Written with Sami Horn (Carnegie Mellon University), Dean Karlan (Northwestern University), and Jonathan Zinman (Dartmouth University)
We randomly assign 240 youth clubs to receive either financial education, access to a subsidized group account, both, or neither. The financial education treatments increase financial literacy; the account-only treatment does not. Both administrative and survey data show that savings after one year increases in all treatment arms relative to the control group. Earned income also increases in the treatment arms, and all of these effects are successfully sustained in a second endline four years later. Despite fairly detailed data, we are unable to draw firm conclusions regarding mechanisms and heterogeneity. Overall we find little evidence that education and account access are strong complements, and some evidence that they are substitutes.