Working Paper: NBER ID: w31682
Authors: Irina Gemmo; Pierrecarl Michaud; Olivia S. Mitchell
Abstract: We study the effect of a financial education intervention on portfolio choices in a unique incentivized setting that allows us to investigate selection into the intervention and treatment effect heterogeneity. After directly eliciting willingness to pay for the financial education, we find that the more financially literate, those confident they can apply the knowledge acquired in the intervention, and those expecting higher returns are willing to pay more. Using portfolio allocation tasks, we show that the financial education leads to better outcomes according to welfare metrics tailored to respondent risk aversion: there is a 20 p.p. increase in the fraction with a welfare gain, representing on average 3.2 p.p. of wealth. Those most willing to participate are those who gain the most from the education, which has important implications for the design of financial education across a variety of settings.
Keywords: financial education; portfolio choice; sample selection; financial literacy
JEL Codes: G11; G41; G53
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Willingness to pay (D11) | Participation in financial education (G53) |
| Selection into treatment (C24) | Characteristics of participants (C90) |
| Cognitive abilities and numeracy (G53) | Benefit from financial education (G53) |
| Selection into treatment (C24) | Estimated effects of financial education (G53) |
| Financial education (G53) | Portfolio efficiency (G11) |
| Financial education (G53) | Welfare (I38) |
| Financial education (G53) | Customization of portfolio allocations (G11) |
| Financial education (G53) | Likelihood of improving expected returns or reducing variance (G11) |