Executive Summary

There is no nationwide system of employer-sponsored or administered retirement savings programs in the United States. Every employer may voluntarily provide such a savings program, but new employers, smaller employers, or employers who have workers that may not stay with the company for very long may find the financial or administrative obstacles to establishing and administering such a program to be difficult to overcome.

America’s voluntary, employer-based, retirement system has left nearly half of American workers without access to a retirement savings plan at work. Those without access tend to be younger, low-earner, lower-educated, female, or minority workers employed in small firms. Evidence shows that having access to a retirement savings plan at work dramatically boosts savings rates.

Some states, recognizing that a large wave of future retirees with little to no savings will increase state costs for means-tested social services, have started to provide a variety of options to increase worker access to and participation in retirement plans. These plans include: multiple-employer plans that reduce the cost of individual firms in providing retirement plans by pooling their resources and sharing costs; voluntary payroll contribution plans for workers; retirement marketplaces where employers can more easily find low-cost retirement plans; and mandatory automatic-enrollment IRAs. The most commonly adopted approach is the automatic-enrollment IRA, or auto-IRA, where those employers who do not otherwise provide a retirement plan for their employees are required to automatically enroll their employees in the state IRA plan. Most of these plans are still relatively new, but we are now able to look across the landscape and start to see some patterns that help shine a light on how these programs are working for those who, without such a plan, would likely not have any retirement savings vehicle.

State auto-IRA plans appear to be successfully targeting those workers without retirement plan access. Savings levels have increased both for auto-IRA participants, as well as for those whose employers have opted to provide access to a 401(k) instead. Workers and employers both express satisfaction with the auto-IRA system. However, challenges remain in securing employer compliance with the auto-IRA mandate, particularly for smaller firms. When individuals are enrolled, as many as one-fifth to one-third opt out of the program, which is voluntary to workers, with the primary reason for opting out being not having enough to save. Among those who remain enrolled, an effective withdrawal rate of 27 percent further reduces the possible investment earnings on contributions. The inability of employers to contribute to these plans also lowers the potential for savings. These factors may leave workers without the savings these auto-IRA plans intend to provide.

To ensure these plans are able to fulfill their potential promise, additional work, both in application on the ground and in research, is required. State plan sponsors need to better understand who is opting out, who is withdrawing funds early from their account and why, and how those withdrawn funds are being used. That information is needed to evaluate what changes could ensure fewer people opt out of these plans and keep their contributions invested for their retirement.

State plan sponsors need to better understand who is opting out, who is withdrawing funds early from their account and why, and how those withdrawn funds are being used.

Nearly half of all working Americans lack access to a retirement plan at work.

Unfortunately, due to the voluntary structure of the employment-based retirement savings system, about 48 percent of working Americans currently lack access to a retirement savings account through work.1

However, this lack of access is not distributed equally among all workers. Of those more likely to work with employers who do not offer a retirement savings plan are:

  • Individuals who are younger (age 18–34, where 57 percent lack coverage);
  • Hispanic (64 percent of whom lack coverage) or Non-Hispanic Black workers (53 percent of whom lack coverage);
  • Those whose education levels are less than a bachelor’s degree, where at least half lack coverage (but particularly those without a high school education, of whom 76 percent lack coverage);
  • Women (49 percent of whom lack coverage);
  • Those who work in firms with fewer than 100 employees, of whom 52 percent or more lack coverage (but particularly those with 10 or fewer employees where 78 percent lacked coverage); and
  • Those in the two lower earnings quintiles (64 percent of the second lowest and 79 percent of the lowest earners lack coverage).2

From Congressional Research Service analysis of the Bureau of Labor Statistics, we also know those who work in part-time positions or in non-unionized jobs also have lower access to retirement plans.3

Access to a retirement plan at work is critical to ensure sufficient retirement savings.

Although workers have the option to establish an Individual Retirement Account (IRA) outside of work and put a portion of their earnings into those tax-advantaged accounts, most do not.4 Only 12 percent of American households save for their retirement outside of the workplace through an IRA.5 Lack of awareness of options and simple inertia can keep workers from establishing such accounts on their own. AARP indicates that workers are fifteen times more likely to save if they can do so through a workplace plan.6 Therefore, the employer link to retirement savings is essential for savings uptake, and those whose employers do not offer such an option may find themselves vulnerable to financial insecurity in their retirement years.

States have begun offering state-sponsored retirement plans.

While efforts to establish a national system of employer administered retirement accounts have met with resistance,7 some states—there are currently nineteen of them, plus two cities—have embarked on a path to establish a state-administered retirement savings vehicle.8 The goals of these plans are both to enhance individual retirement savings and reduce the future costs of retirees who, due to insufficient savings, may ultimately rely heavily on state social service programs.

Filling the Retirement Savings Gap

The goal of these state-administered retirement savings plans is to provide a retirement savings option for private sector workers whose employer does not offer a pension or a defined contribution plan, such as a 401(k) or 403(b). As noted above, nearly half of all Americans work with employers who do not offer a pension or retirement savings plan. States are stepping in to either directly provide such access or to reduce the burden on employers in pursuing such a plan.

Budgetary Concerns About an Aging Demographic

Another driver of the states’ interest in establishing these retirement savings plans is state budget projections. With the aging of the population, there will be a 50 percent increase in the portion of individuals over age 65 from 2020 to 2040.9 Older individuals who are not financially secure may be eligible for a number of Federal and state programs that provide either in-kind or cash income supports. For example, low-income, low-asset individuals age 65 and older can qualify for the Medicaid health program, funded jointly by the federal government and the states. Medicaid provides health coverage to 7.2 million low-income seniors who are also enrolled in Medicare.10 A Pew-commissioned study conducted by Econsult Solutions Inc. (ESI) found that, if baseline savings behaviors continue, the portion of economically vulnerable older Americans would increase by 43 percent, increasing the costs of both federal and state means-tested programs supporting the older population.11 Therefore, if these state retirement savings programs were to increase individuals’ savings above what they would have otherwise saved given the lack of access to an employer plan, both the federal and state governments could reduce projected expenditures for these means-tested social support programs when these workers retire.

The cost of doing nothing, with the projected aging of the workforce, is massive: billions of dollars for the federal government and the states. ESI found that achieving recommended income replacement levels for new retirees who begin with the existing (baseline) income levels in 2020 and transition over time to the recommended (sufficient) levels by 2040 would reduce federal outlays by an estimated $61 billion in 2040, and by $755 billion over the twenty-year period from 2021 to 2040.12 If all retirees, not just new retirees, were to achieve the recommended levels starting in 2020, the cumulative reduction in federal outlays would be $990 billion.

The cost of doing nothing, with the projected aging of the workforce, is massive: billions of dollars for the federal government and the states.

Because ESI projected that virtually every state would have 50 percent or more of their senior population considered economically vulnerable (with the exceptions of Hawaii, Maryland, and New Hampshire) in 2040, they also projected that, under the existing savings trends, every state would expend billions more over the next twenty years in funding programs for its economically vulnerable aged population.13 Overall, ESI projected that the total cost to the states from insufficient savings would grow from $10.6 billion in 2020 to $20.5 billion in 2040, and to a total $334 billion cumulatively over the twenty-year period from 2021 to 2040. Most of these expenditures come from Medicaid, Medicare Part D, the Supplemental Nutrition Assistance Program (SNAP), and various Older Americans Act programs, as these are all means-tested programs, and the states have administrative costs, costs due to required state match formulas, or supplemental state spending.

States have adopted a variety of retirement plan structures.

As with most programs that are operated at the state level, there is variation in how the states have approached filling the retirement savings gap. There are, generally speaking, four different types of state-facilitated retirement savings plans: automatic-enrollment Individual Retirement Accounts, multiple employer pension plans, retirement marketplaces, and voluntary Payroll Deductions IRAs.14

Automatic-enrollment Individual Retirement Accounts are the most common approach.

Automatic-enrollment Individual Retirement Accounts, also known as auto-IRAs, in which employers deduct a portion of pay from an employee’s paycheck and deposit it into the employee’s own IRA, are the most prevalent type of plan structure, having been adopted by fifteen out of the nineteen states with a plan. However, only seven states have active plans as of July 2023 (California, Colorado, Connecticut, Illinois, Maryland, Oregon, and Virginia).15

Auto-IRAs are mandatory for nonexempt employers. Typically, to be exempt from the mandate, employers either are new businesses, have a small number of employees, or already offer an employer-sponsored pension plan, making the state option duplicative. Some programs also allow self-employed workers and those who do not work for a participating employer to self-enroll.

The “automatic” part of the auto-IRA is that non-exempt employers are required to automatically enroll their employees in the state-administered retirement savings program. Typically, eligible employees of participating employers are automatically enrolled in a program but can opt out at any time. About 28–30 percent of participants do opt out of the program, though the amount varies by state.16 Employees who change employers or move out of state can keep the same IRA or transfer savings to a different IRA.

Participants can select the type of IRA, the percentage of their salary they want to contribute to the savings plan, and the investment vehicles. However, if participants do not make an active choice, default selections are made on their behalf. The typical default is for a Roth IRA, wherein a worker contributes earnings post-tax, and withdrawals in retirement are tax-free. However, Roth IRAs are intended to target lower income individuals and, as a result, the income limit for single filers to contribute to a Roth IRA is $153,000 in 2023.17 Those single filers earning over $153,000 could not contribute to a Roth IRA and would either have to select a traditional IRA, if that is an option, or opt out of the program. Under a traditional IRA structure, workers’ contributions to the IRA are pre-tax, but withdrawals in retirement are taxed.

The default contribution rates selected by the states range between 3 to 5 percent of earnings, up to the Federal cap for IRA contributions, which is $6,500 in 2023.18 Several state programs also include an auto-escalation feature, where the default contribution rate increases gradually each year until a maximum contribution rate is reached, typically between 8 and 10 percent of earnings.19

The default investment vehicle for all of the active states is known as a target date fund, which is based on the age of the employee. Employees may choose from other options, but if no choice is made, a target date fund is the default.20

A state retirement board typically oversees each program, is often chaired by the state treasurer or comptroller, and is responsible for making program decisions, such as contracting with an IRA provider.

Voluntary Payroll Deduction IRAs are similar to auto-IRAs, but are not mandatory.

The approach of the voluntary Payroll Deduction IRA is very similar to that of the auto-IRA. The main difference is that employer participation is voluntary instead of mandatory. As a result, it is unclear whether states can auto-enroll workers. Only one state, New Mexico, has adopted this approach, but the program has not yet started.21

Multiple employer pension plans are the second-most-common option.

Under a multiple employer pension plan (MEP), a group of unrelated businesses may voluntarily join together to offer a 401(k) plan. (These are not to be confused with a multiemployer plan, which is negotiated between a union and multiple contributing employers, typically within the same industry.) MEPs can reduce the administrative and fiduciary burdens that small employers may face if they were to set up a 401(k) program on their own. In addition to the higher contribution thresholds permitted under the MEPs compared to the auto-IRA plans,22 MEPs permit employer contribution to the accounts on top of worker contributions. MEPs allow employers to automatically enroll their employees in the plan, but workers retain the ability to opt out of the plan.

The Commonwealth of Massachusetts was the first to start such a plan, but it was targeted to smaller, non-profit organizations with twenty or fewer employees. In Massachusetts, 77 percent of participating employers have elected to offer matching or nonelective employer contributions. Missouri has just started their plan, which applies to all employers with fifty or fewer employees or the self-employed, not just non-profit organizations. Vermont, which had previously established a MEP, has recently switched to an auto-IRA approach instead.23

Retirement marketplaces reduce the burden on employers to identify a retirement savings plan.

Retirement marketplaces are online systems established by the state to lower the burden on employers in identifying a low-cost retirement savings plan. Only Washington State has an active retirement marketplace at this time, applicable to firms with fewer than 100 employees, the self-employed, or sole proprietors. New Mexico has recently approved the establishment of a marketplace, in addition to their voluntary Payroll Deduction IRA.24

Are state-facilitated retirement savings accounts effective?

As is evident from the information above, most of these plans are still fairly new. California, Illinois, and Oregon have been operating the longest, at between four and six years. To evaluate whether these plans will achieve their goal of improved self-sufficiency in retirement, we would need an entire cohort to have had access to these plans throughout their working careers, over a period of forty years or so of work. Although not enough time has passed to provide that kind of analysis, we are able to see some initial results, primarily based on the experiences of these early-adopting states. The remainder of the report focuses on the auto-IRA plans adopted by these three longest operating programs.

What are the success stories?

Early evidence shows that access to a retirement savings plan through state-facilitated auto-IRAs increases the portion of workers who have access to and participate in a retirement plan. In at least one survey, workers participating in the program appear to be happy with auto-IRAs and feel more financially secure. Participating employers also appear satisfied with the program. Additionally, it appears that the presence of a mandatory state-facilitated auto-IRA serves as an inducement for companies to start offering 401(k) plans at work, rather than participate in the auto-IRA program.

Auto-IRAs increase the number of workers who have access to and participate in a retirement plan. According to Georgetown University’s Center for Retirement Initiatives, as of the end of June 2023, the five states with active auto-IRAs had a total of nearly $914 million in savings in the state plans.25 There are a total of nearly 695,000 funded accounts, meaning nearly 695,000 individuals are currently saving for retirement in these programs. These individuals may not otherwise be saving anything at all for retirement.26 Average contribution rates vary across the states, from a low of 3.3 percent to a high of 6.5 percent. The average contribution amount ranged between $108 and $189 per month, and the average total account balance ranged from $422 to $1,754. The lower average numbers tend to be in the newly established programs in Connecticut and Maryland and the higher average numbers tend to be in the longer-lived states of California (active for four years), Illinois (active for five years), and Oregon (active for six years).

Initial data indicates workers participating in auto-IRAs are younger, more diverse, have lower educational attainment levels, are lower earners, and work in service industries. The majority of participants in the Illinois Secure Choice program are younger than age 35, and are overrepresented relative to their share of the Illinois workforce and those eligible to participate in the program.27 Black workers make up about 25 percent of participants and are overrepresented as participants in Illinois Secure Choice relative to their share of those eligible to participate in the program (21 percent), and are overrepresented compared to their share of the Illinois workforce overall (12 percent).28 Hispanic workers comprise 27 percent of program participants, which is larger than their share of the Illinois workforce overall (19 percent), but not as high as the population of Hispanic workers who are eligible to participate in the program (36 percent).29 Those with a high school education or less represent 46 percent of the population participating in Illinois Secure Choice, a level much higher than their 27-percent contribution to the workforce.30

Illinois Secure Choice has begun a data analytics process to match participants in the auto-IRA program with a consumer file from a data analytics firm. This effort is ongoing and results thus far are preliminary based on savers as of December 21, 2022 and a 67 percent match rate between participants and the consumer file. Similar to the Pew study, these preliminary results show 49 percent of savers are below age 35, about 54 percent are women, 41 percent have annual earnings below $50,000, and 22 percent have earnings of less than $30,000 per year.31 The matched data also shows about 23 percent of participants are African-American, but shows that a smaller percentage (18 percent) are Hispanic.32 The overrepresentation of African-American and Hispanic workers in the program is not surprising, given they are also more likely to be employed in industries with employers that do not offer retirement plans. The vast majority work for the service-sector industry, such as restaurants, staffing agencies, and health service companies, and about half live in suburban areas.33

Participating employees and employers appear satisfied with the state-facilitated auto-IRA programs. The Pew Charitable Trusts conducted a survey of individuals eligible to participate in the Illinois Security Choice program over a period of approximately a year, from March 2020 through April 2021, at approximately six-month intervals.34 Of the participants in the program, 96 percent were satisfied or neutral in their evaluation of the program, with almost 40 percent indicating that they feel more financially secure. Pew research conducted on the OregonSaves program indicated that 73 percent of employers participating in the Oregon Auto-IRA program were satisfied or neutral on their experience.35

Auto-IRAs do not crowd out already existing employer retirement plans, and may induce employers to start offering their own plans. An initial concern among some had been whether providing a low-cost, low-administrative burden option within a state might cause some employers to drop their employer-based plans and instead make use of the state auto-IRA option. Recent research indicates that there has not been a decline in pension offerings in the areas where these state-administered programs are in place. In fact, the existence of the auto-IRA mandate may lead some firms to establish 401(k) plans as a preferred alternative, which are exempt from the mandate.36 Pew Charitable Trusts found that in the first year after the Oregon, Illinois, and California plans launched, there was a 35-percent higher growth rate in 401(k) plans in private employers in those states relative to states without an auto-IRA plan. Furthermore, in later years, employers continued to offer work-based plans at higher rates than before the auto-IRAs were launched. Rates of plan termination by employers who were already offering an employment-based plan were either lower than or the same as the national average, indicating the auto-IRAs were not encouraging employers to drop their existing plans.37

What are the possible concerns so far?

There are four primary concerns thus far based on the trends in the longest-running auto-IRA plans: 1) employer noncompliance with the auto-IRA mandates; 2) high opt-out rates among participants; 3) high effective withdrawal rates of assets; and 4) no opportunity for employer contributions.

Employer noncompliance with state auto-IRA mandates remains a concern. Although the auto-IRA programs currently in effect are mandatory for certain employers, depending on whether they already offer a retirement savings vehicle and the number of employees, some states are still seeing a relatively large percentage of employers who have not complied with the state mandate. For example, despite penalties for non-compliance of up to $750 per employee,38 CalSavers has had difficulty getting employers to participate in the program, with about 32 percent either having taken no action, or having registered but not having uploaded their roster of employees or started payroll deductions.39 Although the employer registration rates are high, at a total of about 90 percent across all four waves of the rollout, the portion of registered employers that have uploaded employee rosters remains lower, at about 85 percent, and the portion of registered employers that have started payroll deductions on behalf of employees is much lower, at about 35 percent.40 Similarly, as of July 2023, about 47 percent of Illinois firms that have registered with Illinois Secure Choice have submitted payroll deductions since the program’s inception.41 Illinois assesses a fee of $250 per employee for the first calendar year of non-compliance, and $500 per employee for any subsequent year of non-compliance.42

Employer compliance in CalSavers varies by firm size: it has been relatively high across all of the stages of employee enrollment for larger employers (those with over 100 employees), although most of those larger firms (about 76 percent) are exempt. Still, of those non-exempt employers, about 81 percent have undertaken active payroll deductions on behalf of employees. Compliance drops along with firm size, with the smallest firms (with fewer than five employees) struggling the most with compliance. In these firms, the owner is likely to be handling all of the company’s administrative roles, may use more manual processes for payroll, and experience a disproportionate burden in complying with the auto-enrollment plans. Of these smallest firms, only 29 percent are exempt. Of the remaining 71 percent that are not exempt, only about 19 percent have started payroll deductions on behalf of employees. However, the most common size bracket into which California employers fall is the five to fifty employee range, with 84 percent of firms falling into that size. Among those employers, approximately 46 percent are exempt. Of the remaining non-exempt firms, only about 27 percent have begun making payroll deductions on behalf of employees.43

Issues related to payroll technology have been cited as a challenge with compliance among smaller firms.44 However, so long as full firm compliance with payroll deductions remains in the 19 to 27 percent range among non-exempt firms, that greatly reduces the number of employees who are able to fully participate in the CalSavers program. As a result, most of the CalSavers focus in the coming couple of years will be increasing employer compliance.45

In terms of key lessons learned through the rollout experience that other states may learn from, the CalSavers executive director highlighted the following: 1) the importance of having a dedicated team assigned to the rollout of such an initiative, because the work is simply too large of a lift to add to any existing staff position descriptions; 2) the essential outreach to key stakeholders—such as chambers of commerce and small business groups that represent various demographic or industrial groups—by the leaders of the auto-IRA initiative, to educate and ensure customer service needs will be met; 3) that carrots do not work as well as sticks when it comes to enforcement of the mandate with employers, since financial penalties directly affect their bottom line; 4) the need for sufficient flexibility in the authoring legislation to allow the program to alter implementation strategies as needed to ensure program success; and 5) the advantage of a quicker launch across all sizes and types of businesses instead of a longer rollout.46

High opt-out rates could leave target populations unprotected in retirement. All of the auto-IRA plans are voluntary to employees. They may opt out at the start of the program or at any time they wish. The higher the opt-out rate, the lower the portion of individuals targeted by these programs that will ultimately benefit from savings in retirement. As of June 2023, effective opt-out rates varied between 36 percent in California, 34 percent in Illinois, 23 percent in Oregon, 19 percent in Connecticut, and 24 percent in Maryland.47 As a result, between one-fifth and one-third of all targeted participants are not participating in these savings retirement programs. According to the executive director of CalSavers, the top reason plan participants provide for opting out of the program is that they can’t afford to contribute to the plan.48 This finding is consistent with an analysis of the early years of the Oregon Saves program.49 In Illinois Secure Choice, women comprise a larger portion of the workforce eligible to participate in the program, but are slightly less likely to participate, possibly due to wage differences relative to men.50 Those with dependents under age 18 were also slightly less likely to participate in the plan.51 As Pew’s report notes, “since these participation trends are likely linked to underlying wage differences, broader policies aimed at improving wages may be necessary to improve program participation.”52

High rate of withdrawals relative to contributions may reduce retirement savings. According to the Georgetown University Center for Retirement Initiatives (CRI), the total withdrawals in the month of June 2023 across all five active auto-IRA states was nearly $242 million. Out of a total asset level in June 2023 of $914 million, that is a 27 percent effective overall withdrawal rate.53 The same was true in the month of May 2023. As CRI notes, the ratio of total withdrawals to total contributions in June (since the inception of each program) is about 17 percent in California, 24 percent in Illinois, and 30 percent in Oregon.54 Assets that are withdrawn do not earn interest, and depending on the reason for the withdrawal, individuals may face penalties for accessing their accounts early. Therefore, high withdrawal rates may reduce the effectiveness of these programs in augmenting income in retirement.

Employers are prohibited from contributing to these plans on behalf of their employees. Research shows that when an employer provides an incentive for workers to save by contributing an amount equal to a portion of what the employee themselves saves, employees save more.55 Unfortunately, these state auto-IRAs are prohibited from having employer contributions made to them. As a result, the additional incentives for workers to remain in the plan instead of opting out, or to contribute an amount higher than the default savings rate, are lost. Furthermore, because these workers, by definition, do not have access to a 401(k) or other savings plan through their employer, where such employer contributions are allowed, these workers have to save more from their own earnings to achieve the same level of savings as a similar employee who does have access to such a retirement plan and employer match.

However, the savings levels permitted under federal law are smaller for IRAs than for 401(k)s. In 2023, the contribution limit is $22,500 in 2023 for a 401(k) plan, with the 50-and-over crowd allowed an extra $7,500. For IRAs, however, the contribution is only $6,500, with those 50 and over allowed an extra $1,000, there is only so much an individual can do to make up the difference from not having an employer match. Those with 401(k)s through their work also tend to already be higher earners than those who are enrolled through state auto-IRA plans, and for whom an employer match would be that much more valuable.

What remains unknown?

Who is opting out?

While it is clear from prior research and data that the number one reason participants give for opting out of the auto-IRA program is that they can’t afford to save, we have little information about those of whom this demographic group comprises. If those who are opting out of these retirement savings plans are the same individuals the states hoped would save—and thereby avoid adding to pressures on state budgets—they may be missing the mark. Addressing this question would require a full picture of the earnings of these workers, as well as other sources of household income and expenses. As early analysis of the Oregon Saves program indicated, it may be that those who opt out of these programs do so for economically valid reasons.56 It may be that, for certain groups of workers, the challenge lies not at the end of the decision pipeline with the retirement savings decision itself, but at the front end with earning too low of a wage to be able to afford to save. Thus, the success of these programs may not be fairly evaluated on the entire population, when these programs have no control over wages. It also calls into question the level of earnings that someone working full time requires in order to cover their costs of living and still have sufficient discretionary earnings with which to save for retirement.

Why are individuals withdrawing funds from their auto-IRAs?

States do not have questions on their withdrawal forms for how non-normal distributions from their accounts are being used by plan participants. At most, they know whether the distributions are “normal” or if they will potentially be subject to IRS withholding or penalties. If withdrawals are being made for retirement, then that is not a concern. However, if the withdrawals are being used for other purposes, then these retirement accounts might not ultimately meet their goals of providing retirement income for plan participants. On the other hand, if these withdrawals are being used to handle emergency expenses in lieu of perhaps a more expensive alternative such as same-day loans, and so help participants avoid debt with high interest rates, then perhaps such withdrawals are less concerning. At this juncture though, there is simply no data available on how these non-normal distributions are being used to evaluate if these decisions are economically rational for the participants, even if they run counter to program goals. Because we don’t know why people are making withdrawals, we also cannot modify the program to ensure participants are fully educated about their choices and making optimal decisions.

Who is making non-normal withdrawals of funds from their auto-IRAs?

As with participants who opt out, states also don’t know who is taking full or partial withdrawals of their auto-IRA funds as non-normal distributions. Again, a better understanding of the participants’ annual earnings, total income, and household expenditures would provide greater insight into whether participants are making optimal decisions.

Will the savings amount achieved be sufficient to aid individuals in retirement and reduce state financial burdens?

The answer depends on multiple variables: 1) how many employers are complying with the auto-IRA mandates versus what states had expected; 2) how many employees opt out of the plan; 3) how many employees make significant withdrawals that reduce the size of their savings base, and thus the amount of interest it may earn; 4) of those who don’t opt out or withdraw their contributions, what contribution levels these participants choose; 5) the earnings of auto-IRA plan participants; 6) the number of years individuals contribute to their auto-IRA; 7) which investment vehicles participants choose; and 8) the rate of return on those investments.

As noted above, many states are struggling to achieve full, or even 50-percent, employer compliance with the auto-IRA mandates in firms of certain sizes. If employers are not automatically enrolling and sending payroll contributions to the IRAs on behalf of employees, then states are certainly not going to achieve their goal of increased retirement savings at the levels required to reduce strains on their long-term budgets.

High opt-out rates—ranging from 19 to 36 percent—reduce the portion of employees contributing to the auto-IRAs further. Effective withdrawal rates of between 17 and 30 percent reduce the savings base upon which interest will be paid.

As of June 2023, contribution rates among participants are averaging between 3.3 percent in Connecticut, a newer program, to 5.10 in Maryland, 5.15 in California, 5.87 percent in Illinois, and a high of 6.50 percent in Oregon.57 With monthly contribution amounts ranging between an average low of $108 in Connecticut to an average high of $189 in California, it is possible to back out the average monthly earnings of participants.58 The average monthly earnings run from a low of $2,589 in Illinois, to a high of $3,670 in California. If those monthly averages are representative of what the average individual earns each month in those states, which may or may not be true given the high proportion of part-time workers represented among plan participants, then average annual earnings run from a low of $31,073 in Illinois to a high of $44,039 in California. In most of the states with active auto-IRA programs, the average annual earnings levels of participants fall between the tenth and twenty-fifth percentile of all earnings in the state.59 Because these programs target those who do not have access to an employer sponsored retirement plan, it is expected that most of these participants will work in industries or occupations that have relatively low pay and/or experience higher turnover rates. As a result, contributions levels are likely to remain low in absolute dollar terms, even if percent contribution rates increase over time.

Among the longest-running state auto-IRA programs, the average balance was $1,320 in California, with four years in operation; $1,017 in Illinois, with five years in operation; and $1,754 in Oregon, with six years in operation.60 Even if one assumed that all participants remained in the program (did not opt out) and contributed at the full default rate set by each state—which have auto-escalation features that increase the contribution rate automatically each year—it is unclear whether the amounts saved will have a meaningful impact on individual retirement security. Estimates for an auto-IRA that covers all employers, regardless of size, project an average account balance of $60,600 at age 65 in 2040, due to the large number of low earners being targeted.61 For those individuals, this amount, annuitized today, would provide a monthly income of approximately $377.62 While this amount is certainly better than the $0 amount that would likely have otherwise been saved, it may not be sufficient to ensure individuals do not still require various means-tested social services from the states.

Are there other stakeholder concerns?

Stakeholders have also expressed concern that state-administered plans in some states might lack adequate measures to protect participants’ benefits (e.g., whether deposits would be made in a timely manner, fees would be reasonable, and investment choices prudent) as they are not covered by the Employee Retirement Income Security Act. Depending on the state, existing state law or provisions in authorizing legislation might alleviate some of these concerns.

Employers operating in multiple states could be required to participate in several programs, which could be administratively challenging. For example, employers might have to monitor employee eligibility for different state programs based on residence or office location. However, these employers are already navigating differences in unemployment insurance, workers’ compensation, and paid family leave laws aside from the basic differences in state tax laws.

Some also have concerns about unintended consequences of the mandates, inducing more employers to misclassify workers as independent contractors, which is already an ongoing challenge.

While state auto-IRAs are showing promise at reaching target populations, more work remains if we are to reach the goal of retirement security and reduced dependence on state social services.

Initial results from various studies of state auto-IRAs show that these plans are providing a retirement savings vehicle to those employee populations typically left out of such arrangements in the workplace. Even better, a sizable portion of eligible workers are staying enrolled in these plans. Thus, it seems that the lack of retirement savings is primarily due to the lack of opportunity to do so within the workplace, not due to a lack of desire to save. However, if these plans are to fulfill their promise, additional work, both in application and in research, is required.

As noted by all of the executive directors interviewed for this piece, employer compliance with auto-enrollment mandates declines with firm size, and some problems remain regarding payroll providers providing the necessary assistance, or having the technology, to automatically send deposits from worker paychecks to the third-party processor for investment in the workers’ retirement funds. Furthermore, high opt-out rates and effective withdrawal rates remain a concern, as both reduce the population of savers as well as the amounts accrued in retirement accounts. Additional research is required to better answer the question of what changes are required to ensure fewer people opt out of these plans and keep their contributions invested for their retirement.

Notes

  1. John Sabelhaus, The Current State of U.S. Workplace Retirement Plan Coverage, Pension Research Council Working Paper PRC WP2022-07, Pension Research Council, The Wharton School, University of Pennsylvania, March 2022, https://repository.upenn.edu/server/api/core/bitstreams/53f6523d-7c37-4052-add0-676774a6181a/content.
  2. John Sabelhaus, The Current State of U.S. Workplace Retirement Plan Coverage, Pension Research Council Working Paper PRC WP2022-07, Pension Research Council, The Wharton School, University of Pennsylvania, March 2022, https://repository.upenn.edu/server/api/core/bitstreams/53f6523d-7c37-4052-add0-676774a6181a/content.
  3. John J. Topoleski and Elizabeth A. Myers, Worker Participation in Employer-Sponsored Pensions: Data in Brief, Congressional Research Service, Updated November 23, 2022, https://sgp.fas.org/crs/misc/R43439.pdf.
  4. Anqi Chen and Alicia H. Munnell, Who Contributes to Individual Retirement Accounts?, April 2017, Number 17-8, ,https://crr.bc.edu/wp-content/uploads/2017/04/IB_17-8-1.pdf.
  5. The Role of IRAs in US Households’ Saving for Retirement, 2019, ICI Research Perspective December 2019 (Volume. 25, Number 10), https://www.ici.org/doc-server/pdf%3Aper25-10.pdf.
  6.  Catherine Harvey, Access to Workplace Retirement Plans by Race and Ethnicity,  AARP Public Policy Institute, February 2017, https://www.aarp.org/content/dam/aarp/ppi/2017-01/Retirement%20Access%20Race%20Ethnicity.pdf.coredownload.pdf.
  7. Tara Siegel Bernard, Treasury Ends Obama-Era Retirement Savings Plan, New York Times, July 28, 2017,  https://www.nytimes.com/2017/07/28/business/treasury-retirement-myra-obama.html.
  8. State-Facilitated Retirement Savings Programs: A Snapshot of Program Design Features, Georgetown University, Center for Retirement Initiatives, State Brief 23-03, June 30, 2023, https://cri.georgetown.edu/wp-content/uploads/2023/03/cri-state-brief-snapshot.pdf.
  9. The Cost of Doing Nothing: Federal and State Impacts of Insufficient Retirement Savings, May 2023, ESI Econsult Solutions, https://econsultsolutions.com/wp-content/uploads/2023/05/Impacts_of_Insufficient_Retirement_Savings_May2023.pdf.
  10. Seniors & Medicare and Medicaid Enrollees, accessed September 11, 2023, https://www.medicaid.gov/medicaid/eligibility/seniors-medicare-and-medicaid-enrollees/index.html.
  11. The Cost of Doing Nothing: Federal and State Impacts of Insufficient Retirement Savings, May 2023, ESI Econsult Solutions, https://econsultsolutions.com/wp-content/uploads/2023/05/Impacts_of_Insufficient_Retirement_Savings_May2023.pdf.
  12. The Cost of Doing Nothing: Federal and State Impacts of Insufficient Retirement Savings, May 2023, ESI Econsult Solutions, p. 50, https://econsultsolutions.com/wp-content/uploads/2023/05/Impacts_of_Insufficient_Retirement_Savings_May2023.pdf.
  13. The Cost of Doing Nothing: Federal and State Impacts of Insufficient Retirement Savings, May 2023, ESI Econsult Solutions, p. 53, https://econsultsolutions.com/wp-content/uploads/2023/05/Impacts_of_Insufficient_Retirement_Savings_May2023.pdf.
  14. State-Facilitated Retirement Savings Programs: A Snapshot of Program Design Features, Georgetown University, Center for Retirement Initiatives, State Brief 23-03, June 30, 2023, https://cri.georgetown.edu/wp-content/uploads/2023/03/cri-state-brief-snapshot.pdf.
  15. State-Facilitated Retirement Savings Programs: A Snapshot of Program Design Features, Georgetown University, Center for Retirement Initiatives, State Brief 23-03, June 30, 2023, https://cri.georgetown.edu/wp-content/uploads/2023/03/cri-state-brief-snapshot.pdf.
  16. Sarah O’Brien, As state-run retirement programs become more popular, participants are expected to have $1 billion in savings this year, CNBC, January 15, 2023, https://www.cnbc.com/2023/01/15/state-run-auto-ira-programs-continue-growing-as-more-options-launch.html.
  17. Internal Revenue Service, Amount of Roth IRA Contributions That You Can Make For 2023, accessed September 11, 2023, https://www.irs.gov/retirement-plans/amount-of-roth-ira-contributions-that-you-can-make-for-2023 .
  18. Internal Revenue Service, Retirement Topics – IRA Contribution Limits, accessed September 11, 2023, https://www.irs.gov/retirement-plans/plan-participant-employee/retirement-topics-ira-contribution-limits.
  19. State-Facilitated Retirement Savings Programs: A Snapshot of Program Design Features, Georgetown University, Center for Retirement Initiatives, State Brief 23-03, June 30, 2023, https://cri.georgetown.edu/wp-content/uploads/2023/03/cri-state-brief-snapshot.pdf.
  20. State-Facilitated Retirement Savings Programs: A Snapshot of Program Design Features, Georgetown University, Center for Retirement Initiatives, State Brief 23-03, June 30, 2023, https://cri.georgetown.edu/wp-content/uploads/2023/03/cri-state-brief-snapshot.pdf.
  21. State-Facilitated Retirement Savings Programs: A Snapshot of Program Design Features, Georgetown University, Center for Retirement Initiatives, State Brief 23-03, June 30, 2023, https://cri.georgetown.edu/wp-content/uploads/2023/03/cri-state-brief-snapshot.pdf.
  22. Auto-IRA contributions are capped at $6,500 in 2023 versus 401(k) contributions being capped at $22,500 for those under age 50 and $30,000 for workers age 50 and over.
  23. State-Facilitated Retirement Savings Programs: A Snapshot of Program Design Features, Georgetown University, Center for Retirement Initiatives, State Brief 23-03, June 30, 2023, https://cri.georgetown.edu/wp-content/uploads/2023/03/cri-state-brief-snapshot.pdf.
  24. State-Facilitated Retirement Savings Programs: A Snapshot of Program Design Features, Georgetown University, Center for Retirement Initiatives, State Brief 23-03, June 30, 2023, https://cri.georgetown.edu/wp-content/uploads/2023/03/cri-state-brief-snapshot.pdf.
  25. Total for California, Connecticut, Illinois, Maryland, and Oregon as of the end of June 2023. Website accessed August 1, 2023: https://cri.georgetown.edu/states/state-data/current-year/
  26. Individuals may have more than one employer, but they will only have one funded account which reflects contributions from all of their participating employers.
  27. Alison Shelton, John Scott and Kim Olson, Demographic Overview of Illinois Secure Choice Program Population, August 29, 2023, https://www.pewtrusts.org/en/research-and-analysis/issue-briefs/2023/08/demographic-overview-of-illinois-secure-choice-program-population.
  28. Alison Shelton, John Scott and Kim Olson, Demographic Overview of Illinois Secure Choice Program Population, August 29, 2023, https://www.pewtrusts.org/en/research-and-analysis/issue-briefs/2023/08/demographic-overview-of-illinois-secure-choice-program-population.
  29. Alison Shelton, John Scott and Kim Olson, Demographic Overview of Illinois Secure Choice Program Population, August 29, 2023, https://www.pewtrusts.org/en/research-and-analysis/issue-briefs/2023/08/demographic-overview-of-illinois-secure-choice-program-population.
  30. Alison Shelton, John Scott and Kim Olson, Demographic Overview of Illinois Secure Choice Program Population, August 29, 2023, https://www.pewtrusts.org/en/research-and-analysis/issue-briefs/2023/08/demographic-overview-of-illinois-secure-choice-program-population.
  31. Data received via email to the author on July 13, 2023 from Christine Cheng, Director of Illinois Secure Choice.
  32. Data received via email to the author on July 13, 2023 from Christine Cheng, Director of Illinois Secure Choice.
  33. Data received via email to the author on July 13, 2023 from Christine Cheng, Director of Illinois Secure Choice.
  34. John Scott and Mark Hines, Many in Illinois Retirement Savings Program Feel Their Financial Security is Improving, April 18, 2022, https://www.pewtrusts.org/es/research-and-analysis/articles/2022/04/18/many-in-illinois-retirement-savings-program-feel-their-financial-security-is-improving.
  35. Pew Charitable Trusts, OregonSaves Auto-IRA Program Works for Employers, April 2, 2021, https://www.pewtrusts.org/-/media/assets/2021/04/oregonsaves_retirement_brief_final.pdf.
  36. Kim Olson, New State Retirement Savings Programs Prompt Increased Private Plan Adoption, December 21, 2022, Pew Charitable Trusts, https://www.pewtrusts.org/en/research-and-analysis/articles/2022/12/21/new-state-retirement-savings-programs-prompt-increased-private-plan-adoption?utm_campaign=2023-01-04+Rundown&utm_medium=email&utm_source=Pew&subscriberkey=00Q0P00000qRv57UAC.
  37. Kim Olson, New State Retirement Savings Programs Prompt Increased Private Plan Adoption, December 21, 2022, Pew Charitable Trusts, https://www.pewtrusts.org/en/research-and-analysis/articles/2022/12/21/new-state-retirement-savings-programs-prompt-increased-private-plan-adoption?utm_campaign=2023-01-04+Rundown&utm_medium=email&utm_source=Pew&subscriberkey=00Q0P00000qRv57UAC.
  38. CalSavers, CalSavers to Impose Penalties for Non-compliant Employers Beginning this Month, January 12, 2022, https://www.treasurer.ca.gov/calsavers/penalties.pdf.
  39. CalSavers, CalSavers Retirement Savings Program Participation & Funding Snapshot as of 6/30/2023, https://www.treasurer.ca.gov/calsavers/reports/participation/june_2023.pdf.
  40.  CalSavers, CalSavers Retirement Savings Program Participation & Funding Snapshot as of 6/30/2023, https://www.treasurer.ca.gov/calsavers/reports/participation/june_2023.pdf.
  41. Illinois Secure Choice, Monthly Dashboard – July 2023, https://illinoistreasurergovprod.blob.core.usgovcloudapi.net/twocms/media/doc/secure%20choice%20monthly%20dashboard_july%202023.pdf  It is important to note that the Illinois program is in the process of rolling out the program to firms with at least 5 employees, who are required to register by November 2023.
  42. Illinois Compiled Statutes, “Illinois Secure Choice Savings Program Act,” (Source: P.A. 98-1150, eff. 6-1-15.), Section 85, available at: https://www.ilga.gov/legislation/ilcs/ilcs3.asp?ActID=3588&ChapterID=68
  43. CalSavers, CalSavers Retirement Savings Program Participation & Funding Snapshot as of 6/30/2023, https://www.treasurer.ca.gov/calsavers/reports/participation/june_2023.pdf.
  44. Author interview with former Executive Director of CalSavers, Jennifer Selenski, August 2, 2023. Some firms indicated multiple attempts to contact and work with their payroll processing firms to comply, but had difficulty reaching someone to speak with about making the change.
  45. Author interview with current Executive Director of CalSavers, David Teykaerts, August 23, 2023.
  46. Author interview with current Executive Director of CalSavers, David Teykaerts, August 23, 2023.
  47. Georgetown University, Center for Retirement Initiatives, Program Performance Indicators, Monthly Percent Change, data accessed for each state on August 1, 2023, https://cri.georgetown.edu/states/state-data/current-year/.
  48. Author interview with current Executive Director of CalSavers, David Teykaerts, August 23, 2023.
  49. John Chalmers, Olivia S. Mitchell, Jonathan Reuter, and Mingli Zhong, Auto-Enrollment Retirement Plans in OregonSaves, Ann Arbor, MI. University of Michigan and Disability Retirement Research Center (MRDRC) Working Paper; MRDRC WP 2021-425, April 2021, https://mrdrc.isr.umich.edu/publications/papers/pdf/wp425.pdf.
  50. Alison Shelton, John Scott and Kim Olson, Demographic Overview of Illinois Secure Choice Program Population, August 29, 2023, https://www.pewtrusts.org/en/research-and-analysis/issue-briefs/2023/08/demographic-overview-of-illinois-secure-choice-program-population.
  51. Alison Shelton, John Scott and Kim Olson, Demographic Overview of Illinois Secure Choice Program Population, August 29, 2023, https://www.pewtrusts.org/en/research-and-analysis/issue-briefs/2023/08/demographic-overview-of-illinois-secure-choice-program-population.
  52. Alison Shelton, John Scott and Kim Olson, Demographic Overview of Illinois Secure Choice Program Population, August 29, 2023, https://www.pewtrusts.org/en/research-and-analysis/issue-briefs/2023/08/demographic-overview-of-illinois-secure-choice-program-population.
  53. Georgetown University, Center for Retirement Initiatives, Program Performance Indicators, Monthly Percent Change, data accessed for each state on August 1, 2023, https://cri.georgetown.edu/states/state-data/current-year/.
  54. Georgetown University, Center for Retirement Initiatives, Monthly Performance Trends in State-Facilitated Retirement Savings Programs, June 2023, July 2023, p.2, https://cri.georgetown.edu/wp-content/uploads/2023/07/Monthly-Trends-in-State-Retirement-Plans-May-2023-to-Jun-2023.pdf
  55. Keenan Dworak-Fisher, Matching Matters in 401(k) Plan Participation, U.S. Bureau of Labor Statistics, Working Paper 434, February 2010, https://www.bls.gov/osmr/research-papers/2010/pdf/ec100020.pdf. Note, however, that participants in 401(k) plans may not be comparable to those in these auto-IRA plans, as 401(k) plans are usually available to higher earners.
  56. John Chalmers, Olivia S. Mitchell, Jonathan Reuter, and Mingli Zhong, Auto-Enrollment Retirement Plans in OregonSaves, Ann Arbor, MI. University of Michigan and Disability Retirement Research Center (MRDRC) Working Paper; MRDRC WP 2021-425, April 2021, https://mrdrc.isr.umich.edu/publications/papers/pdf/wp425.pdf.
  57. Georgetown University, Center for Retirement Initiatives, Comparison of Key Program Metrics and Design Characteristics, June 30, 2023, https://cri.georgetown.edu/states/state-data/current-year/#monthly-data-trends. It is important to note that both CT and MD started their programs in 2022, and CT started with a default contribution rate of 3.0 percent, while the other programs began with a default contribution rate of 5.0 percent. Oregon has the longest running auto-IRA program and thus has higher default contributions with auto-escalations being in effect for a longer period of time.
  58. Georgetown University, Center for Retirement Initiatives, Comparison of Key Program Metrics and Design Characteristics, June 30, 2023, https://cri.georgetown.edu/states/state-data/current-year/#monthly-data-trends. Importantly, if employees have additional work with another employer that does offer a retirement savings plan, then those earnings would not be captured in the auto-IRA system and actual earnings levels would therefore be higher.
  59. Authors calculations for California, Connecticut, Illinois, Maryland and Oregon based on the May 2022 State Occupational Employment and Wage Estimates, Bureau of Labor Statistics, available at https://www.bls.gov/oes/current/oes_ca.htm#otherlinks
  60. Georgetown University, Center for Retirement Initiatives, Comparison of Key Program Metrics and Design Characteristics, June 30, 2023, https://cri.georgetown.edu/states/state-data/current-year/#monthly-data-trends.
  61. Angela Antonelli, What are the Potential Benefits of Universal Access to Retirement Savings?, Georgetown University Center for Retirement Initiatives in conjunction with ESI Econsult Solutions, Inc., December 2020, https://cri.georgetown.edu/wp-content/uploads/2020/12/CRI-ESI-Report-Benefits_of_Universal_Access_FINAL.pdf
  62. Based on an immediately payable, single-life annuity for a 65 year old man living in Washington, D.C.in 2023, using the Charles Schwab annuity calculator at https://www.schwab.com/annuities/fixed-income-annuity-calculator.

Tags: ,