Table 2

Balance of Covariates: Observed Properties Uncorrelated with Program Timing

Independent VariableTSC Latrines Per CapitaTSC Started
Mother literate−0.46(0.38)−0.01(0.01)
Mother went to school−0.30(0.44)  0.00(0.01)
Father went to school  0.12(0.43)  0.00(0.01)
Child receives extra tuition  0.26(0.46)  0.00(0.01)
School enrollment−0.24(0.32)  0.00(0.00)
Formal house−0.45(0.50)−0.01(0.01)
Electrification  0.20(0.50)  0.00(0.01)
Village has road  0.18(0.61)−0.03(0.02)
Village has ASHA  0.18(0.64)−0.01(0.02)
Village has school−0.12(0.61)  0.00(0.01)
Village has ration shop−0.25(0.69)−0.02(0.02)
  • Notes: Observations are the same sample of six-year-olds used in our main analysis. Each estimate presents a coefficient and a clustered standard error in parentheses from a separate regression (of the form of Equation 1) of the variables listed as the dependent variable, on either latrines per capita or an indicator for a positive number of latrines per capita (“TSC started”), with district and year fixed effects. Two-sided p-value:

  • p < 0.1.