Intergenerational transmission of human capital in Europe: evidence from SHARE

IZA Journal of European Labor Studies

Table 8 Effects of Parents’ education on Sons and Daughters, IV with country-specific trends

	(1)	(2)	(3)	(4)
Sample	Overall	Overall	Sons	Daughters
Panel A: 2SLS
Dep. Var.: Child’s education
Parental education	0.468	0.421	1.064	0.671**
	(0.334)	(0.272)	(0.973)	(0.334)
Parental educ*female (parent)		0.047***	0.079*	0.052***
		(0.013)	(0.043)	(0.015)
Observations	6,184	6,184	3,117	3,067
Mean of Dep. Var.	13.25	13.25	13.14	13.34
Std. Dev. of Dep. Var.	2.84	2.84	2.89	2.78
Angrist-Pischke first stage F statistic	1.63	8.64	1.39	4.04
Panel B: First stage
Dep. Var.: Parent’s Education
Compulsory education	0.104	0.124	0.153	0.092
	(0.081)	(0.082)	(0.109)	(0.108)
Compulsory educ*female (parent)		-0.051***	-0.055***	-0.045***
		(0.012)	(0.016)	(0.017)
Observations	6,184	6,184	3,117	3,067

Notes: All specifications include controls for country dummies, birth cohort dummies for parents and children (in 1-year intervals), socio-demographic characteristics and country-specific quadratic cohort trends (computed by interacting parental birth cohort and its square with country dummies). The Angrist-Pischke first stage F statistic refers to the first stage regression of parental education; the first stage regression of parental education*female has much stronger power, thus the Angist-Pischke first stage F-statistic is omitted. Standard errors clustered at the parents’ country and cohort level are reported in parentheses. * Significant at 10%; ** significant at 5%; *** significant at 1%.