Citation

Hao L, Xie G. Soc. Sci. Res. 2002; 31(1): 1-28.

Copyright

(Copyright © 2002, Elsevier Publishing)

DOI

10.1006/ssre.2001.0715

PMID

unavailable

Abstract

This study examines the effects of family structure on child misbehavior, paying close attention to the complexity and endogeneity of family structure. We use longitudinal measures of childhood family structure to capture its complexity and fixed-effects models to control for the unobserved, time-invariant family and child characteristics, which contribute to the endogeneity of family structure. Our results suggest that, other things being equal, stability of intact, step-, and single-mother families is beneficial in lowering the risk of misbehavior, with a weaker effect for single-mother families. This finding challenges the consensus about a detrimental effect of alternative family structures and the incomplete institution thesis regarding step-families. Our analysis also provides two methodological findings: (1) duration of different types of current family structure is superior to other longitudinal measures, which, in turn, are superior to snapshot measures of family structure, and (2) conventional OLS models tend to produce downward bias for the effects of family structure on child misbehavior.