Peter Gulia

Yes, the earlier discussion was about a terminated plan, and I saw your question is about continuing plans. The rule is ambiguous in either situation. Even for a continuing plan, there are at least three interpretations. To resolve the ambiguity in particular circumstances, a plan’s administrator should use no less care, skill, prudence, and diligence than would be used by a prudent person experienced in managing an employee-benefit plan of the same kind for the exclusive purpose of providing the plan’s benefits. In considering whether to deliver a summary annual report to someone who no longer is a participant but was at some possibly relevant time, a fiduciary might consider whether the plan is or isn’t burdened by an incremental expense for the delivery. And if a plan’s administrator has contracted a service provider to deliver a summary annual report, the administrator might consider what’s feasible within the contracted services.

Here’s an earlier discussion: After, the Supreme Court decided Intel Corp. Investment Policy Committee v. Sulyma, No. 18-1116, 589 U.S. ___, 140 S. Ct. 768, 2020 Empl. Benefits Cas. (BL) 69188, 2020 U.S. LEXIS 1367, 2020 WL 908881 (Feb. 26, 2020). The Court held that merely receiving a communication does not mean the recipient has knowledge of the information in it. Rather, to get ERISA § 413(2)’s shorter statute of limitations, a defendant must prove the plaintiff’s actual knowledge. The protection a plan’s fiduciaries might get by “over-delivering” a summary annual report is now much less. Here’s the ambiguous rule: https://www.ecfr.gov/cgi-bin/text-idx?SID=0ef9bcffad94c1b0a2d036c698113b2f&mc=true&node=se29.9.2520_1104b_610&rgn=div8

C.B. Zeller, thank you for reminding us about another discussion, and for your answers. Larry Starr’s observations might have been influenced by the unusual situation presented. And if his view was general, it might reflect his clients’ plans, which typically do not provide participant-directed investment, and his firm’s services. The IRS’s without-source-documents method can work if the plan’s administrator and its service providers carefully meet all conditions of the regulations and that method. Others’ thoughts? Is there a downside to limiting hardship distributions to no more than two in a year?

Adding to Pam Shoup’s questions: Why would a plan’s administrator not use the IRS’s method for not receiving source documents and instead relying on the participant’s written statement (made under penalties of perjury)? The hardship self-certification method now is in the Internal Revenue Manual. IRM 4.72.2.7.5.1 (08-26-2020) https://www.irs.gov/irm/part4/irm_04-072-002#idm140377115475856 Under that method, an IRS examiner must not ask for source documents unless: 1) the notice to participants or the claim “is incomplete or inconsistent on its face”; or 2) some participants received at least three hardship distributions in a plan year, there is no “adequate explanation for the multiple distributions”, and the examiner’s manager approves the request for further information. If a plan’s administrator and its recordkeeper or TPA design the software correctly, #1 would never happen (except for a paper claim, and then only if the claims administrator is careless). And #2 seems unlikely unless abuses are bad enough that the examiner is motivated to do the extra work of getting her manager’s approval. Further, may a plan limit hardship distributions to no more than two in a year?

And if the plan administrator's or claims administrator's interpretation treats a previously incurred medical expense as no longer deemed immediate and heavy because the expense was incurred some months ago, is that something a summary plan description must or should describe?

I’d welcome BenefitsLink commenters’ thoughts about these questions: If a plan limits a hardship to an expense incurred no earlier than the past six (or other number of) months, is such a limit a provision a summary plan description must describe? If not, is it a restriction on a benefit that an SPD must describe? How practical would it be to do this disclosure if the plan’s administrator uses a summary plan description computer-generated by the recordkeeper’s plan-documents software?

If there is an ambiguity: An employer might consider also checking with some others: · the COBRA administrator, if the plan’s administrator engaged one; · the health insurer, to the extent (if any) the continuation coverage would be provided through health insurance; · the stop-loss insurer if, regarding a self-funded plan, the employer bought that insurance. Not checking with them could leave an employer exposed to financial consequences that result from offering continuation coverage to someone another decision-maker considers ineligible. Likewise, an employer might consider informing a continuee that the continuee is responsible to pay whatever portion of the continuation premium the U.S. Treasury does not provide.

austin3515, what you describe is mainstream: an employer puts withholding taxes before any other wage reduction or deduction; and within employee benefits, many employers put health before retirement. My example showed enough employer-paid wages to allow, after withholding taxes, health and retirement reductions. But a mix with less in payment-card tips and more in currency tips could make a wage reduction for a 401(k) elective deferral impossible (at least from that paycheck), and perhaps impractical (even if an employer would accrue a wage reduction until a paycheck supports it, if that ever happens). (I suspect you’re right about what seems practical.) If (whether to pursue your prospective client, or sate your curiosity) you want to learn the real-world practicalities, ask someone who manages payroll for a big restaurant group or for an employer similar to your prospective client.

A § 401(k) deferral measured on the reported wages is feasible if the portion paid by the employer is enough. Imagine a tipped employee’s workweek is five shifts of six hours each. Imagine tips in the week is $400 on payment cards and $600 in currency. Imagine the employer allocates none of the credit or payment processing fees to the server. The employer-paid wage is ($2.13 x 30) + $400 = $463.90 Imagine the tipped employee reports to the employer $300 from the $600 in currency tips. Following this, assume the wages the employer will report to tax authorities is $763.90. Employer-paid wages $463.90 FICA taxes - 58.44 7.65% Federal income tax - 88.13 15% (for illustration) of $587.51 State and local income taxes - 29.38 5% (for illustration) of $587.51 Unemployment taxes - 53.47 7% (for illustration) Health insurance - 75.00 to show RBG’s point 401(k) elective deferral - 76.39 10% of reported wages Net pay $ 83.09 While someone who knows hospitality businesses can explain how these overly simplified assumptions depart from reality, the key is that withholding for taxes is based on a wage more than (if there are tips beyond payment-card tips) the employer-paid wages. The mix of payment-card and currency tips changes over time and by particular work settings.

And EBSA's Field Assistance Bulletin No. 2021-01 (Jan. 12, 2021) interprets relevant law to allow a transferring plan’s administrator to “pay [the PBGC’s] fee from the transferred account, unless the plan terms prohibit such payment.” https://www.dol.gov/sites/dolgov/files/EBSA/employers-and-advisers/guidance/field-assistance-bulletins/2021-01.pdf

The PBGC provides (if the distributee’s benefit is more than the ERISA § 203(e) amount [$5,000]) a joint-and-survivor annuity, unless the distributee elects, with the spouse’s consent, a different form of distribution. Pages 60812-60813, 60828-60829 https://www.govinfo.gov/content/pkg/FR-2017-12-22/pdf/2017-27515.pdf For the forms and instructions, https://www.pbgc.gov/prac/missing-participants-program.

Consider that a tipped employee’s compensation might comprise three elements: (1) a base wage, which might be as little as $2.13 per hour; (2) credit-card tips, which might be adjusted by a portion of credit-card processing fees; and (3) tips paid in currency. Recognizing practical difficulties, a plan’s administrator might design its communications and forms to explain exactly which portion of compensation a 401(k) election operates on. It won’t be total compensation because that amount will be an unknown when the 401(k) election must be made (and might remain an unknown until W-2 wages is reported). Further, with some employers and pay practices (and with some deemed currency tips), it might be impractical for a 401(k) election to operate as a percentage of credit-card tips. Considering the variability in a tipped employee’s compensation, some employers prefer that the 401(k) elections specify a dollar amount, rather than a percentage of any measure of compensation. I’m unaware of a publication on this point.

You are seeing the tax-law issue. The IRS’s view was that a cash-or-deferred election can apply only regarding compensation not yet “currently available” to the employee. And the IRS viewed the cash tips as, if not actually or constructively received, at least available.

In the 1990s, another lawyer and I worked on a plan for a company with several national restaurant chains. The IRS reviewer would not approve anything that would allow an employee to make a cash-or-deferred election on the portion of her wages from cash tips. But an employee might consider a wage reduction that’s a big percentage (perhaps up to 100%) of her net wages, after tax withholding, paid by the employer rather than collected as cash tips. But sometimes that net wage after tax withholding is $0.00. The practical challenges vary with the restaurant. If many customers use only payment cards and no currency, the employer might control those payments. But if many pay tips in currency, more is beyond the employer’s control. I don’t know whether the tax law, or the IRS’s views, have changed since I worked on this issue. And here’s a BenefitsLink discussion:

Stating everything about retirement savings in one document can, in the right circumstances, get efficiencies and convenience. But even more important, it can help an employer avoid outright errors in expressing the plans’ provisions. And looking to one document can improve a governmental employer’s administration of its plans. Which methods, including about documenting plans, make sense turns on the particular facts and circumstances of the plans’ provisions, investment alternatives, service providers, and the employer/administrator.

Without expressing any view about the IRS starting or discontinuing a regime for determination letters: Isn’t it sad that a change sets up an incentive for an employer to use a set of inconvenient-to-read documents instead of stating provisions in one document?

I have written documents that state in one document multiple plans and related trusts (or trust substitutes), including for plans under Internal Revenue Code sections 401(a), 401(k), 403(b), 415(m), and 457(b). This can work with careful attention to accounting for: · each trust, and each subtrust; · each kind of plan; · within each plan, each benefit structure; · each source of salary-reduction, matching, nonelective, rollover, and transfer contributions; · each kind of loan, distribution, transfer, or other payout. Using one document for multiple plans might face difficulties with investment issuers and intermediaries. Those difficulties don’t arise or are readily resolved if those businesses trust the plans’ counsel.

That's the general idea. And perhaps all are highly-compensated (or none is highly-compensated).

There never was a public-law requirement to get the IRS’s determination. Does the service provider say ERISA § 402 (or something else in ERISA’s title I) requires an integrated written plan? Or does it say furnishing an integrated written plan is an obligation or condition under the service agreement. If one is eligible and applies for the IRS’s determination, “a restated document is generally required for an individually designed plan’s determination letter submission.” https://www.irs.gov/retirement-plans/determination-letters-for-individually-designed-retirement-plans-faqs

While I lack expertise on tax law’s coverage and nondiscrimination provisions, one who knows those rules might consider these steps: 1) Define the employer, applying IRC § 414’s provisions and rules implementing or interpreting those provisions. 2) Once the employer is defined, define the to-be-measured workforce. That might include USA citizens and resident aliens, and might exclude nonresident aliens. A plan might exclude a nonresident alien who has no USA-source income. Also, a plan might exclude a nonresident alien “if all of the employee’s earned income from the employer from sources within the United States is exempt from United States income tax under an applicable income tax convention.” 26 C.F.R. § 1.410(b)-6(c)(2). 3) Once the workforce is defined, apply § 401’s and § 410’s rules to discern whether coverage and nondiscrimination conditions are met.

Are all five people partners of the partnership (or members of the limited-liability company)? If so, and if there is no employee, consider whether ERISA’s title I governs the plan. 29 C.F.R.§ 2510.3-3 https://ecfr.federalregister.gov/current/title-29/subtitle-B/chapter-XXV/subchapter-B/part-2510/section-2510.3-3 If ERISA governs the plan, the employer/administrator likely should want a good service provider to take fiduciary responsibility for, or provide services for, 404a-5, 404c-1, and other disclosures.

Adi, thank you for reminding us about that January 12, 2021 “Best Practices” document, and for giving us something to think about. EBSA suggests considering a death search “[i]f participants are nonresponsive over a period of time[.]” But that suggestion supposes that the plan’s administrator already has classified some number of participants as nonresponsive. An absence of communication from a participant does not necessarily make her unresponsive because the plan might not have required any election, direction, or other choice. Imagine a participant who leaves her employer at age 42. If she does not take a distribution and the plan follows her continuing direction to invest her account in a target-year fund, 30 years might elapse before the next activity that requires the participant’s choice or act. Or imagine a participant who at age 72 chose to receive automated monthly direct-deposits of her minimum-distribution amounts. If she dies but the bank account is not closed and continues to collect the direct-deposits, the plan’s administrator might have seen nothing that called for treating the participant as unresponsive. Further, even if a participant has not responded to the plan’s communication that called for a response, there remain difficult questions about how much of participants’ money a fiduciary should spend on finding whether a participant is alive or dead. And if there is some age beyond which one might presume some greater likelihood that a participant has died, what age should one use? Under the Treasury department’s life-expectancy table, someone who reaches age 72 has another 17.2 years—that is, to 89.2 (but about half will live longer). Someone who reaches age 85 has another 8.1 years—that is, to 93.1 (and again, about half will live longer). (I’m aware these tables oversimplify mortality measures.) Absent a postal-mail or email undeliverable, or a rejection of a payment (or the absence of a response to something that required a response): Does this suggest checking for deaths, if at all, only among the 90-somethings?

RatherBeGolfing, thank you for guiding us. Revenue Ruling 80-155 restates Revenue Rulings 70-125 and 71-27 and (again) recognizes “allowing interim valuations in addition to an annual valuation[.]” Rev. Rul. 80-155, 1980-1 C.B. 84, 1980-24 I.R.B. 11 [CCH Pension Plan Guide Pre-1986 IRS Revenue Rulings ¶ 19,530]. All those rulings interpret 26 C.F.R. § 1.401-1(b)(1)(ii). https://www.ecfr.gov/cgi-bin/text-idx?SID=a5ac655b7ae110a969c04307a5e2bf7e&mc=true&node=se26.6.1_1401_61&rgn=div8. The 1970 ruling suggests that investments ought to be valued at “fair market value on the inventory date[.]” Rev. Rul. 70-125, 1970-1 C.B. 87 [CCH Pension Plan Guide Pre-1986 IRS Revenue Rulings ¶ 18,801]. None of the three rulings, including the 1970 and 1971 superseded rulings, mentions generally accepting accounting principles or fiduciary accounting principles. While doing no less than yearly valuations and allocations, may a tax-qualified plan’s governing document allocate the trust’s net assets after setting a reasonable accounting reserve for plan-administration expenses? If not, may the trust at least account for payables on expenses incurred and recorded by the inventory date? If a plan must zero-out the unspent plan-expenses account (whether because the plan must not allow any reserve, or because the unspent amount exceeds a reasonable reserve), what fiduciary decision should an administrator make if the plan’s incremental expense for performing the allocation would exceed the amount to be allocated? I suppose my questions are motivated by recognizing a practical point: For a plan that for everything else has participant-directed investments and daily allocations, often measured by investment funds’ share accounting, it’s jarring to have a yearly allocation, by a plan-specified formula, of what remains unspent from a plan-expenses account.

CuseFan, I hope routine death searches, which a defined-benefit plan’s administrator might do to guard against overpayments, need not be done for an individual-account (defined-contribution) retirement plan. I mentioned ten or eleven years because a plan might not compel an involuntary distribution to a beneficiary until ten years after the participant’s death. Internal Revenue Code § 401(a)(9)(H). About the idea that someone might find a participant’s papers, retirement plans’ use of electronic communications means many get little to no paper. And it might take a while for an email address to become discontinued and generate a bounce-back. Or an employer might have assigned its former employee an email address the employer does not discontinue. But I hope neither IRS nor EBSA makes it necessary to spend participants’ money on routine death searches. I’d like to think most death benefits get claimed soon enough after a participant’s death.

And has anyone seen an IRS examiner assert that a plan flunked 401(a)(9) because the plan's administrator had done nothing (beyond deciding claims submitted) to detect participants' deaths?