Housing Instability and Academic Skills: Exploring the Value of a Cumulative Risk Approach

• In Event: Students Experiencing Homelessness: Key Findings for Practice and Policy Decision Making

Abstract

Homelessness, eviction, and frequent moves can negatively impact children’s healthy development (Theodos et al., 2018). However, studies often use a single “flag” of homelessness and high mobility from school records to examine impacts on child outcomes (e.g., Obradović et al., 2009). This approach quickly characterizes those students who are at-risk for academic difficulties, but does not fully capture families’ housing challenges. Research on the consequences of housing instability could benefit from a cumulative risk approach (Wright & Masten, 2015), which takes into account multiple housing risks and occurrences of these risks over time. In the current study, we explored the predictive value of a single dichotomous risk indicator and a cumulative risk indicator of housing instability on children’s academic outcomes.
We used data from the Fragile Families and Child Wellbeing Study, a longitudinal birth cohort study. We included Year 1, 3, 5, and 9 data from 2,121 children who had at least one academic outcome at Year 9 and lived with their biological mother. At each wave, mothers reported on multiple experiences of housing instability: (1) literal homelessness (e.g., stay in shelter); (2) doubled-up with others due to financial problems; (3) eviction; and (4) moved more than once per year. We created two housing instability variables. The dichotomous risk indicator reflected whether or not a child had ever experienced any housing instability. The cumulative risk indicator summed all experiences of housing instability across each wave (range = 0–10). Only 5% of the sample had 4 or more experiences, so we truncated the range to 4+. At Year 9, teachers reported on both children’s literacy and math skills on a 5-point scale. We also included a number of baseline control measures that are confounded with housing instability (e.g., household income, maternal education, relationship status, etc.).
Preliminary analyses explored teacher-reported literacy and math skills by each risk indicator (see Figure 4.1). An independent samples t-test with the dichotomous risk indicator showed the those who had any experience of instability had lower literacy (t(2109) = 3.73, p < .001) and math (t(2104) = 4.22, p < .001) skills compared to those who had no experiences. In comparison, a one-way ANOVA with the cumulative risk indicator showed that only children who had 4+ experiences of instability had significantly lower literacy skills compared to those with no experiences (Mdiff = -.30, p < .05). For math skills, children with 2 and 4+ experiences had significantly lower scores than those with no experiences of instability (Mdiff = -.23, p < .05; Mdiff = -.33, p < .05, respectively). Those with only one experience of instability did not significantly differ from those with no experiences, highlighting the potential value of a cumulative risk approach. The main analyses regressed each academic outcome on one of the housing risk indicators and a set of baseline control variables. Neither the dichotomous nor cumulative risk indicator significantly predicted math or literacy, suggesting that both risk approaches produce similar results when accounting for other risk factors in children’s lives.

Authors

• Rebecca Distefano, Roger Williams University

• Jeanne Brooks-Gunn, Teachers College, Columbia University