We evaluate the impact of a randomized educational intervention on children’s intertemporal choices. The intervention aims to improve the ability to imagine future selves and encourages forward-looking behavior using a structured curriculum delivered by children’s own trained teachers. We find that treated students make more patient intertemporal decisions in incentivized experimental tasks. The results persist almost 3 years after the intervention, replicate well in a different sample, and are robust across different experimental elicitation methods. The effects also extend beyond experimental outcomes: we find that treated students are significantly less likely to receive a low “behavior grade.”

We develop a model of consumption and income that allows for pervasive heterogeneity in the parameters of both processes. Introducing codependence between household income parameters and preference parameters, we also allow for heterogeneity in the impact of income shocks on consumption. We estimate the parameters of the model using a sample from the Panel Study of Income Dynamics, covering the period 1968–2009. We find considerable codependent heterogeneity in the parameters governing income and consumption processes. Our results suggest a great deal of heterogeneity in the reaction of consumption to income shocks, highlighting the heterogeneity in the self-insurance available to households.