What’s Next for the Nonprofit Executive Compensation Excise Tax?

I. Five-Year-Olds

Five is an interesting age. Our first few years are jam-packed with major milestones — sleeping through the night, eating solid foods, crawling, walking, and talking, to name just a few. But by age 5 it seems those once-fantastic feats are all taken for granted. In the developmental lull that follows we don’t delight in the successes of the past; we set our expectations higher for the future.

So, it seems, with section 4960. The excise tax on excess tax-exempt organization executive compensation that was inspired by the $1 million deduction limit under section 162(m) and the golden parachute rules under section 280G, turned 5 last month. And while there are undoubtedly plenty of accomplishments to celebrate from the first few years, expectations are rising. Beginning in 2022, the rather lax reasonable, good-faith interpretation of the statute standard we previously enjoyed is no longer available.¹ Applying the final regulations is now mandatory.²

With mandatory application of the final regulations, applicable tax-exempt organizations (ATEOs) and their related organizations won’t begin filing Form 4720, “Return of Certain Excise Taxes Under Chapters 41 and 42 of the Internal Revenue Code,” in earnest until this spring, making now a good time to focus on two questions: (1) Which aspects of the final rules might revenue agents reasonably expect us to have mastered by the time the IRS moves on from compliance checks to examinations? (2) What interesting ambiguities have arisen that the final regulations don’t address? This article first reviews the basic unambiguous rules finalized in the regulations and then explores some of the ambiguities that remain.

II. Mastery Reasonably Expected

Section 4960 was enacted as part of the legislation commonly known as the Tax Cuts and Jobs Act at the end of 2017.³ About a year later, the IRS and Treasury released Notice 2019-9, 2019-4 IRB 403. Proposed regulations were published about 18 months after that,⁴ with final regulations hot on their heels a mere six months later⁵ — almost a full year before their January 1, 2022, mandatory applicability date and more than three years after the statute’s enactment.

It seems fair to say that the government spent a lot of time thinking about these rules and deliberately gave taxpayers a lot of time to comply with them. It also seems accurate to describe the final regulations as closely tracking the initial guidance provided in Notice 2019-9. Taken together, these observations seem to suggest that the IRS may reasonably expect taxpayers to have mastered the final regulations by the time Forms 4720 for 2022 are filed in 2023.

Granted, some of the rules in the final regulations are more foundational than others. The successor employer rules apply only in limited circumstances and the government received no comments on them despite multiple solicitations,⁶ so it might be unreasonable to expect mastery of those. The more basic rules at the heart of the regulations are what we might reasonably be expected to have mastered by now. Here are some examples of those rules and the resulting implications:

• Every ATEO with employees has “covered employees,” even if none of them earns compensation over $1 million per year. If an employee who made the covered employee list in one year crosses the $1 million threshold in a later year, that remuneration will be subject to the tax. Because it is impossible to predict the future with certainty, it is important to track all covered employees from year to year even if no tax is due under section 4960 in the current year. Names can be added to the list, but they can never be removed. (And whatever you do, don’t lose the list!)⁷

• “Remuneration” is a term of art that is used for identifying covered employees as well as calculating the 21 percent tax. In other words, you can’t know who the covered employees are until you know their remuneration.⁸

• Calculating remuneration is complicated. The rules for determining the timing of remuneration are separate from the rules for determining the amount of remuneration. The amount of remuneration is generally based on the definition of wages subject to federal income tax withholding,⁹ but the timing of remuneration is generally based on vesting, not actual or constructive payment (which is the normal timing rule for federal income tax withholding).¹⁰ Only regular wages (in a word, salary) are included in remuneration upon actual or constructive payment.¹¹ So if you are just making minor adjustments to the amount reported in box 1 on Form W-2 for the year based on the definition of section 3401(a) wages (for example, subtracting the amount reported in box 1 for group term life insurance because it is not included in the definition of wages under section 3401(a) despite being reported in box 1), there is a good chance your calculation of remuneration is wrong in that it does not allocate the remuneration to the correct year. (And, as explained, that also means you may have misidentified your covered employees, a mistake that may have cascading consequences, given that covered employee status lasts forever.)

• Although the vesting-based timing rule referred to above is based on the section 457(f) definition of substantial risk of forfeiture, it is not limited to 457(f) plans.¹² It applies broadly to all sorts of remuneration — essentially anything other than salary — such as bonuses, nongovernmental 457(b) plan benefits, severance pay, and other supplemental wages. Paying severance over a two-year period? The full present value is remuneration in the first year and only the excess over that amount paid the second year is remuneration in the second year — you can’t just include amounts in remuneration as they are paid.

• Earnings and losses on previously included remuneration are combined across all of an employer’s plans in which an employee participates, but not across legal entities and not across employees.¹³ Earnings and losses generally arise from deferred compensation, such as 457(f) plans and nongovernmental 457(b) plans, but any time vesting occurs before payment — which, as noted, might happen with severance pay, for example — there may be earnings or losses arising from the difference between the future value (at payment) and the present value (at vesting) of the remuneration. Separating earnings and losses from other remuneration is important because losses can only offset earnings, not other forms of remuneration (such as contributions to a defined contribution plan or benefit accruals under a defined benefit plan).¹⁴

• Remuneration includes not only amounts paid by an ATEO but also amounts paid by any of its related organizations. This rule is not limited to situations in which the related organization makes payments on behalf of the ATEO for services rendered for the benefit of the ATEO — it explicitly applies to payments a related organization makes to its own employees for the services it receives directly from them having nothing to do with the ATEO.¹⁵

• Proper application of the rules depends on knowing which legal entity (or entities) is (or are) the common law employer (or employers) of each employee who may be a covered employee.¹⁶ While this determination is also relevant in any number of other federal tax contexts, it may not affect tax liabilities the way it can under section 4960. It follows that the IRS may be more motivated to question which entity is the common law employer in the section 4960 context than it has been in other contexts.

You might reasonably question whether these were the only, or even the best, interpretations of the statute. But you will have a hard time convincing anyone — especially anyone at the IRS — that there is ambiguity in the final regulations on these points. And it seems rather unlikely that anyone will decide to litigate these matters.

III. Interesting Ambiguities

A solid understanding of the foundational rules in the final regulations is important, of course, but there is much more to section 4960 than that. Let’s move beyond the basics to some questions the final regulations don’t explicitly answer.

A. Retirement/Deferred Compensation

The applicability of section 4960 to retirement plans has been a source of confusion. You may have noticed that our references to section 457(b) plans were confined to nongovernmental 457(b) plans. That is because governmental 457(b) plans, like section 401(a) plans, are excluded from section 3401(a) wages and thus are exempt from remuneration under section 4960.¹⁷ In other words, you can completely ignore governmental 457(b) plans for purposes of section 4960, but nongovernmental 457(b) plans are essentially treated like 457(f) plans when it comes to remuneration. The governmental versus nongovernmental distinction is thus crucially important for 457(b) plans but, unfortunately, is not explicitly addressed in the final regulations. That’s not to say there is any ambiguity on this point — it’s the sort of technical issue you might need to spend some time working through to understand but for which there is only one correct answer.

A related question that is more open to debate (though perhaps only slightly) is whether section 403(b) plan benefits ever constitute section 4960 remuneration. Section 4960 expressly excludes designated Roth contributions to 403(b) plans from the definition of remuneration.¹⁸ And it would have been arbitrary in the extreme for Congress to have isolated designated Roth contributions from all other forms of section 403(b) benefits to afford them a special exemption from section 4960. Indeed, Congress used precisely the same cross-reference for designated Roth contributions to section 401(k) plans, and no one doubts that all benefits under 401(a) plans (which, of course, include 401(k) plans) are entirely exempt from remuneration. In other words, Congress seems to have taken for granted that section 403(b) benefits, like section 401(k) benefits, generally would be excluded from remuneration so that a special rule was needed only to prevent designated Roth contributions — which are subject to federal income tax withholding — from being the only form of section 403(b), or section 401(k), benefit that would be included in remuneration.

So was Congress mistaken? One inclined to make that case might assert that while section 3401(a) expressly excludes 401(a) plan benefits,¹⁹ it lacks an express exclusion for 403(b) plan benefits; moreover, while the section 3401(a) regulations clearly exempt section 403(b) benefits from withholding, they do not expressly rely on an exclusion from wages to reach that result.²⁰ But for all the same reasons that this regulatory interpretation came about, a contrary interpretation for purposes of section 4960 would be suspect. And when a regulation under section 3401(a) (the statutory definition of wages) itself uses the heading “Wages” and goes on to state that withholding for section 403(b) benefits is not required without stating any other rationale for that conclusion, the natural inference is that section 3401(a) wages do not include section 403(b) benefits even if — and perhaps especially if — the regulation points out, “in general, pensions and retired pay are wages subject to withholding.”²¹ In any event, it seems clear that Congress did not understand the definition of wages under section 3401(a) to include section 403(b) benefits. In sum, at a minimum, there seems to be adequate support in the law to allow us to reach the only conclusion that makes any sense: No section 403(b) benefit is ever included in remuneration.

Before moving on from deferred compensation, there is one more interesting issue worth mentioning. As noted earlier, losses on previously included remuneration can be used to offset earnings on previously included remuneration but not other forms of remuneration. For example, if there are net losses for a particular covered employee for a particular year, those losses cannot be used to reduce other forms of remuneration for the year and instead must be carried forward to offset earnings in a future year. But what happens if there are never any future earnings to soak up those losses? The final regulations do not address this scenario. Some practitioners may conclude that this was merely an oversight and that the only rational way to address this situation would be to allow the losses to be used to offset another form of remuneration in some way. An alternative approach — and likely the one with stronger technical support — would be to focus on the fact that section 4960 is an excise tax and thus should not be expected to comport with income tax norms. Although the result may be counterintuitive — some might even say patently unfair — it seems to be reasonably well established that an excise tax can apply to compensation merely because it has vested, even if the compensation is not ultimately paid.²²

B. Payments to Beneficiaries

Suppose a covered employee dies and amounts that clearly would have been section 4960 remuneration if paid to the covered employee are instead paid to the covered employee’s beneficiary, such as the employee’s spouse. Are those amounts section 4960 remuneration? Oddly enough, the TCJA amendments to section 162(m) included an explicit rule clarifying that “remuneration does not fail to be applicable employee remuneration merely because it is includible in the income of, or paid to, a person other than the covered employee, including after the death of the covered employee.”²³ Some practitioners considered that an entirely unnecessary clarification of an entirely obvious conclusion — of course section 162(m) cannot be avoided simply by paying a covered employee’s compensation to a third party, they thought. The risk of that clarification, particularly when contemporaneous with a similar provision lacking that clarification, is that it gives rise to a negative inference. So what should we make of the fact that section 4960 was enacted as part of the same legislation and does not include an explicit rule like that? Are we to infer that posthumous payments escape section 4960, just not section 162(m)? This issue is not explicitly addressed in any section 4960 guidance, so we are left to figure it out for ourselves.

For starters, as explained earlier, the timing rule for most section 4960 remuneration is based on vesting, so actual payment — whether to the covered employee or another person — might be irrelevant. Sure, the vesting-based timing rule does not apply to salary, but the covered employee’s unpaid salary at death is likely to be a relatively modest amount. Therefore, as a practical matter, this issue is likely to matter most when vesting is accelerated upon a covered employee’s death. For example, a covered employee might have a $5 million unvested section 457(f) plan benefit that vests upon the covered employee’s death and is then paid to a beneficiary. Another, more modest, example is the year-of-death “earnings” on remuneration from prior years under a 457(f) plan.

Of course, the timing of remuneration is only relevant if the amount is remuneration in the first place. As explained earlier, the answer to that question generally depends on whether the amount constitutes wages under section 3401(a). If that were the only rule, the analysis might be rather straightforward because it is well established that amounts paid after an employee’s death are not wages under section 3401(a).²⁴ That exclusion takes care of things like unpaid salary and nongovernmental 457(b) plan benefits. Unfortunately, however, it may not take care of 457(f) plan benefits. Recognizing that benefits under a section 457(f) plan probably are not wages under section 3401(a) under any circumstances,²⁵ section 4960(c)(3)(A) specifically includes in remuneration “amounts required to be included in gross income under section 457(f).” Note that this language does not specify whose gross income is relevant, and section 457(f) explicitly applies not only to plan participants but also to their beneficiaries.²⁶ Moreover, section 4960(a)(1) is similarly broad, referring to “remuneration paid . . . with respect to employment of any covered employee.” Taken together, these rules seem to suggest that section 457(f) plan benefits that vest upon a covered employee’s death and are then paid to beneficiaries generally must be included in remuneration upon vesting. (In some cases, the exception for death benefit plans under section 457(e)(11)(A)(i) may apply, but perhaps only to the extent the benefits paid upon the employee’s death exceed the plan’s lifetime benefits.²⁷)

In light of the seemingly odd inconsistency between section 457(f) benefits and everything else, some taxpayers may be inclined to reject the conclusion that posthumous section 457(f) benefits are required to be included in a covered employee’s remuneration. Because it is hard to conceive of a policy rationale for that disparate treatment, the better view is that this was not the intent. And when we consider the section 162(m) amendment clarifying — perhaps unnecessarily, but perhaps not — that section 162(m) applies to payments includable in the gross income of third parties, doesn’t it make sense to conclude that Congress did intend a different result under section 4960? Perhaps more to the point, it is easy to imagine taxpayers taking this position in the absence of specific guidance. And this isn’t a far-fetched scenario that only a few ATEOs and related organizations will ever face. Everyone dies eventually (at least as far as we know, as of the time of this writing), and covered employee status is unshakable. Perhaps Treasury and the IRS will provide guidance on this issue someday. In the meantime, it seems inevitable that taxpayers will reach differing conclusions on their own.

C. Section 162(m)(6) Interplay

Some ATEOs with related health insurance companies essentially have taken the position that section 4960 does not apply on that basis. Can this position possibly be correct? Once again, the final regulations do not address this issue, so we are left to figure it out for ourselves. Although it takes a bit of work to understand why, it turns out the position is probably correct in most cases.

Section 4960 is based primarily — albeit only loosely and only in part — on section 162(m). Section 162(m) is best known for imposing a $1 million limit on the annual deduction a publicly held corporation may claim for compensation paid to its executive officers. But that’s just paragraphs (1) through (5) of section 162(m). Let’s call those paragraphs “regular section 162(m).” Section 162(m)(6), however, is different from regular section 162(m). In one very important respect, it is narrower than regular section 162(m): It applies only to some health insurers (“covered health insurance providers,” also known as CHIPs) and their aggregated group members. But in virtually every other respect it is broader than regular section 162(m): The limit is $500,000 instead of $1 million; it is not limited to officers (or even employees, for that matter) but rather applies to compensation paid to any individual; and the aggregation rules are based on section 414 instead of the much narrower section 1504 rules adopted via regulation for regular section 162(m).²⁸

The section 162(m)(6) regulations require proration of the $500,000 limit if an individual works for more than one of the aggregated entities.²⁹ For example, if an individual performs 75 percent of her services for one entity and 25 percent of her services for another entity (and she receives compensation from the entities in proportion to the services performed for each), the first entity’s deduction limit is $375,000 (75 percent of $500,000) and the second entity’s deduction limit is $125,000 (25 percent of $500,000). The section 162(m)(6) regulations do not address whether the proration is affected by whether each entity has any use for its share of the $500,000 deduction. Continuing the example above, suppose a taxable CHIP paid $750,000 and a related exempt organization paid $250,000 to the same individual during the year but the EO had no unrelated business taxable income and thus had no use for any compensation deductions. Can the entire $500,000 deduction limit be allocated to the taxable CHIP, since that is the only entity taking any tax deductions? A 2017 chief counsel advice memorandum says no, the 75 percent/25 percent split applies in this scenario exactly as it would if both entities were taxable entities claiming compensation deductions. The key passage reads:

Neither section 162(m)(6) nor the regulations thereunder provide an exception for tax-exempt entities that are members of an aggregated group. Remuneration paid by a tax-exempt entity engaged in a related or unrelated trade or business is considered otherwise deductible regardless of whether the entity may use the deduction (for example, regardless of whether the entity has taxable income, such as unrelated business taxable income, against which the deduction may be taken).³⁰

In other words, the memorandum is saying that a portion of the $500,000 deduction limit must be wasted on the tax-exempt entity because tax-exempt entities are subject to section 162(m)(6) just like taxable entities are, even if they have no practical use for the theoretical deduction available to them. Although the memorandum involved a 75/25 split, it shouldn’t matter what the proportions are. For example, applying the memorandum’s reasoning, even if the tax-exempt entity had paid 100 percent instead of 25 percent of the individual’s remuneration and the taxable CHIP had paid 0 percent instead of 75 percent, section 162(m)(6) technically would apply to the tax-exempt entity, which would be allocated 100 percent of the $500,000 limit despite having no effect on its actual compensation deductions (because it does not claim any).

You may be wondering what any of that has to do with section 4960. We are finally ready to connect the dots. Section 4960 contains a section 162(m) coordination rule: “Remuneration the deduction for which is not allowed by reason of section 162(m) shall not be taken into account for purposes of this section.”³¹ Treasury and the IRS agree that this language incorporates not only regular section 162(m) but also section 162(m)(6).³² If the memorandum is correct that section 162(m)(6) applies to tax-exempt entities even if they have no unrelated business taxable income and claim no compensation deductions, it should follow that those tax-exempt entities are not allowed deductions “by reason of” section 162(m)(6) so that section 4960’s section 162(m) coordination rule applies.³³ The government should not be allowed to whipsaw taxpayers by taking an expansive view of section 162(m)(6) to limit compensation deductions but a narrow view of section 162(m)(6) to increase remuneration subject to tax under section 4960. Simply put, either the memorandum’s section 162(m)(6) conclusion is incorrect or an ATEO with a CHIP in its aggregated group has no section 4960 remuneration — the government can’t have it both ways.

Good for you if you are wondering why there has been no mention yet of the other half of section 4960 — the half loosely based on section 280G. The term “remuneration” is not relevant to that half of section 4960. Does that mean we’re only halfway home?

In a word, yes. Just as there is no stacking of section 4960 and section 162(m), there is no stacking of the two halves of section 4960 — the half inspired by section 162(m) and the half inspired by section 280G.³⁴ So unlike real sections 162(m) and 280G,³⁵ you don’t stack the adverse tax consequences of their section 4960 analogs. For example, if a covered employee with a $1 million base amount vests and receives a severance payment of $5 million, the section 4960 excise tax is 21 percent of $4 million, not 42 percent of $4 million. Stated differently, for reasons that have never been fully explained (perhaps because they are inexplicable), Congress determined that some severance payments are so inherently bad that the excise tax should apply even when they are less than $1 million. For example, if a covered employee with a base amount of $200,000 were to receive a severance payment of $700,000, the section 4960 excise tax would be 21 percent of $500,000, even if that covered employee’s remuneration were well below $1 million. While it is theoretically possible for that to occur, excess parachute payments of less than $1 million are exceedingly rare. (Of course, the same is true of excess parachute payments over $1 million, but those are all but irrelevant because of the no-stacking rule.) If there is a CHIP in the aggregated group, however, the section 280G-inspired half of the section 4960 excise tax would be more likely to apply, because it would no longer be swallowed by the section 162(m)-inspired half. In sum, even if having a CHIP in the aggregated group effectively eliminates the section 162(m)-inspired half of the section 4960 excise tax, it also increases the likelihood that the section 280G-inspired half of the excise tax will apply.

If Treasury and the IRS do not agree with this conclusion, they haven’t said so in any published guidance. Moreover, it is not clear how they can disagree without abandoning the section 162(m)(6) memorandum. In the meantime, taxpayers are taking different positions.

IV. Enforcement Activity

From the very beginning, stakeholders asked the IRS and Treasury to make section 4960 as administratively simple as possible by relying to the maximum extent possible on data from information returns that ATEOs were already required to prepare, such as forms W-2 and 990. Stakeholders pointed out that this approach not only would make compliance less burdensome for taxpayers but also would make enforcement easier for the IRS. Treasury and the IRS concluded that although simpler approaches had a lot going for them, what they didn’t have was any statutory basis. Putting aside whether that conclusion was correct, the stakeholders were undeniably correct on both counts. Complying with section 4960 is indeed much more challenging because nearly all the analysis is entirely bespoke. And because remuneration is so far removed from the amounts reported on Forms W-2 and, in turn, Forms 990, the IRS has no efficient way of verifying Form 4720 reporting without engaging in a full-blown examination.

If you have any doubt about that, consider IRS enforcement activities to date. The IRS has been sending section 4960 “compliance check information requests” to tax-exempt entities. It seems these requests may be sent if an organization’s Form 990 reporting suggests that perhaps the IRS should have received a Form 4720 reporting some section 4960 excise tax, but it didn’t, and the IRS wants to know why. The enormous chasm between the section 4960 rules and the Form 990 reporting that seems to be prompting these requests is exemplified by the following list of some anecdotal real-life explanations for the disparity:

• The entity that filed Form 990 has no employees at all or no employees who receive any remuneration.

• The Form 990 aggregation rules are not the same as the section 4960 aggregation rules.

• The entity has a fiscal year other than the calendar year and thus was not subject to section 4960 for its entire fiscal year that included the statutory effective date.

• The medical services exception under section 4960(c)(3)(B) applies.

The last of those examples is even included in the compliance check information requests themselves among a preprinted list of explanations for why no section 4960 excise tax is due, along with convenient checkboxes that can be used to indicate which of the several reasons may apply. Here are the others:

• Excess parachute payments under qualified plans as defined in section 280G(b)(6).

• Excess parachute payments made under or to an annuity contract described in section 403(b) or a plan described in section 457(b).

• Excess parachute payments to individuals who are not highly compensated employees as defined in section 414(q).

• Other.

Some taxpayers have received identical compliance check information requests year after year. The IRS can’t be happy about having to base its enforcement efforts on information it knows is likely irrelevant. But without a full-blown examination, it’s not clear a better alternative is available. It will be interesting to see whether the IRS shifts to a more thorough examination-based approach to enforcement now that applying the final regulations is mandatory.

Some practitioners are concerned that even in full-blown examinations IRS examiners will lack the sophistication of the chief counsel attorneys who worked on the section 4960 regulations and will mistakenly expect Form 4720 reporting to align with Form W-2 or Form 990 reporting, even though that alignment was explicitly rejected in the rulemaking process. Of course, this raises broader questions about the complexity of federal tax law and the challenge of enforcing complex and ambiguous provisions. In the section 4960 context specifically, it has some practitioners wondering whether taxpayers might be better off aligning their reporting on forms W-2, 990, and 4720 as much as possible, even if that is not technically required — or even technically correct — in hopes of having less to explain to an IRS examiner. For example, some practitioners fear having to explain to an IRS examiner that a covered employee’s common law employer is the correct legal entity to file Form 4720, even if it is a different legal entity than the one filing Form W-2 or the one whose Form 990 reports the covered employee’s compensation. Other practitioners are more optimistic that taxpayers taking technically correct positions should have nothing to worry about. Time will tell whether the optimists or the pessimists have the better crystal ball.

V. Conclusion

Whatever else you may think of section 4960, you would probably agree that it shows no signs of being repealed anytime soon. You might even agree that, as we have seen with section 162(m), it is more likely to be expanded (regardless of which party has control in Washington) than repealed. With these points in mind, now might be a good time to be sure you have mastered the basics of applying section 4960 and are giving some thought to what might come next.