Table 3.

Instrumental variable analysis of peer effects.

Variables	(1)	(2)	(3)
	2SLS	2SLS	2SLS
	All the samples	Peer fathers who worked outside	Peer fathers who worked locally
Drinkpeer	0.105***	0.126**	0.097*
	(0.040)	(0.062)	(0.053)
Aware	−0.214***	−0.201***	−0.223***
	(0.010)	(0.015)	(0.013)
Boy	0.126***	0.137***	0.115***
	(0.010)	(0.016)	(0.014)
Age	−0.000	0.001*	−0.001*
	(0.000)	(0.001)	(0.000)
Preschool	−0.028*	−0.023	−0.033
	(0.017)	(0.027)	(0.021)
Number of siblings	0.023***	0.032***	0.015*
	(0.007)	(0.010)	(0.009)
Local dummy	−0.477***	−0.355	−0.401
	(0.148)	(0.230)	(0.252)
Father’s age	0.032	0.066	−0.001
	(0.046)	(0.066)	(0.064)
Mother’s age	−0.214***	−0.226***	−0.112
	(0.052)	(0.062)	(0.112)
Father’s education	−0.000	−0.003	0.003
	(0.002)	(0.003)	(0.002)
Mother’s education	−0.004***	0.001	−0.008***
	(0.001)	(0.002)	(0.002)
Father drinks	0.066***	0.068***	0.067***
	(0.009)	(0.014)	(0.011)
Asset	0.014***	0.018***	0.010***
	(0.003)	(0.005)	(0.004)
Constant	7.326***	0.126**	4.848
	(2.070)	(0.062)	(4.203)
School fixed effects	Yes	Yes	Yes
Observations	10,772	4,726	6,046
Cragg-Donald Wald F statistic	306.383	126.378	177.466
Hansen J statistic	0.957	1.458	4.994
Permutation test (P-val)		0.316

2SLS means two stages least square method. The values of the robust standard errors are reported in parentheses. The definitions for each of the variables are available in Table 1. ^⁎⁎⁎Indicate significance level of 1%. ^⁎⁎Indicate significance level of 5%. ^*Indicate significance level of 10%.