Peter Gulia

Even if relevant law does not require plan provisions about insurer refunds, might we prefer plan document provisions so that deciding what to do with dividends, demutualization proceeds, medical-loss-ratio rebates, and similar payments need not be a discretionary decision? If so, can a plan provide that all such payments belong only to the employer?

shERPA's observation reminds us of one of the soft spots of the rules for a hardship (or unforeseeable-emergency) distribution - recognizing that the fungible nature of money can interfere with distinguishing between allowed and unallowed purposes and how they affect a participant's claim to use her retirement account. A plan's administrator sees the claim before it, but ordinarily lacks information on the complete picture of the participant's uses of money. A participant who submits an unpaid bill for something that the plan recognizes as a hardship might have used up money that could have gone toward that unpaid bill on paying for something that the plan does not recognize as necessary. But it's the administrator's job to apply the plan's terms to the claim that is submitted.

Here's what 403(b) Answer Book says: Q 6:30 How may a state retirement system fund be a Section 403(b) investment? An investment under a state teachers' retirement system generally is not a permitted Section 403(b) investment. Likewise, an investment under a separately funded employee retirement reserve governed by the state insurance department's supervision generally is not a permitted Section 403(b) investment. [Rev. Rul. 82-102, 1982-1 C.B. 62, revoking Rev. Rul. 67-387, 1967-2 C.B. 153; Rev. Rul. 67-361, 1967-2 C.B. 153] But a contract under a plan that was established on or before May 17, 1982 in good-faith reliance on either of the revoked rulings may continue to cover those participants covered on May 17, 1982 if the plan and the investment meet all conditions required by current Treasury regulations. If either transition rule applies, the contract may continue to cover those participants covered on May 17, 1982, including “an employee who becomes covered for the first time under the plan after May 17, 1982.” [Treas. Reg. § 1.403(b)-8©(3)] As Tom Geer notes, a question about what investment is or isn't permitted is separate from a question about which person may administer a 403(b) plan. As Matt Bozek observes, section 1.403b-3(b)(3)(ii) is mostly about restraining a plan's documents from allocating to each participant alone responsibility to administer the plan. Any non-natural person that considers whether to accept an appointment to serve as a plan's administrator should get its lawyer's advice about how non-tax law might affect the service. To return to what I suspect was joel's original inquiry, the Internal Revenue Code does not preclude a State government agency from arranging a 403(b) plan for those State and local government employees who are public-schools employees (if the arrangement is proper under State law). If the transition rule described above does not apply, the arranger would limit the investments to annuity contracts and custodial accounts that hold regulated investment company shares.

Given the difficulties and frustrations that several posters have described, another step that a TPA might consider to protect itself is to make sure that its service agreement with the plan or its administrator does not obligate the TPA to furnish anything beyond what the TPA has received from the insurers and investment funds. It's not a wholly satisfactory solution. (We recognize that, no matter how clearly and loudly a TPA warns an employer and plan administrator about what services are not included, it's impossible to control what a plan's administrator wants.) But when a client or customer is upset about what didn't get done, it's still better to start that conversation without worrying about the TPA's contract responsibility or other liability.

On #2, the rule puts the duty on the plan's administrator. (That was a change that commenters on the proposed rule requested.) Paragraph © recognize that a plan's administrator may designate a person to act on its behalf in furnishing the information. However, if such an agent doesn't do what's required, the plan's administrator remains responsible.

I consider unwise some of the choices that EBSA made about which pieces of information to present (for all investment alternatives). And the rule's analysis and writing quality isn't 'A' work in very respect. But the reg writers did consider differences between SEC-registered and unregistered investments. For a few of these points, the rule includes specific provisions, and others show clear recognition of differences for unregistered funds. The rule's preamble explains the Labor department's reasoning for seeking symmetry between registered and unregistered investments. Moreover, the ERISA rule doesn't always give in to the securities-law way of doing things, sometimes requiring (for all investment alternatives, including registered investments) that information be presented differently than what securities law calls for. It might be work for trustees who are not in the business of being trustees to pull together the information to be disclosed. But if a plan asks a participant to choose between a trustees-managed fund and other funds, shouldn't a participant get the same base information on every investment alternative so that one could make an informed choice?

On #1, while I don't give advice, let me suggest a way that you and your lawyer might think about it. If the menu of mutual funds has 17 funds, the plan you've described has 18 investment alternatives. The one investment alternative that is not a registered investment company happens to have the plan trustees as its manager. The "pooled account" seems to be within the rule's definition: "Designated investment alternative means any investment alternative designated by the plan into which participants and beneficiaries may direct the investment of assets held in, or contributed to, their individual accounts."

alisonr, it's not clear whether you advise the claimant or the plan's administrator, but either way you might consider giving advice about the plan's claims procedure and the effects of it. For the claimant, working with the claims procedure might be an opportunity to find out what evidence would satisfy the plan's administrator. Or if the administrator's decision doesn't change, exhausting the plan's claims procedure (including all reviews and appeals) might be a condition that must be met before a court would consider a claim for the benefit. For the administrator, using the claims procedure might be an opportunity to double-check the soundness of the decision. And if an administrator has a correctly designed claims procedure and has used it properly and fairly, courts usually defer to the administrator's decision. On the underlying question, what explanation did the decision-maker furnish for why it believes that the evidence the claimant furnished is insufficient to prove the marriage?

By the way, I've been successful in making a bank that received a payment eat the fraud loss if the bank lacked good evidence that the receiving account belonged to the plan's payee, or lacked good evidence that the purported account holder is the same person as the one who opened the account.

How does the plan's administrator know that it was not the participant who changed the banking information? Is there any independent evidence to corroborate the participant's description of what happened? I imagine that the recordkeeper's computer knows that the participant's identity credentials were used, but doesn't know which human caused those digits to become entered into a computer. Or is there more to the mechanics than I know about?

For withdrawal-liability purposes, courts have interpreted ERISA section 4201's (29 U.S.C. 1381) use of the word “employer” to extend far beyond the particular business organization that had an obligation to contribute to a plan. In addition to looking to common-control and affiliated-service-group concepts, courts interpret the word “employer” considering alter-ego, veil-piercing, fraudulent-transfer, and similar theories. And as you already knew, ERISA’s withdrawal-liability part expressly provides: “If a principal purpose of any transaction is to evade or avoid liability under this part [ERISA §§ 4201-4225], this part shall be applied (and liability shall be determined and collected) without regard to such transaction.” ERISA § 4212©, 29 U.S.C. § 1392© (emphasis added). Keep in mind that a business or other person asserted to be part of the employer has to fight the battle with some extra disabilities: ERISA/MPPAA's pay-first, dispute-later rule; arbitration before an arbitrator who's not required to explain his or her decision with a reasoned written opinion; and the reluctance of a Federal court to consider a case until after arbitration is exhausted. That said, there are some opportunities for planning an acquisition to eliminate or reduce the risks. I don't want to discuss them in bulletin-board format, but jpod you're welcome to call me.

Depending on the provisions of earlier documents, it's possible that a pension plan amendment might be another settlor act. See generally Lockheed Corp. v. Spink, 517 U.S. 882 (1996); Curtiss-Wright Corp. v. Schoonejongen, 514 U.S. 73 (1995); Walling v. Brady, 125 F.3d 114 (3rd Cir. 1997); Pope v. Central States, Southeast and Southwest Areas Health and Welfare Fund, 27 F.3d 211 (6th Cir. 1994); Hartline v. Sheetmetal Workers' National Pension Fund, 134 F. Supp. 2d 1 (D.D.C. 2000). However, one wonders what grounds or mode of analysis trustees could have for saying that a plan amendment was a "business" decision. Who or what authorized a trustee to act as a principal or agent to make a business decision about a worker's compensation or an employer's obligation? But if the trustees act within a current plan provision that grants the trustees discretion concerning whether to admit or refuse a participation agreement, exercising that discretion would seem somewhat likely to be a fiduciary function.

Mr. Ecklund, I'd be interested in your thoughts about a follow-on question: Imagine that - before the union's decertification and before the current collective-bargaining agreement expires - the employer proposes a participation agreement that would obligate the employer to pay the multiemployer pension plan enough contributions on enough employees and shifts so that a withdrawal would not be triggered. In evaluating whether to accept or decline the proposed participation agreement, do the plan's trustees have a fiduciary duty to evaluate whether the contribution obligation or the withdrawal liability is in the plan's best interests? If so, what facts and circumstances should the trustees consider to evaluate which right is better? Is it as simple as taking whichever right seems likelier to reduce (or at least not worsen) the plan's under-funding? If triggering the withdrawal would result in the bigger, longer, or more certain right for the plan, may the trustees nonetheless choose the employer's continuing participation if the trustees find that not keeping the employer in the plan would lead to an implosion and mass withdrawal?

austin3515, let's accept your premise that many independent qualified public accountants do a good job; but let's seek to learn a little more. The weakness that I sometimes worry about with a limited-scope audit is that the readers of the CPA's report don't get the CPA's opinion that the plan's financial statements are fairly stated, and so don't get the implied assurance that the plan's true assets are not materially MORE than what is shown. What audit procedures does a CPA perform to consider whether the plan might own assets beyond those that the trustee and plan administrator reported? If the CPA is engaged to do a full-scope audit, would those 'completeness' procedures be more (or more likely to detect what's missing or inconsistent)? Even if those 'completeness' audit procedures for a limited-scope audit are no less than those audit procedures for a full-scope audit, does the plan that receives a limited-scope report get less practical "insurance" because it's more difficult to state the malpractice claim if no opinion was delivered?

Even if both a right and a duty to vote is "passed through" to each participant, beneficiary, and alternate payee, the plan's trustee or other fiduciary might need to vote shares attributable to a participant, beneficiary, or alternate payee who breaches his or her duty to vote. About the need for a special-purpose independent fiduciary, the regularly-serving trustee should ask herself: 'How would I prove that my analysis and decisions were unaffected by my interests and were taken solely to achieve the retirement plan's exclusive purpose?'

The plan’s administrator might consider a possibility that the examiner has a mistaken view simply because the examiner lacks enough knowledge to apply the rule. One possible way to counteract a mistaken view is to get an employer-benefits lawyer’s carefully written explanation of the right interpretation and application of the rule. (Many EBSA examiners don’t get enough guidance from their supervisors and national office, and a better examiner is open to education if the memo is succinct, clear, and fair-minded.) Although a fiduciary might be reluctant to spend the plan’s money on legal advice to educate an EBSA examiner, doing so might be less expensive than “giving in” to a wayward examiner.

If an insurance contract provides only incidental life insurance protection and was issued before September 24, 2007, it can be treated as a contract held under a 403(b) plan. 26 C.F.R. §§ 1.403(b)-6(g), 1.403(b)-11(f). If it is so treated, the contract might continue to receive contributions.

Thanks, Matt; that idea simply hadn't occurred to me. So let me make the hypo a little harder. Assume that there is no explanation for the service provider's failure to furnish information (and that the responsible plan fiduciary is unaware of any circumstances that might explain why the service provider could not furnish information). Could it ever be imprudent to fire a non-responsive service provider?

The ERISA 408(b)(2) rule says that if a service provider has failed to furnish the required information, despite the responsibile plan fiduciary's follow-up efforts, and the missing information relates to future services, "the responsible plan fiduciary SHALL terminate the contract or arrangement as expeditiously as possible, consistent with [the fiduciary's] duty of prudence." A service provider that can't or won't in more than four months' time answer this information request shows by its conduct at least that it doesn't care about following relevant law, and perhaps that it lacks competence or capability for the service that it's supposed to perform. In what circumstances would it somehow not be prudent for the plan fiduciary to get rid of such a weak service provider?

Without staking out any particular view, consider that it’s not entirely right to say that the plan can never have a loss on a participant loan. Although the typical loan terms include repayment through wage deductions, it’s possible for a borrower’s employment to end and for the borrower to be unable or unwilling to repay the loan. A typical individual-account plan allocates both sets of cash flows and accruals on a participant loan to the account of the participant who borrowed. When the repayments the account has collected are less than the loan’s full loan principal and interest, the participant’s account has at least an opportunity loss on that investment. When the repayments the account has collected are less than the loan’s principal, the participant’s account has a negative return on that investment. Consider also that a typical participant loan isn’t meaningfully secured. In theory, a plan could collect on an unrepaid loan by redeeming the participant’s other investments. (And keep in mind that those investments might have declined to a current value less than the outstanding loan receivable.) In practice, a plan’s administrator might decide (or be directed) to leave the “real” investments as-is and treat the defaulted loan as having a cash value of zero. A carefully written plan could provide that a participant directs the plan’s fiduciaries on how to handle his or her defaulted loan. See ERISA § 404©(1)(B). (The practices might be different for a loan made by an insurance company, rather than from direct plan assets.) Moreover, with a participant loan, only 50% (or less) of the account is real security for the outstanding loan. With a mortgage loan, 100% of the real property is security for the loan. ERISA’s rule for what is a reasonable rate of interest – “... commensurate with the interest rates charged by persons in the business of lending money for loans [that] would be made under similar circumstances” – is nonsense. There is no truly similar loan that a commercial lender makes. One might analogize some aspects of a participant loan to a mortgage loan, in which the lender bears some risk that the value of its security interest could have a value lower than the outstanding loan receivable. But what prudent banker should make a loan for which the borrower can, by his or her unwise investment decisions, seriously devalue the lender’s security interest? (There’s a reason why a mortgage lender requires its borrower to maintain fire and homeowner’s insurance.) Besides, what marketplace loan do you know about in which the lender and the borrower are the same person? Despite those and other inherent difficulties of trying to state a hypothetical fair-market-value for something that doesn’t have a market, one can view both laws’ reasonable-interest conditions as an attempt to provide for the loan a potential income meant to approximate what might have been the account’s investment return had the loan not been allowed and the account remained invested in assets other than a participant loan. Because both government agencies’ efforts necessarily deal with hypotheticals, intelligent minds could find a wide range of interpretations. But let’s return to the main purpose: the fiduciary who decides the participant loan interest rate must administer the plan “for the exclusive purpose of providing [the retirement plan’s] benefits to participants and their beneficiaries[.]” Why not provide the loan with an interest rate that helps the participant restore his or her retirement savings?

Even when it's not required, a plan's administrator might prefer to use its claims procedure. With a troublesome inquirer, running the claims procedure is an efficient way to get the inquiries to end. Telling an inquirer 'you're welcome to put in writing your request, your explanation of whatever you believe you should get, and whatever evidence and other information you would like the plan's admministrator to consider' can be a powerful way to say (politely) 'put up or shut up' or 'what happened wasn't the fault of the plan's people'. One reason that it can be so effective is that the procedure might in fact be sufficiently open-minded and fair. In my experience (advising on about 100,000 beneficiary disputes), the inquirer/claimants recognize that the plan has offered a fair procedure and that the denial decisions were fair decisions on the evidence submitted. The procedure also allows an inquirer a moment to reflect on his or her frustruations, which often are really about the death and what the partificipant failed to do. Many claimants go away moderately content, at least feeling that they were heard. Running that kind of claims procedure also gets some legal advantages. If a plan's administrator ran its proper claims procedure, there are significant protections if there is a court proceeding later. A court should require a plaintiff to make his or her case within the plan's administrative record (without any evidence not submitted under the claims procedure). And a court defers to a plan administrator's finding unless it is so obviously wrong that it had to be an abuse of discretion. A participant's quasi-spouse, child, or another person who feels that the participant should have made a beneficiary designation might feel frustrated or even betrayed. Even the most secure customer service representative reacts to an emotional caller. A few dollars and a modest effort put into a procedure can help make an unpleasant situation at least manageable.

Although I often might give different advice based on the particular plan administrator's circumstances, here's an approach that some administrators use for an executor's inquiry. Don't respond to a query simply because the inquirer presents himself or herself as a personal representative. First, require an original or court-certified copy of the order that made the appointment. Second, require the inquirer to give you a written assurance (under penalties of perjury and stating awareness of the criminal punishments for the Federal crime of making a false statement to an employee-benefit plan) that he or she needs the information to prepare an estate tax return that he or she is required to file. Next, read the court order you obtained and consider relevant Federal and State law about the appointee's duties to consider whether his or her statement about the tax-information need is fitting. (Whether through ignorance or deceit, an inquirer often asserts powers and duties that the inquirer doesn't have.) If there truly is a tax-return need, a plan's administrator might furnish information about the account balance, but might not reveal information about the identity of any beneficiary. To appease a personal representative who is worried about his or her duty to collect assets of the decedent's estate, a plan administrator might inform such an inquirer: "The estate is not the beneficiary." To anyone who gets belligerent, a response can be 'you're welcome to file a claim', and the administrator will react to it under the plan's claims procedure. Consider too that an ERISA-governed plan often might legitimately decline to furnish information to anyone other than the beneficiary. And for those who work as a recordkeeper or TPA, rather than as the named plan administrator, consider being careful to do no more than what the plan administrator has instructed you to do (whether under a particular or standing instruction).

QDROphile, jpod, and KJohnson, thank you for the good help. After reading the cases you mentioned (and many others), I'm ready to advise my pro bono client that the uncertainty can't be cured - in either direction. (Even obtaining an ERISA Advisory Opinion wouldn't be a conclusion, because a court need not follow it.)

In ERISA Adivosry Opinion 96-12A, the Labor department advised that a cafeteria plan was not an ERISA-governed plan. The key to the Opinion's analysis is that the cafeteria plan itself does not provide any welfare benefit. The paragraph below quotes the Opinion's analysis but replaces each reference to "Group Health Plan" with "individual health insurance contracts". "[T]he function of the Pre-Tax Plan is to provide a method by which employees may receive tax-favored treatment of contributions that are in any case required under [each individual health insurance contract]. The provision of this tax-favored treatment, however, is not the equivalent of the provision of a benefit enumerated under section 3(1), and it does not appear that the Pre-Tax Plan itself provides any enumerated benefit. It is therefore the position of the Department that the Pre-Tax Plan, as currently structured, does not constitute, in itself, a separate employee welfare benefit plan within the meaning of section 3(1)." I recognize that this is selective quotation because the Labor department also said that the arrangement for making contributions tax-favored was somehow "part of" the group health plan. Even if that "part of" idea made some sense concerning the Topco Associates, Inc. Group Health Plan mentioned in the Opinion, it's hard to see how an employer's payment of an amount to its employee is "part of" an individual health insurance contract - a contract that by its law and terms can have only the insurer and the individual as parties. BenefitsLink mavens, do you think that there is a good argument that: (1) Only each individual insurance contract provides a health benefit of a kind described in ERISA's definition of a welfare plan. (2) The employer's willingness to reimburse its employee's substantiated premium payment provides only money and tax treatment, not any health benefit. (3) Each individual insurance contract is not established or maintained by an employer. Rather, each insurance contract is established and maintained by its two parties - the insurer and the individual, neither of which is the individual's employer. Or what should make me nervous about this argument?

A small employer, in seeking to facilitate health insurance coverage for some of its employees, found that pricing for individual health insurance contracts is more favorable than any group (or association-sponsored) contract. The amount that the employer is willing to pay toward some employees’ health coverage is enough to reimburse 100% of the premium under the insurance contract that each employee selected. This would not involve a § 125 election, because there is no choice to get cash wages or any other taxable benefit. Assume that those entitled to reimbursement, as a subset of the employer’s employees, would meet IRC §§ 105&106 nondiscrimination rules. If the ONLY thing that an employer does is reimburse a substantiated health insurance premium [Revenue Ruling 61-46, 1961-2 C.B. 25] and the employer carefully avoids any ‘involvement’ regarding the individual insurance contracts, does such an employer “maintain” a group health plan? If so, is that ‘plan’ governed by ERISA? If there is an ERISA-governed plan, is it also governed by HIPAA? (COBRA is a non-issue because there are only a few employees.) Being able to arrange this so that it’s not a plan (while still getting ‘health’ treatment for Federal income and FICA taxes) really matters because the employer won’t do anything that requires it to administer a plan.