Peter Gulia

If a client asked me to construe and interpret the document, I would not assume that silence about whether to recognize a disclaimer necessarily precludes recognizing one. Rather, I’d consider the whole document. Also, I might, depending on what one finds in the document, consider Federal common law. Depending on the plan’s text, in considering who is or is not a beneficiary, one might read carefully the document’s definitions and usages to consider the extent to which any of them incorporates by reference 26 C.F.R. § 1.401(a)(9)-4 and, if so, what effect that has. That rule section’s Q&A-4 recognizes a possibility that a disclaimer might affect who is or is not a designated beneficiary, which might matter in how the plan’s provisions apply.

Nothing in ERISA’s title I requires a plan to include a provision for recognizing a disclaimer. In my experience, the IRS’s tax-qualification reviewers express no objection to a document’s detailed provisions for recognizing a beneficiary’s disclaimer and setting conditions on a disclaimer the plan’s administrator will follow. An IRS-preapproved document might lack those provisions. It is unclear whether one could add those provisions without defeating a user’s anticipated reliance on the Internal Revenue Service opinion letter on the preapproved document. A plan’s administrator must obey the plan’s governing document. ERISA § 404(a)(1)(D). Although a document might grant the administrator some power to interpret the document, it is a power to interpret ambiguous provisions, not to rewrite the document. An administrator might disobey the plan’s governing document, perhaps considering that the disclaimant is unlikely to sue on the fiduciary’s breach. If an administrator allows a disclaimer, the administrator might recognize only a document that meets conditions under Internal Revenue Code § 2518. Although that section is in an Internal Revenue Code chapter about gift tax, the Treasury department treats a disclaimer that meets the § 2518 conditions as also effective to remove the refused property from the disclaimant’s income for Federal income tax purposes. Without that, a payer might face difficult questions about whether to tax-report a distribution paid to someone else as a distribution to the disclaimant. The logic path above assumes neither A nor B is (or is deemed) a surviving spouse.

And that's what I ask: could an employer do this within an IRS-preapproved document's tolerance for a discretionary nonelective contribution?

An employer that provides a nonelective contribution for a participant who’s repaying a student loan might do so as a recognition that the worker can’t afford to make the elective deferral that would earn the retirement plan’s matching contribution. This has the desired employee-relations effect only if the employer communicates the provision. If the idea of an allocation group of one for every participant really works, imagine a summary plan description and other participant communications with something like this: Beyond participant contributions and matching contributions, your retirement plan allows us to decide nonmatching contributions (if any) in our business discretion. About each participant (including one who makes no participant contribution), we may decide whether to make a nonmatching contribution, and its amount (if any). We decide this as our business choice. We might consider, among many factors, information we have about whether your financial obligations—including those for a student loan, mortgage, or birth-or-adoption expenses—and payments you made on all or some of those obligations made it unreasonable for you to make participant contributions. We might decide a nonmatching contribution somewhat similar in amount to the matching contribution you could have gotten had you made participant contributions. We need not make these business choices uniformly. What do BenefitsLink mavens think about whether this is practical? Does it vary with the size of the employee population? Is it feasible to implement this using an IRS-preapproved document?

How one reasons a finding about whether a person remains, or is no longer, a participant might turn on the purpose for which a plan’s administrator makes such a finding. Just to pick two of the many purposes that might call for a finding: If the purpose is a count of participants for a Form 5500 annual report, the Instructions state: “An individual is not a participant covered under an employee welfare plan on the earliest date on which the individual (a) is ineligible to receive any benefit under the plan even if the contingency for which [the] benefit is provided should occur, and (b) is not designated by the plan as a participant.” If the purpose is discerning whether someone is a participant with information rights under ERISA § 104(b)(4), one would use the statute’s definition of participant. ERISA § 3(7) defines a “participant” to include someone “who is or may become eligible to receive a benefit[.]” The U.S. Supreme Court held this includes an employee with “a colorable claim that” she will “in the future” fulfill eligibility requirements. Firestone Tire & Rubber Co. v. Bruch, 489 U.S. 101, 10 Empl. Benefits Cas. (BL) 1873, 1881 (Feb. 21, 1989). Justice Scalia’s concurring opinion would include “those who (by reason of current or former employment) have some potential to receive the vesting of benefits in the future[.]”

The Treasury department, acting alone, lacks power to change the Labor department’s rule. The turning point about 100 participants is in ERISA § 104(a)(2)(A): “With respect to annual reports required to be filed with the Secretary under this part, he may by regulation prescribe simplified annual reports for any pension plan which covers less than 100 participants.” ERISA § 3(7) defines a “participant” to include someone “who is or may become eligible to receive a benefit[.]” The U.S. Supreme Court held this includes an employee with “a colorable claim that” she will “in the future” fulfill eligibility requirements. Firestone Tire & Rubber Co. v. Bruch, 489 U.S. 101, 10 Empl. Benefits Cas. (BL) 1873, 1881 (Feb. 21, 1989). Justice Scalia’s concurring opinion would include “those who (by reason of current or former employment) have some potential to receive the vesting of benefits in the future[.]” In Firestone, the Court interpreted “participant” to discern who has a right to information under ERISA § 104(b)(4). That’s in the same part 1 that commands a Form 5500 annual report. And ERISA § 104(b)(4) commands a plan’s administrator to furnish an annual report on a participant’s request. With this background, I doubt an agency could permissibly interpret ERISA to not regard as a participant someone who already met a plan’s age and service conditions designed to meet Internal Revenue Code § 401(k)(2)(D)(ii). There might be room to interpret the verb “covers”. But whatever agency interpretation one might imagine requires (at least) the Labor department’s act. Don’t we guess that 2024 will arrive first?

Gilmore's observations involve policy points for Congress to consider. But until they do, do BenefitsLink mavens think a two-plans solution is less expensive than an independent qualified public accountant's audit?

Some practitioners have suggested: If an employer anticipates a meaningful number of employees will become eligible because of § 401(k)(2)(D)(ii), one might—to facilitate efficient coverage and nondiscrimination testing (including with the relief § 401(k)(15)(B)(i)(II) permits), or for other plan-administration reasons—organize two distinct plans: a plan for those who meet eligibility conditions without any to meet § 401(k)(2)(D)(ii), and another plan for those who are eligible only by meeting eligibility conditions provided to meet § 401(k)(2)(D)(ii). One would design and administer the plans to meet required aggregations and disaggregations, and to rely on only permitted aggregations and disaggregations. Further, each plan might bear its proper share of plan-administration expenses. What do BenefitsLink mavens think about that suggestion?

Although the Notice is not a proposed rule, the Notice's part III expressly invites comments.

If this might be about "Waiting for Green Cards", Amy Peck offers some suggestions: https://www.natlawreview.com/article/waiting-green-cards

This discussion is a nice illustration about why an employer/administrator might prefer to keep its own records, independently from the recordkeepers’ and custodians’ records. Don’t just leave those quarter-yearly transaction reports posted to the recordkeeper’s “sponsor service center” internet site. At least copy each digital file onto the plan administrator’s computer drives and further storage. And use considered file-naming and other methods so you can remember and retrieve what you’ve kept. In the 1980s, it wasn’t easy. But now, we can help lead clients to relatively inexpensive protections.

S Derrin Watson, thank you for your excellent help.

That is a logical view. But I’ve learned enough about the realm of IRS-approved documents to recognize that often logic doesn’t control the answers. So, I’m interested in all observations, but especially would like to learn from those who have experience in the making of IRS-approved documents.

S Derrin Watson, thank you for your excellent explanation. Just curious (and mindful that your thinking doesn’t speak for FIS or another business): Considering the effects and trade-off you describe, could a user’s adoption agreement specify both a “rigid discretionary match” and a “flexible discretionary match”? Would that allow not needing the notice whenever the “flexible” matching contribution is zero?

ESOP Guy gives us good guidance: A plan administrator’s careful reporting to the Social Security Administration lowers how many “MAY be entitled” letters go out. Communication with an inquirer gets rid of most inquiries from those letters. Admittedly, my experiences mostly are about situations in which employers and administrators exhausted, or no longer can use, those opportunities.

While we don’t know your situation’s particular facts, the plan’s administrator might consider asking for its lawyer’s advice about using 29 C.F.R. § 2560.503-1 and the plan’s claims procedure. https://www.ecfr.gov/cgi-bin/text-idx?SID=a4695688324f5ce300c2ed273609e32f&mc=true&node=se29.9.2560_1503_61&rgn=div8 If using it, one might follow the many points a carefully designed procedure ought to provide, including: Furnish, and explain, the claims procedure. Invite the claimant to submit evidence showing that a benefit is owing. Search the plan fiduciaries’ records for evidence about the probability (good or bad) that a benefit was paid. That might include evidence showing whether the administrator followed its procedures to direct payment of involuntary distributions, including on-severance cash-outs and minimum distributions. On a denial: Explain carefully the reasoning. Explain that the claimant must exhaust the plan’s claims procedure. Explain the claimant’s opportunities for further review and appeal under the claims procedure. Explain all time limits in the claims procedure. Explain (again, because it should be in the summary plan description), the plan-imposed “statute of limitations” on claims. There are several advantages to following the claims procedure. They include: If the denied claimant complains to the Employee Benefits Security Administration and EBSA opens an inquiry, the plan administrator’s records should show it acted at least in good faith. (In my experience, EBSA closes an inquiry—even if EBSA dislikes the administrator’s decision—if the record shows that the claimant was afforded her procedural rights.) If there is a lawsuit, the defendants might use the plan-administration record to show that the plaintiff’s assertion is so implausible that the judge dismisses the complaint for failure to state a claim. A court should limit its review to the plan administrator’s claims file. A court defers to the plan administrator’s decision unless it could not have resulted from reasoning.

Are you asking about a § 457(b) eligible plan or a § 457(f) ineligible plan? (Your mention of three years before a normal retirement age suggests § 457(b), but it is not the only possibility.) And if § 457(b), are you asking about a governmental plan or a non-governmental plan?

IRS Notice 2020-68 includes guidance about the tax-qualification condition that a § 401(k) plan (but not any § 403(b) or § 457(b) plan) permit an employee to make elective deferrals if the employee has at least 500 hours of service a year in at least three consecutive years and has met the plan’s age requirement (for example, 21) by the end of the three-consecutive-year period. The guidance includes vesting questions mentioned in this thread. https://benefitslink.com/src/irs/n-20-68.pdf See pages 9-12. Thank you to BenefitsLink for always posting these sources so quickly.

Thank you for your kind words. Lois and David Baker provide a super-useful service. Advertising supports it. If BenefitsLink helps you: Consider advertising for your business. https://benefitslink.com/advertise/ When you’re selecting a product or service to use in your business, get information from those that advertise on BenefitsLink.

ERISA preempts many State laws, but preempts no Federal law. ERISA § 514(a), 29 U.S.C. § 1144(a). If there otherwise might have been some doubt, ERISA provides: “Nothing in this title shall be construed to alter, amend, modify, invalidate, impair, or supersede any law of the United States (except as provided in sections 111 and 507(b)) or any rule or regulation issued under any such law.” ERISA § 514(d), 29 U.S.C. § 1144(d). {My quotation is to the statute, not the compilation in the United States Code. That compilation is rebuttable evidence of the law, but the statute controls.} Internal Revenue Code of 1986 (26 U.S.C.) § 6334(c) provides: “Notwithstanding any other law of the United States (including section 207 of the Social Security Act), no property or rights to property shall be exempt from levy other than the property specifically made exempt by subsection (a).” Following those two quoted statute sections, courts have interpreted ERISA § 206(d) as not precluding an IRS levy, lien, or garnishment on a participant’s, beneficiary’s, or alternate payee’s rights under an ERISA-governed retirement plan. For example, United States v. Sawaf, 74 F.3d 119, 123-24 (6th Cir. Jan. 26, 1996); Shanbaum v. United States, 32 F.3d 180, 182-183, 18 Employee Benefits Cas. (BL) 1929, 74 A.F.T.R.2d (RIA) 94-6292, 94-2 U.S. Tax Cases (CCH) ¶ 50,512 (5th Cir. Sept. 16, 1994).

If the Internal Revenue Service’s levy is legally and procedurally proper, that the property the levy reaches is a right under an annuity contract or custodial account held under an IRC § 403(b) plan does not impair the levy. For some background information, read these two parts of the Internal Revenue Manual: IRM 5.11.5 Levy on Wages, Salary, and Other Income https://www.irs.gov/irm/part5/irm_05-011-005 IRM 5.11.6.2 Retirement Income https://www.irs.gov/irm/part5/irm_05-011-006 When income other than “salary or wages” is levied, “the levy reaches a payment the taxpayer has a fixed and determinable right to.” That’s so even if the payment will be in the future. If a proper levy is made when the taxpayer has no right to an immediate distribution, the levy can remain in effect (unless released) while the IRS waits for the taxpayer’s entitlement.

Do the documents governing the plan and its trust provide for, or at least not preclude, a delivery of property (rather than a payment of money)?

One guesses the recordkeeper assembled the report you describe using facts from its records. That reporting might be designed to report facts with little or no attempt to characterize them beyond instructions the plan administrator had furnished. A service provider might do this for reasons of business efficiency. And a service provider might do this so it avoids discretion that could be argued as suggesting that the provider was a fiduciary, or practiced accounting or law. A Form 5500 annual report is the plan administrator’s report. It should report correctly and fairly. Depending on all the facts and circumstances, that might include reporting on Schedule C the reimbursed amount as the plan’s payments, naming each investment adviser and each other service provider, and showing appropriate service and compensation codes for each.

MoJo, thank you for your information and observations. I appreciate your help in filling-in gaps in my experiences. Bird, thank you for your observations about different kinds of recordkeeping conversions. About situations in which the directed trustee or custodian that comes with a recordkeeper is unwilling to hold money not yet allocated to individuals’ accounts, I assume it is the plan’s sponsor/administrator, acting as an agent of a trustee—perhaps an additional trustee beyond a bank or trust company trustee, that would open a bank account to hold unallocated amounts. Whether that account should be credited interest (or use what would be interest against bank fees) is a fiduciary decision. Likewise, whether gross or net interest (if any) should be allocated among participants’ accounts or used for plan-administration expenses is a fiduciary decision. I recognize that the range of practical choices is fewer with smaller plans. While it might seem a nuisance to set up a bank account for a few days or weeks, segregating amounts from the employer’s assets is an important purpose. austin3515, thank you for helping us see a weakness in the ways small plans arrange services.

austin3515, thank you (!) for this helpful information. I remember the difficulties and issues from my experiences, inside 1984-2005, and outside counsel to a few recordkeepers from 2006. My experiences working for a plan's sponsor/administrator are with plans big enough that, if a blackout isn't done within a weekend, the recordkeeper/custodian would hold the money for not-yet-processed contributions and loan repayments in some temporary account, and allocate the amounts later. Have other BenefitsLink people seen different experiences?