These ceilings not only protect companies from rising health care costs but also provide them with a perverse incentive: A company that has hit its spending cap has little incentive to negotiate the lowest possible prices with medical providers. In fact, it has an incentive not to: Rising expenses not only won’t hurt the company but will tend to drive more retirees from the program.
Meanwhile, retirees with preexisting conditions and serious health problems remain in the health plan, driving up costs further. Robert Eggleston, an IBM retiree in Lake Dallas, Texas, had brain cancer, so even when his monthly costs for retiree coverage exceeded his pension, he remained in the plan, which also covered his wife, LaRue. To get by, he cashed out his 401(k) account and took out a second mortgage on his home. Luckily, he was eligible for free supplies of a tumor-fighting drug through a program for low-income families, but this was an unexpected end for a career IBM employee.
Some employers hastened the death spiral by segregating retirees into their own risk group rather than keeping active employees and retirees in the same “risk pool” and spreading costs among a wider group of people, as had been common practice in the past. When retirees are segregated into their own pool, the per capita costs rise, because an older, sicker population needs more medical care.
After Xerox split its active and retired employees into two pools in 2003, its retirees under sixty-five began paying 50 to 60 percent more than employees their age. Eugene Nathenson, a retired controller of Xerox Financial Services, saw his premiums shoot up from $1,645 in 2003 to $3,196 the following year, while the deductibles he and his wife paid pushed their out-of-pocket costs beyond $6,000 a year.[10]
REPLENISHING THE COOKIE JAR
By the late 1990s and early 2000s, many companies had used up the actuarial gains they’d stockpiled after the early rounds of cost shifting and benefits cuts, just as they had spent the surplus in their pension piggy banks. So they turned to a tried-and-true solution: Cut more benefits.
Companies continued to cut retiree health benefits throughout the 2000s. International Paper cut benefits in 2000, 2001, and 2002, telling retirees that health care costs were spiraling. What had actually happened was that it had used up the pool of accounting gains generated after the company recorded a huge liability in 1991, and then set a cap on benefits. Unlike McDonnell Douglas, which took all the gains at once, International Paper trickled the gains into income in subsequent years, to the tune of $17 million a year. After this stockpile was used up, cutting the benefits generated a fresh pool of accounting gains that added a total of $65 million to its income by 2004. In 2004, the company then closed the plan to salaried employees whose age plus years of employment with the company totaled less than sixty as of January 1, 2004. Another cut in 2009 whittled another $40 million in obligations.
IBM’s cookie jar of earnings enhancements was also running low by the end of the 1990s. IBM’s solution was to establish “health care accounts” for future retirees, which was essentially a way to further limit the amount it would pay for their health benefits. This step, which the company took in 1999, reduced the company’s liability by $127 million and generated a fresh pool of accounting gains that the company added to income over a period of years.
Benefits consultants helped their clients replenish the cookie jars. William Falk, who oversaw the retiree medical consulting practice at Towers Perrin, addressed the problem of diminishing income at an actuarial conference in the late 1990s: “So the clients are saying, ‘Well, what can we do about this? We don’t want our costs to jump up next year. Management and shareholders won’t accept it.’ Well,” he told his fellow consultants at the meeting, “they’re doing a lot of things. They’re looking at reduction in benefits again.”
Towers Perrin sent marketing materials to current and potential corporate clients, saying employers needed to think about “new strategies and approaches to managing health benefits.” Among them: tightening eligibility by requiring higher ages and years of service; getting some dependents out of the plan, such as dependent children; and increasing premiums, deductibles, and out-of-pocket expenses, “especially when you can’t get people [like union retirees] out” of the plan. Companies could also adopt “more aggressive assumptions” for health care inflation, administrative expenses, and participation rates. Health care inflation trends “have been low, but we’ve kept them high. Now we have room to move back down” and generate some new accounting gains.
Of course, companies that announce benefits cuts don’t say they’re cutting benefits because they’ve run out of accounting gains generated by earlier cuts. They generally just blame rising health care costs. That was what Aetna did in 2003 when it announced it would phase out health care benefits for workers who retired starting in 2004. The reality: Aetna’s costs were going up because it had used up the pool of gains generated when it capped the benefits in 1994. It had used as much as $23 million a year from this pool to reduce its annual expense. By 2002 those gains had run out, and to keep its expense from rising, Aetna needed to cut benefits again.
But was Aetna actually spending more money? Not really. Thanks to the ceiling on spending, the amount the company actually paid—that is, dollars out the door—remained roughly flat from 1998 through 2002, ranging between $35 million and $39 million. Asked to explain why it was cutting benefits when it wasn’t spending more, Aetna responded by saying that it needed “to reduce expenses in order to be competitive.” That answer, at least, made more sense: The maneuver increased Aetna’s pretax income by $34 million in the first quarter of 2003. That was 6.4 percent of its earnings. The gains continued to lift Aetna’s pretax earnings about $45 million a year for several more years.
Companies could time the cuts to generate gains at opportune moments. With a little preparation, companies can determine how many cents per share they need to meet earnings targets, and then identify which bits of their retiree plans to trim to generate the gains the company needs to clear the earnings hurdle. Caterpillar increased retirees’ premiums and made other benefit cuts in 2002 that reduced obligations by $475 million. This resulted in a $75 million accounting gain that year, which amounted to 9.4 percent of pretax earnings. The flexibility built into the accounting rules—and employers’ ability to trim benefits—has enabled companies to cannibalize their retiree health plans whenever they need to generate gains to boost income.
At Whirlpool, the retiree health plan has offset the cost of defective appliances over the years. In the second quarter of 2003, trims to its retiree medical benefits generated a one-time gain of $13.5 million. This added 19 cents a share—enough to offset a one-time after-tax charge of 16 cents a share to cover the cost of recalling defective microwaves, with three cents left over. This enabled the company to report earnings of $1.35 a share— beating the $1.31-a-share consensus estimates.
In March 2009, Whirlpool made another benefits trim: It suspended the annual credit it provided to the retiree health savings accounts. The $89 curtailment gain this cut generated neatly offset a one-time charge for a voluntary recall of 1.8 million refrigerators sold in the United States and Canada between 2001 and 2004. And in March and June 2010, the $62 million in gains from other cuts to the retiree health care plan largely offset the $75 million charge for recalling 1.8 million dishwashers sold in the United States and Canada between 2006 and 2010.
Chapter 5
PORTFOLIO MANAGEMENT
Swapping Populations of Retirees for Cash And Profits
ONE OF THE LAST LETTERS Bill Jelly received before he died in 2003 was from his benefits administrator, informing him that the death benefit from his employer, Western Electric, was going to be canceled—in one month. He’d earned the benefit decades ago, when death benefits, like retiree health coverage, were commonly offered to people who’d spent most or all of their careers at a particular company.
The benefit wasn’t much: $39,000, the equivalent of the salary he had the year he retired in 1979. But he had been counting on it to pay his burial costs and the medical expenses the couple faced, thanks to cuts to their
