Association for Public Policy Analysis & Management

Conditional cash transfers (CCTs) have been targeted towards the poor. Thus, their targeting assessments check whether these schemes have been allocated to low-income households or individuals. However, CCTs have more than one goal and target group. Beyond poverty alleviation, these cash transfers seek to increase school enrolment. Hence, students at risk of dropping out of school are an additional target group.

This paper analyses whether one of the most common targeting mechanisms of CCTs, an income proxy means test (PMT), can effectively identify the poor and future school dropouts. I use rich administrative datasets from Chile to simulate different targeting mechanisms. The paper compares the targeting effectiveness of a PMT with other approaches that use the outputs of a predictive model of school dropout. I built this model using machine learning algorithms.

The paper shows that using the outputs of the predictive model in conjunction with the PMT increases targeting effectiveness by identifying more students who are either poor or future dropouts. This joint targeting approach increases effectiveness in different scenarios except when social valuation of the two target groups largely differs. In these cases, the most likely optimal approach is to solely adopt the mechanism designed to find the highly valued group.

The paper provides novel contributions to the social policy targeting field. Overall, the results of the paper emphasise that targeting design, and assessments, must follow the goals of the policy and its consequential definition of target groups. Public officials that value equally the two described target groups may find opportunities for increased targeting effectiveness by modifying the allocation rules of these programs. Additionally, beyond providing one of the first machine learning applications of school dropout in a developing country, the paper shows that appropriate predictive models of this problematic are at hand for public officials. In contexts where countries are improving administrative records, this finding deserves attention.