclusion This study contributes to our knowledge about the employment outcomes of youths who become parents while in foster care. The high rate of early parenthood among youths in foster care means that greater attention should be paid to the outcomes of young parents after they age out in other major life domains such as housing, education, and psychosocial functioning. References An, C., Haveman, R., & Wolfe, B. (1993). Teen out-of-wedlock births and welfare receipt: The role of childhood events and economic circumstances. The Review of Economics and Statistics, 75, 195–208. Ashcraft, A. & Lang, K. (2006). The consequences of teenage childbearing (Working paper #12485). The National Bureau of Economic Research. Cambridge MA: NBER. Baum, S., Ma, J., & Payea, K. (2013). Education pays 2013: The benefits of higher education for individuals and society. New York City: The College Board. College Board. Becker, G. (1993). Human capital: A theoretical and empirical analysis, with special reference to education. Chicago: University of Chicago Press. Bird, B., & Bryant, R. (2015). Investing in boys and young men of color: The promise and opportunity: Issue brief: Focus on employment. Washington, DC: Center for Law and Social Policy. Bloom, D. (2010). Programs and policies to assist high school dropouts in the transition to adulthood. The Future of Children, 20, 89–108. Bureau of Labor Statistics (2015). Employment and unemployment among youth summary, economic news release, august 2015. Washington, DC: Bureau of Labor Statistics. Caspi, A., Entner Wright, B., Moffitt, T., & Silva, P. (1998). Early failure in the labor market: Childhood and adolescent predictors of unemployment in the transition to adulthood. American Sociological Review, 63, 424–451. 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: Chapin Hall at the University of Chicago. Courtney, M., Dworsky, A., Brown, A., Cary, C., Love, K., & Vorhies, V. (2011). Midwest evaluation of the adult functioning of former Foster youth: Outcomes at ages 26. Chicago, IL: Chapin Hall at the University of Chicago. Courtney, M., Piliavin, I., Grogan-Kaylor, A., & Nesmith, A. (2001). Foster youth in transitions to adulthood: A longitudinal view of youth leaving care. Child Welfare, 80, 685–717.

Please cite this article as: Dworsky, A., & Gitlow, E., Employment outcomes of young parents who age out of foster care, Children and Youth Services Review (2016), http://dx.doi.org/10.1016/j.childyouth.2016.09.032

8

A. Dworsky, E. Gitlow / Children and Youth Services Review xxx (2016) xxx–xxx

Courtney, M., Zinn, A., Koralek, R., & Bess, R. (2011). Evaluation of the independent living — Employment services program, Kern County, California: Final report. OPRE report # 2011–13. Washington, DC: Office of Planning, Research and Evaluation, Administration for Children and Families, U.S. Department of Health and Human Services. Danziger, S., & Ratner, D. (2010). Labor market outcomes and the transition to adulthood. The Future of Children, 20, 133–158. Dworsky, A. (2005). The economic self-sufficiency of Wisconsin's former foster youth. Children and Youth Services Review, 27, 1085–1118. Dworsky, A., & Courtney, M. (2010). The risk of teenage pregnancy among transitioning foster youth: Implications for extending state care beyond age 18. Children and Youth Services Review, 32, 1351–1356. Dworsky, A., & Gitlow, E. (2015). What happens to young parents who age out of foster care? Results from the TPSN outcomes study. Chicago: Chapin Hall at the University of Chicago. Dworsky, A., Roller White, C., O'Brien, K., Pecora, P., Courtney, M. E., Kessler, R., ... Hwang, I. (2010). Racial and ethnic differences in the outcomes of former foster youth. Children and Youth Services Review, 32, 902–912. Edelstein, S., & Lowenstein, C. (2014). Supporting youth transitioning out of Foster Care, issue brief 3: Employment programs. OPRE report # 2014–70. Washington, D.C.: Office of Planning, Research and Evaluation, Administration for Children and Families, U.S. Department of Health and Human Service. Feldbaum, M., & Harmon, T. (2012). Using unemployment insurance wage data to improve program employment outcomes: A technical assistance guide for community and technical colleges. Silver Spring, MD: The Collaboratory and FHI360. Fletcher, J., & Wolfe, B. (2009). Education and labor market consequences of teenage childbearing: Evidence using the timing of pregnancy outcomes and community fixed effects. Journal of Human Resources, 44(2), 303–325. Geronimus, A. (1991). Teenage childbearing and social and reproductive disadvantage: The evolution of complex questions and the demise of simple answers. Family Relations, 40(4), 463–471. Geronimus, A., Korenman, S., & Hillemeier, M. M. (1994). Does young maternal age adversely affect child development? Evidence from cousin comparisons in the United States. Population and Development Review, 20(3), 585–609. Goerge, R., Bilaver, L., Lee, B., Needell, B., Brookhad, A., & Jackman, W. (2002). Employment outcomes for youth aging out of foster care. Washington, DC: U.S. Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation. Hofferth, S., & Hayes, C. (1987). Risking the future: Adolescent sexuality, pregnancy and childbearing. Volume. II. Washington, DC: National Academies Press. Hook, J., & Courtney, M. (2011). Employment outcomes of former foster youth as young adults: The importance of human, personal, and social capital. Children and Youth Services Review, 33, 1855–1865. Hotz, V., & Scholz, J. (2001). Measuring employment and income for low-income populations with administrative and survey data. Discussion paper no. 1224–01. Madison, WI: Institute for Research on Poverty.

Illinois Department of Children and Family Services (2016). Executive statistical summary report for 2015. Springfield, IL: Illinois Department of Children and Family Services. Kornfeld, R., & Bloom, H. (1999). Measuring program impacts on earnings and employment: Do unemployment insurance wage reports from employers agree with surveys of individuals? Journal of Labor Economics, 17, 168–197. Lee, D. (2010). The early socioeconomic effects of teenage childbearing: A propensity score matching approach. Demographic Research, 23, 697–736 (6). Levine, D. I., & Painter, G. (2003). The schooling costs of teenage out-of-wedlock childbearing: Analysis with a within-school propensity score matching estimator. The Review of Economics and Statistics, 85, 884–900. Macomber, J., Cuccaro-Alamin, S., Duncan, D., Kuehn, D., McDaniel, M., Vericker, T., ... Barth, R. (2008). Coming of age: Employment outcomes for youth who age out of foster care through their middle twenties. Washington, DC: Urban Institute. Montgomery, P., Donkoh, C., & Underhill, K. (2006). Independent living programs for young people leaving the care system: The state of the evidence. Children and Youth Services Review, 28(12), 1435–1448. Naccarato, T., Brophy, M., & Courtney, M. E. (2010). Employment outcomes of foster youth: The results from the Midwest evaluation of the adult functioning of foster youth. Children and Youth Services Review, 31, 551–559. Office of the President of the United States (2015). Economic costs of youth disadvantage and high-return opportunities for change. Washington, DC: Office of the President of the United States. Pager, D. (2003). The mark of a criminal record. American Journal of Sociology, 108, 937–975. Perper, K., Peterson, K., & Manlove, J. (2010). Diploma attachment among teen mothers. Washington, DC: Child Trends. Putnam-Horstein, E., Cederbaum, J., King, B., & Needell, B. (2013). California's most vulnerable parents: When maltreated children have children. Agoura Hills, CA: Conrad Hilton Foundation. Raphael, S. (2007). Early incarceration spells and the transition to adulthood. In S. Danziger, & C. Rouse (Eds.), The price of independence: The economics of early adulthood. New York: Russell Sage. Singer, A. (2006). Assessing outcomes of youth transitioning from foster care. Salt Lake City, UT: Department of Human Services. U.S. Census Bureau (2015). Poverty thresholds by size of family and number of children. http://www.census.gov/hhes/www/poverty/data/threshld/index.html U.S. Department of Health and Human Services, Administration on Children, Youth and Families, Administration for Children and Families, Children's Bureau (2014o). Highlights from the National Youth in Transition Database (NYTD), Federal Fiscal Year 2013. Washington, DC: Children's Bureau. Ventura, S., Hamilton, B., & Mathews, T. (2014). National and state patterns of teen births in the United States, 1940–2013. National Vital Statistics Reports, 63. (pp. 4). Hyattsville, MD: National Center for Health Statistics, 4.

Please cite this article as: Dworsky, A., & Gitlow, E., Employment outcomes of young parents who age out of foster care, Children and Youth Services Review (2016), http://dx.doi.org/10.1016/j.childyouth.2016.09.032