The U.S. Supreme Court ruled unanimously on May 22, 2026, in M&K Employee Solutions v. Trustees of the IAM National Pension Fund that the Employee Retirement Income Security Act of 1974 does not require actuaries to use assumptions fixed on the measurement date when calculating a departing employer's withdrawal liability under a multiemployer pension plan. Justice Ketanji Brown Jackson authored the opinion.

Under ERISA, a multiemployer pension plan is an arrangement in which a group of employers in a particular industry collectively agree to provide defined benefits to covered employees. When an employer leaves such a plan, it must make a withdrawal liability payment equal to its share of unfunded benefits based on an actuary's calculation as of the measurement date, defined as the last day of the year before the employer withdraws.

The actuary's calculation necessarily occurs after the employer withdraws, creating a gap between the measurement date and the calculation date. Actuarial assumptions, such as the discount rate of interest, are predictive judgments rather than factual inputs and therefore may be selected after the measurement date.

The practical stakes of that gap were on display in the M&K case: the departing employer owed more than three times as much under the interest rate current on the calculation date as it would have owed under the rate set on the prior measurement date. The difference illustrated how the timing of assumption selection can substantially affect the amount a withdrawing employer owes.

The Court noted that ERISA requires actuarial assumptions to be reasonable, take into account plan experience and expectations, and represent the actuary's best estimate of anticipated future experience but does not specify when those assumptions must be selected. Congress specified the relevant assumptions for other ERISA calculations but did not do so for the withdrawal liability calculation under section 1381 of ERISA. The Court found that silence to be consistent with allowing actuaries to select assumptions after the measurement date.