Table B6.

Potential mechanisms: Impacts on consumption per capita (logged), youth 14–21 at baseline by sex and country

	Male		Female
Panel A: Malawi [Social Cash Transfer Program]	Consumption pc (logged) (1) Unadjusted	Consumption pc (logged) (2) Adjusted	Consumption pc (logged) (3) Unadjusted	Consumption pc (logged) (4) Adjusted
Treatment status	0.305***	0.305***	0.311***	0.307***
	(0.0431)	(0.0375)	(0.0433)	(0.0439)
Observations	1,023	1,023	869	869
R-squared	0.219	0.271	0.186	0.228
Endline control mean	10.317		10.339
	Male		Female
Panel B: Zambia [Multiple Category Targeted Grant]	Consumption pc (logged) (1) Unadjusted	Consumption pc (logged) (2) Adjusted	Consumption pc (logged) (3) Unadjusted	Consumption pc (logged) (4) Adjusted
Treatment status	0.259***	0.250***	0.189***	0.219***
	(0.0304)	(0.0298)	(0.0326)	(0.0332)
Observations	1,296	1,296	1,070	1,070
R-squared	0.080	0.161	0.051	0.120
Endline control mean	3.850		3.859

NOTES: Estimations use ANCOVA modeling among panel individuals (follow-up after 30 months in Malawi and after 36 month in Zambia). Robust standard errors in parentheses corrected for clustering.

^***p<0.01,

^**p<0.05,

^*p<0.1.

All controls are measured at baseline and include stratifying indicators used for the randomization (traditional authorities dummies in Malawi and district dummies in Zambia). Inconsistent observations, namely those individuals reporting ever being married/pregnant at baseline but never being married/pregnant at endline were excluded from the analysis. Sampling weights have been applied in Malawi.

Results for female youth are based on the ‘ever married’ panel; results based on the “ever pregnant” panel are fully consistent.