An insurance company faced with continually increasing retiree medical costs recognized that the upward trend in healthcare costs would continue for the foreseeable future. The company was aware that other employers addressed this problem by severely curtailing or eliminating retiree medical benefits, but this option did not coincide with the company’s benefits philosophy—it wanted to preserve a benefit that its workforce highly valued. The company approached Milliman to recommend strategies to reduce current retiree medical costs as well as contain the volatility of future cost increases.
The Milliman team proposed two alternative methodologies. Each alternative specified the amount the company contributed to the plan. The retirees would pay the difference between the company contribution and the total cost of the plan. To illustrate the relative value of each alternative, the team developed a simulation model that quantified the effect of various contribution formulas on company and retiree cost sharing.
Option #1: fixed medical accounts
The first alternative was a fixed medical account from which retiree medical costs would be drawn. The amount for each recipient would be determined at retirement based on the retiree's age, length of service, and Medicare eligibility. This notional (non-funded) lump sum account would represent the company’s total contribution for the retiree’s lifetime. The account would be reduced each year by the cost of the plan. When the account balance reached zero, retirees would have to pay the annual cost of the plan themselves.
While this option fixed the employer's costs, the potential existed for the retiree to exhaust the funds prior to death, an option that did not fit into the company's benefits philosophy.
Option #2: retiree medical accounts
The second alternative the Milliman team proposed used a retiree medical account (RMA) approach that specified a dollar amount the company would contribute to each retiree's medical costs each year, with the amount determined upon an employee's retirement.
The formula reflected years of service and Medicare eligibility. Retirees with spouses would be allocated a larger contribution. For example, the formula could provide an annual RMA amount of $180 per year of service while a single retiree is under age 65, and $60 per year of service for years when the retiree would be eligible for Medicare. For married retirees, the amounts would be increased by 50%. The contribution amount could be increased each year to reflect inflation if conditions warrant such a change.
The RMA amount would be used to pay a portion of the retiree medical costs each year through the retiree's lifetime; retirees would pay the difference between the RMA amount and the medical plan's total cost. Under this scenario, a retiree with 30 years of service would receive an annual RMA amount of $5,400 at ages less than 65, and $1,800 beginning at age 65. If the total cost of the plan for retirees under 65 was $6,000, the RMA would pay $5,400 and the retiree would pay $600 out of pocket.
Using the modeling tool, the RMA methodology was compared to the company's current cost-sharing formula and adjusted to preserve certain company contribution philosophies. The model also illustrated the effect of the RMA program on the company's retiree medical costs under various grandfathering scenarios.
Success on all fronts
The second alternative met all of the company's objectives. The contribution formula the Milliman team devised removed the link between company contributions and retiree medical costs, rewarded long service, and considered Medicare eligibility—all while limiting future contribution increases.
Convinced that it could afford to continue the retiree medical benefits highly valued by its workforce, the client instituted the RMA plan. The firm’s annual contributions to RMA accounts are presently scheduled to increase 5% per year. The company anticipates that its annual retiree medical cost will be reduced by more than 50%, and that it has successfully limited its exposure to future medical inflation.

