Missing 'Inaction'

While it’s hardly a new topic, the subject of missing participants is much in the news today – and arguably a growing concern for plan sponsors, particularly with the expansion of automatic enrollment.

Earlier this year the Government Accountability Office published a report – and some recommendations – on the subject of (re)connecting participants with their “lost” account balances. That report noted that from 2004 through 2013, more than 25 million participants in workplace plans separated from an employer and left at least one retirement account behind, “despite efforts of sponsors and regulators to help participants manage their accounts.” The report acknowledged that there are costs involved in searching for these “lost” participants, going on to note that there are no standard practices for the frequency or method of conducting searches.

Once upon a time the IRS provided letter-forwarding services to help locate missing plan participants, but with the Aug. 31, 2012, release of Revenue Procedure 2012-35, the IRS stopped this letter forwarding program. Moreover, while the Labor Department has provided guidance to plan sponsors of terminated DC plans about locating missing participants and unclaimed accounts, they have yet to do so regarding ongoing plans.

That said, the GAO reported that they had been informed by DOL officials that they are conducting investigations of steps taken by ongoing plans to find missing participants under their authority to oversee compliance with ERISA’s fiduciary requirement that plans be administered for the exclusive purpose of providing benefits.¹

In fact, the Labor Department’s Employee Benefit Security Administration’s Chicago Regional office adopted a “missing participant” regional initiative in fiscal year 2017, and – working with the PBGC, has reportedly recovered nearly $6.3 million for 133 participants, according to Bloomberg Businessweek.

More recently Sens. Elizabeth Warren (D-MA) and Steve Daines (R-MT) reintroduced the bipartisan Retirement Savings Lost and Found Act, noting that many Americans leave their jobs each year without giving their employers directions with what to do with their retirement accounts – a trend the bill’s sponsors say has increased with the expansion of auto enrollment. The legislation calls for the creation of a national online lost and found for Americans’ retirement accounts – and claims to leverage data employers are already required to report to do so (though the devil may lie in the details). The legislation also purports to clarify the responsibilities employers and plan administrators have to connect former employees with their neglected accounts.

Indeed, in such matters, plan sponsors often feel trapped between the proverbial rock and the hard place – pressed hard on the one hand by regulators to locate these former participants (and potential beneficiaries) – on another by state agencies with an avid interest in the escheatment of those funds – and often squeezed by the growing costs not only of trying to locate these participants, but the costs of distributing a wide assortment of plan notices, not to mention the ongoing costs of maintaining these accounts in potential perpetuity.

Little wonder that among its recent recommendations, the GAO recommended that the Secretary of Labor “issue guidance on the obligations under the Employee Retirement Income Security Act of 1974 of sponsors of ongoing plans to prevent, search for, and pay costs associated with locating missing participants.”

Said another way, what’s “missing” is more than former participants – it’s some safe harbor guidance that would provide some comfort and structure to those trying to reasonably fulfill their duties as plan fiduciaries – an ongoing concern for plans² that are an ongoing concern.

- Nevin E. Adams, JD
 
Footnotes

1.  Speaking of missing participants, the headlines of late have focused on issues regarding defined benefit participants. Most notably perhaps, MetLife disclosed last year that it failed to locate some group annuity clients that had likely moved or changed jobs. Nor was this a recent problem – the issue, which the firm said involved some 13,500 pension clients, was attributed to a “faulty system” that the firm had been using for a quarter century – a system that assumed that if the firm was unsuccessful in contacting participants twice that the individual would never respond, and that therefore weren’t going to claim benefits. In fact, the Labor Department’s push for companies sponsoring pension plans to find missing participants cited above reportedly influenced MetLife’s decision to conduct the review. Enter Secretary of the Commonwealth William F. Galvin, who just announced that his office discovered “hundreds of Massachusetts retirees” who are owed pension payments by MetLife. Galvin noted that the regulator planned to look into what MetLife had done in the past to locate and pay the retirees. 

• He also said his office’s investigation has been expanded to look into other firms who provide  retirement payments, including Prudential, Transamerica, Principal Financial, and Mass Mutual. ↩

1. You may recall that last fall the Pension Benefit Guaranty Corporation (PBGC), the nation’s private pension plan insurer, announced the expansion of its Missing Participants Program beyond its historical focus on PBGC-insured single-employer plans as part of the standard termination process to cover defined contribution plans (e.g., 401(k) plans) and certain other defined benefit plans that end on or after Jan. 1, 2018. However, this only deals with terminating defined contribution plans. ↩

Familiar ‘Grounds’?

A recent report by the GAO paints a pretty bleak picture of American retirement. Is it accurate?

For the most part, the report covered familiar ground, bemoaning the “marked shift” away from the traditional defined benefit pension plan (glossing over how few private sector workers were covered by these plans, even in their heyday, and the fraction of those who received a full pension), and highlighting low savings rates, the pervasive lack of broad-based access to workplace retirement plans and the daunting challenges confronting even those who do enjoy that access.

The report also spends several of its 173 pages chronicling (with pictures) the ways in which “leakage” also undermines retirement savings. And for good measure, it invokes the findings of the Melbourne Mercer Global Pension Index  — which the GAO calls the “most comprehensive” — that ranks the U.S. retirement system 20th out of 25 countries surveyed (and gives us a “C” grade). Indeed, the report treads such familiar ground in such a familiar way that it hardly seems controversial.

In fairness, having seen, studied and even commented on the data, reports and surveys cited in the GAO report, the authors can hardly be faulted for their air of pessimism. As they explain in their introduction, “More and more people are retiring, and many are living longer in retirement. Health care costs are rising, Social Security is stretched to the limit, and debt — both personal and public — is a threat to financial security.” But in retreading this “familiar” ground, they also restate as fact some things that have been drawn into question — and gloss over some more recent findings that provide valuable context.

‘Over’ Looked?

For example, the 2013 Survey of Consumer Finance (SCF) is a widely cited report, and is invoked repeatedly by the GAO in its assessment of the resources available to American workers in retirement. This is a reputable and well-regarded source of consumer information, drawn from a sampling of about 6,000 households (different ones every cycle). That said, the information contained is “self-reported,” which is to say that it tells you what individuals think they have (or perhaps wish they had), but not necessarily what they actually have. Now, the GAO has previously relied on this data — and in fact recalls a 2015 report by the GAO that claimed (and was titled) “Most Households Approaching Retirement Have Low Savings.”

The rationale for the “most” in the 2015 report headline appears to come from its focus on households age 55 and older, where the GAO noted that (only) 48% had some retirement savings, and thus one might reasonably assume that the remaining 52% had no retirement savings — and that would seem to be the case. However, 23% of that 52% said they had a defined benefit plan. Now, that assessment may be inaccurate (see above), but if they do, in fact, have a DB plan, that plus Social Security might well be sufficient. So, “most” have no savings, but about half of that “most” might not need savings. Admittedly, that distinction makes for a clumsy headline.

Among those age 55-64 with no retirement savings, the median net worth was $21,000 (about half of these had no wage or salary income), while among those in the same age bracket with any retirement savings, their median net worth was $337,000. Compared to those with retirement savings, these households (those aged 55-64 with no retirement savings) have about one-third of the median income and about one-fifteenth of the median net worth, and are less likely to be covered by a DB plan. The bottom line is that, even accepting the self-reported data of these individuals, there is a considerable disparity, and one that suggests that a more targeted analysis (and dare I suggest remedy) might be more in order.

In evaluating things like retirement income and coverage, the GAO report draws on information from, among other sources, the Current Population Survey (and in some cases other reports based on that information). While it is one of the most-cited sources of income data for those whose ages are associated with being retired (typically ages 65 or older), and has also been used to provide annual estimates of employment-based retirement plan participation, a 2014 redesign of the questionnaire has resulted in much lower estimates of the percentages of workers who participate in an employment-based retirement plan. In fact, the non-partisan Employee Benefit Research Institute (EBRI) has cautioned that it has resulted in historically “sharp and significant” reductions in the levels of worker participation in employment-based retirement plans.

Missed ‘Out’?

Not mentioned in the GAO report (but cited in a recent Forbes article by Andrew Biggs of the American Enterprise Institute) is an analysis by Census Bureau economists Adam Bee and Joshua Mitchell, who used IRS data to measure the share of new retirees receiving benefits from private retirement plans. Biggs notes that in 1984, only 23% of new retirees received any sort of private pension benefits, but by 2007, 45% of new retirees received private pension benefits.

As for the aforementioned ranking of the U.S. retirement system, it’s really hard to compare apples to oranges, as such comparisons inevitably do. But in the Forbes article noted above, Biggs reminds us that Mercer measures adequacy by virtue of things like tax preferences for retirement savings, ages at which participants can access their savings, whether savings must be annuitized, etc. As things to consider, perhaps — but a ranking based on subjective weightings and criteria that includes certain qualitative factors doesn’t necessarily produce an objective result.

‘Post’ Retirement

Another recent analysis, “Using Panel Tax Data to Examine the Transition to Retirement” — conducted by Peter J. Brady and Steven Bass of the Investment Company Institute and Jessica Holland and Kevin Pierce of the IRS — found that most individuals were able to maintain their inflation‐adjusted net work‐related income after claiming Social Security. Looking only at how much individuals reported as net income on their taxes the year before they started drawing Social Security benefits, compared with the three years after they began that draw, they found that, looking at working individuals age 55 to 61 in 1999 who did not receive Social Security benefits that year, three years after they started claiming Social Security (which could be viewed as a proxy of sorts for entering retirement) that median ratio of net work‐related income at that point compared to net work‐related income one year before claiming was 103% — which means, of course, that three years later, they are actually reporting (slightly) higher income levels than they were prior to retirement. Does that mean they will still be doing so a decade later? No — but why not even an acknowledgement that such results have been documented with actual IRS data?

Other Points

The GAO report does remind us that where you work matters (in 2016, 89% of workers in information services had access to an employer-sponsored plan, compared with 32% of workers in the leisure and hospitality industry), and how you work matters (“one reason lower-income workers lack access to employer-sponsored retirement plans is that they struggle to meet plan eligibility requirements related to sufficient tenure and hours worked”) in terms of having access to a retirement plan, and how much you make really matters (“…workers in the lowest income quartile were nearly four times less likely to work for an employer that offered a retirement plan, based on our analysis of 2012 SIPP data, controlling for other factors”). And it highlights the critical importance of, and the very real danger posed to the nation’s retirement security by the projected shortfalls in Social Security.

On the other hand, for some reason the GAO report cites the creation of the QDIA safe harbor as a failure of sorts, in that no surge in new plan adoption accompanied it (completely disregarding the huge boost to diversified savings and increased participation via automatic enrollment that has resulted). Ditto the demise of the MyRA, whose dismal take-up rate stood in some contrast to its shockingly high cost. It had a different — and more sympathetic — perspective on the state-run programs for private sector workers, decrying the uncertain status of such offerings after the signing of legislation that overturned the safe harbor rule from the Obama administration.

Ultimately, the GAO report makes one very simple recommendation: the appointment of an independent commission to “comprehensively examine the U.S. retirement system and make recommendations to clarify key policy goals for the system and improve how the nation can promote more stable retirement security.” All well and good.

But here’s hoping that, should such a committee be formed, it will look beyond the all-too-familiar ground that the GAO chose to tread.

- Nevin E. Adams, JD

Retirement Plans and Retirement Income: It’s Complicated

One of the great concerns of our industry — when we aren’t worrying if people have saved enough for retirement — is worrying about how those savings are going to last through retirement.

Enter to that debate a recent report from the Government Accountability Office (GAO) that basically takes the Labor Department to task for not doing enough to encourage the use of lifetime income options in workplace retirement plans.

Sure enough, it’s been hard for lifetime income options to get traction with retirement plans. The GAO rightly outlines a number of the concerns typically articulated with these options — which are well known to those who have looked to remedy the situation (see “5 Reasons Why More Plans Don’t Offer Retirement Income Options”).

GAO Alternatives

So, what suggestions does the GAO have for the DOL? Well, the GAO has several specific suggestions, including that the DOL:

• do more to clarify the safe harbor for selecting an annuity provider;
• consider providing legal relief for plan fiduciaries offering an appropriate mix of annuity and withdrawal options;
• help encourage the use by incorporating references to selecting a lifetime income/annuity provider in both its current publications, or by issuing new guidance to do so;
• include participant access to advice on the plan’s lifetime income options from an expert in retirement income strategies; and
• consider providing required minimum distribution (RMD)-based default income plan distributions as a default stream of lifetime income based on the RMD methodology beginning, unless they opt out, when retirement-age participants separate from employment, rather than after age 70½.

The GAO goes so far as to suggest that the Labor Department encourage plan sponsors to use a record keeper that includes annuities from multiple providers on their record keeping platform (as if one wasn’t complicated enough), to encourage them to offer participants the option to partially annuitize their account balance, and to take into account changing providers that could put the current lifetime income products at risk.

The Logic

The thinking, of course, is that if you made it (feel) safer for plan sponsors to offer lifetime income options, more would do so. And that if you had more plan sponsors offering the options, more participants would have a chance to, and ultimately would, choose them. And that if you set up those options as a default, even more participants would “choose” them, or at least wind up with them. It’s hard to rationally argue with that logic. And yet, for years I’ve watched a number of truly innovative and objection-responsive solutions come to market — and, for the most part, fail to live up to expectations. Why? Well, it’s complicated.

Consider that, unlike automatic enrollment — around which the Pension Protection Act so carefully crafted provisions that provided an exit path for participants who had second thoughts — unwinding lifetime income options is generally seen as more… complicated. Not impossible, mind you — but “complicated.”

As for the enhancements to the current safe harbor that GAO recommends — well, if plan sponsors are to be believed, that would certainly help. But let’s not ignore the reality that the DOL has already taken some pretty significant steps in that direction (and seems reluctant to go further). Despite those steps, I think it’s fair to say that plan fiduciaries still find it their responsibilities with regard to those options… complicated, certainly compared to the other providers/services for which they are accountable.

Behavioral Barriers

There’s little question that participants need help structuring their income in retirement — and little doubt that a lifetime income option could help them do that. That said, the biggest impediment to adoption may simply be that (industry surveys notwithstanding) participants don’t seem to be asking for the option — and when they do have access, mostly don’t take advantage, even when defined participants have a choice, they opt for the lump sum. Not that those dynamics can’t be influenced by plan design or advisor input, but justifiable concerns remain about fees, portability and provider sustainability. Moreover, there are significant behavioral finance impediments — be it the overweighting of small probabilities, or mental accounting — or simply the fear of losing control of finances, a desire to leave something to heirs, or simple risk aversion. It’s often not just one thing. It’s… complicated.

Ultimately, while the GAO’s recommendations would surely expand the availability of these options, would it have an impact on participant adoption? An income option with all the features participants want will be even more complex (and expensive) — and, if there were to be a default withdrawal option into a lifetime income option, I suspect that the opt-out rates would be high, for (all) the reasons noted above.

Why? Well, because it’s (still) complicated.

- Nevin E. Adams, JD

5 Things You Need to Know About Retirement Readiness

Recently the Government Accountability Office (GAO) released a report titled with its conclusion: “Most Households Approaching Retirement Have Low Savings.”

The title was no surprise — though the definitions of “most” and “low” bear further understanding.

The GAO Analysis
The GAO based most of its conclusions on findings from the 2013 Survey of Consumer Finance (SCF), conducted by the Federal Reserve every three years. It is a reputable and well-regarded source of consumer information, drawn from a sampling of about 6,000 households (different ones every cycle). That said, the information contained is “self-reported,” which is to say that it tells you what the individual thinks they have (or perhaps wishes they had), but not necessarily what they actually have. More on that in a minute.

The rationale for the “most” in the headline appears to come from its focus on households age 55 and older, where the GAO noted that (only) 48% had some retirement savings, and thus one might reasonably assume that the remaining 52% had no retirement savings — and that would seem to be the case. However, 23% of that 52% said they had a defined benefit plan. Now, that assessment may be inaccurate (see above), but if they do, in fact, have a DB plan, that plus Social Security might well be sufficient. So, “most” have no savings, but about half of that “most” might not need savings. Admittedly, that distinction makes for a clumsy headline.

What about the 29% who were age 55 or older, and who had neither a DB plan nor retirement savings? Well, the median annual income of that group was (just) $18,932. It’s not hard to imagine why that income bracket might not have set aside any savings for retirement. On the other hand, that group is probably well served by Social Security. Despite that income level, more than a third (35%) of those in this group say they own a home with no debt. Assuming that is the case (see “self-reported” above), that could make a big difference in their post-retirement expenses and/or represent an additional resource they could draw on for retirement income.

The GAO conclusions notwithstanding, here are five things you need to know about retirement readiness (and retirement readiness projections):

1. Social Security matters — especially for lower-income individuals.
Social Security remains the largest component of household income in retirement. In fact, GAO noted that for households age 65-74 with no retirement savings, Social Security makes up 57% of their household income on average. In fact, a quarter of this group rely on Social Security for more than 90% of their income.

For all households age 65-74, median annual income is about $47,000, and Social Security makes up on average 44% of income for households in this age group, larger than any other income source. About 90% of all households in this age range receive some Social Security income, and the median amount they receive is approximately $19,000.

The median income for households age 75 and older is about $27,000, and the median Social Security income is approximately $17,000. In fact, when compared to younger households age 65-74, Social Security makes up a larger share of household income for retirees age 75 and older, with 62% of these households relying on Social Security for more than 50%, and more than one-in-five (22%) relying on Social Security for more than 90% of their income.

2. Average or even median savings amount in the abstract tell you very little about retirement income adequacy at an individual level.

People live in different places and have different lifestyles and lifestyle expectations. They may have resources available beyond that reported in surveys such as the SCF, and individual health circumstances can make a huge difference in the adequacy of reported income sources vis-à-vis actual retirement spending requirements.

3. Pre-retirement income (or a percentage thereof) may not be a reliable proxy for post-retirement needs.

Many of these studies — and those cited by the GAO were no exception — use a “replacement rate” standard which, while it may be a convenient metric to use to convey retirement targets to individuals, has some serious shortcomings when applied to these large-scale policy models for determining whether an individual will run short of money in retirement.

The reality is that what and how we spend money pre-retirement often has little to do with our actual financial needs post-retirement.

As EBRI’s Research Director Jack VanDerhei points out, “…simply setting a target replacement rate at retirement age and suggesting that anyone above that threshold will have a ‘successful’ retirement completely ignores longevity risk, post-retirement investment risk, and long-term care risk.”

4. Many retirement projection models assume a 50% probability of success.
Aside from the shortcomings inherent in relying on a replacement rate for these projections, if you try to factor in longevity risk (the risk of running out of money in retirement), and post-retirement investment risk, VanDerhei notes that if you use a replacement rate threshold based on average longevity and average rate of return, you will, in essence, have a savings target that will prove to be insufficient… about 50% of the time.

Of course, when it comes to their retirement security, individuals tend to prefer something much closer to 100%.

5. Those who have a retirement plan are much better off than those who don’t.
Those with no retirement savings had a median income of approximately $29,000, while those in the same age range who have some retirement savings have a median income of $76,000. Among those age 55-64 with no retirement savings, the median net worth was $21,000 (about half of these had no wage or salary income), while among those in the same age bracket with any retirement savings, their median net worth was $337,000. In that group, the median income of those with no retirement savings was $26,000; among those with some retirement savings, the median income was $88,000.

Compared to those with retirement savings, these households (those aged 55-64 with no retirement savings) have about one-third of the median income and about one-fifteenth of the median net worth, and are less likely to be covered by a DB plan.

All of which suggests that it would probably be more meaningful to examine the retirement readiness of those without access to a retirement plan separately from those who do.

- Nevin E. Adams, JD