5 Things You May Not Know About Roth 401(k)s

I made the switch to Roth for my retirement savings years ago – and I have long counseled younger workers, who were likely paying the lowest tax rates of their working lives, to do so as well.

Pre-tax treatment has, of course, been the norm in 401(k) plans since their introduction in the early 1980s. On the other hand, the Roth 401(k) wasn’t introduced until the Economic Growth and Tax Relief Reconciliation Act of 2001, and even in that legislation wasn’t slated to become effective until 2006 (and at that time was still slated to sunset in 2010).

Despite that late start, according to a variety of industry surveys (Plan Sponsor Council of America, Vanguard, PLANSPONSOR), roughly 60% of 401(k) plans now offer a Roth 401(k) option, and PSCA data shows that 28.6% of 403(b) plans already allow for Roth contributions. Participant take-up, which just a few years ago hovered in the single digits, is now in the 15-20% range.

Here are five other things you may not know about the Roth 401(k):

To allow designated Roth contributions, a plan also has to offer traditional pre-tax 401(k) contributions. 

A plan can only offer designated Roth contributions if it also offers traditional, pre-tax elective contributions. Oh, and they can be offered by either a 401(k), 403(b) or 457 plan.

You can auto-enroll designated Roth contributions. 

The plan must state how the automatic contributions will be allocated between the pre-tax elective contributions and designated Roth contributions.

You can “catch up” on a Roth basis. 

Individuals can make age-50 catch-up contributions as a designated Roth contribution to a designated Roth account.

You have to include Roth contributions in the 401(k) plan annual nondiscrimination testing. 

Designated Roth contributions are treated the same as traditional, pre-tax elective contributions when performing annual nondiscrimination testing. Moreover, just like other elective deferral, they must be included when calculating the top-heavy ratio each year. You cannot allocate forfeitures, matching or any other employer contributions to any designated Roth accounts.

You can take loans from designated Roth accounts. 

As long as the plan permits, you can identify from which account(s) in your 401(k), 403(b) or governmental 457(b) plan you wish to draw your loan, including from a designated Roth account.

Retirement plan education has long had as a basic tenet that individuals will be paying lower tax rates in their retirement future. Of course, tax rates – and future income levels – are hard to predict. Both can rise more than anticipated in the future – and a Roth option can provide some essential flexibility to diversify tax treatment, as well as investments.

- Nevin E. Adams, JD

For more information, see Retirement Plans FAQs on Designated Roth Accounts from the IRS.

3 (More) Pervasive Retirement Myths - an Academic Perspective

Last week I highlighted three “myths” about the retirement system that will not die — all the more distressing because they are perpetuated by retirement industry pundits. Here are three more that (mostly) aren’t perpetuated by those who actually work with retirement plans, but by academics.  Academics who, sadly, are often listened to, and cited by those who regulate and legislate these programs.

The Match Doesn’t Matter

This doubtless comes as a surprise to those of us who work with retirement plans and retirement plan participants. More precisely, there are a few studies out there that suggest that a matching contribution doesn’t have a very strong impact on participation. The study that is cited most often in support of this one is one conducted several years ago in conjunction with H&R Block, where individuals were offered a financial incentive to take their tax refund and deposit it in an IRA. Most didn’t — and thus the notion that the promise of the match doesn’t produce (much) in the way of participation.

Now, if you find yourself saying, “But that’s not really an analogous situation to 401(k) saving” — well, you wouldn’t be the only one to think so.

But even if it has only a modest impact on the decision to participate, the data clearly suggests that it has a big impact on the rate of savings — and on the savings accumulations itself.

You Can Plug the Leakage Problem by Limiting Loans

Preventing “leakage” has become a constant drumbeat of many in our industry, their passion fueled by estimates of the enormity of the problem that seem limited only by the imaginations of those clamoring to solve it. As recently as a week ago, I was at a conference where a respected researcher went so far as to suggest that we are losing as much as $1 in leakage for every $2 contributed into the system. I kid you not.

Now, the “out” for these wild exaggerations is that nobody knows for sure, and so — if you’re looking for them — you see those estimates footnoted with disclaimers like “author’s calculations.” Or sometimes they come right out and admit that it’s based on a “host of simplifying assumptions.”

Don’t get me wrong — “leakage” can be a threat to an individual’s retirement security. But when the nonpartisan Employee Benefit Research Institute (EBRI) looked into the matter in testimony provided to the ERISA Advisory Council last year, it was cashouts — not loans or hardship withdrawals (even including the impact of a six-month suspension of contributions) — that turned out to be approximately two-thirds of the leakage impact.

So, if you want to solve the real leakage problem, you might want to start by accurately assessing the size of the problem, rather than just making numbers up. And then you might look for ways to make it easier for folks to keep their savings in the system at job change.

And let’s not forget — locking people’s money up with no access until retirement may solve the leakage problem. But it could also lead to a reduction in the amount of money people save in the first place.

The 401(k)’s Tax Incentives Are ‘Upside Down’

One of the comments you hear from time to time is that the tax incentives for 401(k)s are “upside down” (or as it has been more colorfully described, “out of whack”) — that is, they go primarily to those at higher income levels, who perhaps don’t need the encouragement to save. And from a pure financial economics perspective, those who pay taxes at higher rates might reasonably be seen as receiving a greater benefit from the deferral of those taxes.

Indeed, if those “upside-down incentives” were the only forces at work, one might reasonably expect to find that the higher the individual’s salary, the higher the overall account balance would be, as a multiple of salary. However, drawing on the actual administrative data from the massive EBRI/ICI 401(k) database, and specifically focusing on workers in their 60s (broken down by tenure and salary), EBRI Research Director Jack VanDerhei has found that those ratios hold relatively steady. In fact, those ratios are relatively flat for salaries between $30,000 and $100,000, before dropping substantially for those with salaries in excess of $100,000.

In other words, while higher-income individuals have higher account balances, those balances are in rough proportion to their incomes. They are not “upside down.” Now this is the outcome — the balances — not just focused on the contributions, the amount(s) going in, which is what most of the criticism focuses on.

As retirement plan advisors are well aware, these programs are subject to a series of limits and nondiscrimination test requirements: the boundaries established by Code Sections 402(g) and 415(c), combined with the ADP and ACP nondiscrimination tests. Those plan constraints were, of course, specifically designed (and refined) over time to maintain a certain parity between highly compensated and non-highly compensated workers in the benefits available from these programs. The data suggest they are having exactly that impact.

More recently, the “fairness” of these incentives has been questioned since they “only” go to workers who are covered by workplace retirement plans (though, as I pointed out last week, those numbers are often distorted as well). A better solution might be to expand those covered by such plans, rather than to undermine the diligence of those who are saving.

You’re less likely to hear these statements in your earshot — they tend to show up in academic forums, and sometimes in legislative hearings — forums where “decorum” might suggest avoiding confrontation.

But this is not just an academic exercise. And no one is well served — even the often well meaning academics who are trying to help improve the current system — by perpetuating misunderstandings.

- Nevin E. Adams, JD

“Breaching” Out?

“401(k) breaches undermining retirement security for millions,” was the headline of a recent article in the Washington Post.(1) No, we’re not talking about some kind of data hacking scandal, nor some new identity theft breach. Rather, those “breaches” are loans and withdrawals from 401(k)s.

Providing the impetus for the article is a new report by HelloWallet(2) that indicates that “more than one in four workers dip into retirement funds to pay their mortgages, credit card debt or other bills.”

While the article primarily deals with the potentially negative aspects of loans and withdrawals, it also touches on some broader concerns—and does so with some factual inaccuracies. For example, the article incorrectly states that “in 1980, four out of five private-sector workers were covered by traditional pensions;” in fact, only about half that many actually were at that point (a correct statement would be that 4 out of 5 covered by a plan at that time were in a traditional pension, which works out to about 2 in 5 of all private-sector workers—which puts the article’s “now, just one in five workers has a pension” statement in a far more accurate context).

The article notes that the “most common way Americans tap their retirement funds is through loans,” although U.S. Department of Labor data indicate that loan amounts tend to be a negligible portion of plan assets and that very little is converted into deemed distributions in any given year. That finding is supported by hard data from the EBRI/ICI 401(k) database, the largest micro-database of its kind: Among participants with outstanding 401(k) loans in the EBRI/ICI database at the end of 2011, the average unpaid balance was $7,027, while the median loan balance outstanding was $3,785.(3)

The Washington Post article cautions that these loans “must be repaid with interest,” but fails to mention that the 401(k) participant is paying interest to his/her own account: Those repayments represent a restoration of the retirement account balance by the participant, as well as a return on that investment. Sure, that interest payment is coming from the participant’s own pocket, but it’s generally being deposited into their own retirement account, and (if it was being used to pay down debt) likely at a much more favorable rate of interest.

The article also notes that those who withdraw money from their 401(k) early (before the authorized age) face “hefty” penalties, and certainly there is a financial price. However, those penalties consist of the 10 percent early withdrawal penalty (that was put in place long ago to discourage casual withdrawals by those under age 59 -½), and the income taxes they would be expected to pay on the receipt of money on which they had not yet paid taxes (which they would likely pay eventually).(4)

The HelloWallet report describes defined contribution (DC) retirement plans as a “marginal contributor to the actual retirement needs of U.S. workers.” However, an EBRI analysis of the Federal Reserve’s Survey of Consumer Finance (SCF) data, looking at workers who are already retired, finds that DC balances plus IRAs—which for many retired individuals were funded by rollovers from their DC/401(k) plans when they left work—represent nearly 15 percent of their total assets (which includes things like houses, but does not include Social Security and defined benefit annuity payments, although DB rollovers are included), and nearly a third of their total financial assets, as defined in the SCF.

The dictionary describes a “breach” as an “infraction or violation of a law, obligation, tie, or standard.” The article quotes the HelloWallet author as noting that “What you have is 401(k) participants voting with their wallets saying they would much rather use this money for other purposes.” However, these reports can’t always know, and thus don’t consider, how many participants and their families have been spared true financial hardship in the “here-and-now” by virtue of access to funds they set aside in these programs(5) (an AonHewitt study,(6) cited both in the article and in the HelloWallet report, notes that just over half of the hardship withdrawal requests were to avoid home eviction or foreclosure).

It’s hard to know how many of these “breachers” would have committed to saving at the amounts they chose, or to saving at all, if they (particularly the young with decades to go until retirement) had to balance that choice against a realization that the funds they set aside now would be unavailable until retirement. We don’t know that individuals who chose to save in their 401(k) plan did so specifically for retirement, rather than for interim (but important) savings goals—such as home ownership or college tuition—that, sooner or later, make their own contributions to retirement security. Indeed, what appears to a be a short-term decision that might adversely affect retirement preparation may actually be a long-term decision to enhance retirement security with a mortgage paid off or higher earnings potential.

In sum, we don’t know that these decisions represent a “breach” of retirement security—or a down payment.

Nevin E. Adams, JD


(1) The Washington Post article is online here.

(2) You can request a copy of the HelloWallet report here.

(3) For an updated report on participant loan activity from the EBRI/ICI database, see “401(k) Plan Asset Allocation, Account Balances, and Loan Activity in 2011,” online here.

(4) While the Post article states that these taxes would be paid at capital gains rates, in fact, withdrawals are taxed as ordinary income.

(5) See “’Premature’ Conclusions,” online here.

(6) See “Leakage of Participants’ DC Assets: How Loans, Withdrawals, and Cashouts Are Eroding Retirement Income 2011,” online here.

“Premature” Conclusions

A couple of months back, my wife noticed a water spot on the ceiling of our dining room. Now, it didn’t look fresh, but considering that that ceiling was directly underneath the master bath, she had the good sense to call a plumber. Sure enough, there was a leaky gasket—and from the look of it, one that had been there for some time before we took ownership. Fortunately, the leak was small, and the damage was minimal. Even more fortunately, we took the time to have the plumber check out the other bathrooms, and found the makings of similar, future problems well before the “leakage” became serious.

Homes aren’t the only place with the potential for problems with leakage. A recent report on 401(k) loan defaults suggests that “leakage”—the money being drawn out of retirement plans prior to retirement —is a lot larger than a number of industry and government reports have indicated. In fact, the report (online here) claims that “the leakage could be as high as $37 billion per year,” although it completes that sentence by acknowledging that the estimate depends “…on the source of the data on loans outstanding and the assumed default rate.”

The paper promotes a recommendation that ERISA be amended so that plans could choose to allow those who take out 401(k) loans to be defaulted into insurance that would repay those loans on default. It looks at a number of different sources to conclude that the available data do not really capture all the loan leakage (because some of it is obscured as part of distribution upon termination/separation from service), and that the available data do not (yet) capture the impact of the prolonged economic slowdown that is evidenced in other, non-401(k) loan trends.

Setting aside the validity of those conclusions, and the scale of their impact on the analysis, the issue of “leakage” remains a focus for many in our industry.

Late last year, an EBRI Issue Brief examined the status of 401(k) loans, noting that in the 2010 EBRI/ICI 401(k) database, 87 percent of participants in that database¹ were in plans offering loans, although as “has been the case for the 15 years that the database has tracked 401(k) plan participants, relatively few participants made use of this borrowing privilege.”
Indeed, from 1996 through 2008, on average, less than one-fifth of 401(k) participants with access to loans had loans outstanding. At year-end 2009, the percentage of participants who were offered loans with loans outstanding ticked up to 21 percent, but it remained at that level at year-end 2010 (see the full report, online here). This hard data, by the way, measuring activity by more than 23 million 401(k) participants.

If loan levels and amounts outstanding have remained relatively constant during this period (which included the “Great Recession”), one might nonetheless wonder about the overall impact on retirement readiness.

If you define “success” as achieving an 80 percent real replacement rate from Social Security and 401(k) accumulations combined, looking at workers ages 25–29 (who will have more than 30 years of simulated eligibility for participation in a 401(k) plan), then the decrease in success resulting from the COMBINATION of cashouts, hardship withdrawals, and loans is just 6.1 percent.² The impact when you add in the impact of loan defaults is less than 1 percentage point higher (approximately 7.1 percent for all four factors combined).

Looking at the overall impact another way, more than three-fifths of those in the lowest-income quartile³ with more than 30 years of remaining 401(k) eligibility will still be able to retire at age 65 with savings and Social Security equal to 80 percent of their real pre-retirement income levels, even when factoring in actual rates of cashout, hardship withdrawals, and loans—INCLUDING the impact of loan defaults.

The impact at an individual level can, of course, be more severe—something that will be explored by future EBRI research.

A Problem to Fix?

There is, however, a potentially larger philosophical issue: whether the utilization of these funds prior to retirement constitutes a “leakage” crisis that cries out for a remedy. We don’t know how many participants and their families have been spared true financial hardship in the “here-and-now” by virtue of access to funds they set aside in these programs. Nor do we know that individuals chose to defer the receipt of current compensation specifically for retirement, rather than for interim (but important) savings goals—such as home ownership or college tuition—that make their own contributions to retirement security. It’s hard to know how many of these participants would have committed to saving at all, or to saving at the amounts they chose, if they (particularly the young with decades of work ahead of them) had to balance that against a realization that those monies would be unavailable until retirement.

In fixing the recent leakage problem in our home, the plumber replaced the worn gaskets, but at the same time sought to improve on things by tightening (as it turns out, over-tightening) some of the connections further up the line. That extra step produced an unanticipated outcome that didn’t show up until the next day, in dramatic fashion. Like my plumbing problem, retirement plan “leakage,” unminded, has the potential to cause damage—to deplete and undermine retirement savings. However, a view that all pre-retirement distributions from these programs are a problem that requires redress not only ignores the law and regulations as written, it also has the potential to create unanticipated changes in savings behaviors.

And the data—based on hard data from actual participant balances and activity—indicate that such concerns are at least somewhat premature.

- Nevin E. Adams, JD

Notes

¹ The EBRI/ICI Participant-Directed Retirement Plan Data Collection Project is the largest, most representative repository of information about individual 401(k) plan participant accounts in the world. As of December 31, 2010, the EBRI/ICI database included statistical information about 23.4 million 401(k) plan participants, in 64,455 employer-sponsored 401(k) plans, representing $1.414 trillion in assets. The 2010 EBRI/ICI database covered 46 percent of the universe of 401(k) plan participants.

² Workers are assumed to retire at age 65 and all 401(k) balances are converted into a real annuity at an annuity purchase price of 18.62. Additionally, the projections assume no break in contributions occurs with a change in employers, that the maximum employee contribution is 6 percent of compensation.

³ Those in the higher income quartile have more trouble reaching the success threshold, given the PIA formula in Social Security. Cashouts, loans and hardship withdrawals have approximately the same impact as for those in the lowest income quartile.

Loan Arranging

The headlines last week were all about how new legislation had been introduced to cut down on the “leakage” from 401(k) plans.

That leakage, roughly defined as pulling money out of retirement savings before retirement, has been a concern of many for some time. About once a year, someone puts out a report about the number of outstanding loans from these programs and/or the uptick in hardship withdrawals (see IMHO: Double Dipping) ; and, from 2008, see IMHO: Safety “Net”). Not surprisingly, with the sluggish U.S. economy and the so-called jobless recovery, people seem to be dipping into these accounts at higher rates. And yet, for all the hyperbole around such things, the actual rate seems to be pretty consistent (allowing for differences in the data samplings).

That said, last week, Senators Herb Kohl (D-Wisconsin) and Mike Enzi (R-Wyoming) introduced the Savings Enhancement by Alleviating Leakage in 401(k) Savings Act of 2011 (the SEAL Act), which, among other things, was designed to “[limit] the most 401(k) loan practices that provide easy access to retirement funds but adds costs and fees to pension plans” (see “Legislation Aims to SEAL 401(k) ‘Leakage’”).


From the bill’s summary, the only real “limitation” on loans was capping the number outstanding at any one time to…three. Beyond that, it gave participants longer to stave off having an outstanding loan turned into a “deemed distribution” (a euphemistic term for treating that loan you previously took out as a distribution and subjecting it to income taxation), and it allows participants to continue making contributions after taking a hardship (the current six-month restriction being a vestige of the sense that if you could afford to keep contributing after taking a hardship, you really weren’t in a hardship situation). The SEAL act would ban 401(k) loan debit cards—but did anyone ever think THAT was a good idea?

Now, I’m sure that keeping those repayment windows open longer will compound someone’s recordkeeping headaches, but it seems to me a reasonable means of giving participants time to restore that money to their retirement accounts, rather than simply forking over a big chunk of it to the IRS. Moreover, I think we all know that a properly administered hardship request can alleviate the hardship without imposing a six-month “penalty” on retirement savings (and matching contributions). As for “only” letting participants have three loans out at one time—well, IMHO, that’s hardly a limit at all.

In fact, headlines notwithstanding, the bill might have been more accurately (if less acronymically melodic) titled the “Making It Easier for Participants To Repay Outstanding Loans and Keep Saving” bill.

All in all, the legislation may or may not actually forestall 401(k) leakage, but IMHO, it’s certainly a plumber’s helper.


—Nevin E. Adams, JD

Double Dipping?

Last week, Fidelity published some data from their quarterly analysis of participant activity.

For the very most part, the data (see Fidelity Finds Q2 Uptick in 401(k) Withdrawals) revealed what we have come to expect from such reviews: Participants continue to stay the course, deferral rates are largely unchanged (average was 8%), and more increased their rate of deferral (5.3%) than decreased it (2.9%) in the most recent quarter. Additionally, the balance recovery continues apace, with the average account balance up 15% (though for this good news, you have to reach for a year-over-year comparison, because Q2 figures to be a pretty rough one for participant accounts—down 7.6% in Fidelity’s database).

However, this report drew more than the customary amount of coverage for its finding that rates of hardship withdrawals and loans were higher—and by some measures, significantly higher—than they were a year ago.1

My initial reaction was that this latter finding was something of an outlier; after all, in recent weeks we have seen other, similar reports from other providers that indicate that withdrawal volumes were down, certainly compared with the dark early days of 2009. But then, this was more recent data, and with the report of weekly jobless claims back up to half a million, well, let’s just say that it didn’t require much imagination to see a link between more people out of work and an uptick in “premature” distributions.

But later, as I considered the data, the findings didn’t seem so out of line. While the Fidelity survey found that 11% of participants took out a loan this quarter (9% did a quarter earlier), neither the average initial loan amount ($8,650) nor the loan term (3.5 years) seemed out of line. The report did, however, contain at least two alarming statistics. First, 22% of participants recordkept by Fidelity now have an outstanding loan, and while that isn’t much beyond the one in five that did a year ago, it nonetheless represented a decade-long record for Fidelity’s database. Indeed, this was the item that most media outlets qravitated toward (based on the calls I got).

In fact, IMHO, a more troubling trend lies in the fact that nearly half (45%) of participants who took hardship withdrawals one year prior also took a hardship withdrawal in the 12-month period ending in the second quarter of this year.

While participant loans are (too) frequently touted as borrowing from yourself, they put the participant in the position of having to replace—on an after-tax basis—the contributions you’ve withdrawn and a rate of “return” in the form of interest on the loan. Still, at least you’re on a schedule to replace the money you’ve withdrawn.

Now, properly administered, it’s not easy to obtain a hardship withdrawal. You actually have to be experiencing a financial hardship, for one thing, and you often not only have to prove that, but also that you have exhausted other sources. Worse, that hardship withdrawal is reduced not only by the need to pay taxes on the pre-tax monies you’ve now tapped, you have to fork over another 10% as a penalty to Uncle Sam (unless, of course, you’re older than 59 ½). What that means, of course, is that what you think you are seeing isn’t exactly what you are going to “get.”2 In fact, it could easily be just half what the requesting participant might be expecting.

The reality is that people are surely hurting, and if the money they have set aside in their 401(k) helps them through this period, puts food on the table, or keeps the family in their home, that’s, IMHO, a good thing. After all, there’s not much point in taking out a loan when you don’t have a source of income to repay it.

But what concerns me about the trends in the Fidelity report, aside from the implications that so many are struggling financially, is that, in my experience, hardship withdrawals, unlike loans, are not only gone for a while, they are savings that are often “gone” forever.

—Nevin E. Adams, JD

1Loans initiated over the past 12 months grew to 11% of total active participants from about 9% one year prior. The portion of participants with loans outstanding also increased two full percentage points in the second quarter to 22%. The average initial loan amount as of the end of the second quarter was $8,650, with an average loan duration of three and half years, according to Fidelity.

According to Fidelity’s data, 62,000 of the participants recordkept by the firm initiated a hardship withdrawal in the second quarter, compared with 45,000 participants who did so the quarter before. As of the end of Q2, 2.2% of Fidelity’s active participants took a hardship withdrawal, compared with 2.0% a year earlier.

2 Those taxes aren’t always obvious at the point of distribution, unfortunately, but due at the individual’s next tax filing.