Peter Gulia

I have no access to IRS gossip that might help answer RayRay’s question. About the retroactive-amendment cycles, consider the effects of waiting until 2026 to make plan amendments that reach back to 2019. A trade-off for not expecting a plan’s sponsor to document a change promptly after it has been put into effect is dealing with a work compression when a documenting cycle arrives. I recognize a TPA often works with processes one had little or no practical ability to decide or even influence. Could a further delay of either SECURE Act’s or cycle 4’s retroactive-amendment cycle lead to other problems?

Almost always, a recordkeeper’s service agreement excludes tax or other legal advice. Further, there often is a warning that no one may rely on the recordkeeper’s explanation even of mere information about tax or other law. You’re smart to pursue your own research.

Paul I, thank you for adding information to aid my thinking. BenefitsLink neighbors, here’s a follow-up question: The saver’s-match changes apply to tax years that begin on or after January 1, 2027. An eligible individual would not specify her choice of her applicable retirement savings vehicle to receive the US Treasury’s saver’s-match contribution until, in early 2028, she files her 2027 income tax return. Imagine plan-document amendments and restatements done in 2026 (and through 2027) have no adoption-agreement item, no other check-a-box item, and nothing else in a service provider’s plan-documents regime to specify whether the plan accepts or refuses a saver’s-match contribution. Imagine the base plan too is silent about whether the plan accepts or refuses a saver’s-match contribution. (Maybe the plan sponsor’s choice, retroactive to 2028, becomes documented in 203n.) Imagine the most recent (in early 2028) summary plan description or summary of material modifications is generated from the plan documents. So, imagine the SPD or SMM communicates nothing about whether the plan accepts or refuses a saver’s-match contribution. How would a participant, seeking in early 2028 to complete her 2027 Form 1040 or a Schedule of it, know whether her employer’s retirement plan would accept a saver’s-match contribution? Do my guesses bear mistaken assumptions? What service methods might recordkeepers build? (For a plan with tens of thousands of participants, including many who might be § 6433’s eligible individuals, telling an inquirer to “ask human resources” is an unwelcome service.)

For the 2022-enacted “saver’s match”, Internal Revenue Code § 6433 provides the Treasury pays the credit “as a contribution” to the eligible individual’s applicable retirement savings vehicle. Whether (i) an eligible retirement plan or (ii) an Individual Retirement Account or Individual Retirement Annuity, a plan or an IRA is not an applicable retirement savings vehicle unless it “accepts contributions made under this section[.]” I.R.C. (26 U.S.C.) § 6433(e)(2)(C) https://www.govinfo.gov/content/pkg/USCODE-2024-title26/html/USCODE-2024-title26-subtitleF-chap65-subchapB-sec6433.htm. Could a plan sponsor design a plan not to accept a § 6433 contribution? If so, what factors might influence a plan sponsor’s decision-making about whether to allow or refuse saver’s-match contributions? A § 6433 contribution gets some treatments and restrictions that could be different from those of other elective-deferral amounts. Does this affect anyone’s analysis and decision-making? Accepting saver’s-match contributions likely requires yet more separate subaccounting. Does that affect decision-making? Imagine a plan easily would meet all coverage and nondiscrimination conditions without accepting § 6433 contributions. Might that affect decision-making? What else should an employer—or a service provider—think about?

As mentioned above: “If the administrator finds the worker is ineligible, the plan’s sponsor might consider whether applying the plan as written might result in a failure of a tax-qualification condition, including those about coverage and nondiscrimination. If so, and if the plan’s sponsor prefers a tax-qualified plan, the plan’s sponsor might consider a retroactive amendment to increase coverage.”

Before considering other steps, the plan’s administrator (even if that is the same person as the plan’s sponsor) might, with its lawyer’s advice, consider whether the worker was or is eligible for the retirement plan. That a worker is, or is treated as, an employee under one or more public laws, even a Federal law, does not—at least not by itself—mean the worker is an eligible employee, or even an employee, within the meaning of a retirement plan’s definitions and provisions. Consider also that if ERISA’s title I governs the plan, ERISA supersedes and preempts the legal effect of a State agency’s finding that a worker is or is not an employee. Or, if a plan is not ERISA-governed, the plan might be governed by the internal laws of a State other than the State that made a finding about who is or is not an employee. Or, the plan’s provisions might make a State agency’s finding—even if the same State’s law governs the plan—unimportant or even irrelevant. Many retirement plans’ governing documents include a definition or provision that a worker is not an employee for the retirement plan unless the service recipient classifies the worker as an employee. That can be so even if the service recipient’s classification of a worker as not an employee is contrary to all public laws. (Some lawyers started writing plans this way beginning in 1974; others began soon after reading Vizcaino v. Microsoft Corp., 120 F.3d 1006, 21 Empl. Benefits Cas. (BL) 1273 (9th Cir. July 24, 1997) (After finding that a class of workers Microsoft had treated as nonemployees had been employees, the appeals court remanded to the trial court the question of what benefit ought to have been provided under a non-ERISA plan, and remanded to the plan’s administrator questions about what benefit, if any, ought to have been provided under an ERISA-governed plan.). RTFD—Read The Fabulous Document. If the administrator finds the worker is ineligible, the plan’s sponsor might consider whether applying the plan as written might result in a failure of a tax-qualification condition, including those about coverage and nondiscrimination. If so, and if the plan’s sponsor prefers a tax-qualified plan, the plan’s sponsor might consider a retroactive amendment to increase coverage. Further, if a change results in a plan its administrator had reported as not governed by ERISA’s title I having existed as ERISA-governed for a period before 2026, the administrator might consider what Form 5500 reports to amend or make. If it becomes needed or helpful to record the worker’s employment commencement date, might that be the first day the worker first performed an hour of service as an employee (even if that date is before the worker became, or could have become, eligible)? Might that date be even earlier than 2023? This is not advice to anyone.

If an individual gets § 403(b) distributions while she remains subject to New Jersey’s income tax, there is some recovery for previously taxed amounts. But if an individual becomes a domiciliary or resident of another State and subject to its income tax, the other State may tax § 403(b) distributions, and need not provide any relief regarding amounts previously taxed by sister States. See 4 U.S.C. § 114 https://www.govinfo.gov/content/pkg/USCODE-2024-title4/pdf/USCODE-2024-title4-chap4-sec114.pdf. New Jersey is not alone in setting up a “double-taxation” risk for someone who retires elsewhere. A participant contribution—whether § 401(k), § 403(b), § 457(b), or something else—is not any exclusion from compensation for Pennsylvania’s income tax (and Philadelphia’s wage tax).

If the absence is “(i) by reason of the pregnancy of the individual, (ii) by reason of the birth of a child of the individual, (iii) by reason of the placement of a child with the individual in connection with the adoption of such child by such individual, or (iv) for purposes of caring for such child for a period beginning immediately following such birth or placement,” ERISA §§ 202-203 set up mandated provisions for crediting service and not having a break in service. And the U.S. Family and Medical Leave Act might provide some benefit-continuation rights. These might bear on how a plan’s fiduciary reads and interprets a last-day condition.

Numbers789, when you asked other TPAs, did you make clear that your situation is not about a partner of a partnership or a member of a limited-liability company treated as a partnership, but rather about a shareholder of an S corporation? If your client (whichever person that is) asks you to perform services assuming the plan administrator’s reckoning of compensation, consider, with your lawyer’s advice, whether to ask your client to indemnify you against your losses and expenses from having followed your client’s instruction—if such a provision is not already in your service agreement. This is not advice to anyone.

That legislation is proposed does not mean it will be enacted. Before 1984, people complained about New Jersey law’s income tax treatment of § 403(b) participant contributions. Criticisms became more focused when New Jersey enacted an exclusion from income for § 401(k) deferrals, but not for § 403(b) or § 457(b). After 42 years’ asymmetry, one might wonder about the legislative prospects. If the NJ-burdened employee prefers non-Roth elective deferrals and the charity is amenable to helping her, the charity might consider establishing a plan with a § 401(k) arrangement. That plan might be available to an employee who is a resident of New Jersey. Conversely, a § 403(b) plan might exclude an employee who is eligible for the employer’s plan that includes a § 401(k) arrangement. An employee who is eligible to make a § 401(k) cash-or-deferred election under a plan of the employer may be excluded from § 403(b)(12)(A)(ii)’s universal-availability condition. 26 C.F.R. § 1.403(b)-5(b)(4)(ii)(B) https://www.ecfr.gov/current/title-26/part-1/section-1.403(b)-5#p-1.403(b)-5(b)(4)(ii)(B).

When the Internal Revenue Service had some humans who would read a few individuals’ income tax returns, one might look at an S corporation’s tax return information about the business and an individual’s description of her occupation to consider whether a shareholder-employee’s wages was reasonable compensation for her work. I imagine now many tax preparers, even some Certified Public Accountants and Enrolled Agents, no longer worry that the IRS might challenge the reasonableness of a shareholder-employee’s compensation. If a shareholder-employee is tempted to declare wages less than reasonable compensation, does it make sense to declare at least the amount that supports her desired retirement contributions?

Beyond tax law, consider whether a suggested plan provision would result in participant loans that “[a]re available to all . . . participants and beneficiaries on a reasonably equivalent basis[.]” 29 C.F.R. § 2550.408b-1(a)(1)(i) https://www.ecfr.gov/current/title-29/section-2550.408b-1. That’s a condition of the statutory prohibited-transaction exemption.

If an S corporation pays its shareholder employee wages no less than $360,000, does it matter whether another element of income is or isn’t countable?

ReallyChill, in my experience, an attachment posted in a BenefitsLink discussion is not available until one has signed in with one's username and password.

If the employee’s pay includes pay for January 1-2, the rule cited above might suggest counting a time when the employee otherwise would work but does not work because of a holiday. “An hour of service is each hour for which an employee is paid, or entitled to payment, by the employer on account of a period of time during which no duties are performed (irrespective of whether the employment relationship has terminated) due to vacation, holiday, illness, incapacity (including disability), layoff, jury duty, military duty[,] or leave of absence.” 29 C.F.R. § 2530.200b-2(a)(2) https://www.ecfr.gov/current/title-29/part-2530/section-2530.200b-2#p-2530.200b-2(a)(2). If a term to be applied is something other than an employment commencement date that refers to a first hour of service, the plan’s administrator might read carefully the plan’s definition. Or if the plan uses a word or phrase but none of the plan documents and none of ERISA title I, the Internal Revenue Code, or other relevant Federal law sets the in-context meaning of the word or phrase, a plan’s administrator might use its discretion to interpret the plan and its discretion to find facts. This is not advice to anyone.

Consider also that “[a]n hour of service [might include] each hour for which an employee is paid, or entitled to payment, by the employer on account of a period of time during which no duties are performed . . . due to vacation[] . . . or leave of absence.” 29 C.F.R. § 2530.200b-2(a)(2) https://www.ecfr.gov/current/title-29/part-2530/section-2530.200b-2#p-2530.200b-2(a)(2). While a norm for many employees is that one must work a while to accrue vacation days, an executive or valued knowledge worker might get a different arrangement. I’ve seen employments begin with a paid vacation or mini-sabbatical—often, two months or more.

So, the recordkeepers that do this can limit the percentage of each deferral or other contribution directed to a to-be-limited designated investment alternative, but don’t limit an account balance? That’s helpful information, thank you.

If an employer doesn’t volunteer to pay plan-administration expenses, expenses necessarily are borne by participants (and beneficiaries and alternate payees). The question is how to allocate expenses among individual accounts. I can see how someone who keep one’s addresses tidy and up-to-date might feel she should not be charged for expenses made necessary by others who did not maintain one’s addresses.

Does any recordkeeper offer (assuming the customer plan has enough purchasing power) services to support the professors' "Retirement Guardrails" idea? For example, imagine a plan's sponsor/administrator specifies that a mutual fund investing in securities about gold or precious metals is a designated investment alternative but only for no more than 5% of a participant's account. Has any recordkeeper developed the systems to apply that "guardrail" constraint?

ESOP Guy, your information aids my thinking; thank you. I advise plans, typically with tens of thousands of participants, for which a plan’s administrator does almost no work beyond instructing and overseeing the recordkeeper’s services. Because a recordkeeper wants no discretion, everything has to be specified as rules-based procedures, often so a computer-system record can apply the procedure, or at least signal beginning the procedure.

If you want to look for courts’ decisions that might help support or attack an interpretation, consider: Mark W. Dundee, Qualified Domestic Relations Answer Book https://law-store.wolterskluwer.com/s/product/qualified-domestic-relations-order-qdro-answer-bk-pen-3-mo-subvitallaw-3r/01t0f00000J3FByAAN?srsltid=AfmBOoqOrI9sEYDQdNBQKGeaMsNfmf1vqBytfIqS2U9Ir2r27Y6QV2bE. There are not many appeals opinions. A Federal district court’s opinion is not a binding precedent anywhere, not even in the same district or even with the same judge. E.g., Camreta v. Greene, 563 U.S. 692, 709 n. 7 (2011) (“A decision of a federal district court judge is not binding precedent in either a different judicial district, the same judicial district, or even upon the same judge in a different case.”, quoting Moore’s Federal Practice § 134.02[1][d] (3d ed. 2011). See also Bryan A. Garner et al., The Law of Judicial Precedent § 3 (Horizontal Precedents) at page 40 (2016) (collecting citations and quotations).

Pam Shoup, yesterday’s notice of proposed rulemaking does not answer, at least not clearly, your question. https://www.govinfo.gov/content/pkg/FR-2026-02-25/pdf/2026-03723.pdf ERISA’s § 105(a)(2)(E) is susceptible to several possible interpretations. That might be specially so at least until the rules and other guidance Congress directs in SECURE 2022 § 338 are published, effective, and applicable. About looking to an employer for an individual’s postal address if the individual still is the employer’s employee, many employers have records that are no more complete and no more accurate than the retirement plan’s recorkeeper’s records. Some employers’ records might be more outdated, and some might have false records. ESOP Guy, for a plan that has no small-balance cash-out and has an involuntary distribution only to the extent of a § 401(a)(9)-required distribution, what do you think about a plan-administration regime that doesn’t begin its extra efforts to refresh a no-longer-employed participant’s postal address until three years before one’s applicable age—thus, for someone born in 1960 or later, don’t search until the participant reaches age 72?

If you are an attorney or lawyer for the would-be alternate payee, consider evaluating (and then advising your client) on some possibilities and probabilities about whether getting discovery so a domestic-relations order would state a percentage might be less expensive and more effective than trying to persuade the plan’s administrator to approve an order that states a time-rule formula. You might consider this even if you have no doubt that the plan’s administrator is wrong. I won’t speculate about what ERISA § 206(d)(3)(C) means, what a Federal court might say it means, or what might persuade a Federal court to not defer to a plan administrator’s interpretation. Rather, my practical point is that challenging the plan’s administrator could be an uphill fight. And even if a plaintiff wins an ERISA litigation, that does not assure an award of attorneys’ fees. Under ERISA § 502(g)(1), a court may, not must, award attorneys’ fees. And it’s in the court’s discretion. A Federal judge might wonder why a plaintiff pursued litigation when a little discovery could have accomplished what the alternate payee needed, without bothering the Federal court’s attention. (I have met judges who would think that way.) This is not advice to anyone.

CuseFan, thank you for helping me think about this. The offer is to ERISA-governed and non-ERISA plans. To plans with or without a cash-out. The $30-a-year fee ends when the individual is found. The brochure does not spell out the details of how that is determined. Perhaps many in a to-be-located class will be found in a first year’s searches. The steps for trying to find an individual are more than records searches and written communications. It can include, if needed, telephone calls to the individual’s spouse, children, other relatives, and named beneficiaries, whether primary or contingent. You’re right that a responsible plan fiduciary would consider the particular fee in a context of the whole of all direct fees and indirect compensation. Yet, adding an incremental service a plan’s current service agreement does not provide might be worth some compensation (or might not). A fiduciary must consider what’s reasonable in light of all the facts and surrounding circumstances. And different plans might have different answers to those questions. The recordkeeper’s offer is a new launch. In my work for my client, I’m not yet evaluating the service. If my client dislikes charging someone for a service she didn’t request, it might be wasteful to look into the service.

blguest, are you certain the plan is an ERISA-governed plan? Might the plan be a governmental plan? If a governmental plan recognizes a domestic-relations order, the plan might require conditions much tighter than ERISA § 206(d)(3) sets. And there can be applicable law beyond the thing that to pension practitioners looks like “the” plan document. Even if an ERISA-governed plan must recognize, or a non-ERISA plan provides that the plan recognizes, not only an order that specifies “the amount or percentage of the participant’s benefits to be paid by the plan to each such alternate payee” but also an order that specifies “the manner in which such amount or percentage is to be determined”, a plan’s administrator might insist that an order “clearly specifies” that “manner”. ERISA § 206(d)(3)(C). Further, consider that a Federal court might defer to an administrator’s exercise of discretion about what “clearly specifies” means, unless one’s interpretation of law or finding of facts is too obviously capricious. For a governmental plan, State law often prescribes in which court and with what special notices and procedures one may sue a governmental actor. And a plan’s administration might be entitled to an attorney general’s or other government-engaged lawyer’s defense with no expense borne by a decision-making official or officer. This is not advice to anyone.