New evidence on the efficacy of state-based retirement programs: The case of OregonSaves

Highlights

• Seventeen US states have launched automatic-enrollment retirement plans for workers lacking access to employer-sponsored plans.
• The oldest state plan, OregonSaves, significantly expanded access to savings via payroll deduction for low-income workers.
• Low earnings, high opt-out rates, and high job turnover rates hinder asset accumulation.
• While the program provides access to emergency savings, persistent withdrawals erode retirement savings.

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.

Introduction

Approximately one-third of American workers lacks access to employer-sponsored retirement plans (ESRPs), and many of these work in jobs with low wages and high job turnover.¹ Though such workers could establish and save via individual retirement accounts (IRAs), only 15% of US households contribute to IRAs (Investment Company Institute, 2023). The Survey of Income and Program Participation (SIPP) reports that only 7.6% of workers without access to an ESRP actively contribute to IRAs (see Appendix Table 1); also the 2019 Survey of Consumer Finances (SCF) found that only 10.7% of households with earnings levels similar to those of OregonSaves participants had any retirement assets (see Appendix A.2). Whether this group’s lack of retirement savings deserves the attention of policymakers is a matter of debate: some characterize low-paid workers’ low retirement saving as a “crisis” (e.g., Miller et al., 2015), while others assert it is not, since low-earner households receive the highest Social Security replacement rates (e.g., Biggs, 2019a, Biggs, 2019b). To expand access to retirement savings via employer payroll deductions, 17 states have implemented “automatic IRA” programs requiring employers to automatically enroll employees into state-created IRAs (Georgetown University, 2025); and proposed federal legislation seeks to extend these plans to the national level (Sullivan, 2024). Consequently, it is important to determine whether the nation’s oldest automatic IRA program, OregonSaves, is boosting retirement savings or providing other potential benefits.

OregonSaves’ design was strongly influenced by academic research on 401(k) eligibility (Benjamin, 2003) and automatic enrollment (Madrian and Shea, 2001, Carroll et al., 2009). Oregon law requires private-sector firms lacking an ESRP to automatically enroll their employees in OregonSaves, with a before-tax default contribution rate of 5% and annual automatic escalation of 1%.² If employees do not opt out, employers transfer payroll deductions to OregonSaves which (opens and) makes deposits into participants’ Roth IRAs. The fact that participants can withdraw funds from their Roth IRAs with minimal tax implications reduces the cost of opting out and withdrawing assets, allowing OregonSaves to function both as a retirement savings plan and a liquid savings account. Unlike with 401(k) plans, there are no employer matches, but there also are no forced withdrawals following job turnover (Hung et al., 2021).

OregonSaves had 124,570 funded accounts and $243.6 million in total assets as of year-end 2023, with a corresponding average account balance of $1,956. Using administrative data, Quinby et al. (2020) reported that 43.3% of actively employed employees had both positive contributions and positive account balances in September 2019. Similarly, we find that only 39.6% of employees ever have a positive OregonSaves account balance during their first 12 months of eligibility. Quinby et al. noted that 20% of the accounts with positive balances in September 2018 experienced withdrawals by September 2019, and that the likelihood of withdrawal rose significantly with job turnover during that 12-month period. Chalmers et al. (2022) used administrative data through April 2020 to document low earnings and high turnover rates for the average employee served by OregonSaves. The authors also found that the top two reasons employees gave for opting out of OregonSaves were “I can’t afford to save” (27.8% of those opting out) and “I have my own retirement plan” (23.3%). Both studies confirmed that OregonSaves participation rates were far below those in most ESRPs (Beshears et al., 2023).

The present paper expands prior research four ways. First, to quantify the extent to which OregonSaves targets different industries and job types relative to those offering 401(k) plans, we incorporate US Department of Labor (DOL) data on ESRPs. For example, the top three industries served by OregonSaves are food services (37.9%), healthcare (11.4%), and business support (9.7%), while the corresponding market shares for ESRPs are 2.9%, 20.7%, and 2.1%. When we examine healthcare at a more granular level, we show that OregonSaves and 401(k) plans serve distinct employer types (e.g., assisted living facilities versus hospitals).

Second, we provide evidence on which employees are most likely to stop contributing to OregonSaves and why. Following employees over their first 12 months of eligibility, we discover that opt-out and job turnover rates, as well as earnings levels differ meaningfully across industries. While almost 70% of eligible workers either opt out from OregonSaves or experience job turnover by month 12, the fraction ranges from 58% for business support, to 85% for agriculture. Such high attrition rates have important implications for calculations of annual contribution rates and asset accumulation patterns. When we limit attention to the 30% of employees still actively employed and making a positive plan contribution in month 12, industry-level effective annual contribution rates range from 4.2% for construction, to 5.0% for real estate (where rates near 5% reflect widespread acceptance of the 5% default contribution rate and few withdrawals). Conditioning instead on employees having made at least one contribution to OregonSaves within 12 months of eligibility, the effective annual rate falls, varying from 2.9% for transportation/storage, to 3.9% for professional/scientific. When we expand the sample to include the 60.4% of eligible employees who never have a positive account balance, the effective annual contribution rate falls to 1.3%, varying from 0.6% for agriculture to 1.9% for management.³ These low estimates highlight the difficulty of relying on automatic enrollment to increase retirement savings for this group of workers.

Third, we shed new light on the role that earnings play in employee retirement plan participation decisions. We find that workers with higher industry-level earnings have modestly higher opt-out rates, likely due to a higher likelihood of having pre-existing IRA savings. Yet holding industry constant, employees who earn less than their coworkers are more likely to stop contributing quickly. Workers hired after their firms registered for OregonSaves are also less likely to opt out, perhaps because OregonSaves contributions do not cause a salient change in their take-home pay. Workers being exposed to the program for a second or third time are also less likely to opt out. These last two findings suggest that, when the OregonSaves program reaches steady state, participation rates may be higher than those measured during the program’s rollout.

Fourth, because OregonSaves’ ultimate goal is to increase savings, we study the evolution of account balances. A year after making an initial contribution, the average participant account balance is $699 (median is $354). Excluding the 10% of accounts where all contributions are withdrawn by year-end, the average increases to $777 ($453), but it is still an order of magnitude below that of ESRP participants with “0–1” year of tenure (Vanguard, 2023). OregonSaves participant account balances are limited by several factors, including workers’ lower wages, higher job turnover, the lack of employer matches, and steady outflows. Regardless, if current patterns hold, it appears that OregonSaves is unlikely to generate retirement savings comparable to those of a typical 401(k) plans for more than a small fraction of eligible employees. Whether the outflows are used to smooth consumption or reduce debt are important questions for future research.