Playing Catch-up with Health Savings

• In Event: Income Sources in Retirement

Abstract

This paper investigates how individuals respond to contribution limits for Health Savings Accounts (HSAs), a tax-advantaged vehicle designed for both short-term medical spending and long-term saving. Using restricted-access IRS administrative data covering the universe of HSA-linked tax forms, we provide the first causal evidence on behavioral responses to HSA policy using a regression discontinuity design centered on the age-55 “catch-up” provision.

We find large increases in HSA contributions at age 55 among individuals who previously contributed at least 90% of the limit, consistent with standard life-cycle models. Smaller but statistically significant increases also occur among unconstrained savers, suggesting possible behavioral mechanisms such as anchoring or heuristic decision-making. Responses are strongest among individuals contributing via payroll deduction, those in family plans, and those who previously saved without withdrawing.
We also examine complexity in the spousal catch-up rules and find that many married couples fail to fully optimize joint contributions, likely due to informational frictions or household coordination costs. Analyzing distributions, we find no immediate increase in HSA withdrawals at age 55 and only modest increases the following year, implying that induced catch-up contributions represent genuine saving rather than short-term liquidity use.

Finally, event-study evidence reveals that while average HSA balances rise over time, median users often disengage within a few years. These findings highlight how tax incentives, constraints, and behavioral cues jointly shape saving behavior, with implications for designing effective and accessible savings policy as HSAs continue to grow in prominence.

Authors

• Jacob Berman, Treasury Dept

• Adam Bloomfield, adam.bloomfield@gmail.com

• Sita Slavov