Closing the Gap between Evidence and Policy in Latin America: What Works?

Abstract

Many governments and international organizations call for more efforts to integrate scientifically rigorous impact evaluations into policy streams, though government officials frequently self-report that they rarely engage such evidence (Newman et al. 2017; Migone and Brock 2017). Previous research has found that small “nudges” in communications content, style, or source can affect behavior in a wide variety of contexts (Thaler and Sunstein 2008). To test how nudges might impact bureaucrats’ willingness to engage with policy-relevant academic evidence, researchers sent email invitations to 130,000 public officials in Latin American countries with easily available names or contact information. This study probes engagement with a website designed to provide academic evidence in a convenient and accessible format. The study tests the effects of four treatment arms: nudges, messenger nationality, direct relevance of an example study, and saturation (the percent of bureaucrats invited from a given ministry). Results indicate that greater saturation had the strongest positive effects. Findings for messenger nationality were mixed: the Chinese researcher significantly decreased uptake generally, the Colombian appreciably increased engagement in Colombia, and the others had null effects. Motivational nudges had generally negative effects, and the relevant example made no significant difference. Broadly, these results suggest that encouraging the use of impact evidence is a difficult problem that is insensitive to most behavioral nudges but may be receptive to leveraging group dynamics.