Abstract: Seventeen US states have mandated employers to facilitate auto-enrollment retirement saving for workers lacking access to employer-sponsored plans, and proposed federal legislation seeks to extend these plans to the national level. This paper examines the experience with OregonSaves, the country’s longest-running plan, documenting that the program did prompt some participant savings. Median account balances were about $600 by mid-2023, but opt-out rates were above 50%, especially for the low-paid. Repeated exposure to the plan slightly reduced opt-outs, though withdrawals remained common. While modest savings accrued for many, it remains unclear whether these accounts will grow large enough to significantly increase retirement consumption.
Abstract: We examine debt delinquency among U.S. adults aged 50+, using administrative data from 4 million borrowers. In August 2022, about one in five older adults had delinquent debt, signaling financial challenges in retirement. Delinquency rates decreased with age, affecting 25 % of those aged 50–61 and 16 % of those 62+. Racial/ethnic disparities were evident: consumers in majority American Indian/Alaska Native, Black, Hispanic, or Asian American/Pacific Islander areas faced higher probabilities of delinquency and larger median debt amounts compared to majority-white areas across various debt types. These findings highlight disparities in financial security among older adults.
Abstract: We examine the effects of the Universal Credit expansion and mortgage forbearance on the financial well-being of United Kingdom (UK) residents during the pandemic. Using anonymized individual-level consumer financial data on 2 million UK consumers, each with one or more defaulted accounts accrued before the pandemic, we found that average nonmortgage debt increased by 17% from October 2019 (£5497) to December 2021 (£6456). Using a difference-in-difference approach, we found mixed policy impacts on the debt people carried. Although the expansion of Universal Credit was intended to help financially vulnerable families, consumers who were more likely to benefit from the Universal Credit expansion took on 1% more total nonmortgage debt after the policy expansion. By contrast, during the period of mortgage forbearance, mortgage holders accumulated 1% less total nonmortgage debt compared to nonmortgage holders. These results suggest that policies implemented in the UK to protect financially vulnerable families were insufficient to prevent beneficiaries from accumulating additional debt during the pandemic.
Abstract: Oregon recently launched an automatic-enrollment retirement savings program for private sector workers lacking access to other workplace retirement plans. We analyze participation choices, account balances, and inflow/outflow data using administrative records between August 2018 and April 2020. Within the small to mid-sized firms served by OregonSaves, estimated average after-tax earnings are low ($2,365 per month) and turnover rates are high (38.2 percent per year). Younger employees and employees in larger firms are less likely to opt out, but participation rates fall over time. Overall, we conclude that OregonSaves has meaningfully increased employee savings by reducing search costs.
Abstract: We document the prevalence and amount of debt delinquency in the US among older adults, overall and by racial/ethnic group characteristics. Using administrative data on credit use for a sample of 4 million older adult borrowers, we show that about one in five consumers age 50+ with a credit bureau record had delinquent debt in August 2022, suggesting difficulties meeting financial obligations in retirement. The prevalence of debt delinquency decreased with age, with about one in four adults age 50-61 and one in six adults age 62+ holding delinquent debt. Consumers living in local areas where a majority of residents identify as American Indian or Alaska Native, Black, Hispanic, or Asian American or Pacific Islander were more likely to have delinquent debt and/or higher median amounts of delinquent debt, relative to consumers in majority-White areas, for various types of delinquent debt.
Abstract: To design policies that ensure justice for all, we need tools for estimating not just a proposal’s price tag but its fairness. Part of a set of demonstration analyses that use the Equity Scoring Initiative’s dimensions of equity improvement to structure equity assessments, this report shows the potential impact of the Setting Every Community Up for Retirement Enhancement 2.0 Act (SECURE 2.0) for different groups. It discusses how equity scoring can inform the design, debate, and implementation of access- and intermediary-oriented policies, especially those focused on improving retirement security. We project that SECURE 2.0 would significantly improve access to retirement savings overall. We also find that SECURE 2.0 would likely generate greater absolute gains for full-time workers, who have been disproportionately favored when it comes to employment benefits, than for part-time workers, thereby widening the gap in retirement savings access between those groups. Based on the mixed results, we find that SECURE 2.0 would not improve equity across all three of our dimensions.
Abstract: Retirement savings adequacy remains a challenge in the United States, notably with defined contribution plans in which participants are responsible for funding adequacy. In some retirement savings plans, employers increase contributions for employees after reaching a certain age. The effectiveness on whether this change increases total retirement saving depends on whether employees offset these increases by reducing their own contributions at these designated age thresholds. Analyzing individual-level administrative data and varied age-based employer contribution rules, we find that these age-based contribution increases do not crowd-out employees’ own contributions, and total retirement savings increase. Our research highlights the potential positive impact of employer-initiated interventions on improving retirement readiness.
Abstract: Understanding wealth is central for uncovering the barriers to wealth-building and designing policies that unlock opportunities for everyone. However, household wealth data at the local level are generally not widely available, especially statistics disaggregated by race and ethnicity. In this research report, we document how we use machine learning to estimate net worth and emergency savings data at the local, city, state, and national levels. We also disaggregate our estimates by racial and ethnic groups at the city, state, and national levels. Using a random forest model, we predict whether households in the American Community Survey have $2,000 in emergency savings and their net worth. We then aggregate this household-level data to produce statistics at different geographic levels and by racial and ethnic groups.
Conferences: Social Security Administration Retirement and Disability Research Consortium Annual Meeting (2019), NBER Conference on Incentives and Limitations of Employment Policies on Retirement Transitions (2019)
Media coverage: Brookings, MarketWatch, NBER Digest, ThinkAdvisor, 401kspecialist
Abstract: Oregon recently launched an automatic-enrollment retirement savings program for private sector workers lacking access to other workplace retirement plans. We analyze participation choices, account balances, and inflow/outflow data using administrative records between August 2018 and April 2020. Within the small to mid-sized firms served by OregonSaves, estimated average aftertax earnings are low ($2,365 per month) and turnover rates are high (38.2% per year). Younger employees and employees in larger firms are less likely to opt out, but participation rates fall over time. The most common reason given for opting out is “I can’t afford to save at this time,” but the second most common is “I have my own retirement plan.” As of April 2020, 67,731 accounts had positive account balances, holding$51.1 million in total assets. The average balance is $754, but with considerable dispersion; younger workers accumulating the fewest assets due to higher job turnover. Overall, we conclude that OregonSaves has meaningfully increased employee savings by reducing search costs. The 34.3% of workers with positive account balances in April 2020 is comparable to the marginal increase in participation at larger firms in the private sector. Employees opting out of OregonSaves are often doing so for rational reasons.
Dissertation Fellowship: Social Security Administration and the Center for Retirement Research at Boston College, Robert R. Nathan Fellowship
Grants: Social Security Administration, AARP, Pew Charitable Trusts, Wharton Boettner Center/Pension Research Council
Conferences (including scheduled): World Risk and Insurance Economics Congress (2020), RAND Behavioral Finance Forum (2020), National Tax Association's Annual Conference on Taxation (2020), Association for Public Policy Analysis & Management (APPAM) Fall Research Conference (2020), National Association for Business Economics (NABE) Tech Economics Conference (TEC2020), ASSA/AEA Virtual Annual Meeting (2021), Western Economic Association International (WEAI) Virtual International Conference (2021), Midwest Finance Association Annual Conference (2021), Southwestern Finance Association Annual Conference (2021), Midwest Economics Association Annual Conference (2021), Eastern Economic Association Conference (2021), NBER Aging Program Meeting (2021), Western Economic Association International (WEAI) Virtual 96th Annual Conference (2021)
Abstract: I theoretically analyze and empirically identify the optimal default savings rate in automatic enrollment retirement saving plans. I derive a formula for the optimal default as a function of sufficient statistics that can be empirically identified. I estimate individual adherence to the default using exogenous increases in the default rate of OregonSaves, the first state-sponsored auto-enrollment plan in the U.S. I also use survey data to infer the degree of undersaving if workers actively switch to a non-default rate. Combining estimates from administrative and survey data with the optimal default formula, I find the optimal default is 7% of income.