When Information is Not Enough: Evidence from a Centralised School Choice System

We implemented a large-scale randomised controlled trial encompassing 900 junior high schools in Ghana, a country with universal secondary school choice, to study whether providing students and parents with information on school characteristics and selection strategies improved outcomes in a centralised school selection mechanism. Information changed households’ preferences and the characteristics of schools to which they applied. Students gained admission to higher value-added schools, yet they were not more likely to matriculate on time or at all. Incomplete school information was not the only friction. Household shocks and inaccurate preference forecasting likely contributed to continued admission deviations.