Table 7:

Impact on Categorical Indices

	Life Satisfaction Index	Income Index	Migrant Index	Consumption Expenditures Index	Goals Index	Savings Index	Borrow and Lend Index	Financial Literacy Index
	b/se/p/q	b/se/p/q	b/se/p/q	b/se/p/q	b/se/p/q	b/se/p/q	b/se/p/q	b/se/p/q
Treatment 1: Financial Education Only	0.064	0.073	0.024	−0.054	−0.078	0.123*	−0.056	0.067
	(0.069)	(0.066)	(0.067)	(0.063)	(0.070)	(0.070)	(0.069)	(0.071)
P-values for the coefficients	0.353	0.264	0.718	0.395	0.266	0.078	0.415	0.343
Q-values for all 8 hypotheses	0.475	0.475	0.718	0.475	0.475	0.475	0.475	0.475
Treatment 2: Financial Services Only	0.056	0.038	0.004	0.036	0.205**	0.117	−0.093	0.007
	(0.074)	(0.072)	(0.076)	(0.058)	(0.101)	(0.075)	(0.071)	(0.077)
P-values for the coefficients	0.451	0.597	0.957	0.538	0.043	0.121	0.189	0.928
Q-values for all 8 hypotheses	0.796	0.796	0.957	0.796	0.344	0.484	0.504	0.957
Treatment 3: Financial Education and Services	0.072	0.115*	0.090	−0.129	0.060	0.148**	−0.044	0.136*
	(0.073)	(0.068)	(0.067)	(0.084)	(0.080)	(0.069)	(0.069)	(0.073)
P-values for the coefficients	0.32	0.090	0.181	0.126	0.453	0.031	0.523	0.063
Q-values for all 8 hypotheses	0.427	0.240	0.290	0.252	0.518	0.240	0.523	0.240
Mean in Control Group	0	0	0	0	0	0	0	0
N	1558	1560	1558	1558	1531	1557	1557	1560
R Squared	0.059	0.099	0.074	0.151	0.039	0.082	0.065	0.037
P-values for the following tests
T1 = T2 (p-value)	0.904	0.607	0.774	0.113	0.005	0.935	0.590	0.417
T1 = T3 (p-value)	0.907	0.529	0.285	0.358	0.057	0.709	0.861	0.331
T2 = T3 (p-value)	0.823	0.279	0.229	0.037	0.163	0.672	0.483	0.092
T1 + T2 = T3 (p-value)	0.636	0.969	0.530	0.268	0.577	0.362	0.284	0.555

Note: Data from endline survey. Each column is a separate regression. To create categorical indices, for each outcome, we create a z-score by subtracting the control group mean and dividing by the control group standard deviation. Then, we average z-scores by category and standardize again following the same procedure. We estimate treatment effects on these category indices. Simultaneous inference is corrected for using multiple inference testing. The q-values are calculated using the Benjamini-Hochberg step-up method to control for the false discovery rate (FDR). We follow the procedure outlined in Anderson (2008), and test α at all significance levels (1.000, 0.999, 0.998… 0.000). The q-value is the smallest α at which the null hypothesis is rejected.

^*p<0.10,

^**p<0.05,

^***p<0.010