Events at CERGE-EI

Prof. Martina Björkman Nyqvist

Stockholm School of Economics, Sweden

Join online
Meeting number: 2741 092 1050
Meeting password: 273695

Authors: Martina Björkman Nyqvist, Phyllis Awor, Andrea Guariso and Jakob Svensson

Abstract: Many interventions that demonstrate large impacts in pilot settings lose effectiveness when implemented at scale. We study this challenge using an iterative trial embedded in the national expansion of a community health worker program in Uganda. The evaluation incorporated structured feedback into scale-up, generating interim evidence on performance that was shared with implementers to guide adaptation. A year into the expansion, results showed limited gains. Consistent with a simple framework of scaling under uncertainty, we find that attenuation at scale need not be permanent: whether impact is restored depends on how organizations respond to performance signals and the trade-offs they face during expansion. One organization undertook substantial revisions that recaptured pilot-level mortality reductions, while the other introduced limited adjustments and achieved smaller improvements. We interpret this divergence as arising from differences in organizational opportunity costs, as intensifying one program may crowd out alternative uses of scarce resources. Embedding structured learning into expansion can shape not only how programs evolve, but also how scarce resources are allocated across competing interventions.