Conditional cash transfers for primary education: Which children are left out?

Conditional cash transfer (CCT) programs to increase primary-school enrollment and attendance among low-income households have been shown to benefit children and households, but to date little is known about who joins such programs. We test three hypotheses about predictors of CCT program participation in indigenous societies in Bolivia, focusing on attributes of the household (ethnicity), parents (modern human capital), and children (age, sex). We model whether children receive a transfer from Bolivia's CCT program (), using data from 811 school-age children and nine ethnic groups. Children from the group least exposed to Westerners (Tsimane') are 18-22 percentage points less likely to participate in the program than children from other lowland ethnic groups. Parental modern human capital and child sex do not predict participation. We discuss possible mechanisms underlying the findings and conclude that the Tsimane's current lower returns to schooling are the most likely explanation.