Search

Existing studies estimate the negative effects of family dissolution; however, evidence regarding when such effects appear is lacking. The effect is hard to isolate for at least two reasons: families select into dissolution; and effects might be observed before, during and after the dissolution. This paper presents evidence regarding the timing of dissolution effects. Family fixed effects (FEs) and a value-added model handle selection; timing is addressed by exploiting repeated observations of student performance before and after dissolution. I present evidence of weaker performance among students from dissolving families, an effect mainly driven by the deterioration of family relations before the break. The divorce itself seems to be less important. [ABSTRACT FROM AUTHOR]