Peter Gulia

If the employer/administrator does not count hours of service each day, and instead counts or approximates hours by a wider period, consider (among many points): “In the case of hours of service to be credited to an employee in connection with a period of no more than 31 days which extends beyond one computation period, all such hours of service may be credited to the first computation period or the second computation period. Crediting of hours of service under this paragraph must be done consistently with respect to all employees within the same job classifications, reasonably defined.” 29 C.F.R. § 2530.200b-2(c)(4), https://www.ecfr.gov/current/title-29/part-2530/section-2530.200b-2#p-2530.200b-2(c)(4). The rule section of which that quoted text is a subpart includes at least six admonitions that administrative-convenience rules must be “consistently applied.” If one’s client seeks to make a service-crediting rule the administrator could apply, uniformly, to all situations that involve December and January, what rule would that be? How confident are you that the employer/administrator’s computer system can apply that rule? Is crediting a pay period’s hours of service to the computation period in which the pay period ends simpler than crediting the hours to the computation period in which the pay period begins? This is not advice to anyone.

Without remarking on the many other issues: A plan’s administrator might want its lawyer’s advice about whether—before beginning a further evaluation of whether the writing presented meets conditions to be a domestic-relations order and, if so, a qualified domestic-relations order—the administrator might first take prudent steps to confirm that the writing is a court’s order. Consider, after considering surrounding facts and circumstances, asking the court itself for a certificate that the writing is the court’s order. And consider prudent steps to detect, independently, whether a certificate is a forgery, or was unauthorized. While doing that might not be a plan’s regular procedure for an order the plan received reasonably promptly after the order’s date, a delay of 30 years might suggest a presumption of regularity no longer is fitting. And while a suggestion to get one’s lawyer’s advice often is unheeded, this situation seems to involve unusual risks (and so more value in careful procedure and careful communication). This is not advice to anyone.

Many States’ laws governing a personal power of attorney (rather than a power coupled to a business stake or position) include clear-statement or “are you sure you mean to provide that” rules. Many restrict gifts, or a gift more than a specified amount. Some restrain a power to make one’s principal’s beneficiary designation, or otherwise to change beneficial interests. Even when an ERISA-governed plan’s administrator is unconstrained by a State’s law, an administrator still must read the power-of-attorney document, and must construe and interpret the document’s effect and meaning. To do so, an administrator might interpret the effect and meaning of a document and the powers it grants by looking—at least for some aspects—to the State law the document specifies as the power’s governing law. Or, if the document has no choice-of-law clause, the law of the place where the document was made. (Often, that’s knowable the principal’s acknowledgment or the notary’s certificate.) Construing and interpreting a document in accordance with the law the document’s maker at least impliedly assumed as relevant regarding the document seems a logical way to discern what powers the principal granted or omitted. Using North Carolina law as an illustration, the statute enumerates ten things an agent is not empowered to do unless the power-of-attorney document “expressly grants” the power. N.C. Gen. Stat. § 32C-2-201(a), https://www.ncleg.gov/EnactedLegislation/Statutes/PDF/ByChapter/Chapter_32C.pdf. Using AlbanyConsultant’s story, if the participant’s sister presented as the primary beneficiary’s power-of-attorney document one made following North Carolina’s statutory short-form power of attorney and the principal had not initialed the line for “Disclaim or refuse an interest in property”, a plan’s administrator might find that the agent lacks power to disclaim the primary beneficiary’s benefit. Even if an agent generally has a power to disclaim, an ERISA-governed plan’s administrator might refuse a disclaimer unless the agent can show that the disclaimer is in the principal’s best interest, is not a fraud on any creditor or healthcare-financing regime, and does not involve the agent’s self-dealing (even if the principal had expressly authorized the self-dealing). The situation AlbanyConsultant describes might be untroubled by any of those issues. Rather, an agent might submit the primary beneficiary’s claim, and would deposit the retirement plan’s payment to the primary beneficiary into a bank account the agent set up for her principal’s money. This is not advice to anyone.

If a plan’s administrator recognizes an agent’s authority to submit the primary beneficiary’s claim and approves the claim, the plan pays the distribution to the primary beneficiary. (The agent would deposit the payment into a bank account the agent set up for her principal’s money.) A starting point is, as many BenefitsLink neighbors remind us, RTFD—Read The Fabulous Documents. A plan’s administrator might recognize an agent’s authority to submit her principal’s claim if the plan allows (or at least does not preclude) recognizing the beneficiary’s power of attorney, and the plan’s administrator finds that the power grants the might-be agent authority to do the thing she asks to do. An ERISA-governed plan may (but need not) state provisions for recognizing or refusing an act carried out by a person’s agent. Likewise, a plan may (but need not) state provisions about how the plan’s administrator decides whether it will recognize a person’s power-of-attorney document as sufficient for the administrator to recognize the agent and the agent’s authority. If not inconsistent with the plan, a plan’s administrator may adopt written procedures to guide its exercise of discretion. Unless the plan provides otherwise, ERISA alone does not require a plan to recognize a power of attorney or other agency. See, e.g., United Refining Co. Incentive Sav. Plan for Hourly Emp. v. Morrison, No. 1:12-cv-238, 2013 U.S. Dist. LEXIS 166186, 2013 WL 6147672, at *17 (W.D. Pa. Nov. 22, 2013) (“In order to honor Morrison’s beneficiary designation, the Plan Administrator would be required to determine the meaning and validity of the [power of attorney], which is an exercise explicitly rejected by the court in Kennedy.”). Some administrators interpret plan documents’ silence about recognizing a power of attorney and other provisions that grant discretionary authority as allowing an administrator to recognize or refuse an agent. Discretion must be exercised with loyalty, impartiality, and prudence. Even when an ERISA-governed plan’s administrator recognizes a person appointed an agent, the plan’s governing documents and ERISA govern the meaning and effect of the agency regarding the plan, including whether the agent has or lacks authority to do the thing the agent would seek to do. See, e.g., Taylor v. Kemper Fin. Servs. Co., No. 98 C 0929, 1999 U.S. Dist. LEXIS 14989, 1999 WL 78207 (N.D. Ill. Sep. 24, 1999); Pension Comm. Heileman-Baltimore Loc. 1010 IBT Pension Plan v. Bullinger, No. 1:92 Civ. 00204, 16 Empl. Benefits Cas. (BL) 1024, 1992 U.S. Dist. LEXIS 17325 (D. Md. Oct. 29, 1992); Clouse v. Philadelphia, Bethlehem & New England R.R. Co., 787 F. Supp. 93 (E.D. Pa. 1992); see also In re Shafer, No. 2:13-cv-00405, 2014 U.S. Dist. LEXIS 156622, 2014 WL 5599064 (S.D. Ind. Nov. 4, 2014). With ERISA’s supersedure of States’ laws, a plan’s administrator may interpret a power-of-attorney document in ways that need not follow any particular State’s law. But many fiduciaries consider the meaning and effect of a power-of-attorney document under a relevant State’s law. If a plan’s administrator recognizes an agent’s authority to submit the beneficiary’s claim and approves the claim, the plan pays the distribution to the named beneficiary. (The agent would deposit the payment into a bank account the agent set up for her principal’s money.) A plan would not pay a contingent beneficiary unless the primary beneficiary is dead, or disclaimed, which seems unlikely if the might-be disclaimant has diminished capacity. Even if a plan allows a beneficiary’s disclaimer and might allow a beneficiary’s agent to disclaim, an attempted disclaimer might not be an I.R.C. § 2518 qualified disclaimer, which many plans require, or might be invalid as a fraud on creditors. And at least one court has interpreted that unless a plan states that a power to disclaim can be exercised by a beneficiary’s agent, only the beneficiary personally may exercise the power to disclaim. R. Scott Nickel, as Plan Benefit Adm’r of the Thrift Plan of Phillips Petroleum Co. v. Estate of Lurline Estes, 122 F.3d 294, Pension Plan Guide (CCH) ¶ 23937U (5th Cir. 1997). A fiduciary might consider whether ERISA’s exclusive-purpose loyalty and prudence call the fiduciary to take steps to protect a beneficiary. AlbanyConsultant, you might help the plan’s administrator use its lawyer’s time efficiently by collecting the documents governing the plan and relevant procedures, and noting provisions that might matter for the analysis. And if the plan’s recordkeeper or third-party administrator has procedures about recognizing or refusing a power of attorney, those procedures might aid the plan administrator’s decision-making. This is not advice to anyone.

About whether a severance-from-employment of the third worker results in a partial termination, consider whether amending the plan so that participant is immediately 100% vested might be less expensive than the legal advice the plan’s sponsor/administrator might get to support why the severance-from-employment (if one assumes the change from employee to nonemployee) does not result in a partial termination. This is not advice to anyone.

Paul I, thank you; your explanation is valuable information. I like your observations about being mindful of opportunities for misuse within a client. BenefitsLink neighbors: About a disclaimer or warning, is it common for a TPA’s written material to include a warning that it is only for the named client’s use, and no one else may use it or rely on it? Is a warning like that enough to be “reasonable steps” to guard against risks of misuse? Is there something more a professional ought to do? (I promise not to misuse anything you share with me. And I ask only to support my teaching.)

Have you ever had a client misuse your work to unfairly influence a third person? The professional-conduct code of the American Society of Pension Professionals and Actuaries and other divisions of the American Retirement Association includes this: “A Stakeholder [a Member or a Credential Holder] shall not perform Professional Services when the Stakeholder has reason to believe that they may be altered in a material way or may be used to violate or evade the Law. The Stakeholder should recognize the risk that materials prepared by the Stakeholder could be misquoted, misinterpreted, or otherwise misused by another party to influence the actions of a third party{,} and should take reasonable steps to ensure that the material is presented fairly and that the sources of the material are identified.” Am. Ret. Ass’n, Code of Pro. Conduct, Control of Work Product (amended May 2026), https://fcwpol.files.cmp.optimizely.com/download/cea33626560611f18c27b2e7a7a4a6b0. In your real-world experience, how often does it happen that something you wrote or compiled was used with a person beyond your client? Was your writing misused? Did someone use your materials to persuade a person beyond your client that you support a conclusion, opinion, or advice that’s not your advice? When someone used your materials to persuade a person beyond your client that you support a conclusion you did not express, do you feel you had failed to prepare for the risk that your work could be misused? Or, would the misuse have happened no matter how carefully you expressed your work? Do you think what the rule asks is fair to the professional? Should a professional have a duty to guard against the possibility that someone other than one’s client misunderstands your work you presented to your client? And, most important, why or why not?

Following this rule change, a securities broker-dealer need not require its worker to report one’s opening of a securities account beyond one’s employer if the account is restricted to § 530A accounts (and other securities excused under the rule). Likewise, other banking, commodities, insurance, investment-advice, securities-related, and other financial-services business that supervise all or some workers’ personal transactions might excuse reporting a § 530A account. Yet, a financial-services business might in its procedures require more disclosure and reporting than public law, including self-regulatory organizations’ rules, requires. This is not advice to anyone.

30Rock, imagine some further possibilities: The workers of the transferred business still are the seller’s employees, are leased to the buyer’s new subsidiary, and expenses allocable to those workers are paid by the buyer’s new subsidiary (or the buyer parent or an affiliate). Or, the workers of the transferred business are the buyer’s new subsidiary’s employees, and the buyer’s new subsidiary is a participating employer under the seller’s plan, maybe for a transition period (even if that might result in a multiple-employer plan). Or, another of many ways to allocate economic and accounting consequences between the seller and the buyer. You might get more information when each plan’s administrator reads all the documents, not only all documents governing the plan it administers but also all documents about the deal between the seller and buyer, including related agreements. This is not advice to anyone.

I don’t know what EPCRS or anything of tax law suggests for a situation like this. Might the employer that paid purported contributions ask the receiving plan’s administrator and trustee to recognize the employer’s mistake of fact? Might the employer’s assumption that the employer’s employees could accrue further benefits under a retirement plan of which the employer was not a participating employer be a mistake of fact? Also, might the receiving plan’s administrator’s acceptance of the purported contributions be a breach of that administrator’s fiduciary responsibility? One imagines the receiving plan’s administrator knew, or had it used ERISA § 404(a)(1)(B) prudence would have known, that the payer was not a participating employer (and that the payer’s employee were not eligible for accruals attributable to amounts paid by a nonparticipating employer). If there was a mistake, ERISA’s title I does “not prohibit the return of [a mistaken] contribution to the employer within one year after the payment of the contribution[.]” ERISA § 403(c)(2)(A)(i). The receiving plan would return to the nonparticipating employer the amounts mistakenly paid in, each adjusted for investment loss but not for investment gain. The receiving plan’s net-breakage gain might be allocated to the receiving plan’s account for plan-administration expenses. The employer would pay its affected employees the wages due for the amounts that were not elective deferrals. Next January, the employer would report correctly W-2 wages paid in 2026. The employer would pay each affected employee an interest or time-value-of-money amount on the wages not timely paid, or, if the greater, the amounts each applicable State wage-payment law provides. About the amounts paid for what was not a matching contribution under the mistakenly-receiving plan, the employer might use that money toward any nonelective or matching contribution obligation (if any) the employer has under a retirement plan of which the employer is a participating employer. This is not advice to anyone.

Here’s my follow-up: If a working partner is a less-than-5%-owner “with respect to the plan year ending in the calendar year in which the [deemed] employee attains the applicable age [73 or 75]” (and, to simplify our questions, is a less-than-5%-owner on every day of that year), does this means she is forever a not-5%-owner—even if her capital interest or profits interest later becomes more than 5%? 26 C.F.R. § 1.401(a)(9)-2(b)(3)(ii) https://www.ecfr.gov/current/title-26/part-1/section-1.401(a)(9)-2#p-1.401(a)(9)-2(b)(3)(ii). Does a fortuity that a working partner’s capital interest and profits interest both are no more than 5.0000% for one measurement year mean that no § 401(a)(9) minimum distribution is required until the partner stops working?

Unless a plan sponsor’s consultant is such a big player in the recordkeeper’s business that the firm gets special access (my firm doesn't have that pull with either Empower or Voya), an inside lawyer or other expert won’t take a call not introduced by a sales executive (or an existing customer's relationship manager). So, one would start with the recordkeeper's sales manager for the plan sponsor’s region. One way to get a salesperson to bump a question to the expert is to say, quickly, that you want to not waste the salesperson’s time and effort on a prospect that couldn’t pan out. For example, a salesperson shouldn’t waste her time on a government-sector proposal if the prospect might be nongovernmental and the recordkeeper does not offer services for a small organization’s even smaller select-group § 457(b) plan. Or, even if the recordkeeper has capabilities and offers for both business lines, a proposal is generated from or for a particular business line. So, it might be in a salesperson's interest to get help on a classification question.

That alone doesn’t resolve the classification question. That a State’s statute created an organization does not by itself mean the organization is a government’s agency or instrumentality. And that an organization is distinct from the State or a political subdivision does not by itself mean the organization isn’t a government’s agency or instrumentality. To discern this, one would read the creating or enabling statute and other law to consider the organization’s powers and other facts and circumstances. And would consider how these relate to IRS interpretations about what is or isn’t a government’s agency or instrumentality. This is not advice to anyone. If the organization’s executive is unsure about whether the organization is governmental, your acquaintance might consider reaching out to Empower or Voya. While each would deny that it provides tax or other legal advice, either has inside lawyers and other businesspeople with knowledge and practical experience to help sort out whether an employer is § 457(e)(1)(A) or § 457(e)(1)(B). Another way to get some partial information about governmental or not is to ask whether the organization is required or permitted to participate to participate in a State employees’ retirement system. Or, if the State maintains a § 457(b) plan that admits governmental employers beyond the State itself, ask that plan’s executive or service provider whether the plan would admit the organization as a participating employer. That plan’s counsel, inside or outside, might do the legal analysis. Consider also that a State’s § 457(b) plan might have purchasing power a “tiny” organization lacks. (In my 42 years’ experience with governmental plans, I’ve seen many super-micro employers’ workers enjoy pricing that can be had only with a mega plan’s purchasing power.) Alternatively, being denied participation under a State’s plan might be a partial clue that the organization might not be governmental.

Even if a nonexempt prohibited transaction has been corrected, the plan’s administrator’s Form 5500 report should disclose the transaction, at least to the extent Form 5500’s instructions call for. A statute-of-limitations period for a disqualified person’s excise tax does not begin to run until an excise tax return is filed. Unless the recordkeeper also is a law firm that stands behind its legal advice, an employer or plan administrator gets no reasonable-cause relief for relying on the recordkeeper’s advice. Yet, an employer or plan administrator might consider whether to use (or omit) an IRS or EBSA correction procedure. This is not advice to anyone.

Is the employer a State, a State’s political subdivision (for example, a county, city, town, village, borough, or other municipality), a State’s agency, a State’s instrumentality, or a political subdivision’s agency or instrumentality? [I.R.C. § 457(e)(1)(A)] Or, is the employer a tax-exempt organization (other than a governmental unit)? [I.R.C. § 457(e)(1)(B)] Some service providers have capabilities for one kind or the other, but not for both.

For a third-party administrator, some hard questions in designing a records-retention (and records-destruction) plan are about considering public-law record-retention requirements imposed on the plan’s administrator or the plan’s trustee, and deciding the extent to which the TPA wants to serve as a backup to help the plan’s administrator or trustee meet one’s duties. (Some TPAs perceive that one’s service recipients often are inept in meeting a fiduciary’s records-retention duties, and will be thankful when the TPA has preserved a record the plan’s administrator or trustee ought to have kept.) Another factor might be keeping a record until it no longer could be needed to disprove a not yet time-barred breach-of-contract or tort claim against the TPA. Or, discarding a record so its discovery could not be used to help prove a claim against the TPA. Tightening-up might include rewriting a TPA’s service agreement to narrow obligations, and widen a responsible plan fiduciary’s permissions granted to the TPA. This is not advice to anyone.

Thank you for the always wise reminder that a starting point for thinking about an employee benefit, fringe benefit, or convenience is whether it helps the employer attract or keep workers.

Artie M., I suspect we share some observations. In designing an ERISA-governed plan’s provisions, a plan sponsor need not be burdened by a fiduciary responsibility to decide in the plan’s participants’, beneficiaries’, and alternate payees’ interests. (When the plan sponsor is a business organization, a decision-maker might have some responsibility to the organization’s shareholders, partners, members, or other owners of capital interests or profits interests. When the plan sponsor is a charitable organization, a decision-maker might have some responsibility to the organization’s charitable purposes. Either kind of organization interests can be in tension with participant interests.) One doubts an ERISA-governed plan’s fiduciary has an ERISA-imposed duty to inform the plan sponsor about the fiduciary’s perception that a different plan design would help participants. But ERISA might not preclude a fiduciary from volunteering information to the plan sponsor, if the fiduciary can do so without incurring an expense that burdens the plan. If a plan’s administrator seeks to discern whether the plan’s loan provision is or isn’t “operating as [the plan sponsor] intended”, the administrator would first need to know what the plan sponsor intended (or now intends). Some plan sponsors don’t want a participant-loan provision that “function[s] more like a revolving credit facility than a retirement savings vehicle.” But some plan sponsors don’t object to, or even welcome, a participant-loan provision that some or many participants use to borrow, even frequently, against one’s retirement plan right. And some plan sponsors do not worry about a plan-administration burden if much of the work is within the recordkeeper’s services obligated under its contract, especially if paid for with fees, direct or indirect, from the plan’s assets, not from the employer. In my experience, many employer paymasters prefer—if the plan allows participant-loan repayments collected from an employee’s wages—restricting a participant loan to one at a time. If I advise an ERISA-governed plan’s sponsor (sometimes I advise only a plan’s administrator, and not the plan sponsor), I might suggest that the plan sponsor decide all details of the participant-loan provision, specify these in settlor plan documents, and not grant any discretionary authority to the plan’s administrator. Even if only one human is the decision-maker for the plan sponsor and for the plan’s administrator, I advise that it matters to set clear distinctions between plan-design decisions and plan-fiduciary decisions. But my way of thinking about plan-design choices might be awkward for many employers. A set of IRS-preapproved documents might call for some provisions about a participant loan to be set not by the base plan document and not by the adoption agreement but rather by a loan “policy” or loan “procedure”, often with ambiguity about whether the person that decides those provisions might be responsible as the plan’s fiduciary for those decisions. Some plan sponsors and some plan administrators lack a lawyer’s advice, and might not fully consider the consequences of settlor-or-fiduciary distinctions. Artie M., your suggestion about a best practice makes sense if the committee is a plan-sponsor committee or is a shared sponsor-and-fiduciary committee. Also, it can make sense for a fiduciary committee that has some discretionary power to set some of the participant-loan provision’s terms, or that volunteers to present information to the plan sponsor. My thoughts above are about an ERISA-governed individual-account retirement plan that requires participant-directed investment, and provides a loan to a participant is a participant-directed investment that affects only the borrower’s account. Different ways of thinking often apply for a church plan, if it has not elected to be ERISA-governed, or for a governmental plan. This is not advice to anyone.

Brenda Wren, does allowing multiple loans somehow allow a participant to exploit a weakness in Internal Revenue Code § 72(p)’s loan limit? Or, is using a next loan first to pay off the preceding loan’s outstanding amount, with an amount not so consumed paid to the participant, similar to the effect of multiple loans?

Please help me learn something about plan design. Wouldn’t it be simpler to provide that a participant may have only one outstanding loan at a time? That each next loan must first repay the whole outstanding amount of the preceding loan. Is there a plan-design reason (beyond inattention) for a plan sponsor not to provide this? I confess to not having done the arithmetic about participant loan limits. Am I ignorant about how participants use loans? Am I ignorant about how recordkeepers account for participant loans?

Assume an employer lacks money to make employer-provided contributions to Trump accounts. Assume the employer, in its circumstances, does not fear any effect about coverage or nondiscrimination for any retirement plan, health plan, other employee-benefit plan, or fringe-benefit plan. Should an employer provide the convenience of an employee’s voluntary payroll-deduction contributions to Trump accounts? Why or why not? Should an employer provide information about Trump accounts? Is it best for an employer deliberately to do nothing about Trump accounts? Your thoughts?

From context in your query, I'm guessing the investment adviser did not accept, or has restored to the plan trust, the mistaken amount. Consider crediting to each participant's, beneficiary's, and alternate payee's individual account the amount the account was incorrectly charged, with reasonable interest (or, if greater, the investment adviser's gain allocable to having had the use of the mistaken amount). After all corrections are complete, the plan's administrator should evaluate its procedures and controls, particularly about how the administrator did not instruct the recordkeeper about the change in investment-adviser fees. After discerning the weakness, the administrator might tighten the procedures and document that change. (But the administrator should not write a procedure it won't follow.) This is not advice to anyone.

I too have seen a range of different ways to document a plan's discontinuance, termination, and final distribution. But the key is for a decision-maker to get advice from an adviser responsible to the decision-maker.

Here’s the rule Paul I mentions: 26 C.F.R. § 1.401-10 https://www.ecfr.gov/current/title-26/section-1.401-10.

Does this help? “In the case of an employee of two or more corporations which are members of a controlled group of corporations (as defined in section 414(b) as modified by section 415(h)), the term compensation for such employee includes compensation from all employers that are members of the group, regardless of whether the employee’s particular employer has a qualified plan. This special rule is also applicable to an employee of two or more trades or businesses (whether or not incorporated) that are under common control (as defined in section 414(c) as modified by section 415(h)), to an employee of two or more members of an affiliated service group as defined in section 414(m), and to an employee of two or more members of any group of employers who must be aggregated and treated as one employer pursuant to section 414(o).” 26 C.F.R. § 1.415(c)-2(g)(2) https://www.ecfr.gov/current/title-26/part-1/section-1.415(c)-2#p-1.415(c)-2(g)(2).