Association for Public Policy Analysis & Management

This paper illustrates the application of a sequential targeting mechanism in the context of ACHAB, a randomized field experiment aimed at assessing the impact of a matched savings program on college participation among children of low-income households in Italy. In the demonstration, target applicants are those students who would enroll at the university only if receiving the program’s incentive. Their empirical identification is achieved through a sequential targeting mechanism based on: (1) household income and; (2) a predictive model of applicants’ probability of enrolling at the university. The performance of the proposed targeting mechanism is assessed by comparing effectiveness and cost-effectiveness estimates against an alternative “no targeting” scenario (i.e., a scenario in which applications are admitted randomly to the study). According to our most-conservative estimates, the proposed targeting mechanism yields a 12% larger impact and a 30% cost-effectiveness gain. The paper also shows that a simplified and more easily implementable targeting schedule could reach satisfying results (23% cost-effectiveness gains). The paper discusses how to correctly define and identify target beneficiaries when designing financial aid interventions and how to test the performance of the adopted targeting strategy in an experimental setting. The implications for financial aid programs and the conditions for streamlining this targeting mechanism into standard financial aid policy are discussed.