Fittingly, the United States’ seminal law regulating employee benefits, the Employee Retirement Income Security Act of 1974 (ERISA), became law on Labor Day fifty years ago. This new regulation of employee benefits represented a “decade of congressional study of the Nation’s private employee benefit system.” Such a lengthy study produced no small product. ERISA is an “enormously complex and detailed statute that resolved innumerable disputes between powerful competing interests.” But a hefty legislative accomplishment need not be a sloppy one, as courts have recognized. ERISA is also a “comprehensive and reticulated statute,” one designed carefully and with whose enforcement scheme courts are “reluctant to tamper.”
A live disagreement over two interacting provisions in ERISA, however, tests this impression. Section 406 proscribes “prohibited transactions” between a fiduciary and a “party in interest”—including a party “providing services to” an employee benefit plan. This is sweeping language: § 406 prohibits payments by a plan to any entity providing, for example, essential services like recordkeeping.Section 408, however, can exempt many of these prohibited transactions. Section 408 allows a plan fiduciary to engage in an otherwise prohibited transaction if it is reasonable in compensation and necessary for the plan’s operation.
Sections 406 and 408 thus comprise one of ERISA’s many instances of balancing competing interests—here, interests of a plan fiduciary and a plan participant alleging that a prohibited transaction occurred. The question dividing courts is whether § 406’s broad application prohibits seemingly permissible, even “necessary” or “reasonable,” transactions conducted at arm’s length, like recordkeeping arrangements with third parties, or whether § 406 does not apply as broadly as its text suggests. The Second, Third, Seventh, and Tenth Circuits believe the latter; the Eighth and Ninth Circuits believe the former. The answer, in turn, influences what a plaintiff must plead to plausibly allege that a plan administrator engaged in a prohibited transaction. If § 406 means plainly what it says, a plaintiff need only plead (subject to the modern pleading standard) facts alleging that a plan administrator transacted with a third party. If § 406 does not apply as broadly as its text suggests, however, a plaintiff must plead more—for example, the elements in § 408: that the transaction provided unreasonable compensation and that it was unnecessary for the plan’s operation. Within this divided terrain, the Supreme Court granted certiorari for the Second Circuit’s opinion in Cunningham v. Cornell University. With oral argument scheduled for January 22, 2025, the Court will soon resolve yet another question involving pleading standards under ERISA.
This paper sides with the Eighth and Ninth Circuits and argues that § 406 means what it says: a prohibited transaction involves any party “providing services to” an employee benefit plan and extends as broadly as its plain text permits. This paper proceeds as follows. Part I lays out the relevant statutory provisions and provides some context for fiduciary duties. Part II explains the Second, Third, Seventh, Eighth, Ninth, and Tenth Circuits’ holdings. Part III provides several considerations for resolving their disagreement. Specifically, Part III parts ways with the Second Circuit, which held that the § 408 defenses of reasonableness and necessity are elements of a claim that a plaintiff must plead. Building on that point, Part III argues that § 408 contains affirmative defenses that plaintiffs need not anticipate during the pleading stage, as this construction enjoys support from courts construing prohibitory provisions in other statutes, particularly the Age Discrimination in Employment Act of 1967 and the Fair Debt Collection Practices Act, that resemble § 406 and § 408 structurally. Additionally, Part III explains, consistent with the modern pleading standard, what a plaintiff must allege in a complaint to show a prohibited transaction occurred. Finally, Part III collects familiar legal mechanisms that help divide the “plausible sheep from the meritless goats”—that is, plausible complaints under § 406’s broad language against meritless claims that could discourage fiduciaries from entering into third-party transactions to administer a plan effectively.
A central component of ERISA’s comprehensive scheme is the enforcement of fiduciary duties. One part of the Act’s regulatory provisions is entitled “Fiduciary Responsibility,” which provides plan fiduciaries with a “number of detailed duties and responsibilities.” Section 404 lays out fiduciary duties, which have been called the “highest known to the law.” Two such duties in § 404 are relevant to this paper. First, § 404(a)(1)(B) imposes a prudence requirement: plan fiduciaries must discharge their duties “with the care, skill, prudence, and diligence under the circumstances then prevailing that a prudent [person] acting in a like capacity and familiar with such matters would use.” Second, the duty of loyalty requires plan fiduciaries to “discharge [their] duties with respect to a plan solely in the interest of the participants and beneficiaries.”
Section 406 supplements these fiduciary duties by specifically prohibiting certain transactions between plan fiduciaries and parties in interest—that is, certain transactions “deemed ‘likely to injure the pension plan.’” The statute provides:
A fiduciary with respect to a plan shall not cause the plan to engage in a transaction, if he knows or should know that such transaction constitutes a direct or indirect—
(A) sale or exchange, or leasing, of any property between the plan and a party in interest;
(B) lending of money or other extension of credit between the plan and a party in interest;
(C) furnishing of goods, services, or facilities between the plan and a party in interest;
(D) transfer to, or use by or for the benefit of a party in interest, of any assets of the plan; or
(E) acquisition, on behalf of the plan, of any employer security or employer real property in violation of section 1107(a) of this title.
Moreover, ERISA defines, in relevant part, a “party in interest” as “a person providing services to such plan.” With § 406’s text and the definition of “party in interest” taken together, § 406’s sweep becomes clear—a fiduciary cannot transact with someone providing services to a plan, which include the direct or indirect furnishing of goods, services, or facilities. On its face, § 406 prohibits fiduciaries not only from entering transactions likely to injure a plan (transactions not conducted at arm’s length), but also from paying third parties to perform essential services supporting a plan.
Section 408, however, can exempt many of these otherwise prohibited transactions. Specifically, § 408(b)(2) exempts from § 406 contracts and “reasonable arrangements” with a party in interest for “services necessary for the establishment or operation of the plan” if the party in interest also receives “reasonable compensation.” Additionally, the accompanying regulation defines “necessary” with similar breadth as § 408: a service is necessary if it is “appropriate and helpful to the plan obtaining the service in carrying out the purposes for which the plan is established or maintained.” Thus, § 406 and § 408 are similarly broad—what is prohibited wholesale may be exempted by similar magnitude.
Confronted with a broad prohibition and a similarly broad exemption, courts have reached different conclusions about how these provisions interact. This Part first describes the Eighth and Ninth Circuits’ holdings, which have interpreted § 406 to apply broadly. It then describes the Second, Third, Seventh, and Tenth Circuits’ holdings, which have restricted § 406’s reach.
The Eighth Circuit was the first in this group of courts to embrace an expansive reading of § 406. In Braden v. Wal-Mart Stores, Inc., Jeremy Braden, a Wal-Mart employee, alleged that Wal-Mart executives managing his (and others’) pension plans entered into a prohibited transaction under § 406 by having Merrill Lynch & Co., the plans’ trustee, hold the plan assets in a mutual trust, in which plan participants could invest, and provide administrative services necessary to the plans’ operation. Braden alleged that Wal-Mart’s selection of the mutual funds failed to consider Merrill Lynch’s interest in including funds that shared Wal-Mart’s fees with Merrill Lynch. Braden alleged that this failure cost the plans $60 million. Additionally, he alleged that Wal-Mart’s transactions with Merrill Lynch gave Wal-Mart “undisclosed amounts of revenue sharing payments in exchange for services rendered to the Plan.” The district court granted Wal-Mart’s motion to dismiss, concluding that Braden failed to plausibly allege that Wal-Mart had paid unreasonable fees to Merrill Lynch and thus engaged in a prohibited transaction.
The Eighth Circuit confronted § 406 with an eye first toward pleading standards. As an initial matter, the Eighth Circuit concluded that § 408’s exemptions for reasonableness and necessity are affirmative defenses. In so holding, the Eighth Circuit noted that the “burden of proving justification or exemption under a special exception to the prohibitions of a statute generally rests on one who claims its benefits.” As such, § 406 allocates the burdens of pleading and proof, and Braden did not need to show the latter. Instead, a plaintiff need only plead that the defendant entered into a prohibited transaction. Here, that meant Braden simply had to show—and did show—that Merrill Lynch “furnish[ed]” services to Wal-Mart. Pleading this much, the court said, aligns with Rule 8 of the Federal Rules of Civil Procedure, which requires that a complaint present a “short and plain statement” showing the plaintiff deserves relief. And because Braden plausibly alleged that a prohibited transaction occurred, Wal-Mart then had to prove Merrill Lynch’s services were reasonable and necessary for the plan under § 408.
Moreover, the Eighth Circuit concluded that its holding aligns with traditional trust law principles, which, as the Supreme Court has noted, may inform interpretations of ERISA. At common law, fiduciaries bore the burden of justifying the type of transaction § 406 prohibits—that is, one in which a fiduciary “might be inclined to favor [a party in interest] at the expense of the plan’s beneficiaries.” Drawing from this tradition, the Eighth Circuit concluded that § 406 always, for purposes of fairness, places the burden of proof on the party in the presumptively self-dealing transaction. Fairness, the court noted, was pertinent in Braden’s case, given the information asymmetry plaintiffs often face in the pleading stage. So early into litigation, Braden could not yet show that Wal-Mart’s transactions with Merrill Lynch were unreasonable or unnecessary because the trust agreement between Wal-Mart and Merrill Lynch required the amounts of the payments to be kept secret. As such, Wal-Mart bore the burden of proof for § 408’s exemptions.
Over a decade later, the Ninth Circuit similarly held that § 406 applies broadly to transactions with parties in interest. In Bugielski v. AT&T Services, Inc., Robert Bugielski, a former AT&T employee, alleged that AT&T entered into prohibited transactions. Fidelity Workplace Services (Fidelity) had served as the plan’s recordkeeper since 2005. In 2012, AT&T amended its contract with Fidelity to allow plan participants to access Fidelity’s brokerage account platform, BrokerageLink. For a fee, BrokerageLink permitted participants to invest in mutual funds not otherwise available through AT&T’s plan. Fidelity, besides the fees it collected from participants, also received revenue-sharing fees from the mutual funds available through BrokerageLink. After plan participants invested billions into BrokerageLink’s mutual funds, Fidelity made millions in revenue-sharing fees.
Moreover, in 2014, AT&T contracted with Financial Engines Advisors to provide optional investment advisory services to plan participants. Because Financial Engines Advisors needed access to participants’ accounts, AT&T amended its contract with Fidelity to provide Financial Engines Advisors that access. Financial Engines Advisors and Fidelity entered into a separate agreement in which Fidelity received a significant portion of fees (millions of dollars) that Financial Engines Advisors earned from managing participants’ investments. Because of this string of events, Bugielski alleged that AT&T’s amendment of its contract with Fidelity to incorporate BrokerageLink’s and Financial Engines Advisors’ services constituted a prohibited transaction.
Unlike the Eighth Circuit, which considered pleading standards and common law fiduciary principles, the Ninth Circuit stuck primarily to § 406’s text. There was “no dispute” that Fidelity “provid[ed] services to” the plan, as defined in 29 U.S.C. § 1002(14)(B), and was thus a party in interest. Additionally, AT&T’s act of amending the contract between it and Fidelity constituted a “furnishing of services” under that phrase’s ordinary meaning. Put together, these facts meant the contract amendment was a prohibited transaction under § 406. To that end, the Ninth Circuit refused to trim § 406’s scope, explaining that § 406 creates a “broad per se prohibition of transactions.” Because Congress provided an exemption in § 408, the Ninth Circuit saw no need to “fashion a judge-made” one by trimming § 406’s scope.
The Ninth Circuit confirmed its textualist approach with an advisory opinion by the Department of Labor’s Employee Benefits Security Administration (EBSA) explaining why it amended § 408(b)(2)’s implementing regulation. EBSA’s explanation contemplates arm’s-length transactions as prohibited transactions under § 406 and specifically contemplates a revenue-sharing arrangement like the one at issue in Bugielski:
The furnishing of goods, services, or facilities between a plan and a party in interest to the plan generally is prohibited under section 406(a)(1)(C) of ERISA. As a result, a service relationship between a plan and a service provider would constitute a prohibited transaction, because any person providing services to the plan is defined by ERISA to be a “party in interest” to the plan. However, section 408(b)(2) of ERISA exempts certain arrangements between plans and service providers that otherwise would be prohibited transactions under section 406 of ERISA.
Given this explanation, the Ninth Circuit concluded that § 406 must apply broadly.
Branching out a bit, the Ninth Circuit supported its holding with Supreme Court precedent. The Ninth Circuit distinguished Bugielski from Lockheed Corp. v. Spink. Lockheed involved an employer who offered increased pension benefits, payable out of the plan’s surplus assets, to employees retiring early if they agreed to drop employment-related claims against the employer. The Supreme Court held that this arrangement was not a prohibited transaction under § 406(a)(1)(D), which prohibits a fiduciary from causing a plan to engage in a transaction that constitutes a “transfer to, or use by or for the benefit of a party in interest, of any assets of the plan.” The Court said that the commonality of transactions prohibited under § 406(a) is that they “generally involve uses of plan assets that are potentially harmful to the plan.” However, the Ninth Circuit refused to apply this statement to Bugielski’s claim because § 406(a)(1)(C) directly encompasses AT&T’s prohibited transaction, whereas § 406(a)(1)(D) did not “in direct terms” address what an employer can ask an employee to do in return for benefits. Additionally, the court distinguished Lockheed by noting that AT&T’s amended contract could generate millions of dollars for a party in interest—a windfall that Lockheed’s arrangement did not involve because the employer was compensating employees. As for other precedent, the Ninth Circuit rejected the Third and Seventh Circuits’ holdings regarding § 406(a)(1)(C) because they did not “follow the statutory text.” This paper considers those cases, and those with which they align, next.
The Third Circuit was the first of these four circuits to trim § 406’s scope to non-arm’s-length transactions. In Sweda v. University of Pennsylvania, Jennifer Sweda and other plaintiffs alleged that the University of Pennsylvania entered into prohibited transactions with Vanguard Group (Vanguard) because Vanguard charged recordkeeping fees as a percentage of various options that plan participants could choose. Like the third parties providing similar services in Braden and Bugielski, the Third Circuit noted that § 406’s plain language, combined with the definition of a party in interest, meant that Vanguard was a party in interest providing services to the University of Pennsylvania’s pension plan.
Yet the Third Circuit concluded this could not be right, or else § 406(a)(1)(C) would produce “absurdity.” A broad reading of § 406(a)(1)(C) would prohibit ubiquitous service transactions and expose fiduciaries to liability for “every transaction whereby services are rendered to the plan.” The Third Circuit held that such a result would “miss the balance that Congress struck,” noting that ERISA balances numerous disputes between competing interests. To confirm its understanding, the Third Circuit drew on Lockheed, the case that the Ninth Circuit distinguished in Bugielski. The Third Circuit likened Lockheed to Sweda’s complaint by arguing that the Supreme Court “similarly avoided absurdity” in § 406(a)(1)(D) by holding that an employer’s act of conditioning payment on participants’ performance was not a prohibited transaction.
All of this, the Third Circuit concluded, shows that § 406 really prohibits transactions that present “legitimate risks to participants and beneficiaries” to the “advantage of a party-in-interest.” Given the need to stop such advantageous transactions, the Third Circuit trimmed § 406(a)(1)(C)’s application to cases involving an “intent to benefit a party in interest.” That is what a plaintiff must plead under § 406(a)(1)(C) to avoid an “absurd” reading of that provision.
The Tenth Circuit followed the Third Circuit’s lead in Ramos v. Banner Health. A class of plaintiffs alleged that Banner Health entered into a prohibited transaction with Fidelity to receive recordkeeping services for Banner Health’s 401(k) plan. The class sued Banner Health because its agreement with Fidelity allowed Fidelity to received uncapped, “excessive” fees from revenue-sharing arrangements. The Tenth Circuit’s analysis was brief, focusing mostly on the effects of reading § 406(a)(1)(C) as broadly as its text might allow. The Tenth Circuit limited § 406(a)(1)(C)’s reach to transactions that contain “some prior relationship . . . between the fiduciary and the service provider to make the provider a party in interest.” This is because ERISA, in the Tenth Circuit’s view, “cannot be used to put an end to run-of-the-mill service agreements” from which fiduciaries would retreat if § 406(a)(1)(C) were so litigious. For the same reasons as the Third Circuit, the Tenth Circuit cited Lockheed for permission to read § 406 to avoid “absurd result[s].”
Not too long after Ramos, the Seventh Circuit joined the Third and Tenth Circuits in Albert v. Oshkosh Corp. Like the cases already described, Andrew Albert sued Oshkosh Corporation for entering into a prohibited transaction with Fidelity by paying Fidelity excessive fees for plan administration services. The Seventh Circuit, like the Third and Tenth Circuits, declined to read § 406(a)(1)(C) as broadly as its text might permit. Instead, the Seventh Circuit glossed § 406(a)(1)(C) to say that it prohibits only those transactions that “look[ ] like self-dealing,” as opposed to “routine payments for plan services.” The Seventh Circuit concluded that this reading aligns with ERISA’s purpose: plan participants’ well-being. If § 406(a)(1)(C) prohibited “essential” transactions like recordkeeping, plan participants would be harmed.
Notably, the Seventh Circuit focused only on § 406, not the § 408 defenses, because Oshkosh Corporation did not assert a defense under § 408. Because that provision was not at issue, the Seventh Circuit distinguished Oshkosh from the court’s earlier precedent that did focus on § 408. Specifically, in Allen v. GreatBanc Trust Co., the Seventh Circuit had held that § 408 contains affirmative defenses that a plaintiff need not anticipate during the pleading stage. In Allen, the court found that the transactions at issue, which the defendant tried to defend under § 408, were “indisputably prohibited transactions within the meaning of section 406.” Yet because Oshkosh did not involve a claim under § 408, but rather how broadly § 406’s prohibition should extend, the Seventh Circuit refused to apply Allen. The court also refused to apply Allen because the transactions in Oshkosh were far more routine. If Allen’s holding about § 408’s affirmative defenses could apply to a case involving only § 406, the court in Oshkosh reasoned, a plaintiff could plead a prohibited-transaction claim simply by identifying a party in interest and alleging that it engaged in one of the transactions contained in § 406(a)(1). That would produce an “absurd” result.
In total, the Third, Seventh, and Tenth Circuits, respectively, offer the following restrictions on § 406(a)(1)(C)’s scope: a prohibited transaction occurs if a plaintiff alleges that the fiduciary intended to benefit the party in interest, that the transaction looks like self-dealing, or that the fiduciary and third party have a prior relationship to make the third party a party in interest. Otherwise, the transaction is not a prohibited transaction.
Finally, and most recently, the Second Circuit trimmed § 406(a)(1)(C)’s scope. Cunningham v. Cornell University involved an allegation that Cornell University entered into a prohibited transaction with CapFinancial Partners to receive its recordkeeping services and that, among other things, this arrangement resulted in higher fees than necessary.
To answer this allegation, however, the Second Circuit took a different approach by asking whether the § 408 exemptions of reasonableness and necessity are affirmative defenses. Parting with the Seventh and Eighth Circuits, it held that they are not. As such, for a prohibited transaction claim, a plaintiff must plead that the transaction was “unnecessary or involved unreasonable compensation.” In other words, § 408 contains elements of an offense that must be pleaded. The court began by noting that § 406’s language explicitly incorporates § 408: it begins with the carveout, “[e]xcept as provided in section 1108 of this title . . .” The court noted that exemptions usually are understood as defenses that the defendant must raise. However, some exemptions are so integral to the offense that a complaint must account for them. For this proposition, the Second Circuit relied on the criminal law context, where an exemption should be understood as an element of the prohibited conduct when one cannot refer to the conduct without referring to the exemption. The court applied this presumption to ERISA: § 406’s wording encompasses § 408 such that a plaintiff must articulate what § 406 prohibits by referencing § 408. Besides, the court said, if § 406 were read in isolation, it would prohibit many routine transactions—similar to the reasoning that the Third, Seventh, and Tenth Circuits used to curb § 406’s scope.
The Second Circuit expanded upon its statutory interpretation by noting that § 406(b), not at issue in Cunningham, “on its face is restricted only to transactions carrying indicia of a conflict of interest” and, as such, does not incorporate § 408’s language with a statutory carveout. Given that § 406(a) does contain a statutory carveout referencing § 408, the Second Circuit concluded that this was an intentional difference that raises the pleading standard for § 406(a). Heightening the pleading standard under § 406(a), the Second Circuit noted, does not change the fact that the defendant must ultimately prove that the § 408 exemptions apply. The bottom line, then, is that the burden of pleading and the burden of proof do not follow each other in the Second Circuit for a § 406 claim. There, plaintiffs must allege that the fiduciary entered into a prohibited transaction that was either unreasonable in compensation or unnecessary, and defendant fiduciaries retain the ultimate burden of proving the transaction’s appropriateness.
With the circuits’ positions explained, this paper now offers additional considerations for a solution.
Courts recognize that ERISA is a complex and “carefully crafted” statute. As such, they have tried to avoid altering its enforcement scheme. This is especially because ERISA resolved “innumerable disputes between powerful competing interests—not all in favor of potential plaintiffs.” The Supreme Court in particular has capitalized on this point, declining, for example, to extend remedies not specifically authorized by the text. Indeed, interest balancing animates the circuits’ disagreement here. ERISA’s mechanism for enforcing fiduciary duties is perhaps the most carefully crafted.
Given the careful craftsmanship of fiduciary duties, the logical place to begin is § 406’s text. Recall that § 406 states that a fiduciary cannot engage in a transaction if the fiduciary knows or should know that the transaction involves the “furnishing of goods, services, or facilities between the plan and a party in interest.” This text reveals two features. First, § 406 creates a broad, per se rule against prohibited transactions. Second, nothing in the text limits the prohibition to non-arm’s-length transactions.
For those interested, the legislative history supports this understanding because it requires the fiduciary to investigate whether a party-in-interest relationship exists:
The type of investigation that will be needed to satisfy the test of prudence will depend upon the particular facts and circumstances of the case. In the case of a significant transaction, generally for a fiduciary to be prudent he must make a thorough investigation of the other party’s relationship to the plan to determine if he is a party-in-interest. In the case of a normal and insubstantial day-to-day transaction, it may be sufficient to check the identity of the other party against a roster of parties-in-interest that is periodically updated.
This legislative history suggests that a case-by-case investigation is appropriate regarding whether the transacting party is a party in interest, given how “significant” or “insubstantial” the transaction is. However, it does not suggest that whether § 406 is violated must also be a case-by-case inquiry in terms of how “significant” or “insubstantial” the transaction is. As the Senate Report explained, certain third-party transactions must be permitted, but the “breadth of proscriptions” will “prohibit transactions which are deemed desirable” to the plan’s functioning. This suggests that § 406 presumes a transaction is prohibited, even if it seems sensical. In fact, the Senate removed a provision from the section outlining fiduciary duties that would have expressly permitted a party in interest to provide “multiple services to a plan, regardless of whether the ‘party in interest’ was also serving in a fiduciary capacity and receiving fees or compensation for the performance of discretionary functions with respect to plan funds.” If the legislative history carries any weight, it bends toward prohibiting more transactions, not less.
Courts interpreting ERISA not long after its enactment confirmed that whether § 406 is violated does not depend on the transaction’s significance (as the House Report above discussed) or even whether the transaction creates any harm at all. In Marshall v. Kelly, for example, a corporation’s lending of credit, even indirectly, was enough to create a prohibited transaction. McDougall v. Donovan offers another example. The court found a prohibited transaction where trustees of a pension fund purchased a jet aircraft from a third party, who had purchased the aircraft from union members who were covered by the trust. In other words, a third party’s involvement did not convert the transaction into an arm’s-length transaction, and the transaction was prohibited despite the aircraft coming from members that the plan covered. This expansive reading of § 406 continues today. EBSA’s explanation for amending § 408(b)(2)’s implementing regulation, which the Ninth Circuit cited approvingly in Bugielski, contemplates arm’s-length transactions that § 406(a)(1)(C) generally proscribes.
It makes sense that § 406 should operate broadly. As at least three courts have noted, § 406 “gloss[es]” the duties of loyalty and prudence that ERISA requires of fiduciaries in § 404. If § 406 supplements a general, longstanding duty of prudence, then that duty suggests a fiduciary must investigate whether a party-in-interest relationship exists in a transaction. The legislative history says as much, as do a few courts’ interpretations of § 404 shortly after ERISA’s adoption. In other words, § 406 presumes a third-party transaction is a prohibited one unless fiduciaries satisfy the party-in-interest inquiry or unless § 408 says otherwise.
Central to the circuit disagreement is the interaction between § 406’s proscriptive sweep and § 408’s broad exemption from prohibited transactions. To review, the Ninth Circuit’s conspicuously textualist approach in Bugielski permitted § 406’s breadth to stand, though it did not consider whether § 408 contains affirmative defenses due to Bugielski’s procedural posture (motion for summary judgment). The Seventh and Eighth Circuits, conversely, called § 408 an affirmative defense. In its own camp, the Second Circuit held that § 408 contains elements of a claim (specifically, necessity and reasonableness) that must be pleaded. Finally, the Third, Seventh, and Tenth Circuits took their own approaches by trimming § 406 through variations of an intention test (that is, looking for the fiduciary’s intent to benefit a third party, thus making the transaction a prohibited one).
The question, then, is whether § 408 contains an affirmative defense or elements of a claim that must be pleaded. As a matter of civil procedure, plaintiffs need not anticipate affirmative defenses in their complaint—here, the “absence of exemptions” contained in § 408. But defendants must raise an affirmative defense at the pleading stage if they want to use it. So, resolving this question shapes not only the pleading stage for an alleged § 406 violation but also helps define how § 406 and § 408 interact.
To begin with, no firm test exists for identifying what constitutes an affirmative defense, at least when “both a statute and its legislative history are silent on the question” of the burden of proof. Other courts have identified clues indicating that an exemption is an affirmative defense—for example, when the exemption begins with “unless” or “except.”
Some courts have simply stated that an exemption is an affirmative defense without much elaboration. These courts have concluded as such by noting that the defendant bears the burden of proof for an exemption, which suggests that the plaintiff need not anticipate it. Policy considerations also support this. As the Seventh Circuit has stated, “[a] plaintiff is not required to negate an affirmative defense in his or her complaint . . . for the painfully obvious reason that the defendant will not have pleaded any affirmative defenses until it files its answer or a motion to dismiss.” Requiring otherwise would “remind” the defendant of affirmative defenses available, which frustrates the liberal complaint regime.
In keeping with the absence of a single test, the Supreme Court has elsewhere employed a circumstantial approach to determine what constitutes an affirmative defense. In Meacham v. Knolls Atomic Power Laboratory, the Court held that exemption from liability for a disparate-impact claim under the Age Discrimination in Employment Act of 1967 (ADEA) regarding “reasonable factors other than age” is an affirmative defense. The Supreme Court found it as “no surprise” that the exemption is an affirmative defense. The Court noted that it has long operated on the principle that when a “proviso . . . carves an exception out of the body of a statute or contract, those who set up such exception must prove it.” Courts should apply this convention unless they have “compelling reasons to think that Congress meant to put the burden of persuasion on the other side.”
To confirm this framework, the Court noted that the ADEA’s exemptions explicitly reference what is “otherwise prohibited” in the statute’s section on prohibited conduct. The Court noted that when it tried to read the same language—“otherwise prohibited”—in the neighboring provision of § 623(f)(2) under the ADEA (which also expressly references prohibited conduct) as elements of an offense, rather than as an affirmative defense, Congress overrode its interpretation and reinstated that defenses for “otherwise prohibited” conduct are affirmative defenses under the ADEA.
The Court also rejected the notion that an exemption set apart from the prohibited activity is an “elaboration on an element of liability” that a plaintiff must plead. In a typical ADEA complaint where a plaintiff alleges an adverse employment practice on a factor other than age, that practice is at the very center for liability in the first place, not a “negation of it or a defense to it.” Under the modern pleading standard, where a court takes all the facts as true in a complaint, a court “assume[s]” that an ADEA complaint involves a factor other than age that harmed the plaintiff. Because of this, the defendant must claim the exemption—that the employment practice was reasonable. Requiring a plaintiff to plead the reasonableness exemption would eliminate much of the point of pleading. On this point, the Court asked, why make the plaintiff develop a case for reasonableness when doing so is “good enough to avoid liability?”
The Court recognized that requiring more of defendants could corner them for their business practices, but it concluded that Congress should address that concern. The Court offered some reassurance: the more “plainly reasonable” the employer’s practice is, the less it should have to persuade the factfinder. But where the practice’s reasonableness is more “obscure,” the employer will have more work.
Meacham helps illuminate the prohibited-transactions battle under ERISA for three reasons. First, the ADEA’s structure is similar to ERISA. That is, the ADEA prohibits age discrimination in employment, “with exemptions laid out apart from the prohibitions (and expressly referring to the prohibited conduct as such).” This is just like § 408 in ERISA, which is set apart from § 406 with explicit reference to that section. Second, no material difference exists between what the ADEA’s exculpatory section references (“otherwise prohibited” activity) and what ERISA’s exculpatory language in § 408 references. Section 408(b) is entitled, “Enumeration of transactions exempted from section 1106 prohibitions.” Appealing to a section heading for statutory interpretation is appropriate, as the Supreme Court has confirmed. Such reliance on headings enjoys more support where, as here, the heading was enacted with the statutory text rather than being added during codification. As such, the § 408 exemption expressly references § 406 just as the ADEA exemption expressly references the section on prohibited conduct. Third, avoiding unnecessary pleading under the ADEA applies similarly to prohibited-transaction complaints. Just as it would be “substantially redundant” for an ADEA plaintiff to plead an alternative, reasonable employment practice, so too would it be redundant for an ERISA plaintiff to plead an alternative method for a third-party transaction that would be reasonable and necessary.
In light of these considerations, consider again the Second Circuit’s opinion in Cunningham. The Second Circuit focused on the carveout language in § 406: “[e]xcept as provided in section 1108 of this title . . . .” This emphasis is misplaced with Meacham in mind. In Meacham, the Court focused on the language of the exemption, not the language of the provision detailing the prohibited activity. As part of its justification for emphasizing the carveout, the Second Circuit appealed to the criminal law context. When a criminal law defines an offense but contains an exception, the pleadings must allege conduct that the exemption does not cover. However, the criminal law context, where due process concerns and the risk of liberty deprivation are heightened, is not directly translatable to the civil context, where those concerns are generally lower. Moreover, it is unconvincing to use criminal law concerns about due process to avoid requiring fiduciaries to defend their business decisions.
Additionally, the Third Circuit has rejected the idea that carveout language in a statute’s prohibitory provision (that references the statute’s exemptions) turns the exemptions into elements of the offense to be pleaded. In Evankavitch v. Green Tree Servicing, LLC, the Third Circuit considered an exculpatory provision in the Fair Debt Collection Practices Act (FDCPA). Under the FDCPA, a debt collector is liable to a consumer for contacting third parties in pursuit of that consumer’s debt unless the communication falls under a statutory exemption. Specifically, § 1692c(b) states that “[e]xcept as provided in section 1692b . . . a debt collector may not communicate[ ] in connection with the collection of any debt” with third parties.
Under the Supreme Court’s guidance that no strict test exists to delegate the burden of proof in an ambiguous statute, the Third Circuit employed a five-factor test. First, it looked for statutory exceptions and whether the offense they exempt can be referenced without the exceptions. In the FDCPA, “no reference to the Act’s exceptions is necessary to discern that calls to third parties” to collect a consumer’s debt are prohibited.
Second, the Third Circuit looked at the statutory structure to clarify where the burden of proof should fall. Under the FDCPA, the exceptions to communication with third parties in § 1692b are set apart from the general prohibition in § 1692c(b). Placing the exception and the general prohibition in different parts of the statute “has been recognized by the Supreme Court as indicative of an affirmative defense,” to which the Court cited Meacham.
Third, categorizing an exception as an affirmative defense helps “avoid unfair surprise and undue prejudice” toward the plaintiff. The Third Circuit said this means asking whether a defendant’s failure to raise the exemption would “deprive[ ] [a plaintiff] of an opportunity to rebut that defense or to alter her litigation strategy accordingly.” If a debt collector defends its allegedly prohibited call under the exemption, plaintiffs would alter their litigation strategy compared to if the defendant did not invoke the exemption. Avoiding unfair surprise, the court suggested, means the plaintiff should not have to anticipate the defense during the pleadings.
Fourth, the party with better access to relevant information generally bears the burden of proof. Of course, a plaintiff must adequately plead a violation, but the plaintiff need not plead more facts, including those implicating the exemption, that lie “peculiarly” with the defendant. In Evankavitch, the defendant was the only party with “any realistic ability” to produce exculpatory evidence, and even then the defendant failed to produce sufficient evidence at trial showing that the exemption should apply. If the defendant struggles to produce exculpatory evidence, then the plaintiff will struggle more.
Finally, the Third Circuit considered whether Congress’s concerns about plaintiffs in creating a statute support putting the burden of proof on the defendant. If making the plaintiff carry the burden would frustrate the statute’s remedial purpose, as the Third Circuit found it would regarding the FDCPA, then the defendant bears the burden of proof.
Applying Evankavitch’s five-factor test to ERISA suggests that § 408 is an affirmative defense. First, § 406(a) contains express language creating a carveout, just like the FDCPA: “[e]xcept as provided in section 1108 of this title . . . .” And as in the FDCPA, where parties can reasonably know what is prohibited without referencing the exemption, so too can ERISA parties reasonably know what § 406 prohibits without referencing § 408. That is, parties know that a fiduciary should not enter into a transaction that furnishes goods or services between the plan and a party in interest without an applicable exemption.
Notice, also, that this reasoning works against the Second Circuit’s opinion in Cunningham, which focused on the carveout language in § 406. Although the prohibitory language in § 406 references the exemption in § 408, that does not mean the exemption must be referenced to discern what § 406 generally prohibits, as the Second Circuit reasoned. If this were the case, any general prohibition in a remedial statute that references an exception could turn the exception into an element of the defense that a plaintiff must plead. Construing generally prohibitory statutes this way would frustrate their remedial goals. And, as noted, the same concerns that attend badly written laws in the criminal context do not rise to the same level with how § 406, a civil statute, is written. In fact, the Third Circuit in Evankavitch considered, but rejected, transporting similar concerns about criminal laws into the FDCPA (also a civil statute).
Regarding the second prong of the Third Circuit’s five-point test, the statutory structure of § 406 and § 408 mirrors § 1692c(b) in the FDCPA: exemptions are set apart from the general prohibition. The same structure appears in the ADEA as well.
Third, construing § 408 as an affirmative defense avoids worries about “unfair surprise[s]” for the plaintiff. If fiduciaries defended their transactions with a party in interest based on reasonableness and necessity, plaintiffs would likely need to explore the fiduciary’s knowledge and intent. In that case, plaintiffs would need to change their litigation strategy to prove that a fiduciary was not entering into the transaction for reasonable compensation or for necessary reasons.
Under factor four, fiduciaries have easier access to relevant information that helps determine whether a transaction with a party in interest was reasonable and necessary. This is similar to the Third Circuit’s conclusion in Evankavitch that debt collectors have better access to information for whether their contacts with third parties can be exempted. The reasonable assumptions courts draw in favor of plaintiffs during the pleading stage support this. In Allen, the Seventh Circuit characterized § 408 as an affirmative defenses because “exemptions from prohibited transactions . . . assume that a transaction in the prohibited group occurred, and they add additional facts showing why that particular one is acceptable.”
Relatedly, under the Third Circuit’s fifth factor, ERISA’s carefully crafted structure and remedial nature support reading § 408 as an affirmative defense. ERISA’s regulation of prohibited transactions is evident through § 406’s broad sweep, which guards against the “misuse and mismanagement of plan assets.”
All told, § 408 of ERISA can be fairly characterized as an affirmative defense. Plaintiffs need not anticipate it during the pleading stage.
Since Twombly and Iqbal, the pleading standard has been a plausibility standard. That is, a complaint must “allow[ ] the court to draw the reasonable inference that the defendant is liable for the misconduct alleged.” The pleading standard is not a probability requirement, and a complaint can survive “even if it strikes a savvy judge that actual proof of [the facts alleged] is improbable, and ‘that a recovery is very remote and unlikely.’” Additionally, on a motion to dismiss, courts draw inferences in favor of the nonmoving party (the plaintiff), which Twombly and Iqbal did not change. Where an “obvious alternative explanation” for the defendant’s actions exists, plaintiffs may have to plead additional facts to rule out the alternative.
Moreover, the facts alleged must give the defendant “fair notice of what the . . . claim is and the grounds upon which it rests.” As applied to ERISA, a complaint alleging that a defendant transacted with a party in interest, even if that transaction is exemptible under § 408, lets the defendant know what the accusation is. Such facts would be sufficient to shift the burden of persuasion to the defendant to show that “no more than reasonable compensation [was] paid” for the third party’s services. Importantly, giving the defendant fair notice about an accusation does not require a plaintiff to plead “[s]pecific facts” explaining how the fiduciary’s conduct was unlawful. Because courts draw reasonable inferences from a complaint in favor of plaintiffs, plaintiffs need only plead facts “indirectly showing unlawful behavior.”
On the other side of the pleading standard, defendant fiduciaries must do more than show that their transactions are consistent with legal conduct. A fiduciary could have many reasons to choose funds with higher fees or generally to transact with a third party. Most obviously, those reasons may well involve reasonableness and necessity to qualify under § 408’s safe harbor. In many cases, the fiduciary’s reasons for engaging in an alleged prohibited transaction will likely not present an “obvious alternative explanation” that requires more facts to be pleaded. What makes a transaction with a third party reasonable and necessary will vary by context.
The modern pleading standard does not burden plaintiffs with ruling out all possible reasons—including reasonableness and necessity—for a fiduciary’s third-party transactions. Requiring otherwise not only contravenes the convention to construe a complaint in a plaintiff’s favor: it constructs a “probability requirement” that Twombly and Iqbal rejected. If plaintiffs had to account for reasonableness and necessity (as elements of a cause of action) in their complaints, they would need to allege facts that “tend systematically to be in the sole possession of defendants.” Section 406 does not require plaintiffs to explain fiduciary’s presumptively unlawful actions; § 408 obliges fiduciaries with that task. Principles of trust law also inform this understanding. Section 406 proscribes transactions where a fiduciary “might be inclined to favor [a party in interest] at the expense of the plan’s beneficiaries.” Trust law obliges fiduciaries to justify such self-dealing transactions.
Yet some may worry about “nudg[ing] plaintiffs with marginal cases into court” under this position. Where a plaintiff can allege a third-party transaction, defendants will have to defend every service-provider transaction, the concern goes. However, courts have rejected this concern as a reason to curb ERISA’s remedial nature. In Braden, the Eighth Circuit hardly opined on this issue because it found the burdens of proof and persuasion so clearly marked out between § 406 and § 408. In Meacham, the Supreme Court noted that putting the burden of persuasion on defendants will sometimes affect their business strategies. Nevertheless, the Court concluded that those concerns must go to Congress when the provision is decisively an affirmative defense. In fact, the Court in Meacham defended its holding by noting that Congress responded to the Court by making a provision in the ADEA an affirmative defense after the Court held otherwise. The corollary, as applied to ERISA, is that if § 408 contains an affirmative defense, Congress may step in and say that it does not. It may, for example, declare, as the Second Circuit did, that § 408 contains elements of an offense that must be pleaded. But as the case law stands, § 408 is most naturally read as an affirmative defense along the same lines of reasoning used in Meacham and the Third Circuit’s decision in Evankavitch.
More specifically, within the ERISA context, the Supreme Court has refused to curb ERISA’s remedial provisions based on concerns about frivolous litigation. In Fifth Third Bancorp v. Dudenhoeffer, the Court dismissed such concerns when it held that ESOP (employee stock ownership plan) fiduciaries have the same duty of prudence as non-ESOP fiduciaries under § 404. Fifth Third worried that if it were not entitled to a “presumption of prudence,” the threat of lawsuits would deter it, and other companies, from offering ESOPs to their employees. But the Court unanimously dismissed this concern. It acknowledged that ERISA carefully balances plaintiffs’ and defendants’ interests. Moreover, the Court held as such despite recognizing that fiduciaries are often stuck “between a rock and a hard place” when making investment decisions (for example, making such decisions in a volatile stock market). But concerns about frivolous lawsuits by themselves are not a valid factor to curb remedial statutes. Instead, weeding out meritless claims is better accomplished through “careful, context-sensitive scrutiny of a complaint’s allegations.”
Indeed, several effective tools exist to weed out meritless claims. For one, the pleading standard for a motion to dismiss must be context specific because a complaint involving the duties of loyalty and prudence, which underlie § 406, depend on “the circumstances then prevailing” when the fiduciary acts. District courts can also order the plaintiff to reply to the defendant’s or third party’s answer to the complaint. Moreover, a district court can grant the defendant’s motion for a more definite statement if a plaintiff presents “vague or ambiguous” allegations. Federal Rule of Civil Procedure 11 imposes sanctions for fraudulent claims. And summary judgment is the greatest guard against trial.
Finally, not all plaintiffs have an incentive to sue a fiduciary. As long as a plaintiff knows that § 408 exists, the plaintiff would have to believe that § 408 does not exempt the transaction. The plaintiff would also have to identify some harm the transaction caused, which would already likely be difficult in the pre-discovery phase. If a plan operates effectively and gives plaintiffs no reason to suspect something is afoul, then a plaintiff likely will not, and likely could not, sue.
This paper has explored a live disagreement regarding what a plaintiff should have to plead for a prohibited-transaction claim under § 406 of ERISA. The Eighth and Ninth Circuits have held that a plaintiff need only allege that such a transaction occurred, while the Second Circuit held that § 408’s exemptions of reasonableness and necessity constitute elements of a claim that must be pleaded. Still others—the Third, Seventh, and Tenth Circuits—have devised variations of an intent test to curb § 406’s broad scope. However, this paper argues that § 408’s exemptions constitute an affirmative defense that a plaintiff does not have to anticipate during the pleading stage. At bottom, ERISA is a “remedial statute to be liberally construed in favor of employee benefit fund participants.” This paper’s position is consistent with that long-held understanding.
