More than eight years ago, Oregon became the first state to formally stand up its auto-IRA program, called OregonSaves. Since then, a growing number of states have either copied the auto-IRA approach or prescribed a different form of “mandatory” retirement savings.
In a new report, analysts from the Center for Retirement Research at Boston College examine a subset of these programs. Their results suggest that early fears that state-run individual retirement accounts would stop employers from starting retirement plans appear to have been overblown.
In fact, there is strong evidence that the presence of auto-IRAs has boosted the share of firms offering retirement savings options, with a significant number of employers starting their own 401(k)s rather than directing workers into state-run programs.
Auto-IRAs, then, have expanded access to savings in two ways, the authors detail: through participation in auto-IRAs themselves and through increased adoption of employer plans.
New Retirement Framework
The subset of auto-IRA policies covered by the study were rolled out between 2017 and 2022 in Oregon, Illinois, California and Connecticut.
The programs have two main features. First, the state contracts with a private-sector financial services company to create Roth IRA accounts for the employees of participating employers. Firms that participate automatically enroll their workers and make the payroll deductions specified by the state, set at either 3% or 5% in the four states studied.
The assets are initially invested in a fund, typically a target-date fund, recommended by the plan manager and approved by the state administering the program. Like 401(k) plans with auto-enrollment, employees are free to opt out of the auto-IRA, change their contribution amount or change their investment fund to other plan options.
Second, the report explains, employers must either participate in the state auto-IRA program or adopt a retirement savings plan. The states studied have applied these employer mandates on a rolling basis, starting with large firms, while enforcing compliance with the possibility of financial penalties.
Running the Analysis
To gauge the effect of state auto-IRAs on employer plans, the Center for Retirement Research uses administrative tax data to follow firms annually from 2012 to 2023.
The analysis starts with IRS Form 941, which is used to identify a firm’s state and the number of people it employs. This form, though, does not include information on whether an employer offers a retirement plan.
So, the authors use a firm’s employer identification number to link each business to their employees’ W-2s. This allows the study to use Box 12 on the W-2 — which includes retirement plan contributions — to see whether at least one of the firm’s employees contributes to an employer plan. The study considers such a contribution within a given year as evidence that the firm offered a plan.
With data on firms offering a plan in hand, the next step is to create a panel of “treated” and “control” firms for each of the seven policy changes relating to the state auto-IRA policies. Treated firms are in the indicated size range and located in the adopting state. Control firms are in the indicated size range but from states that never adopted an auto-IRA policy.
What the Numbers Show
With this framework, the authors use a regression analysis to examine how the share of firms offering their own plan changed once a mandate came into effect.
To illustrate that, they consider a hypothetical example for Oregon firms with 20 to 99 employees, whose mandate came into effect in 2018. They suppose that 30% of these firms offered a plan in 2016 and 36% in 2020.
In such a case, the share of treatment firms offering a plan increased by 6 percentage points two years after the policy. They then suppose that these numbers are 32% and 35% for control firms — representing a 3-percentage point increase over the same period.
“Our analysis assumes that, in the absence of the policy, firms in Oregon would look like these control firms,” the authors explain. “That is, their offer rate would increase by 3 percentage points between 2016 and 2020. Since Oregon firms increased their offer rate by 6 percentage points rather than 3 percentage points, we would estimate that the policy caused a 3-percentage-point increase in the offer rate in 2020 compared to 2016."
After digging into the real-world numbers, the authors find no evidence of any major shift toward fewer employer plans. Instead, firms without a retirement plan subject to an auto-IRA mandate were more likely to begin offering a retirement plan relative to similar firms in other states during the study period.
Other Conclusions and Considerations
The authors also explore several possibilities for why some firms in this situation would adopt their own plan.
For example, they hypothesize that employers might find the administrative burden of auto-IRAs higher than an employer plan, which they can offload to a paid third party. Alternatively, some workers value an employer plan relative to the auto-IRA because of higher contribution limits or the employer match.
Surprisingly, the authors observe, the study did not find compelling evidence that any reasons related to perceived financial costs and benefits of the program play a significant role. Instead, the authors maintain, more behavioral factors seem to play a role.
“For example, since state auto-IRA policies required firms without a retirement plan to make a choice, they may have given third-party plan vendors a good opportunity to market their products,” the report concludes. “These efforts may have convinced employers to start their own plans rather than participate in the state auto-IRAs.”
© Touchpoint Markets, All Rights Reserved. Request academic re-use from www.copyright.com. All other uses, submit a request to TMSalesOperations@arc-network.com. For more information visit Asset & Logo Licensing.