The Tortuous History of HRAs: What Does the Future Hold?

paul-fronstin-webBy Paul Fronstin, EBRI

In October 2018, regulations were issued by the departments of Treasury, Labor, and Health and Human Services — at the direction of an Executive Order by President Trump — to expand the use of stand-alone Health Reimbursement Arrangements (HRAs) by employers of all sizes.

This is another twist in the complicated history of HRAs.

HRAs first became available around 2001, when a handful of employers paired such arrangements with high-deductible health plans.  These employers offered HRAs under then-existing law — which was unclear at the time.  In 2002, the Internal Revenue Service (IRS) released Revenue Ruling 2002-41 and Notice 2002-45 (published in Internal Revenue Bulletin 2002-28, dated July 15, 2002) to provide guidance clarifying the conditions under which HRAs could be provided to workers on a tax free basis, and addressed questions related to the benefits offered under an HRA, the interaction between HRAs and cafeteria plans, FSAs, COBRA coverage, and other matters.  One of the provisions in Notice 2002-45 allowed HRAs to be used on a stand-alone basis to reimburse workers for eligible health insurance premiums, which included premiums for health insurance purchased by workers directly from insurance companies in the individual or non-group market.

Employers never made this provision widely available and, in 2013, the Obama Administration banned the practice in the Department of Labor Technical Release No. 2013-03.

Then in December 2016, the 21st Century Cures Act included a provision known as the Qualified Small Employer Health Reimbursement Arrangement (QSEHRA), which allows certain small employers to use HRAs on a tax-preferred basis to reimburse workers for health insurance purchased on the individual market, as well as deductibles, and cost-sharing more generally.   A year later, President Trump issued his executive order.

Under the newly-proposed regulations, two types of HRAs would be allowed.

  1. A stand-alone HRA that could be used to purchase coverage in the non-group market, with no contribution limit. The HRA must be used to purchase ACA-compliant plans and must meet ACA affordability requirements in order for the employer to meet the shared responsibility requirement.  The HRA will not be subject to ERISA if certain conditions are met.
  2. An “excepted benefit HRA,” where employers would be able to contribute up to $1,800 that workers could use to pay their out-of-pocket cost sharing and/or certain premiums, such as those for short-term health insurance, COBRA, disability insurance, and dental and vision insurance.  When offering an excepted benefit HRA, employers must also offer a group health plan, but workers could decline it and get coverage in the non-group market.  Employers are not allowed to offer both the stand-alone HRA and excepted benefit HRA to the same class of workers.

The administration expects around 800,000 employers will offer stand-alone HRAs by 2024 and beyond.  As a result, some 10.7 million individuals would be covered by such an HRA by 2027 and 6.8 million fewer workers (and their dependents) would have traditional employment-based health coverage.  It is no surprise that there would be fewer people with employment-based health coverage.  Employers have been interested in the concept of “defined contribution (DC) health” coverage and giving workers an HRA that they can use to purchase coverage in the non-group market may be an attractive means of moving to DC health.  Employers never moved in the direction of giving workers a fixed contribution to purchase health insurance for a number of reasons.  Historically, they were hesitant to drop group coverage in favor of offering individual policies because the non-group market was not considered a viable alternative to the employment-based system.  And, more recently, even with the advent of private health insurance exchanges, employers did not embrace them as the initial hype would have expected us to believe.

There is no question that the HRA provision gives employers the means to drop traditional health coverage and go to a “DC health”-type plan.  One of the concerns is that employers will try to structure their plans in such a way to send high-risk employees to the individual market.  The regulations include a number of provisions to prohibit such a discriminatory practice.  However, what if only employers facing the highest premiums in the group market adopted HRAs?   If such a phenomenon occurred, the non-group market would not become more stable, and may see average premiums increase, which the group market would see a reduction in average premiums as higher risk groups left.

Employers may require that there be a viable non-group market for their employees to go to before moving to HRAs.  Stability in premiums may be one requirement, which may make it less likely that multi-state employers move to HRAs given the variation in premium growth across states.  The quality of the benefits offered is another consideration.  For instance, the prevalence of narrow network plans in the non-group market may be something that continues to hold employers back from HRAs.  And of course there is always the uncertainty of future changes in the non-group market, as employers have no control over that marketplace and would have no control over how workers spent HRAs in the non-group market.

There are a number of unanswered questions.  Which employers would go in this direction, under what circumstances, and for which employees?  Would it vary by firm size?  Do the strength of the economy and labor market conditions factor in?  The next recession will be the first recession in history since the insurance market reforms were put into place by the ACA.  It will be the first time in history that a recession was paired with the inability of insurance companies to deny people coverage for pre-existing conditions or to charge them more for such conditions.  It will also be the first time that health insurance premiums for people under 400 percent of the federal poverty level would be subsidized by the federal government during a recession in a meaningful way.  Employers may decide that they no longer need to offer health benefits to be competitive in the labor market during the next recession, and the combination of the insurance market reforms and the ability to give workers tax-free money to purchase health insurance on their own may finally put the future of employment-based health coverage to the test.

The new regulations are expected to be finalized early next year and take effect for plan years beginning on or after Jan. 1, 2020.

The 401(k) Turns 40: It Has Come a Long Way, but What Is Next?

The 40th anniversary of the introduction of Sec. 401(k) to the Internal Revenue Code is particularly significant to me on several levels. As a retirement professional, I have seen my career closely track the ups and downs of 401(k) plans. I’ve watched them evolve from tiny supplemental savings plans, to useful benefits for attracting and retaining workers, to the primary employer-sponsored retirement plan of corporate America. But now as 401(k) plans turn 40, plan sponsors, policy makers, and the industry must decide how they fit into the decumulation phase of workers’ lives.40years

I started saving in my own 401(k) around 1990. It was a brand-new offering by my employer, and I had previously only had an IRA for my retirement planning. At that time, there were just under 98,000 401(k) plans and about 20 million active plan participants. Total assets in these plans at the time were $385 billion.

In the early 1990s, Sec. 404(c) was written into the Employee Retirement Income Security Act of 1974 (ERISA). This new safe harbor allowed plan fiduciaries to not be liable for investment losses suffered by plan participants who self-direct their investments, provided that the plan sponsor met certain conditions. One was offering basic information about plan investment options. Nonetheless, in 1993, when I was still a pretty new investment consultant, a plan sponsor nearly hung up the phone on me when I suggested he offer help to participants on selecting their investments. Like many back then, he feared this would be construed as fiduciary investment advice — for which he could be held responsible.

The Department of Labor (DOL) then issued Interpretive Bulletin 96-1, giving plan sponsors much more latitude to educate participants. However, there was still no safe harbor for advice. I experienced the challenge of this every time I conducted a participant meeting. For example: Standing in front of a refrigerator in a break room at 7 a.m. at a manufacturing plant, I was trying to teach a room full of women (for whom English was a second language) asset allocation. At the end of the presentation, one of the women raised her hand and asked me if I could just tell her which investments to select. Of course, I could not. (Actually, before I could answer to tell her that, her supervisor instructed her to ask her husband. But that is another story.)

The dawn of automatic enrollment came in the year 2000, when the IRS issued guidance on “negative elections.” People like the woman in the break room at the manufacturing plant now could be enrolled into a 401(k) plan without having to figure out their own investment elections. This seemed like a pretty big deal to me, and now that I was director of retirement research at 401(k) recordkeeper Hewitt Associates, I could examine the impact of automatic enrollment on 401(k) participation. Working with David Laibson and James Choi at Harvard University, we found that few people opted out under automatic enrollment. At the same time, few people also increased their contribution rate, even if it was a very low rate such as 2 or 3 percent of pay. I presented this research at various conferences, and invariably someone in the audience explained how they would never offer automatic enrollment because of concerns about state wage garnishment laws. So, adoption remained low.

401(k) plans received a big boost with the Economic Growth and Tax Relief Reconciliation Act of 2001 (EGTRRA), which increased the amount participants could contribute to their plans and created catch-up contributions, among many other provisions. That same year, the SunAmerica Advisory Opinion was issued by the DOL, permitting asset allocation services to be provided to participants in ERISA-covered plans under certain conditions. This launched managed accounts, a type of early robo-advice for 401(k) participants.

Of course, 9/11 also happened that year, and with it, a market collapse. Plan participants generally stood firm: I was in charge of the Hewitt 401(k) index at the time, which saw net transfers of just 1.4 percent of balances in the month of September 2001.

In 2004, the number of plans with a 401(k) feature clocked in at 418,553. There were 44,407,000 active participants in 401(k) plans with total assets of $2.19 trillion. There was also a new type of investment option available: target-date funds. These allowed plan participants to invest in professionally managed portfolios tied to their retirement time horizon. However, few sponsors used them as the plan’s default investment. Again, there was no safe harbor protection around defaults, and many sponsors were counseled that stable-value funds were a more prudent option because the possibility of loss was remote.

The Pension Protection Act of 2006 (PPA) changed many things. It created a safe harbor for automatic enrollment and sanctioned the use of automatic contribution escalation. The latter was the brainchild of Richard Thaler of University of Chicago and Shlomo Benartzi of UCLA. Originally named Save More Tomorrow, automatic contribution escalation—like automatic enrollment—leveraged behavioral finance. In this case, it recognized that the same people who are reluctant to increase savings today may allow their contributions to be automatically increased over time. (Note: Research shows they generally do.) Under the final regulations issued by the DOL in 2007, target-date funds were deemed to be a qualified default investment alternative within the PPA’s automatic enrollment safe harbor. The PPA also provided ERISA pre-emption of state wage garnishment laws. Use of automatic enrollment, automatic contribution escalation, and target-date funds subsequently soared.

The year 2006 saw another milestone as well: Schlichter Bogard & Denton filed its first lawsuits alleging excessive fees in 401(k) plans. As the DC practice leader at Callan Associates, I saw requests for fee analysis by plan sponsors skyrocket. Many wanted to understand how they could eliminate the practice of revenue sharing. A once-popular way to pay for plan administration, revenue sharing allocates an agreed-upon portion of an investment fund’s expense ratio back to the plan’s recordkeeper. The fee lawsuits took issue with this approach.

The market collapse of 2008 was as tough on 401(k) plans as it was on everything else. During the market decline of 2001–2002, people had joked that their plan was now a 201(k). No one joked about the 2008–2009 collapse. Times were so hard that some plan sponsors were forced to eliminate their employer matching contributions. One plan sponsor’s comment summarized the angst of that period: “I just want to know what is going to blow up in my plan next.” Although the number of plans with a 401(k) feature was now up to 511,583 and the number of active participants was 59,976,000, total assets slumped from nearly $3 trillion in 2007 to $2.23 trillion at the end of 2008.

The markets — and 401(k) plans — have since bounced back with a vengeance. In 2016, a Department of Labor survey found that 62 percent of workers had access to some type of defined contribution plan, most likely a 401(k) plan. Of those with access, 72 percent were participating. Assets in 401(k) plans as of the end of 2017 stand at $5.28 trillion.

My personal journey with 401(k) plans finds me celebrating their 40th anniversary as president and CEO of the Employee Benefit Research Institute. And we are celebrating our own 40th anniversary. EBRI was conceived in 1978 for the purpose of researching not only 401(k) plans, but all employee benefits. Today, the EBRI/ICI 401(k) database is the largest of its kind in existence, housing 27 million participants’ data on balances, investments, and activities. The database shows that workers with the greatest proportion of assets in 401(k) plans are those in their 50s. Forty-three percent of assets reside with workers who are mostly in the final stretch of their career before retirement. These workers are likely to rely on their 401(k) assets as their primary, or potentially their only, retirement nest egg. They face the daunting challenge of figuring out how to draw down their assets so that their nest egg lasts their full retirement. They don’t know how long that retirement will be, what health care costs they will face, or how the markets will perform.

Many do not use their 401(k) plan as their drawdown vehicle. EBRI finds that in 2016, the amount of dollars moved to IRAs through rollovers was more than 16 times the amount contributed directly to IRAs. And for the most part, retirees’ drawdown strategy is simply to take the required minimum distribution. Our research shows that, depending on the size of the nest egg, only between about 12–27 percent of assets are drawn down over the course of the typical retirement. When asked, retirees say they would rather preserve assets for a rainy day rather than spend them, even if it means living well below their ability.

As the 401(k) celebrates its 40th anniversary, the question arises: Ultimately how useful is a retirement account that people are afraid to spend? And does the 401(k) need to do more to benefit the huge new wave of retirees that will depend upon it?

EBRI has catalogued these changes and their implications in its new Fast Facts. Also, see our timeline: The Rise of the 401(k) Plan – Zero to $5.28 trillion in 40 years.

Will Small Employers Widely Adopt More Accessible Multiple Employer Plans?

Blog9.20President Trump’s August 31 Executive Order (EO) on Strengthening Retirement Security in America could be a game changer for the retirement prospects of workers at small businesses. According to the Bureau of Labor, while 89 percent of workers of larger employers (500 or more employees) have access to retirement plans, for workers at smaller employers (less than 100 employees) that proportion drops to 53 percent¹. This clearly contributes to the $4.13 trillion aggregate retirement savings shortfall that EBRI estimates American workers face, according to its Retirement Security Projection Model.®

First, some background: Why do small employers refrain from offering retirement plans? Cost is a key factor. A Pew Charitable Trusts survey found that the top reason small business owners give for not offering a retirement plan is cost (followed by resources).² More widely available multiple employer plans (MEPs) could alleviate cost burdens by allowing small employers to band together and gain economies of scale. Trump’s EO calls for expanding access to multiple employer plans by addressing issues that have prevented widespread adoption of such plans. Namely, the EO instructs the Secretary of Labor to consider new rules around when a group or association of employers qualify for a MEP (the “common nexus” requirement). It asks the Secretary of the Treasury to consider amending regulations to address consequences if employers within MEPs fail to meet plan requirements (the “one bad apple” rule).

Whether both the common nexus and one bad apple rules can be properly addressed is the subject of a different blog. Further, we’ll let others assess whether the hoped-for economies of scale will actually materialize. But assuming the answer to all of the above is “yes,” a further question is: Will small employers adopt more widely available MEPs?

The data are mixed. In one provider survey, 33 percent of small employers (5–99 employees) indicated that they would be likely to consider a MEP if easily accessible. Meanwhile, the Transamerica Center for Retirement Studies finds that among companies that say they are not likely to offer a 401(k) plan — many of which cite not being big enough — a quarter would consider joining such a MEP.³ However, 23 percent were “somewhat likely” to do so — only 2 percent said they were “very likely.” The Transamerica study also points out that less than a third of employers view an employee-funded retirement plan as very important in attracting and retaining employees.
At the same time, among small companies that do not offer a 401(k) or similar plan, Pew found that about a quarter were likely to begin sponsoring a plan in the next two years — and this was before the EO. It is possible that a MEP gives such plan sponsors an easier pathway to offering benefits, allowing them to more easily change their intentions into action. And indeed, nearly half of those in the study said that availability of a plan with reduced administrative requirements would increase the likelihood of their organization offering a retirement plan in the future.

Assuming that a third of small employers that do not currently offer DC plans adopt MEPs going forward, preliminary estimates from EBRI’s Retirement Security Projection Model® indicate that the $4.13 trillion aggregate retirement savings shortfall would be reduced by $65 billion.

But this assumes 100 percent participation among eligible employees, and that small businesses joining MEPs will implement plans that look similar to existing plans of companies of their size. The majority of small employers, according to the Pew survey, do not offer automatic enrollment or automatic contribution escalation, although many offer employer contributions. If cost is truly the reason small employers don’t currently offer such plans, it is possible that fewer employers that participate in MEPs will offer employer contributions than average. Without automatic enrollment and automatic contribution escalation, participation in such plans may be low.

Importantly, in the Transamerica survey, 88 percent of workers believed that the value of a 401(k) or similar plan is an important benefit, and 81 percent agreed that retirement benefits offered by a prospective employer were a major factor in the final decision to accept a job.

Other possible proposals and alternatives to increase coverage exist:

  • President Obama’s proposed Automatic IRA program: estimated to reduce the retirement savings shortfall by $268 billion4.
  • The Automatic Retirement Plan Act (ARPA) of 2017: estimated to decrease the deficit by $645 billion5.
  • Universal DC plans: estimated retirement savings shortfall would decline by $802 billion6.

1March 2017
2Small-Business Views on Retirement Savings Plans: Topline Results of Employer Survey. 2016 Pew Charitable Trusts report.
3Striking Similarities and Disconcerting Disconnects: Employers, Workers, and Retirement Security. 18th Annual Transamerica Survey. August 2018.
4Under the Obama proposal, the model assumes a 3 percent default contribution rate and no opt-outs. It further assumes that there were no employer contributions and that no current defined-contribution-plan sponsors decided to discontinue their current plan and shift to the auto-IRA.
5ARPA assumes triennial automatic enrollment with a default contribution rate of 6 percent, and auto contribution escalation at 1 percent per year [up to 10 percent]. Assumes no opt-outs for this calcuation.
6This analysis assumes that all employers offer a type of plan and a set of generosity parameters similar to employers in their size range. Assumes observed contributions and opt-outs.

Modernizing the U.S. Retirement System: It Takes a Village

I recently moderated a discussion entitled “Modernizing the U.S. Retirement System” at a meeting by the American Academy of Actuaries. The panel consisted of some of the best thinkers in retirement: Ted Goldman of the American Academy of Actuaries; Stephen Goss of the Social Security Administration; Mark Iwry of The Brookings Institute, and former Senior Advisor to the Secretary of the Treasury; and Steve Vernon of the Stanford Center of Longevity.

For them, the challenge is clear: The onus is increasingly on American workers to secure their retirement future, and many are not up to the task. Ted Goldman compared it to asking drivers to assemble their own cars before going to work in the morning — most would simply stare at the pile of auto parts in confusion. Yet American workers are expected to manage their retirement plans as if they are financial professionals.


Car Parts                                                                    Retirement Parts

In addition, the panel concluded, our retirement system has serious holes around coverage and adequacy. Yet policymakers continue to address these gaps in only a fragmented way.

The panelists listed nearly two dozen ideas for moving our retirement system into the future, including reducing defined contribution (DC) plan leakage, creating emergency savings funds, auto-IRAs, personalized auto features, risk pooling, retirement income portfolios, and required minimum distribution reform.

Some of the ideas from the panelists were controversial, but many were just common sense, such as creating protections and incentives for employers and providers to offer solutions that help individuals do a better job of assembling their retirement parts.

How can we do this? Mark Iwry gave the example of how he used rulings and general information letters starting in 1998 to make employers comfortable with offering automatic enrollment in DC plans. His efforts were clearly critical in putting automatic enrollment on the map when it comes to DC plan design. However, the reality remains that automatic enrollment was still hampered by considerations such as state wage laws and lack of a safe harbor; take-up rates remained low despite his efforts. Then came the Pension Protection Act of 2006. Once plan sponsors had the protections they needed, adoption soared. Today the majority of large 401(k) plans have adopted automatic enrollment, and most with automatic enrollment also have automatic contribution escalation in place as well. There is no doubt this has led to higher participation and contribution rates by American workers.

The clear message is that it takes creative solutions as well as a partnership between the private sector, regulators, and legislators to develop a holistic approach that will ultimately lead to a retirement system that does a better job of securing American workers’ financial future. The dialogue started at the American Academy of Actuaries needs to continue.

Lori Lucas, CFA, President and CEO

Just the Facts? No. Analysis Matters!

My theme for the year seems to be: “Getting the facts right matters.” But a corollary to this is: “Interpreting analysis correctly is important, too.”

A case in point is the coverage of the Employee Benefit Research Institute’s (EBRI) Issue Brief, “Cumulative Out-of-Pocket Health Care Expenses After the Age of 70,” published on April 3, 2018. Based on data from the Health and Retirement Study, we reported that while for some retirees, costs such as out-of-pocket nursing home expenses can be substantial, the majority of older people pay modest out-of-pocket health care expenses in retirement.


Some in the press and within the industry interpreted this to mean that previous studies projecting the amount of savings required for couples to cover their health care expenses, including premiums, in retirement, have been overstated. For example, some note that health care cost projection studies have found that an average couple retiring today at age 65 needs between $280,000 and $370,000 to cover health care expenses in retirement. These numbers, it is concluded, are not the modest out-of-pocket health care expenses calculated in EBRI’s April research.

This conclusion simply doesn’t hold up: The two analyses have materially different intent, scope, data, and assumptions.

The intent of the studies that project health care costs in retirement is to help workers understand how much they may need to save for a financially secure retirement that includes adequate health care coverage. These studies use assumptions about types of Medicare coverage people will have during retirement, and to a large extent, reflect the premiums for such coverage.

In contrast, the out-of-pocket health care costs study EBRI issued in April was intended to provide a good understanding of the risks of out-of-pocket health care expenses faced by retirees—beyond health care premiums. That study used self-reported out-of-pocket costs including hospital stays, nursing home stays, outpatient surgery, doctor’s visits, prescription drugs, dental services, home health care, and hospice care after age 70 from a household survey in order to show the magnitude of such costs during retirement.

Other key differences between the April 2018 report and the two other reports are that the April 2018 study:

• Examines individuals, not couples
• Takes into account expenses for people after the age of 70 (not including those ages 65-69).

We also noted in the April report that, because of data limitations, reported cumulative out-of-pocket medical expenses should be interpreted as the lower bound of such expenses, rather than the true estimates of the means or medians. Further, we noted that health care expenses could be catastrophic for some individuals, and ranged from just under $172,000 to just over $269,000 for a single person at the extreme end of the distribution. We also noted that it is not easy to predict in advance who will actually have high medical expenses, and that as a result, the risk of high medical expenses remains a significant one.

It is critically important that workers preparing for retirement understand the amount they will likely need to save in order to have adequate health care coverage in retirement. But workers and retirees also should understand the magnitude of the out-of-pocket costs that they may experience beyond medical premiums due to catastrophic health care needs such as long-term care in retirement.

Both are valuable insights—but conflating the two can be misleading for both workers and policymakers. This is a good lesson about interpreting research in general.

Who Are Gig Workers and How are They Faring: Some Surprising and Some Not-so-surprising Findings from the New Bureau of Labor Statistics Survey


While the term “gig economy” may conjure images of Millennials driving for Uber, the reality is that workers in alternative employment arrangements are far more likely to be Gen Xers or Baby Boomers. This is the evidence from the Bureau of Labor Statistics’ (BLS) recently released Contingent and Alternative Employment Arrangements supplement. The survey finds that 10.6 million workers identified themselves as independent contractors—6.9 percent of total employment. Of this group, by far the largest cohort is workers ages 55 or older: 37 percent. This compares to the 22 percent of workers with traditional work arrangements that are in this age cohort. Granted, many of the independent contractors do not have mobile app-based employment: 25 percent describe themselves as being in professional and business services. However, a 2018 Financial Attitudes & Behaviors Toward the Gig Economy survey by T. Rowe Price shows that older – not younger – workers are more likely to have gig work: 9 percent of Millennials identify themselves as earning income in the gig economy, compared to 19 percent of Gen Xers and 11 percent of Baby Boomers.

On the positive side, the BLS reports that 79 percent of independent contractors prefer their work arrangements to traditional jobs. However, the report also notes that, in general, the proportion of workers in alternative employment arrangements who actually participate in employer-provided retirement plans is lower BlogPostPicthan for those in traditional arrangements. The 2017 Gig Worker On-demand Economy survey by Prudential confirms this, finding that only 16 percent of gig-only workers say they have access to employer-sponsored retirement plans (and this could also be from previous employers or via their spouse), compared to 52 percent of full-time workers. Traditional workers are nearly six times as likely to have access to benefits such as life insurance as gig-only workers. As Jake Biscoglio of Prudential Financial pointed out at EBRI’s “Exploring the Gig Economy” panel at its 38th Policy Forum, access to benefits is the number one challenge reported by gig workers.

Chances are, these older gig workers are not securing their financial wellbeing on their own. For example, only 10.2 percent of workers in that age category contribute to an IRA, with a median IRA balance for such workers standing at $51,400.

At EBRI’s recent Policy Forum, Julie Stitzel, managing director of Policy and Strategic Initiatives at the U.S. Chamber of Commerce attributed the lack of benefits for gig workers to outdated labor laws, noting that Chamber members are hesitant to push the envelope when it comes to providing benefits because of reclassification risk. She gave the example of one well-known gig employer that has partnered with online financial advisor Betterment to help contractors save their money. The Betterment service is provided for free in the first year and is discounted to independent contractors in the second. But the gig employer does not believe it can go beyond providing this simple integration with Betterment out of concern of running afoul of reclassification risk, according to Stitzel. In other words, gig employers may want to provide benefits to their contractors—and may even seek creative solutions to do so—but they feel hamstrung in their efforts given current labor laws.

In the Policy Forum, a wide range of possible solutions was discussed for drawing gig workers into the retirement system, including a possible federal retirement marketplace solution similar to what is being offered in New Jersey and Washington State; more widespread availability of open multiple employer plans; or employer-sponsored defined contribution plans that can be ported between traditional and nontraditional employers.

But one widely agreed-upon theme at the Policy Forum was that any good policy solution will require good data. As such, the fact that the BLS has even fielded the recent Contingent and Alternative Employment Arrangements survey is a step in the right direction: this is the first time the survey has been conducted since 2005. We might note that good policy solutions also require good analysis. The Employee Benefit Research Institute (EBRI) will continue to work with available data such as that of the BLS survey to bring into focus the true state of gig workers’ overall wellbeing.

How is income in retirement changing?

On April 24th, EBRI released its 28th annual Retirement Confidence Survey (RCS) produced in conjunction with Greenwald & Associates.

Among the findings, we learned that retirees’ confidence in their ability to afford certain retirement expenses declined. While 85 percent of retirees surveyed in 2017 expressed confidence in their ongoing ability to afford basic expenses, only 80 percent expressed such confidence this year. And, confidence in their ability to afford medical expenses is also down – 77 percent in 2017 vs. 70 percent this year.

While retiree confidence in these aspects of the affordability of retirement declined, the confidence workers have in living comfortably throughout retirement increased in 2018.

Even so, workers appear to be facing their own challenges, as they expect much more of their retirement income to be from sources that do not pay a guaranteed monthly income.

The RCS shows how workers’ expectations of the sources of income in retirement might differ from the sources of income for current retirees: 26 percent of retirees report receiving income from work, while 68 percent of workers expect working for pay to provide them income in retirement; 2-in-3 retirees report Social Security is a major source of income, while only about a third of workers believe Social Security will be a major source; and more than 4-in-10 retirees report income from a defined benefit (DB) or pension plan is a major source of income, while only 32% of workers expect a DB plan to be a major source for them in retirement.

Most current workers expect to rely on “income” sources that require both a savings strategy during their working years and a withdrawal strategy to create income in retirement. Workers see their workplace defined contribution (DC) retirement plans as a source of income in retirement far more than retirees report they have. Eight-in-ten workers say they believe their DC plan will be a major or minor source of income in retirement, compared to just 50 percent of retirees. Six-in-ten workers say they also expect income in retirement from an Individual Retirement Account (IRA).

As we transition to a generation of retirees that will be primary dependent on DC plans, how do workers plan to convert savings to income? In 2018, for the first time, DC plan participants were asked about their plans for the money accrued in their plan when they retired. Only half of workers are confident that they know how much income they will need each month in retirement or how to withdraw income from their savings and investments, with only 1 in 8 very confident.

In short, future retirees’ potential greater reliance on unpredictable sources of income combined with their lack of confidence in withdrawing income from their savings and investments is noteworthy—especially in light of its juxtaposition with their increased confidence in living comfortably throughout retirement.

The RCS is produced annually through the generous support of our funding members: AARP, Conduent HR Services, FINRA , J.P. Morgan, Lincoln Financial, Mercer, MetLife, Nationwide Financial, Principal Financial Group, T. Rowe Price, The Segal Group, U.S. Chamber, Vanguard, and Wells Fargo.