Table 5.

Regression 1 and 3 covariates’ estimated effects

	Years of education			Days attended
	(19–21-year-olds)			(6–10-year-olds)
Variable	Coef.	Robust std. err.	t	Coef.	Robust std. err.	t
Household wealth	1.366a	0.080	17.18	0.115a	0.026	4.41
Household wealth sqd.	−.194a	0.032	−6.12	−.031a	0.011	−2.75
Number of other youths in hours	−.130a	0.024	−5.39	0.001	0.011	0.11
Age	0.099b	0.052	1.91	0.055a	0.009	6.22
Head primary incomplete	0.247b	0.129	1.92	0.098c	0.043	2.29
Head primary complete	0.867a	0.121	7.17	0.160a	0.042	3.84
Head some secondary	1.486a	0.126	11.83	0.109a	0.042	2.60
Head secondary or more	2.425a	0.140	17.31	0.268a	0.046	5.87
Head non-std. schooling	−0.527	0.794	−0.66	0.376a	0.126	2.99
Head education missing	0.024	0.838	0.03	0.657a	0.126	5.23
Heard of arsenic prob.	0.557a	0.118	4.75	0.073b	0.038	1.92
Mean subdistrict wealth	−.812a	0.260	−3.12	−0.154	0.107	−1.44
Std. dev. sub-district wealth	0.202	0.333	0.61	0.018	0.139	0.13
Flood prone	0.224b	0.135	1.66	−.118c	0.055	−2.13
Garbage pile	−0.059	0.258	−0.23	0.125	0.211	0.60
Landslide prone	−0.570	0.679	−0.84	0.027	0.331	0.08
Industrial pollution	1.359a	0.324	4.20	−0.201	0.325	−0.62
Safe from eviction	0.443c	0.198	2.23	−0.001	0.055	−0.01
Squatter household	−1.062	0.666	−1.60	−0.092	0.153	−0.60

^aStatistically significant at 0.01 level

^bStatistically significant at 0.10 level

^cStatistically significant at 0.05 level