Extracting the causal component from the intergenerational correlation in unemployment

This paper examines the intergenerational correlation in unemployment in Norway and, by use of the sibling-difference method, separates that correlation into its causal and non-causal parts. Detailed register data covering the entire Norwegian population provide the long panel of data this method requires, and allow experimentation with different definitions of parental unemployment. Confirming existing evidence, I find a substantial intergenerational correlation in unemployment. Approximately half of this is due to observed family heterogeneity. The estimated causal effect is non-significantly negative for all measures of parental unemployment.