Who Moving Does (or Doesn't) Hurt: Residential Mobility Among U.S. Children

Residential mobility is a common event in many children’s lives, with approximately 15 percent of children moving in 2011, and it is consistently linked with a range of adverse outcomes (e.g., Jelleyman & Spencer; Leventhal & Newman, 2010); however, it remains unclear if associations vary over the course of childhood and for particular subgroups of children.  Moreover, the majority of the extant research suffers from the threat of selection bias because families have some choice as to when and where to move.  Addressing variation in associations for policy-relevant subgroups, and using rigorous methods to account for selection bias is critical for making sound and targeted policy recommendations.

Accordingly, the goal of this study is to explore associations between residential mobility and children’s achievement and behavioral outcomes.  We conducted all analyses by distinct developmental periods (early childhood, birth to 54 months; middle childhood, kindergarten to 11 years old; and early adolescence, 11 years old to 15 years old) and examined analyses stratified by gender, maternal education, family income, and parental marital status.  Finally, we employed analytic techniques designed to mitigate the threat of selection bias.

We used data from the NICHD Study of Early Child Care and Youth Development (SECCYD; NICHD Early Child Care Research Network, 1997), a birth-cohort study in 10 sites across the U.S. that followed children and their families over a 15-year period. Data were gathered across developmental periods from parent and child assessments and interviewer observations, in the home, lab, child care, and school settings for the SECCYD.

To answer our research questions, we employed propensity score matching and weighting and fixed effects analyses to examine associations between residential mobility (comparing mobile and stable children) and children’s outcomes.  All models used a diverse range of child, parent, and contextual covariates with additional variables incorporated in propensity models. 

Results from SECCYD suggest that residential mobility was not universally adversely associated with children’s development. Instead, children from disadvantaged backgrounds who moved, namely from low-income families, demonstrated lower achievement and more behavior problems than comparable children who did not move. Findings are discussed in terms of implications for programs and policies, particularly among our nation’s military, migrant, and homeless children.