Top Cats in Top Hats


mostly, but not entirely, exempts from its provisions unfunded plans “maintained by an employer primarily for the purpose of providing deferred compensation for a select group of management or highly compensated employees.” In the informal taxonomy of employee benefits speak, these plans carry the label “top-hat” plans. (I personally would have preferred “top-cat plans,” finding a Hanna-Barbera feline a more visualizing arresting image than the pinnacle of Fred Astaire’s formalwear.)

Top-hat plans are not, as I just noted, entirely exempted from ERISA. The statute exempts them from ERISA’s funding, vesting, accrual, and distribution minimum standards, and from ERISA fiduciary regulation. They are subject to minimal reporting and disclosure requirements and are also subject to Part 5 of Title I, which houses ERISA’s enforcement and preemption provisions.

Given this odd patchwork of coverage and exclusion, top-hat plans occupy a somewhat harsh regulatory environment. Participants in such plans get little substantive help from ERISA, either with respect to vesting or fiduciary conduct. At the same time, participants lose the pre-ERISA recourse they had to state law remedies (because top-hat plans are subject to ERISA preemption).

So what is the idea underlying the limited ERISA exceptions for top-hat plans? The legislative history is somewhat sparse, but a few observers, and the Department of Labor, suggest that top-hat plans are exempt from important ERISA protections because their participants can protect themselves by affecting or influencing the design and operation of the plan, through negotiation or otherwise. A few courts have elevated this observation into the statutory definition of top hat plan by examining whether participants had the ability to influence or affect the design of the plan.

Other courts, though, have focused on two facts: the percentage of employees covered by the plan and whether the employees are, in fact, management or highly compensated. To the extent there is a magic percentage, beyond which a plan is not a top-hat plan, the number appears to be somewhere around 15%. (At least one court has suggested that if a plan’s primary purpose is to cover a select group of management and highly compensated employees, the inclusion of a few non-management and non-highly compensated employees does not deny the plan a top-hat exemption.)

But I wonder whether the term select is actually intended to mean a group limited by number. The dictionary definition of “select,” when used as an adjective, is not one of maximum numerical limitation, but rather one of identifying the most fit members of a particular group. Thus, a relevant question, as a matter of statutory interpretation, is how to determine who of a group is the most fit to participate in a plan without most ERISA substantive protections. I would suggest that fitness here should mean that an employee has the ability to influence the design and operation of the plan. This would comport with the Department of Labor’s supposition concerning the purpose of the exemption for top-hat plans.

I’ve become interested in top-hat plans for two reasons, one of which may reveal a bias in my analysis. The first reason is that there is little writing about what a top-hat plan is. (There is, of course, a lot of writing about when a plan for executives is unfunded, but that is because of the tax treatment of non-qualified deferred compensation, an area of law that has just gone through statutory restatement and revision. But after 35 years of ERISA, there has not been a lot written on how we know when a plan wears a top hat.) The second reason is that I have consulted on a case involving a plan that was, by self-characterization and, unfortunately (here is my bias), by judicial characterization (in the district court), a top-hat plan.

It is an interesting case, worth discussing. The following account of the case somewhat simplifies fairly complex facts, but it does provide a reasonably accurate context in which to present the legal arguments that were resolved by the district court:

The case involved a teaching hospital, whose compensation structure was subject to a university-wide salary cap, which was below the earning level of certain physicians that the hospital wished to attract to its staff (or to retain). To attract such physicians to the hospital (and other physicians who hoped in the future to exceed the salary cap), the hospital adopted a “deferred compensation” plan. (There were actually two plans, but for purposes of this account, I will treat them as a single plan, though their features were not identical.)

A physician employed at the hospital was paid a salary that was partly based on projected practice income. The salary itself could not exceed the salary cap. But the doctor was also paid a portion of his net practice income. If the doctor’s total compensation for a year exceeded the salary cap, a portion of the net practice income would be credited to the plan. (The doctor could direct how the money, held by the hospital, would be invested. Creditors of the hospital could, of course, reach the money, so the plan was technically unfunded.) The doctors paid current tax on plan contributions but not on investment income once the contributions were in the plan.

Although the plan provided for deferred compensation until retirement, it also provided for partial forfeiture if a plan participant left the hospital in a year in which he ran a practice deficit. The plan also provided for circumstances in which the plan could pay a participant current compensation if his or her salary declined substantially or to pay a practice income deficit (where salary anticipated more practice income than was actually produced) to avoid financial hardship. (Otherwise, a deficit was recovered by docking future salary by 10% per year.)

The plaintiff in the plan was a doctor who left the hospital in a year in which he ran a practice deficit. The hospital thus docked his plan account by a significant percentage of its value. (Keep in mind that the forfeited amounts included already-taxed dollars.) Could the plan do this? Not if it were subject to ERISA’s vesting rules. But top-hat plans are not subject to these rules.

But was the plan a top-hat plan? The doctor had several interesting arguments about why the plan was not a top-hat plan, all of which were rejected by the district court.

The first argument was that the doctor had little power to influence the design or operation of the plan. This seems obvious. Who would want to participate in a plan that taxes you on contributions and then provides a forfeiture condition? I think the answer is no one. (By the way, at least so far as I can tell, the hospital did not have a contractual right to recover a practice deficit from a doctor who left the hospital but did not have an account in the plan. In any event, the hospital could not have used a self-help remedy if the plan had not been in existence—it would have had to go to court to enforce whatever contractual right it had.) This suggests that the individual doctor, and probably the doctors as a group, did not have much of an opportunity to influence the design or operation of the plan. And indeed, the facts of the case suggest that the doctors, neither individually nor collectively, could have influenced the hospital to change this basic but alarming plan feature. (In some sense, the proof is in the pudding: if they could have, they would have.) The court, though, essentially ruled that a plan’s top-hat status does not turn on whether a participant has the power to influence plan design. According to the court, Congress intended high compensation levels and/or management status to be a proxy for the ability to influence plan design and operations.

The second argument was that there were too many doctors who participated in the plan for it to be a top-hat plan, i.e., far more than 15% of the workforce. But many of the doctors never exceeded the salary cap and thus never received contributions. Should these doctors be considered part of the select group? Well, if the idea behind the top-hat exemptions is that a participant can protect himself against “abuses” that would otherwise be curtailed by ERISA (or at least understand what the risks of the plan actually are), the hospital employees who participated in the plan were (according to the doctor’s argument) all the hospital’s physicians, since in any given year any physician could have had practice income that exceeded the salary cap and thus received a contribution to the plan. The physicians vastly exceeded 15% of highly paid and management employees and also vastly exceeded 15% of the entire workforce. However, the physicians that actually received a contribution numbered fewer than 10% of the employees. The court held that the only physicians who counted were those who actually received contributions, not those who at the beginning of the year were eligible for contributions depending on their practice income during the year.

The third argument was that the plan’s primary purpose was not deferred compensation, a statutory condition for top-hat status, but rather to negotiate the problems that were presented by the university’s salary cap and the year-to-year fluctuations in a doctor’s practice income. (Remember that the plan not only provided a mechanism for the hospital to recover a practice deficit of a doctor who left the hospital during the year, but also permitted a doctor whose salary declined to receive current distributions from the plan.) Deferred compensation was certainly a result and one purpose of the plan, but perhaps not the primary purpose. The plan would likely have looked different if deferred compensation had been the primary purpose.

The district court disagreed with these arguments but straightforwardly described the issues. If the case is appealed, it might provide the appellate court a good platform to think through the issues of what types of plans should be able to dance around ERISA’s substantive rules while twirling a top hat on the cane of ERISA preemption. My own view (although, as I said, I am biased) is that the district court’s take was wrong. But the issues are interesting and, I think, important.