As more states enact retirement savings programs for private sector workers who can’t save through their jobs, policymakers and analysts have speculated about the potential impact on employers: Would these state programs “crowd out” the private market for plans such that businesses would not adopt their own 401(k)s or comparable alternatives? Would some employers decide to no longer offer their own plans at all? Or, alternatively, could these programs encourage new plan growth?
Preliminary data from annual filings to the U.S. Department of Labor by employer-sponsored plans suggests that in states that have created what are known as auto-IRAs, employers with plans continue to offer them and businesses without plans are still adopting new ones at similar or higher rates than before the state options were available. Typically, the state-sponsored individual retirement accounts (IRAs) automatically enroll workers at eligible employers, though employees can choose to opt out.
The ability to participate in such workplace retirement savings plans has proved to be one of the most significant factors in helping people save for life after their working years. According to the Center for Retirement Research, very few workers without access to a workplace plan save on their own, such as through traditional IRAs. And a Social Security Administration analysis based on tax records indicates that about a quarter of private sector workers nationwide don’t have access to a plan through their job. The most recent complete data is from 2015, but researchers don’t think there have been significant changes since.
Research by The Pew Charitable Trusts, meanwhile, has found that workers without access to a savings plan are often lower-paid and more likely to be people of color. Pew work has also shown that nontraditional or contingent workers have lower rates of access than workers in traditional employment relationships.
To address these issues, eight states in recent years have sought to boost access by establishing auto-IRA programs, which typically are state-sponsored and privately managed. Employees contribute a portion of their regular pay; they can raise or lower the starting amount or choose not to participate at any time. The first of these programs was enacted in Illinois in 2015, followed by Oregon, California, Connecticut, Maryland, New Jersey, Colorado, and Virginia.
Oregon, Illinois, and California have begun enrolling workers and putting contributions into the accounts. These three states are still scaling up their programs, but they have all been in operation long enough to provide indications of employer trends.
Earlier research shows general support among small- and midsize-business owners for providing such savings options. In 2017, Pew published the results of a national survey of small-business owners and benefits managers that detailed their views of hypothetical auto-IRA programs. Among those with retirement plans and five to 250 workers, only 13% said they would drop theirs and enroll in such a program if launched in their state.
Among small and midsize employers without plans, 51% said that they would start their own plan rather than enroll workers in the state-sponsored program. That survey also suggested one reason that these employers might be prompted to adopt their own plans: Most of them won’t offer retirement benefits until they are financially stable and already offering other benefits. The availability of a statewide auto-IRA might encourage those employers that have the means but have not decided to sponsor their own plans.
Early federal data on employer-sponsored plans in the three states gives a sense of the new landscape and shows that employers are making decisions in line with those in other states. Since 2013, before the first state auto-IRA programs were introduced, the percentage of new plans as a share of all employer-sponsored plans increased nationwide from roughly 6% to nearly 8% by 2019.
The three states implementing state auto-IRA programs show a similar trend—with the proportion of new plans holding steady or increasing in each. In 2019, for example, Oregon and California had some of the highest proportions of new plan adoption, with Illinois’ proportion just slightly lower than the national average.
During the same period, the proportion of employer sponsors terminating or ending their plans was consistently about 4% of all plans, both nationwide and in the states implementing auto-IRA programs. The share of plans that were ended began to trend down slightly toward the end of the period: The U.S. average and the proportion of plans terminated in California, Oregon, and Illinois fell to just 3% in 2019.
This early evidence from California, Oregon, and Illinois indicates that auto-IRAs appear to complement the private sector market for retirement plans such as 401(k)s. Some employers may be moving toward plan sponsorship in response to the state auto-IRA programs. Meanwhile, those that cannot afford their own plans can take advantage of a no-cost, basic savings program for their workers. Further research and study will be needed to better understand how employers and plan sponsors in states implementing auto-IRA programs respond to such initiatives and if they do so by adopting their own plan or doing away with one they already have.
How the Retirement Plan Adoption Rates Were Calculated
The data comes from federal Form 5500 Annual Reports, a required regulatory filing to the U.S. Department of Labor for sponsors of retirement and welfare benefit plans. These reports are the primary source of information about the operations, funding, and investments of these plans. Pew drew on data from Form 5500 and Form 5500-SF, or short form—typically for plans with fewer than 100 participants—from 2013 to 2019 to develop a comprehensive picture of retirement plans over this period.
Using the benefit codes provided by the filers, Pew could limit the review to employers offering retirement plans, including defined benefit and defined contribution plans. Plan sponsors are required to file only once per plan, which means that a plan covering participants at different locations across several states would be represented only in the state where it was filed. As such, there are likely plans attributed to one state that include workers in one or more additional states.
Additionally, only data collected through 2008 is considered complete and will not change. Data from 2009 continues to be updated (typically around the beginning of each month). Data from 2013 to 2016 was downloaded in February 2018. Data from 2017 and 2018 was downloaded in February 2020, and data from 2019 was downloaded in October 2020.
John Scott is the director of and Theron Guzoto is a principal associate with The Pew Charitable Trusts’ retirement savings project.