Peter Gulia

May a plan charge only those participants who are no longer employed? For some plans, the plan’s participating employers pay some or all of the plan’s expenses. Sometimes, an employer prefers to pay expenses for its current employees, but not expenses attributable to beneficiaries, alternate payees, and participants who are no longer employees. Unless the plan’s documents expressly obligate the employer to pay the plan’s expenses, a plan may charge the plan’s reasonable expenses against the individual accounts of the plan’s participants and beneficiaries. “Nothing in Title I of ERISA limits the ability of a plan sponsor to pay only certain plan expenses or only expenses on behalf of certain plan participants. [S]uch payments by a plan sponsor on behalf of [some] plan participants are equivalent to the plan sponsor providing an increased benefit to those employees on whose behalf the expenses are paid. Therefore, [a] plan[] may charge vested separated participant accounts the account’s share ([for example], pro rata or per capita) of reasonable plan expenses, without regard to whether the accounts of active participants are [not] charged such expenses[.]” Field Assistance Bulletin 2003-3. However, a retirement plan must provide that a vested benefit that exceeds $5,000 may not be distributed without the participant’s consent. ERISA § 203(e)(1), 29 U.S.C. § 1053(e)(1); accord I.R.C. (26 U.S.C.) § 411(a)(11)(A). Interpreting both this ERISA provision and a related tax-qualified-plan condition, a Treasury department interpretation provides that a participant’s “consent” to a distribution isn’t valid if the plan imposed a “significant detriment” on a participant who doesn’t consent (and thus leaves his or her account invested under the plan.) 26 C.F.R. § 1.411(a)-11(c)(2)(i). To interpret this significant-detriment interpretation, the Treasury department stated its view that a plan may charge the accounts of former employees (even while not charging current employees) as long as the expense otherwise is proper and a severed participant’s account bears no more than its “fair share” of the plan’s expense. Rev. Rul. 2004-10, 2004-7 I.R.B. 484-485 (Feb. 17, 2004). (However, the ruling’s reasoning suggests some possibility that an expense allocation that’s more than the “analogous fee[] [that] would be imposed in the marketplace . . . for a comparable investment outside the plan” might be a precluded “significant detriment”.) To illustrate the “fair-share” idea, the Treasury department’s ruling expressly cautions that former employees’ accounts must not subsidize current employees’ accounts: “[A]llocating the expenses of active employees pro rata to all accounts, including the accounts of both active and former employees, while allocating the expenses of former employees only to their accounts” would be an improper allocation. Following this, the Treasury department said that a plan doesn’t impose a “significant detriment” because it charges beneficiaries’, alternate payees’, and severed participants’ accounts “on a pro rata basis”. The Treasury department ruling’s reference to “a pro rata basis” doesn’t mean that a plan can’t allocate expenses among accounts under what the Labor department calls a “per capita” method. The Labor department’s Bulletin uses “per capita” to express the idea of charging an account an amount that’s the same for each account in the class and that doesn’t vary based on the account balance, and uses “pro rata” to express the idea of charging an account a percentage of an account (or subaccount) balance. The Treasury department’s ruling doesn’t draw this distinction, and instead uses “pro rata” only to express the idea of allocating to accounts of former employees (or persons other than current employees) no more than those accounts’ proportionate share. Nothing in the Treasury department’s ruling says that these proportionate shares could not be computed regarding all accounts by dividing the expense by the number of accounts or allocating the expense as a percentage of plan account balances.

Consider also pointing out that not only Form 5500 but also a trustee's, recordkeeper's, and other investment and service providers' systems would treat whatever nine-digit number is furnished as an EIN, and so might expose it more widely than one wants to expose a Social Security Number or Individual Taxpayer Identification Number.

That’s an important part of why my rhetorical question italicizes the word why? The employer’s reasoning for what it does or seeks to do matters. If a (nongovernmental) employer discontinues salary-reduction contributions to a particular insurer or custodian because the employer considers that vendor’s contracts an unwise investment choice, that reason suggests an employer ought to bear fiduciary responsibility for its decision-making. But if an employer’s reason for discontinuing a vendor is about limiting the work burden of the employer’s payroll function, applying neutral limits unrelated to evaluating investment choices might be a “reasonable” constraint that neither establishes nor maintains a plan. (For example, a charitable-organization employer might limit vendors to a specified number, selecting the vendors with the most employee contributors.) The ERISA rule recognizes “[t]he administrative burdens and costs to the employer” as a factor that might be relevant in evaluating whether an employer’s non-plan payroll practice affords a “reasonable choice”. 29 C.F.R. § 2510.3-2(f)(3)(vii)(E). https://ecfr.federalregister.gov/current/title-29/subtitle-B/chapter-XXV/subchapter-B/part-2510 Issues about which decisions do or don’t “establish” or “maintain” a plan are highly fact-sensitive. Further, one can’t get an ERISA Advisory Opinion or other advance ruling. Even if one could, it wouldn’t constrain a court’s decision-making.

That’s an important caution. A plan’s administrator should not present a statement of the kind I describe unless the administrator’s uniform practice is not to look at beneficiary designations and the statement is factually true and not misleading.

Questions about why and how much an employer may limit § 403(b) investment choices while neither establishing nor maintaining a plan are fact-sensitive. The last time I researched this point was a long time ago. If you were to look now, I doubt you’d find a court’s decision that sets a “bright line” in either direction. The employer, with its lawyer’s help, might ask this rhetorical question: If we want no responsibility for maintaining a plan, why do we want to restrain our employee’s investment choice?

Everyone concurs that a plan’s administrator must carefully check everything when (after the participant’s death) there is a claim to decide. kmhaab’s originating post describes a threat that someone might assert it was a fiduciary-responsibility breach for the plan’s administrator to receive a beneficiary designation without warning its maker that the designation would be ineffective to the extent that it attempts to provide benefits to a beneficiary other than the participant’s surviving spouse without the consent of the person who is, on the participant’s death, his then surviving spouse. Such a failure-to-warn fiduciary-breach claim is not about how to decide to whom the plan provides a death benefit. Rather, it is about whether the administrator breached a fiduciary responsibility in its administration of the plan, and, if so, what equitable relief (if any) might remedy the breach. (The harm that might result from the administrator’s failure to warn is the participant’s loss of his opportunity to make a beneficiary designation his spouse consents to.) We can imagine, and Former Esq. describes, some ways a judge or arbitrator might find that a plan’s administrator ought to have protected a participant from an expectation that putting a designation in the plan’s records might have more effect than the plan provides. The first of my posts describes an attempt, perhaps void (if the plan or ERISA requires the administrator to check beneficiary designations) or otherwise ineffective, to warn a participant against such an expectation. About both those points, we don’t know what a particular Federal judge would decide on a particular case’s facts. My experiences (some with a retirement-services provider for tens of thousands of plans with about 12 million participants, and more recently with big employers’ plans) are that many plans’ administrators do not check a beneficiary designation until (after the participant’s death) a claim (rather than an imagined assumption of facts) requires the administrator to decide the correct beneficiaries. If that is a particular plan’s administrator’s practice, does it harm participants to tell them? Even if a warning is void or ineffective to excuse an administrator from a responsibility it had, would having an opportunity to read the warning harm a participant? I don’t suggest a warning is required or needed, only that it might help and should not hurt.

That an activity or power is custody under a Federal or State statute or rule that regulates an investment adviser does not by itself mean that the activity or power results in the person handling, in the meaning of ERISA § 412 [29 U.S.C. § 1112], an employee-benefit plan’s money or other property. Rather, you’d consider all the facts and circumstances about what the investment adviser does, or has authority to do, and evaluate whether it results in “handling” for ERISA § 412. Leaving aside questions about what legal effect the temporary bonding rules might have, at least one of them treats actual or apparent authority to transfer money, rights, or other property to “a third party” as handling that property. 29 C.F.R. § 2580.412-6(b)(3) https://www.ecfr.gov/cgi-bin/text-idx?SID=9ccdd2f6ed4def0e3fed2ceed96cb727&mc=true&node=se29.9.2580_1412_66&rgn=div8 If a person does not handle a plan’s assets but is the plan’s fiduciary (for example, because the investment adviser renders investment advice), ask for your lawyer’s advice about whether such a person might need fidelity-bond coverage no less than $1,000, even if the amount the fiduciary handles is zero. ERISA § 412(a) provides: “The amount of such bond shall be fixed at the beginning of each fiscal year of the plan. Such amount shall be not less than 10 per centum of the amount of funds handled. In no case shall such bond be less than $1,000[.]”). See also DoL-EBSA, Guidance Regarding ERISA Fidelity Bonding Requirements, Field Assistance Bulletin No. 2008-04 (Nov. 25, 2008), at Q 35 (“Generally, each plan official must be bonded in an amount equal to at least 10% of the amount of funds he or she handled in the preceding year. The bond amount cannot, however, be less than $1,000[.]”). https://www.dol.gov/sites/dolgov/files/ebsa/employers-and-advisers/guidance/field-assistance-bulletins/2008-04.pdf

David Rigby, thank you for your further thinking and smart observations. A plan’s administrator (or whichever fiduciary has authority to decide a claim for a benefit after the participant’s death) must consider whatever is relevant in deciding the claims, including consider what effect an attempted beneficiary designation has under the plan’s provisions. Based on my experiences, there might be tens, and perhaps hundreds, of thousands of plans for which no plan fiduciary checks a beneficiary designation until after the participant’s death. And unless a plan’s governing document imposes an obligation or responsibility beyond those ERISA’s title I would require if the documents provide no more than ERISA § 205 requires, it’s not obvious that a plan’s administrator must evaluate the effect of what a participant wrote or omitted until doing so is necessary to decide a claim. Were a plan’s administrator defending an actual or threatened ERISA fiduciary-breach action of the kind kmhaab describes, the administrator might like having in the evidence or complaint a warning of the kind I suggest. That way, even if the litigant has standing to pursue the participant’s rights, one could show a Federal judge (or, even better, persuade a would-be plaintiff’s attorney about what the judge would find) that the participant could not have reasonably believed that the plan’s administrator had approved the attempted beneficiary designation as proper. I recognize, however, that some, perhaps many, might perceive the warning as indecently hard.

What do BenefitsLink people think about using a warning something like this: That this form was received and processed does not mean the plan’s administrator or anyone approved a beneficiary designation. These forms are recorded with no review. If you did not meet your plan’s requirements and conditions to make a valid beneficiary designation (or your designation, valid when made, becomes invalid), your plan’s administrator will follow the plan’s provisions ignoring your attempted designation.

If one edits away portions of the text about kinds of plans other than pension (retirement) plans: Subject to the statutory duration limitation in ERISA section 518 and [Internal Revenue] Code section 7508A, all . . . employee pension benefit plans subject to ERISA or the Code must disregard the period from March 1, 2020 until sixty (60) days after the announced end of the National Emergency or such other date announced by the Agencies in a future notification (the ‘‘Outbreak Period’’) for all plan participants, beneficiaries, . . ., or claimants wherever located in determining the following periods and dates— (5) The date within which individuals may file a benefit claim under the plan’s claims procedure pursuant to 29 CFR 2560.503–1, (6) The date within which claimants may file an appeal of an adverse benefit determination under the plan’s claims procedure pursuant to 29 CFR 2560.503–1(h)[.] Example 7 (Internal appeal—employee pension benefit plan). (i) Facts. Individual F received a notice of adverse benefit determination from Individual F’s 401(k) plan on April 15, 2020. The notification advised Individual F that there are 60 days within which to file an appeal. What is Individual F’s appeal deadline? (ii) Conclusion. When determining the 60-day period within which Individual F’s appeal must be filed, the Outbreak Period is disregarded. Therefore, Individual F’s last day to submit an appeal is 60 days after June 29, 2020, which is August 28, 2020. https://www.govinfo.gov/content/pkg/FR-2020-05-04/pdf/2020-09399.pdf

My suggestions yesterday, to a lawyer who will himself analyze the law and evaluate the facts, suggested one might consider when the money was under the control of the plan’s trustee or its agent. With that suggestion, I invited Griswold to consider whether the TPA was the trustee’s agent. And the context was only the narrow purpose of answering the Form 5500 question.

Just curious, does the IRS's software allow a user to save, in the system, a draft and return to it when the user is ready to make the draft final, pay, and submit?

The Form 5500 question asks: “Was there a failure to transmit to the plan any participant contributions within the time period described in 29 CFR 2150.3-102?” Under that rule, a participant contribution (or a participant loan repayment) is a plan’s asset no later than “the earliest date on which such contributions or participant loan repayments can reasonably be segregated from the employer’s general assets.” https://ecfr.federalregister.gov/current/title-29/subtitle-B/chapter-XXV/subchapter-B/part-2510/section-2510.3-102 One might answer the Form 5500 question by interpreting the nebulous phrase “transmit to the plan” to focus on when the money left the employer’s control and was under the control of the plan’s trustee or its agent. (Was the TPA the employer’s agent, or the trustee’s agent? Or if the TPA was an agent of both, what exactly were the TPA’s obligations in the situations you describe?) If the plan’s fiduciaries, including the administrator and the trustee, treated amounts paid to the TPA as plan assets, and the failures were that some amounts were not promptly credited to participants’ individual accounts, that might be a distinct breach. If it is, the breaching fiduciary might owe restoration. But it is not always and necessarily the same breach (and usually prohibited transaction) that calls for a Yes answer to the Form 5500 question quoted above.

Some of my client plans with Empower Retirement use an identity-proofing service to facilitate online enrollment—for elective contributions—of a participant on whom the recordkeeper lacks both an email address and a telephone number. Many recordkeepers, and some third-party administrators, use identity-control procedures (or the plan administrator’s instruction) for an address change. Many plans restrict a distribution, and even a loan, for some interval after an address change. Some plans require a Medallion Signature Guarantee for a distribution of, for example, $100,000 or more. About identity-control and knowledge-based-authentication services, several suppliers, including Experian, offer services. Here’s my question for BenefitsLink neighbors: Imagine a plan for which the employer pays no plan-administration expense; all expenses must be borne by participants’ (and beneficiaries’) accounts. The plan’s administrator uses identity-proofing services. Assume those services are not embedded in a third-party administrator’s or recordkeeper’s fee. How should the administrator allocate among individuals’ accounts the plan’s expenses for the identity-proofing services? 1. to only those individuals on whom the administrator needed to test identity? 2. to all individual accounts “per capita”? 3. to all individual accounts in proportion to account balances? 4. something else?

The case QDROphile mentions—Templeman v. Dahlgren, Civil No. 89-667-FR, 12 Employee Benefits Cases (BL) 2275 (D. Or. July 31, 1990)—now is over 30 years old. The Federal court decided only that ERISA did not preempt the divorce lawyer’s State-law claim against an employer/administrator. (Considering the facts described in Judge Frye’s opinion with some we might infer, the decision might have been incorrect.) The court remanded a removed-to-Federal action to the State court that claim came from. Many of us read the case as impliedly recognizing at least some possibility to pursue a State-law claim (for example, the tort of negligent communication) against a nonlawyer who gave legal advice to a lawyer. An element of that tort is that the relying person’s reliance on the communication must be reasonable. The Federal court’s opinion does not even mention that it might be unreasonable for a lawyer to rely on a nonlawyer’s legal advice. BenefitsLink friends, have you seen a case in which a divorce lawyer sued a retirement plan’s administrator, third-party administrator, recordkeeper, or other service provider for (allegedly) giving incorrect or incomplete advice about how to prepare a domestic-relations order?

And under many States' corporation laws, even a dissolved corporation might have some powers as needed to wind up the corporation's duties and obligations.

In some recent years, practitioners have had a few successes in presenting an issue with enough time for IRS people to work on it with a view to an announcement during a late February Joint TE/GE Council Employee Plans meeting. https://tegecouncil.org/ Although it’s likely too late for an issue presented now to get an announcement in 2021’s meeting on February 25-26, you might get some “seminar law” remarks, and perhaps some attention for something more formal in early 2022.

ACK, thank you for sharing the information. I’ve never been faced with that choice. But if I were, I suspect I’d likely recommend against providing the one-year-marriage condition. It seems a plan-administration complexity. An exception might be if the plan’s sponsor is owner-dominated and the owner herself wants the condition.

While far from a scientific sample, I checked an IRS-preapproved document a client uses. That set’s adoption-agreement form has nothing to specify that a one-year-marriage condition is provided or omitted. The basic plan document has no mention of a one-year-marriage condition. Has anyone seen a § 401(k)/profit-sharing document that: allows a user to specify a one-year-marriage condition? provides a one-year-marriage condition, without giving the user a choice to omit it?

Here’s a link to the notice: https://www.govinfo.gov/content/pkg/FR-2011-11-08/pdf/2011-28853.pdf.

If one assumes the plan is ERISA-governed: The plan’s administrator might consider whether the plan provides or omits a one-year-marriage condition for a spouse to get whichever ERISA § 205 right—qualified preretirement survivor annuity, or whole account—the plan provides. ERISA § 205(f)(1) states: “[A] plan may provide that a qualified joint and survivor annuity (or a qualified preretirement survivor annuity) will not be provided unless the participant and spouse had been married throughout the 1-year period ending on the earlier of— (A) the participant’s annuity starting date, or (B) the date of the participant’s death.” The statute does not state, at least not directly, that a plan may provide such a one-year-marriage condition to limit a spouse’s ERISA § 205(b)(1)(C) right to get the participant’s account. But consider this rule: Q-26: In the case of a defined contribution plan not subject to section 412, does the requirement that a participant’s nonforfeitable accrued benefit be payable in full to a surviving spouse apply to a spouse who has been married to the participant for less than one year? A-26: A plan may provide that a spouse who has not been married to a participant throughout the one-year period ending on the earlier of (a) the participant’s annuity starting date or (b) the date of the participant’s death is not treated as a surviving spouse and is not required to receive the participant’s account balance. The special exception described in section 417(d)(2) and Q&A-25 of this section does not apply. 26 C.F.R. § 1.401(a)-20/Q&A-26 https://ecfr.federalregister.gov/current/title-26/chapter-I/subchapter-A/part-1/subject-group-ECFRa9e579304da97df/section-1.401(a)-20. The Treasury department’s rule interprets not only the Internal Revenue Code of 1986 but also ERISA § 205. Reorganization Plan No. 4 of 1978 (Aug. 10, 1978), 43 Fed. Reg. 47,713 (Oct. 17, 1978), 92 Stat. 3790 (1978), § 101. See also 5 U.S.C. App. 237. But a court defers to an agency’s rule only if the statute is ambiguous and the rule is a permissible interpretation of the statute. For a situation of the kind Santo Gold describes, a plan’s administrator might consider these steps (with others): 1. Does a governing document state a one-year-marriage condition? 2. If so, does the document’s provision comport with ERISA § 205? Or must the administrator ignore the purported provision because it is contrary to ERISA’s title I? 3. If the plan provides a one-year-marriage condition, how does it apply to the participant’s and the spouses’ facts? BenefitsLink neighbors, do the designers of IRS-preapproved documents give a user a choice about whether to state or omit a one-year-marriage condition?

The point Former Esq. describes might be in ERISA Advisory Opinion 89-06A: “The Department of Labor . . . would consider a member of a controlled group which establishes a benefit plan for its employees and/or the employees of other members of the controlled group to be an employer within the meaning of section 3(5) of ERISA.”

Under Reorganization Plan No. 4 of 1978, the Labor department’s rule to interpret ERISA § 408(b)(1) also controls the interpretation of Internal Revenue Code of 1986 § 4975(d)(1). And until the Supreme Court changes the Chevron precedent, a Federal court must defer to the rule if it is a permissible interpretation of the statute. If a participant loan is a non-exempt prohibited transaction, the § 4975 excise tax is imposed only on a disqualified person, such as the participant. The excise tax never is imposed on a plan, and is not imposed on a fiduciary that acted only as a fiduciary. The IRS could show that a loan with too-generous interest is a deemed distribution, with whatever consequences that brings. Also, the IRS might, in some circumstances, show that a loan with too-generous interest allowed the participant to evade deferral or contribution limits, or otherwise get more tax deferral than was proper. Executive agencies make difficult choices about how much law enforcement to pursue. The IRS is no stranger to those choices (and often describes them in government reports).

A general principle for a prohibited-transaction exemption is that an employee-benefit plan should pay no more, or get no less, than what would result from even-handed dealing. Under the rule to interpret and implement the ERISA § 408(b)(1) exemption [hyperlink above], “[a] loan will be considered to bear a reasonable rate of interest if such loan provides the plan with a return commensurate with the interest rates charged by persons in the business of lending money for loans which would be made under similar circumstances.” While this requires that a lender plan get no less than an imaginary bank lender would get, some might read this text not to preclude a plan getting more interest. Beyond the rule’s text, the Labor department’s explanation of its rulemaking supports such a reading. Loans to Plan Participants and Beneficiaries Who Are Parties in Interest With Respect to the Plan, 54 Federal Register 30520, 30524-30525 (July 20, 1989). Tax law might treat a loan with more interest than is “commercially reasonable” as a deemed distribution. See 26 C.F.R. § 1.72(p)-1 (flush language before the Q&As). https://ecfr.federalregister.gov/current/title-26/chapter-I/subchapter-A/part-1/subject-group-ECFRac15f73ecf59a99/section-1.72(p)-1 But I have never seen the Internal Revenue Service assert this. And my experience includes working inside what then was the #2 recordkeeper with tens of thousands of plans, many of which had situations of the kind MoJo describes. 485073054_participantloanexemptionexplanation30515-30531.pdf

Bird, thank you for your helpful information. About ownership percentages, people get it wrong if they don’t carefully distinguish a partner’s or member’s capital interests, profits interests, and loss interests. And a count of a self-employed partner’s or member’s deemed compensation often is wrong if someone fails to account for the relation between guaranteed-payment rights and other elements.