John Beshears is a Professor at Harvard University. James J. Choi is a Professor at the Yale School of Management. Joel Dickson is the Global Head of Enterprise Advice Methodology at The Vanguard Group. Aaron Goodman is a research economist at The Vanguard Group. Fiona Greig is Global Head of Investor Research and Policy in Vanguard’s Investment Strategy Group. David Laibson is a Professor at Harvard University.
In the United States, employer-sponsored defined contribution (DC) plans such as 401(k)s have come to serve at least two purposes. First, employees save money in these plans to fund their retirement spending. Second, individuals may access plan balances before retirement age to cover short-run expenses. In our recent paper for the Pension Research Council, we examined the extent to which using DC plans for the second purpose undermines the first purpose. Our results suggest that most individuals who access DC balances before retirement age are able to continue contributing to the plans at their prior rate, while simultaneously putting extra money into the plan to replenish the balances they took out.
How 401(k) loans work
To study what happens when employees access their DC balances before retirement age, we analyzed 401(k) loans, which are permitted under the rules of most plans. When a participant takes out a 401(k) loan, some of the account balance is paid out as a cash lump sum. The worker then must repay the principal borrowed plus interest over the following months or years, via direct payroll deductions. If employees leave their employers before the loans are repaid, the full loan balances are typically due by the following April 15. Failure to repay causes the remaining balances to be treated as plan withdrawals for tax purposes.
Key research findings
We analyzed data on DC plan activity from Vanguard, which manages the records for more than 1,700 plans with approximately 5 million participants. We tracked a group of 188,686 participants who took out plan loans in the year 2021 and did not leave their employers over the next two years. Among this group, 26% of these participants decreased their contribution rates to their plans, and a similar percentage boosted their contribution rates, within those two years. An even larger fraction (38%) passively experienced contribution rate increases due to auto-escalation, which raises the contribution rates of employees enrolled in the auto-escalation program on a prespecified schedule without requiring those employees to take any action at the time of the increase.
Some of these contribution rate changes would likely have occurred even if the participants had not taken out plan loans. To provide a benchmark, we identified a group of participants who did not take out loans but who were otherwise similar to the loan takers. Specifically, for each loan taker, we searched for a ‘control’ participant in the same plan, who remained at the employer for the two-year follow-up period, who did not take out a loan or a withdrawal in the same month or in the prior six months, and who was a good match in terms of tenure, age, income, and prior plan contribution rate. Relative to this control group, matched loan takers had contribution rates that were 0.8 percentage points lower on average by the end of the follow-up period. This effect is a reduction of approximately 10%, relative to the control group’s mean contribution rate.
We conclude that taking out DC plan loans had very little impact on participant contribution rates to their retirement plans. Indeed, because loan takers immediately began repaying their loans via payroll deduction, most of them directed a larger fraction of every paycheck to their DC accounts. Similarly, we found very little impact on contribution rates when we focused only on small loan amounts or on large loan amounts. Moreover, results were comparable when we examined loans initiated during the year 2017 (addressing the concern that our results for the sample of loans initiated during 2021 could have been unique to the COVID period), and when we studied people who took “hardship withdrawals,” generally subject to a tax penalty and allowed by some plans if a participant documents a financial hardship.
Implications
In 2022, the U.S. Congress passed the SECURE 2.0 Act. Among many other provisions, SECURE 2.0 allows employees to withdraw up to $1,000 per year from their DC plans to cover emergency expenses without paying financial penalties. Our results suggest that most plan participants who take a penalty-free withdrawal have the capacity to increase the rate at which they contribute to the plan, covering both their existing savings rate and additional contributions to replenish the savings withdrawn.
Plan sponsors seeking to facilitate the replenishment of withdrawn savings could make such a contribution path the default option for participants who take emergency withdrawals.
Views of our Guest Bloggers are theirs alone, and not of the Pension Research Council, the Wharton School, or the University of Pennsylvania.
