Estimating the Effects of Residential Mobility: A Methodological Note

ABSTRACT

This article reviews the methodological challenges of estimating a causal association between mobility and children's cognitive, social, emotional, and physical development. Utilizing a comprehensive set of empirical articles published in the past 25 years that employ quantitative methods, it describes the limitations of previous studies and the innovative ways that researchers have attempted to deal with them. The concept of mobility is inconsistently operationalized along four dimensions: school versus residential, distance, timing, and frequency. Imprecise operationalization conflates different forms of mobility, which have differential effects on development. Attempts to estimate a causal association between mobility and development suffer from three sources of bias: selection, contextual shifts, and contemporaneous instigating events. (a) Methods that account for unobserved differences between mobile and nonmobile children have consistently shown smaller or even positive effects of mobility. (b) Moving can have a positive or negative effect on children's ecological contexts in ways that are systematically correlated with child development. (c) Moves are frequently catalyzed by changes in family structure and employment. The article concludes with recommendations for future research. Researchers should continue to engage fixed- and random-effect, matching, and instrumental variable techniques, each of which makes the question of causality explicit.

KEYWORDS:

• Mobility
• child development
• methodology
• causality

Funding

The authors gratefully acknowledge support for this work from the Annie E. Casey Foundation.

Notes

1. Given statistical power limitations, an increasingly insignificant and/or smaller coefficient does not necessarily suggest that an effect is zero, only that the estimates found in earlier studies may have been biased by selection.

2. For Black students, no association between mobility and school quality was found regardless of the distance moved.