Original article can be found here.

by The Center Square

On Oct. 19, the Supreme Court asked for responses to petitions from FedEx, Southern Company Services and the committee that runs a Kellanova pension plan. Retirees revived their lawsuits against all three in the lower courts, and the cases ask whether the math behind a spouse’s survivor benefit has to reflect how long people actually live. The Court hasn’t agreed to hear them yet. It should, and it should say that federal pension law requires assumptions a reasonable actuary would use.

I come to this as a small business owner with 30 years in institutional investment management in addition to expert-witness work in fiduciary cases. Both taught me that a number is only as good as the assumptions under it.

What the Fight Is About

A retiree with a traditional pension can take a single-life benefit that ends at death or a smaller monthly check that keeps paying the spouse afterward. Congress required that smaller check to be “the actuarial equivalent of a single annuity for the life of the participant.” The Sixth Circuit gave a simple illustration: a $1,000 single-life pension converts to a $900 joint pension, with $450 going to the survivor (Reichert v. Kellogg Co.). Two inputs get a plan from $1,000 to $900, an interest rate and a mortality table. Change either one and the check size changes.

The Tables in Question

The Kellanova plans use the UP-84 table, which the Sixth Circuit described as “based on mortality data from the 1960s and 1970s.” FedEx’s plan uses it as well, alongside a 1971 table built on 1960s data. In Southern Company’s plan, according to the Eleventh Circuit, a 1951 mortality table for men and a 5% interest rate fed the plaintiff’s survivor-benefit math. None of these tables has heard of the twenty-first century.

Meanwhile, people live longer. The Social Security Trustees put remaining life expectancy at 65 at 13.1 years for men and 17.1 for women in 1970, versus an estimated 18.3 and 20.9 in 2024 (Trustees Report, Table V.A4). Husbands gained more than five years and wives nearly four. A survivor benefit exists to cover the years after one spouse is gone, so those years matter.

What the Words Mean

The employers answer that ERISA names no table for these conversions, so whatever the plan document specifies sets the equivalent. Congress did prescribe an applicable mortality table for other present-value calculations, such as lump sums, and left it out here. Judge Nalbandian pressed that point in dissent at the Sixth Circuit, arguing that the majority added a requirement Congress never wrote. As a conservative, I take the argument seriously. Courts shouldn’t write statutes.

Still, “actuarial equivalent” is a term of art, and the Sixth Circuit read it “as it was understood by actuarial scientists when ERISA was enacted.” Treasury’s regulation says equivalence may be determined “on the basis of consistently applied reasonable actuarial factors.” The Eleventh Circuit rejected the employers’ “anything goes (as long as you wrote it down)” reading and held that plans “must base that calculation on the kind of actuarial assumptions a reasonable actuary would use.”

Cicero tied justice to promise-keeping: “The foundation of justice, moreover, is good faith; that is, truth and fidelity to promises and agreements” (De Officiis, Book I, section 23). A survivor benefit is a promise to a spouse. ERISA already asks the people who run plan money to show “the care, skill, prudence, and diligence” of a prudent man familiar with such matters. Committees that scrutinize every line of the investment menu can scrutinize the tables behind the survivor check.

The Employers’ Case

The cost argument deserves a hearing. The ERIC trade group, which backs the employers’ petitions, says the Sixth Circuit’s decision “gives employers no clear way to comply and, left in place, puts more than 20,000 defined benefit plans, with roughly $4.3 trillion in liabilities, at risk of ‘battle of experts’ litigation.” Those are ERIC’s figures, and its announcement doesn’t say how they were measured. ERIC’s Doug Hinson adds that employers who adjust their assumptions “will necessarily violate Congress’s rule that benefits can’t go down, because what’s good for some participants isn’t always good for all.” That argument turns on ERISA’s anti-cutback rule, and it needs an answer. A Missouri federal court sided with the employers in March. In Landel v. Olin Corp., it wrote that it “does not interpret 29 U.S.C. §§ 1055(d) and (e) to contain specific substantive standards beyond its requirement of ‘actuarial equivalence,’” and it dismissed the suit with prejudice. Plans in the Sixth Circuit and in Missouri now operate under different readings of the same sentence, and uncertainty costs money.

The ruling is narrower than that alarm suggests. The Sixth Circuit didn’t order a particular table or award damages, and it didn’t decide that any plan’s assumptions fail. It let the retirees try to prove their case and left plan actuaries room within professional standards. A newer table can move a survivor check up or down, depending on both spouses’ ages, the survivor percentage and the interest rate. Whether these plans shortchange anyone is a question for discovery, and I’d rather see it answered with numbers than adjectives.

A Fix Both Sides Can Live With

The Court should take these cases and settle the standard, but Washington can supply what courts can’t, which is a bright line. The IRS already updates its mortality tables for defined benefit plans, most recently in Notice 2026-27 for 2027. Treasury, or Congress if it wants to own the decision, should give plans a safe harbor for survivor conversions that use those tables or another table of comparable currency, with a reasonable transition period. It should also say how plans can update without cutting anyone’s benefit, for example by paying the greater of the old and new conversion for benefits already earned, which answers ERIC’s anti-cutback concern. Plans that stay with older tables should show their math. Sponsors get a clear rule, and retirees get a check calculated for the century they’re living in.

A pension is a promise priced in advance. Keep the promise, price it with current arithmetic and show the math.

Original article