Loading...
Please wait, while we are loading the content...
Similar Documents
What Explains the Relation between Foster Care and Children ’ s Academic Achievement ?
| Content Provider | Semantic Scholar |
|---|---|
| Author | Berger, Lawrence M. Cancian, Maria Aparecida Eva Noyes, Jennifer L. Ríos-Salas, Vanessa |
| Copyright Year | 2017 |
| Abstract | Children experiencing foster care placement exhibit poorer average educational outcomes than other children. However, recent evidence suggests this largely reflect pre-existing disadvantage as children appear to score worse on standardized achievement tests prior to entering care than while in care. We compare child achievement prior to and during a foster care placement lasting a full school term or more, and examine three potential mediators—school attendance, school instability, and school quality—that may explain differences in achievement in the two periods. We use linked Wisconsin administrative data on children observed in 1 337 foster care spells lasting a full school term or more. For 869 of these spells, children were also observed in-home for a full year prior to placement (total N = 2 206). We use ordinary least squares regressions with and without child-specific fixed effects to estimate associations of foster care placement with academic achievement, school attendance, school instability, and school quality. We then examine whether the latter factors mediate associations of placement with achievement. We find little evidence that foster care is associated with achievement as measured by reading and math test scores. Moreover, foster care was associated with lower absenteeism. We do not find consistent evidence suggesting that attendance, school quality or school instability mediate the association between out-of-home placement and achievement. PRELIMINARY DRAFT: Not for citation/circulation 2 INTRODUCTION Children in out-of-home placement (OHP) as a result of child protective services (CPS) involvement exhibit poorer average educational outcomes than other children (Barrat and Berliner, 2013; Fantuzzo and Perlman, 2007; Scherr, 2007; Smithgall et al., 2004; Stone, 2007). Yet, potential practice and policy responses rest on whether OHP causes poor educational experiences and outcomes or is better understood as a marker of pre-existing disadvantage and associated adverse experiences. If differences in educational outcomes between children experiencing OHP and other children reflect that OHP does not compensate for pre-existing disadvantage, then policy and practice may be better served by addressing these pre-existing factors than by focusing on the role of OHP itself. Recent research indicates that the association between OHP and poor school achievement is unlikely to be causal. Rather, differences in achievement between children experiencing OHP and economically disadvantaged children (those receiving Supplemental Nutrition Assistance Program or SNAP benefits) are considerably smaller than differences between children experiencing OHP and the general population of children. Moreover, children experiencing OHP and other CPS-involved children (who did not experience OHP) exhibit similar levels of achievement. In fact, children generally score worse on standardized achievement tests in the years prior to entering care than in the years in which they are in care (Berger et al., 2015). This latter finding raises questions regarding a potential positive relation between OHP and academic achievement in some circumstances. To begin to address this possibility, we analyze the relation between OHP and academic achievement for children experiencing a full school term (September to June) in care using newly available Wisconsin longitudinal administrative data. We compare children’s reading and PRELIMINARY DRAFT: Not for citation/circulation 3 math standardized test scores after spending an entire school term at home to their later scores after spending an entire school term in OHP. We then examine the role of three potential mediators: school attendance, school instability (number of schools attended in a school term), and school quality. Each of these factors may (1) be influenced by OHP status, (2) be associated with academic achievement, and (3) fully or partially explain associations between OHP and achievement. For example, OHP may increase school attendance if it results in a less chaotic home environment that is more conducive to consistently attending school. Increased attendance may, in turn, lead to improved achievement (Conger and Rebeck, 2001). In addition, although OHP may be associated with increased school moves if placement itself (or changes in placement) mechanically necessitates school changes, children who experience OHP already tend to be quite disadvantaged and to experience considerable residential and housing instability. Finally, to the extent that OHP results in a (or more than one) school move, it may be associated with either an increase or a decrease in school quality. These potential mechanisms have received limited attention in the existing literature linking OHP with academic achievement. However, some implications for our work can be drawn from both the OHP and child homelessness literatures. The few studies to examine school attendance have produced mixed results, suggesting that OHP is associated with increased attendance for some children—particularly those experiencing longer and more stable placements—and decreased attendance for others—particularly those experiencing shorter and unstable placements (Conger and Rebeck, 2001; Larson, 2010). Of additional relevance, recent research indicates that attendance mediates the link between child homelessness and achievement (Tobin, 2016). Turning to school stability, OHP is associated with changes in schools, particularly in a context of placement instability (Conger and Finkelstein, 2003; Fries, PRELIMINARY DRAFT: Not for citation/circulation 4 Klein, and Ballantune, 2014; Smithgall et al., 2004). Whereas poor achievement for children in OHP is widely thought to reflect such instability, recent research suggests that, among children in OHP, school stability is more closely associated with child behavior than achievement (Fries, Klein, and Ballantune, 2014; Leonard and Gudino, 2016). Finally, OHP may be linked to school quality. Recent evidence suggests that children in OHP are less likely to live in poor households than otherwise similar children (Pac, Waldfogel, and Wimer, 2017). This may imply that OHP is associated with living in lower poverty neighborhoods and, relatedly, attending higher quality schools. This appears to be the case for black and Hispanic children, but not for white children, experiencing OHP (Fries, Klein, and Ballantune, 2014). It is unclear, however, whether school quality is necessarily associated with achievement among disadvantaged children. For example, recent research finds no link between school quality and academic achievement for homeless children (Tobin, 2016). Given the relatively high frequency of CPS involvement and OHP—for example, among black children, over 1 in 5 will be confirmed for maltreatment, and over 1 in 10 will experience OHP by 18 years of age—and the fundamental public responsibility for children placed in state care, it is critical to better understand whether, how, and under what conditions OHP may be beneficial, harmful, or neutral for child development (Wildeman et al., 2014; Wildeman and Emanuel, 2014). This study begins to address these gaps by focusing on key potential mechanisms linking OHP to school achievement, which have largely been unexamined. DATA We use longitudinal linked administrative data from the Wisconsin Department of Public Instruction (DPI) and the 2014 Multi-Sample Person File (MSPF) data system housed at the Institute for Research on Poverty at the University of Wisconsin–Madison. The DPI data include PRELIMINARY DRAFT: Not for citation/circulation 5 information on academic performance, attendance and behavior, school quality, and basic demographics for all children in Wisconsin public schools from the 2005–2006 through 2013– 2014 school years. The MSPF includes linked individual-level administrative data from a host of public social welfare programs including Wisconsin’s CPS system. As such, we have data on the entire population of children in Wisconsin public schools who experienced OHP between 2007 and 2014. We examine OHP spells for Wisconsin public school students who (1) experienced an OHP spell lasting an entire 10-month school term (September through June) for the first time between 2007 and 2014 and (2) in that year, were enrolled in grades 3 through 7, when our academic performance measures—the reading and math tests from the Wisconsin Knowledge and Concepts Examinations (WKCE)—are administered. From 2 064 OHP spells in our data, we excluded 250 spells for which children were not observed in their own home during our study period, and 477 spells for children who were missing educational data during the term spent in OHP, resulting in a sample of 1 337 OHP spells. We observe pre-OHP test scores for only 869 of these spells because a large number of children were in the second grade, and therefore were not administered achievement tests, in the pre-OHP period. Thus, our analytic sample includes an unbalanced panel in which not all students were observed prior to OHP, and a smaller, balanced panel. Our primary analytic sample consists of 2 206 observations: 1 337 observations during an OHP spell lasting a full school term and 869 observations in the last year a child spent entirely in-home prior to that OHP spell. Some children experienced more than one OHP spell; as such, the sample includes 1 321 children. We cluster the standard errors to adjust for multiple observations PRELIMINARY DRAFT: Not for citation/circulation 6 We use mean imputation to replace missing values for attendance, number of schools attended during the term, and school quality. Data were missing for less than 1% of observations for each measure. We also use mean imputation to replace missing values on the control variables. Here, missing data ranged from 5% for mother’s age at birth to 16% for having a child disciplinary incident. We include indicators that a given value was |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | https://dpi.wi.gov/sites/default/files/imce/policy-budget/research_reports/OHPachievement_Mediators_APPAM_.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
