Table 3.

Age-specific household wealth gradients in early childhood health (By Wealth Quintiles)

	Panel A
	Bangladesh (Dependent Variable: HAZ Scores)
	Poorest	Poorer	Middle	Richer
Age 0–11 M	–0.579***	–0.451***	–0.250*	–0.157
	(0.146)	(0.132)	(0.134)	(0.136)
Age 12–35 M	–0.749***	–0.522***	–0.485***	–0.456***
	(0.105)	(0.107)	(0.099)	(0.094)
Age 36–59 M	–0.643***	–0.513***	–0.428***	–0.283***
	(0.100)	(0.090)	(0.088)	(0.087)
	Weighted Sample: 7505, R-Squared: 0.209; F: 43.48, Mean VIF: 2.83
	Panel B
	Bangladesh (Dependent Variable: WAZ Scores)
	Poorest	Poorer	Middle	Richer
Age 0–11 M	–0.581***	–0.373***	–0.159	–0.105
	(0.109)	(0.109)	(0.102)	(0.112)
Age 12–35 M	–0.573***	–0.430***	–0.330***	–0.302***
	(0.087)	(0.086)	(0.084)	(0.081)
Age 36–59 M	–0.320***	–0.283***	–0.322***	–0.217***
	(0.085)	(0.080)	(0.082)	(0.076)
	Weighted Sample: 7504, R-Squared: 0.205; F: 38.98, Mean VIF: 2.85
	Panel C
	Nepal (Dependent Variable: HAZ Scores)
	Poorest	Poorer	Middle	Richer
Age 0–11 M	–0.503**	–0.659***	–0.018	–0.070
	(0.226)	(0.222)	(0.207)	(0.238)
Age 12–35 M	–0.649***	–0.396**	–0.247	–0.113
	(0.176)	(0.180)	(0.207)	(0.177)
Age 36–59 M	–0.384**	–0.228	–0.165	0.054
	(0.161)	(0.178)	(0.178)	(0.169)
	Weighted Sample: 2342, R-Squared: 0.224; F: 17.24, Mean VIF: 2.89
	Panel D
	Nepal (Dependent Variable: WAZ Scores)
	Poorest	Poorer	Middle	Richer
Age 0–11 M	–0.130	–0.532**	0.061	0.074
	(0.194)	(0.222)	(0.180)	(0.194)
Age 12–35 M	–0.344**	–0.254*	–0.199	0.001
	(0.141)	(0.146)	(0.164)	(0.127)
Age 36–59 M	–0.214	–0.310**	–0.371**	–0.118
	(0.154)	(0.154)	(0.155)	(0.145)
	Weighted Sample: 2339, R-Squared: 0.193; F: 11.78, Mean VIF: 2.90

Note: The numbers in the parenthesis refer to the linearized standard errors that account for survey design and sample weight. ***indicates p< 0.01, **indicates p < 0.05 and *indicates p <0.1. Each panel presents a separate regression. The table reports regression coefficients of the interaction terms of dummies for each wealth quintile with dummies for child age groups where the richest quintile is the reference category. We control for same set of control variables as mentioned in Table 2. The regressions for HAZ scores control for the same set of control variables as displayed in Table 2 columns (1) and (3) for Bangladesh and Nepal respectively. The regressions for WAZ scores control for the same set of control variables as displayed in Table 2 columns (2) and (4) for Bangladesh and Nepal respectively. Each regression also includes dummies to capture regional effects. The regional dummies for Bangladesh are based on the seven major divisions, and for Nepal the regional dummies are based on the five developmental regions