Peter Gulia

When in 2018 did the participant get the hardship distribution?

ESOP Guy, thank you. (I have thought about logical consistency with the employer’s reporting for other purposes.)

Kevin C, jpod, and Bird, thank you for your further observations. Because I must not reveal my client’s confidences, I can’t explain to BenefitsLink readers why in the particular circumstances the questions are real and innocently raised, and why seeking careful answers to them really matters. And I hope it won’t surprise anyone that, despite my burdens for this project, I volunteered to write off my time. I haven’t yet formed conclusions. Again, thank you, everyone, for helping me expand my thinking.

Yes, and the plan's fiduciary will use my written advice to show that the fiduciary acted loyally and prudently. Thank you, everyone, for helping me test my thinking.

And whatever the employer or fiduciary decides (or both decide), each might want its lawyer’s advice about: whether and how to document the record of what information was considered and what analysis was done; communicating the decision to meet whatever duties might apply; even without or beyond a duty of communication, communicating to start the running of a statute-of-limitations period on a disappointed person’s claim.

And what if Congress might provide that no 401(a)(9) minimum distribution is required for the Roth portion of a participant's 401(k), 403(b), or 457(b) plan account. For counting U.S. Government revenue gains and losses for tax or budget legislation, anything the scorers assume will motivate increased use of Roth pushes revenue gains into earlier years, and the scoring frame is no more than ten years.

An employer might not pay to avoid a threatened change to the participants’ tax treatment. Not every employer needs a deduction for the contributions. But even if an employer would “own” all tax treatments, my observation is that it’s awkward for an employer to be in the role of tax law enforcer when it has no other stake in deciding whether its employee is retired. Why is an employer the arbiter of whether someone no longer gets tax deferral on portions of her retirement savings?

One fact common to the situations I’ve described is that the employer has no money at stake, only its role as an enforcer of Congress’s tax policy.

Four years ago, a postdoctoral student in my course on Professional Conduct in Tax Practice chose for his semester paper to write about whether it’s feasible for a lawyer or a certified public accountant (he is both) to render advice to a marijuana business. Many associations’ opinions say it doesn’t break the profession’s conduct rules merely to render advice. But those opinions disclaim considering anything beyond the professional-conduct rules. Challenges under other law—including crimes of aiding another’s crime or engaging in a monetary transaction that involves proceeds of a crime, and risks of forfeitures (which might undo a fee)—have scared off many professionals. Even a client’s engagement of a professional might be vulnerable to an argument that the agreement is not a legally enforceable contract. What I heard about Empower was last month. But I’m not aware of anyone checking on it. A roomful of practitioners, mostly TPAs, said they wouldn’t accept a marijuana business, even for work strictly limited to a retirement plan. If anyone wants Empower’s answer, perhaps it’s straightforward to call there and ask.

I heard (but have not checked) that Empower Retirement / Great-West Life & Annuity Insurance Company is willing to serve some marijuana businesses.

Luke Bailey, thank you for reminding us about using an administrator’s powers and discretion. Recognizing a fiduciary’s duty to use those powers impartially is why my client seeks advice. My originating example is based on just one of several kinds of situations the administrator must interpret. And the number of people and situations involved is more than count-’em-on-your fingers. So, I’m still hoping BenefitsLink mavens will help us discover some principles or reasons to apply. One way of looking at the one-month-a-year situation is to assume that wages 1/12 of full-time is not enough to live on, and from that assume the worker is retired. What about someone who works throughout the year, but averages 13 hours a month? Or someone who works every week, putting in 3 or 4 hours a week? Those situations produce wages similar to the one-month-a-year worker, but is the work somehow more regular? If regularity matters, what about someone who works every month, but only one day a month? Does it matter whether the work is physical labor, office skills, or knowledge work? And for all these, why does it matter?

CuseFan, thank you. Yes, the participant is eligible to make elective deferrals (for any pay period that has wages).

As I mentioned, the participant has a right to take a distribution, but no obligation to receive a distribution unless the plan’s IRC § 401(a)(9) provision compels it. In the original situation—in which the employee consistently works one month every year (and the employee and the employer both expect their relationship to continue), what fact, if any, would trigger a plan administrator’s finding that the employee became “retired” or severed from employment?

jpod, thank for your helpful explanation. BenefitsLink mavens, am I right in guessing there is also some for treating at least the first situation as not requiring a plan's administrator to compel a minimum distribution?

ESOP Guy, thank you for the caution about difficulties that might result if someone is treated as retired more than once. This individual-account plan has no matching or nonelective contribution, only elective deferrals. David Rigby, thank you for the Gray Book information. jpod, thank you. Other than a once-a-year W-2 report of the wages paid in the preceding year, there might be little or no indicia about whether the worker is an employee. The employer does not provide security badges for any of its employees. Likewise, the employer does not furnish e-mail addresses or direct-dial telephone numbers for others, even full-time, in the worker’s group. The employer “self-insures” worker’s compensation, so there is no list or number furnished to an insurance company. The employer does not provide health coverage for any worker. All, the employer would prefer not to compel an involuntary distribution unless doing so is necessary to avoid a tax-qualification failure based on a failure to meet IRC § 401(a)(9). (The plan already permits an in-service distribution grounded on the participant’s age.) Since there seems to be a consensus answer about the not-hypothetical situation, let’s ask about a variation from it. What happens if there is a year in which the worker is not needed? For example, imagine the worker worked each January in 2011-2019, but in 2020 there is no work for this worker to do. Would we treat him as severed from employment on December 31, 2020? Or is it enough that the worker remains willing to work, and the employer will call him to work, if there is work for him to do?

An employer maintains a retirement plan that provides no involuntary distribution except as required to meet Internal Revenue Code § 401(a)(9). The employer has a non-owner worker, older than 71, who works in only one month of each year. The work is real, not a subterfuge. Should the employer/administrator treat that worker as “retired” to compel a minimum distribution? Why or why not? Does it matter whether the worker is or isn’t available to work in the other eleven months (if the employer wanted services of the kind the worker performs)?

Some administrators' QDRO procedures give the participant and each proposed alternate payee an opportunity to submit whatever information one wants the administrator to consider in evaluating whether an order submitted for treatment as a QDRO is a QDRO. In an ERISA litigation, some judges use a litigant's failure to present an argument at an administrative stage as support for dismissing a complaint. And if an argument is presented and the administrator renders a reasoned decision, a Federal court often defers to an administrator's decision (unless it was an abuse of discretion).

And to further support the reasoning about why the employer's hiring practice does not abuse the nondiscrimination provision, one might record the employer's independent business reason for selecting the particular worker. For example, someone who previously served as a regular worker in the business and, even when not working, has informal communications with its chief executive might have business knowledge or skills superior to those of others who are available to work on a temporary basis.

In nonrule guidance, the Labor department stated a view that a plan’s administrator need not question a domestic-relations court’s description of a person as the participant’s former spouse, even if the same court in the same proceeding decided that there was no marriage. ERISA Adv. Op. 92-17A (Aug. 21, 1992) https://www.dol.gov/sites/default/files/ebsa/employers-and-advisers/guidance/advisory-opinions/1992-17a.pdf If there is a proceeding about whether a plan’s administrator met or breached its responsibility to administer the plan, a court need not defer to (and might not be persuaded by) that interpretation. In addition to other steps in evaluating whether an order is a QDRO, a plan’s administrator might evaluate whether the order “relates to the provision of child support, alimony payments, or marital property rights to a spouse, former spouse, child, or other dependent of a participant[.]” ERISA § 206(d)(3)(B)(ii)(I). Please understand that I don’t express a view in either direction.

A Labor department rule distinguishes compensation for services, overhead, and direct expenses. 29 C.F.R. § 2550.408c-2 https://www.ecfr.gov/cgi-bin/text-idx?SID=24dba48a3e2a9983025132fa53bf4ce0&mc=true&node=se29.9.2550_1408c_62&rgn=div8 For a fiduciary who gets full-time pay from the employer, that rule allows only a “reimbursement of direct expenses properly and actually incurred and not otherwise reimbursed.” Even then, a fiduciary must not benefit herself or a person in which the fiduciary has an interest that could affect the exercise of the fiduciary’s best judgment as a fiduciary. And there can be more restraints than those.

At least one IRS-preapproved document appears to allow an involuntary cash-out if “the vested amount of an Account payable to a Participant or Beneficiary does not exceed $5,000 . . . at the time such individual becomes entitled to a distribution hereunder[.]” (Your plan's governing document might state different provisions.) But ERISA § 404(a)(1)(D) commands a plan’s fiduciary to follow the plan’s governing documents only insofar as the documents “are consistent with” ERISA’s title I. Even if one may rely on the Internal Revenue Service’s opinion letter for some assurance that a document is not contrary to Internal Revenue Code § 411, there is no assurance about anything for ERISA’s title I. ERISA § 203(e)(1) [29 U.S.C. §1053(e)(1)]: “If the present value of any nonforfeitable benefit with respect to a participant in a plan exceeds $5,000, the plan shall provide that such benefit may not be immediately distributed without the consent of the participant.” Consider also 26 C.F.R. § 1.411(a)-11(c)(3)(i): “Written consent of the participant is required before the commencement of the distribution of any portion of an accrued benefit if the present value of the nonforfeitable total accrued benefit is greater than the cash-out limit . . . on the date the distribution commences.” I recognize I have not answered your question. Perhaps others on BenefitsLink might steer you toward more knowledge than I have. And whatever you learn, you should evaluate with your lawyer's advice.

cpc0506, if you suggest the plan’s sponsor search for its written plan, consider these possibilities: An investment issuer or custodian might never have had possession of its customer’s written plan, nor an obligation to cause its account holder to make one. Some financial-services businesses might set up an account based on an applicant’s statements, with little checking on the accuracy of those statements. In my experience, even organizations with reasonable know-your-customer procedures might not detect that a customer lacks a written plan.

QDROphile, thank you for your helpful observations. All the people I'm thinking about are non-owners, and also have no way (beyond good work) to influence whether the employer continues the employment. So, it's about how much work means still-employed rather than retired.

If designed to require no more than IRC 401(a)(9) requires, a retirement plan (other than an IRA) need not compel a distribution to a participant who is not a 5% owner until after "the employee retires." Lacking a detailed rule about when for 401(a)(9) purposes an individual-account plan's participant "retires", many administrators treat severance-from-employment as the dividing line. But is there any range in which someone who remains on the employer's roster as an employee works so little that she should be treated as retired to invoke a required beginning date? For some examples, how about an employee who works: 20 days in one month (with no work in the other 11 months)? one day every month? a half-day every month? one day each quarter-year? a half-day each quarter-year? (All these describe real situations.) Is it good enough for 401(a)(9) purposes to treat an employee as not retired until a calendar year's W-2 wage report shows zero wages?

If a plan's fiduciary finds the expense of allocating an amount among participants overwhelms the amount of the settlement proceeds received, what instruction do you ask for? That the settlement proceeds be allocated to a plan-expenses account? If the plan does not already have a plan-expenses account, do you establish one? Are there other practical solutions?