By Jeff Petertil
Thirty years ago, The Wall Street Journal published on its editorial page an article I wrote. This opinion piece made a bit of a stir at the time, which was intended, and got me fired, which was not. The piece also had a lengthy afterlife. As such, it provides perspective on how complex ideas of measurement and value make their way into public discourse and either sink or swim. In this case, it sank … but some flotsam remains worth examining.
Beyond my personal reflection, the story is an example of the role actuaries may play in societal decision-making and the need to continually refine the tools available.
To my knowledge, the February 1992 op-ed piece is the only time in at least the past 50 years that either The Wall Street Journal or The New York Times (the two most influential papers of record of the period for the U.S.)—featured an article by an actuary on its editorial pages. What did it say, how did it come about, and what happened after?
My WSJ article was a critique of a corporate accounting rule that was taking effect and which had significant actuarial ramifications. The standard was having substantial impact on the business world, with the social and financial consequences characterized as anything from a ripple to a tsunami. Fortune magazine wrote of “The Killer Cost Stalking Business,” while a WSJ headline lamented “Firms Stunned by Retiree Health Costs” in the years leading up to what became known as “FAS 106.” The maritime metaphor used more often than “tsunami” or “ripple” was “iceberg.”
Corporations had been promising employees that health insurance would be continued into retirement. Cash-basis accounting was the usual treatment, which worried accounting standards professionals. They did not want to be caught unawares about retiree health care costs, which had hit their radar because of its similarities to retiree pensions. Enactment of the Employee Retirement Income Security Act of 1974 (ERISA) gave some certainty over funding, protecting, and taxing pension plans, but did not address the accrual accounting in financial statements. The Financial Accounting Standards Board (FASB), founded in 1972 with responsibility for establishing rules for corporate accounting, had a major project from 1974 to 1985 to grapple with accrual measurement of pensions—a benefit paid in an employee’s retirement but promised much earlier.
Actuaries, with their expertise in financial models contingent on future events, had been at the center of pension measurement for decades. ERISA sanctioned that role and FASB sought actuarial advice, while not ceding too much authority to actuaries. The American Academy of Actuaries, which itself had only been founded in 1965, was one of many organizations consulting with FASB and monitoring progress.
While the pension standards issued in December 1985 were being hashed out in the ’70s, some employers were making another promise—to continue into retirement the health insurance coverage they offered to employees. Similarities to pensions were obvious because long-term benefits were going to the same people from the same employers. The tax, accounting, and security rule-makers needed to understand the differences between pensions and retiree health insurance and tweak the pension rules to account for those differences. The actuaries would help—and I was one of those actuaries who seemed to know about retiree health insurance.
‘The Risks Seem Infinite’
I started my career at a health insurance company and joined an actuarial consulting firm in 1981. Like many actuarial firms that had begun with pension work, this consultancy had a small practice with insurance companies where I was placed, but most of its revenue was related to its pension business, which had been booming as ERISA compliance became compulsory for pension sponsors.
To anybody who worked in health insurance, the idea that private insurance would contract to cover a person for life was unfathomable. Few health insurers, then or now, want to guarantee a premium rate for more than one year, two at the most. Beyond that, the risks seem infinite, the underwriting data, if available, innumerable. Mortality statistics used for pensions, annuities, and life insurance are finite and choices are binary—you are either alive or dead and everybody in the census is eventually the latter. Morbidity statistics, on the other hand, can include hundreds of medical records and billings. These are not only unique to each individual but are relatively poor predictors of future morbidity. The uncertainty surrounding the future health prospects of a group of people, much less an individual, kept health insurance on a short leash.
And yet major corporations, as well as state and local governments and nonprofit entities such as universities, were promising to provide, finance, and subsidize health insurance for significant portions of their work force into their retirement years—essentially lifetime coverage, which meant not a few years of guaranteed health insurance but decades.
Concerned about retiree health costs, FASB started by requiring disclosure of the benefit. When it began considering accrual accounting, corporate financial people had impetus to conduct their own internal investigations as to the ultimate cost of the health insurance promise. Many were only dimly aware of the tip of the retiree health insurance iceberg, much less what lay beneath. These corporate treasurers, controllers, and CFOs, however, were aware of the large annual increases in employee health insurance attacking their budgets and bottom-line profit. Increasingly they turned to self-insurance for their employee coverage, to avoid risk charges of insurers and the annual hassle of insurance renewal. Employee health insurance dollars might be large and foreboding but the volatility risk was not. Insurance brokerage houses pivoted from brokering health insurance to providing self-insurance services. A few offered actuarial analyses of health costs, as did some pension consulting firms.
I became the person within my consulting firm specializing in employer self-insured coverage of medical benefits and, in time, retiree health benefits. The Academy restructured to put more emphasis on health benefits while the newly formed Interim Actuarial Standards Board recognized the need for actuarial standards in retiree health practice. Each formed a committee I was invited to join, from which I participated in educational presentations to the IRS and FASB on actuarial, financial, and design aspects of retiree health care. The Society of Actuaries (SOA) included retiree health topics on its fellowship exams, and I was asked to write some pieces for the syllabus. From 1984 on, I was busy being an “expert”—working with clients nationwide, supervising actuarial valuations, educating others, and representing the actuarial profession on these issues.
In 1990, I left the consulting firm for “down time” to reflect on some theoretical considerations bothering me regarding the measurement of retiree health liabilities. From this came a paper on factors that affect “trend”—long-term growth in health costs—which won an award from the SOA, and an article published in Contingencies delving into differences between pension and retiree health cost projections, particularly the portion near term versus far in the future. An example for a 65-year-old man showed fully half the retiree health value would be beyond 10 years while only a quarter of the pension value would be. Thus, while his current year’s pension might be several times his current health premium, the accounting and actuarial present values would put the two liabilities much closer together. Little recognized, such measurement differences were significant, given the uncertainty of longer projections. Current cash flow for health might seem inconsequential compared to pension payments, but long term the retiree health promise became quite important. Pensions, however, had legal protections that retiree health did not.
Meanwhile, FASB was moving toward its final standard, releasing for comment a February 1989 exposure draft for Other Postretirement Benefits accounting. It received no shortage of comments, many doubting there even was a liability. Actuaries expressed concerns about how the exposure draft dealt with insurance cost-sharing, such as deductibles, in the projection measurement. There was a rigidity complicating the work of actuaries without improving the accuracy of the projection of future costs. The Academy submitted written comments and, at a public hearing that fall, I was among the testifying Academy representatives. FASB deliberations continued through 1990. In conversation with FASB staff, I suggested ways the standard might address the likelihood that sponsors would drop or otherwise reduce their retiree medical plans long before the multi-decade escalation the accountants were anticipating. (These appeared in a sidebar in the Contingencies article.) I was told the use of a higher discount rate was not off the table but was being considered separately by a FASB task force looking at the use of discount rates.
In December 1990, FASB released its final statement, Statement of Financial Accounting Standards No. 106, with an effective date a couple of years later. Quickly dubbed “FAS 106,” its modifications permitted actuaries flexibility in modeling future benefit payments, but there was no allowance for later major revisions to the plan. The actuary would project the costs of the current substantive plan into the future, which would be discounted to an actuarial present value using a rate based on high-quality fixed-income investments available at the time. Companies had the option of booking the entire huge amount as a liability, even if they had no funding and no legal compulsion to continue the plan.
My Contingencies article, written before release of FAS 106, was published a few weeks later, in the January 1991 issue. It ran to several pages; my paper on trend had been even longer, as had the study notes written for the SOA and comments to FASB from the Academy that I had co-authored. Feeling the need for a succinct and direct demurral to FAS 106 for a general business audience, I wrote something shorter, and submitted it to The Wall Street Journal in January 1991.
Actuaries don’t do such things, but my background and position could be considered unique. I had written for school and local newspapers and had letters to the editor printed in Time, Business Week, and the WSJ itself. I had also been interviewed by the WSJ on retiree health topics in previous years and been quoted in a couple of their articles. Before leaving the major consulting firm I had considered submitting a WSJ piece and been encouraged by the communications director. All this gave me the confidence and contacts to think a major news forum would be interested.
I was then self-employed, with midsize actuarial and pension consulting firms seeking my help interpreting FAS 106 and valuing the retiree health plans of clients. A major accounting firm was interested in hiring me and while I considered that, I helped their actuarial arm prepare for audit and consulting work. As 1992 began, they were my major client and source of income.
WSJ editors spoke with me in January 1991 and were interested in publishing my submission as an opinion piece on their editorial page when space permitted, but after a week or so passed with other issues taking precedence, they sidelined it. A year later, updating changes, I resubmitted the article and the WSJ published it as an op-ed piece in February 1992.
The WSJ suggested cuts but otherwise took my article pretty much as I submitted it. When waiting for op-ed space to open had gone on a bit, however, I said I would take the article to another publication, at which point they said they would print it within the week. I then faxed the clients I was working with a copy of the draft with notice that it would be appearing within a few days, without mentioning where. On Thursday, February 20, while at a downtown Chicago law firm seeking participation on a project, I checked for messages and heard one asking me to contact the WSJ. They were printing the article the next day and wanted to know a few things, including what headline to use. I had not included one in my submission; my experience with papers was that editors usually picked the headlines, so this question had not occurred to me. The Chicago lawyers were waiting for me, the WSJ was waiting for me, and I did not have an answer. The editor suggested, “Ignore the Retiree Health Benefit Rule.” I hesitated, thinking, “Uh-oh, it’s not that simple,” but had nothing else to offer and, it was pointed out to me, I had used “ignore” in the last sentence of the article.
The next day my piece appeared on the editorial page of the newspaper. Mid-morning, I received a call from the actuary who was my contact at the accounting firm client. Apologetic and uncomfortable, he said my contract was being dropped. We had recently discussed whether my approach to valuations could work in the firm’s practice and were doubtful, so the severance was not unexpected, but this wasn’t mutual. The abrupt nature of the firing certainly took the glow off later congratulatory calls I received. From other sources I later learned he had orders from above; his boss had the draft but never read it and easily succumbed to the pressure from higher up. Later sources indicated FASB was well-pleased I had gotten the axe.
As well they might. FASB staff and board had worked conscientiously over a decade to hone a standard responsive to a somewhat intractable accounting problem and seen the completed result pilloried with the word “ignore” in a headline in the newspaper most read by those they influenced. It had to seem unfair.
The WSJ reporter who covered accounting issues had discussed the topic with me and said the board would not be happy with the piece. This was hardly a surprise to me but does indicate the paper scrutinized the relevance of my opinion and its contextual importance before publishing it. Writing from a perspective broader than actuarial present values or accounting standards, I tried to show the numbers generated did not make sense in the larger world of getting and spending. Opening paragraphs noted the accounting charge taken by General Motors was almost equal to its equity and that taken by IBM had led to its first ever quarterly loss, both without much change in market value. I also observed that court cases were upholding employers’ right to terminate the benefit and that there was potential for manipulating expense accounting. I considered publication by the WSJ as acknowledgment that my perspective was of interest to their readership.
There was never a written response by anyone, as far as I know. My article said publicly what most of those close to the subject knew—stating an employer’s liability in FASB terms involved much uncertainty even if a company acknowledged a liability, which many did not. No one rushed to defend the standard, but it was not about to be repealed and there was no consensus for an alternative. Those charged with issuing a company’s financial statements could not actually ignore the rule. My article was aimed at a broader audience, including the last sentence. “Unless a 50-year commitment to retiree health benefits has been made, the best thing a manager can do with the FASB rule is what the financial analysts have done so far – ignore it, and get on with the duty of deciding just what the corporate commitment to health care is.”
I had seen that commitment as shaky. Continuing employee health insurance into retirement appeared financially inconsequential to corporations after the 1965 advent of Medicare, which seemed to presage the eventual end of private health insurance and a socialistic approach to health care in the U.S. By the mid-’80s, however, the country was in the thrall of Reaganomics, socialism was dead, Medicare was in trouble, and the placid waters of retiree health insurance were now seen as strewn with dangerous ice flows commonly acknowledged by business journalists to be “only the tip of the iceberg.”
But my 1992 WSJ article pointed out that the cost could be easily reduced. It became evident in the 1980s that the federal government was not going to extend employee protections for retiree health benefits or add tax incentives for funding. Nevertheless, FASB saw the need for accounting similar to pensions and for understanding how promised medical claim payments might most accurately be projected. Actuaries with the same aim collaborated with them.
Writing the 1988 SOA study note, we included all the factors that might affect the future cost of health care—expanding utilization of new and improving medical services, continuing aging of retiree populations, Medicare reductions, macroeconomic spending limitations, retirements before Medicare eligibility, family coverages, etc. By showing benefit managers scenarios of how these factors would play out over the years, actuaries could aid in refining the design of these plans. For the accounting liability, those projections would compact into a single deterministic present value, but I felt the discounted present value and allocation of what might prove to be phantom liabilities could wait until the future of payments was established.
Though I did not restrict discount rates, what I had not recognized was that few in the pension and employee benefits community had experience with “risky” discount rates. They may have understood that some invested assets expected higher returns than others—i.e., stocks more than bonds—but the connection with actuarial present values was undiscussed. Present values were the end game, but discount rates, though highly influential, were secondary issues.
As to accounting alternatives, I dropped into the middle of the 1992 WSJ piece the irony that on the day FASB approved the rule, the Nobel Prize in Economics was awarded to advocates of an asset pricing model that accounting could have used. I was particularly aware of that because I had studied with one of the Nobel recipients and done so largely because of job responsibilities valuing proposed projects on a cost/benefit basis using discounted cash flow. In doing so, we used a discount rate above the current yield of high-quality fixed-income investments—the benchmark FASB chose for accounting discount rates. These projects were venture capital opportunities, not benefit valuations. The income expected was speculative, not fixed, and getting that income was riskier than for a high-quality investment. Our investors sought a stock return, not a bond return, and use of a higher discount rate reflected the risk.
If retiree health benefits were an employer liability to employees, then employees with the benefit must have an asset, but my background assessing financial value of things besides employee benefits told me that asset was speculative and risky. The actuarial present value could reflect that risk in 1992 with a discount rate more like 15% than the 8% to be found in the bond market and required for FAS 106 valuations. A more careful matching of the discount rate with the risk and uncertainty level of payments long term would have led to a more relevant liability figure for those plan sponsors who were not funding the retiree health benefit and had reserved the right to transform or terminate it.
The use of discount rates in determining present value depends upon the purpose and is complex both in concept and practice. The accountants really did not want to be the arbiters of this. Years later the head of accounting in the U.K. was asked at an actuarial forum about the use of discount rates and he replied to the effect that “I thought you took care of that.” But actuarial literature gives limited guidance beyond regulatory requirements as to discount rate selections, although mathematical formulas for interest rates are a key part of actuarial education. Following my WSJ piece, an Academy task force formed to examine use of those rates across all areas of actuarial practice, not just employee benefits. The group recommended a study to the SOA and there was a grant for research, but I was told the researcher never made much headway.
In 2002, the SOA called for papers for a Vancouver conference on The Great Pension Controversy. Although I am not a pension actuary, I responded with a paper on the use of discount rates for terminable benefit plans. The paper was accepted and praised by one of the main organizers and was eventually published in 2005 in the North American Actuarial Journal (which also published my paper on aging statistics that year). I joined a financial economics task force cosponsored by the SOA and Academy. With urging from this group, the SOA championed a request for proposals for a financial economist and/or academic accountant to do research. We thought a Stanford team would take it on, but they declined.
As to the WSJ piece, after it was published, I continued to consult on health benefits and participate in valuations, many under FASB rules. The Governmental Accounting Standards Board, undeterred by my FASB fracas, asked me to help with their retiree health standard. In a 1998 interview with a benefits monthly, I noted being fired by the client accounting firm in the immediate wake of the 1992 publication. This eventually came to the attention of Ellen Schultz, an investigative reporter for the WSJ. She called me and asked about the incident. Her story exposing corporate chicanery that left many a retiree without promised health care benefits was published on WSJ’s front page in 2000 and included in her 2012 book, Retirement Heist, on the corporate manipulation of retirement plans at employee expense. The story quoted from my 1992 op-ed and noted my firing by an accounting firm. The morning the story was published, the first call I received was from the actuary who had been required to fire me in 1992. “It looks like you were right, Jeff.”
Maybe. The Schultz reporting brought to a wider audience what was evident for some time: Corporations were no longer fulfilling the retiree health care promise. Her examples featured what I had hinted at—the personal impact on despairing retirees while accounting dollars were released, book earnings increased, and corporate executives were rewarded. Whether things would have been different with the alternatives I suggested, however, is an open question. The accounting numbers would have been smaller, both going in and coming out, but the corporate retreat from paying for retiree health care was based more on the prospect of ever-increasing benefit cash flow than on the accounting liabilities. My article’s last sentence challenged the corporate commitment to health care; the ensuing years answered that the commitment did not extend to lifetime coverage.
A Challenge for 2022
What did I expect from publication? In a different setting, Wallace Stegner wrote of the writer who, “drops his feather into the Grand Canyon and stands expectantly, waiting for the crash.” There was a story not being reported, connections not being made, so I wrote about the disconnect between the accounting and the financial reality that stock prices seemed unmoved by retiree health cost news.
Early on it was evident few financial analysts believed any iceberg from retiree health benefits would sink corporate ships. Around 1988, House Ways and Means Committee staff ventured to New York to understand the impact of the impending accounting change. On the first day the congressional staffers spoke with actuaries and consultants, myself included, at major benefit consulting firms and on the second day with people at investment banks. Sometime later, I had a chance conversation with an actuary employed by an investment bank who was at the second day’s last meeting. The House interviewers told him, “Everybody on the first day said it would be a major catastrophe and everybody on the second day said it would be a minor ripple.” Those on the second day may have been underinformed about the particulars of retiree health plans, but they knew how companies responded to risk. Analysts at bond rating houses such as Moody’s made their own adjustments to the FAS 106 numbers, although not divulging the details of their skepticism.
The shifting emphasis among an individual’s rights and responsibilities during the 1980s made corporate moves away from providing lifetime security seem not only defensible, but better for all concerned. Newer companies shied away. IBM wrestled with its retirement policy while also deciding whether it was a hardware manufacture or a service provider, but upstart software provider Microsoft minimized pensions and probably never considered retiree health benefits. Avoidance became the private sector model; pension and retiree health benefits now reside mostly with governmental employees. The poster child for the private/public schism in retirement policy is the U.S. Post Office, which after it was cut loose from the federal government was required to prefund its retiree health benefits. Some have attributed its constant battle for solvency to that requirement—we will see what happens after much of that burden was lifted by the recent Congress.
I took pride in what we had achieved with the retiree health projection model. It encompassed aspects that simpler models ignored. I was among those who, in writing the actuarial standard for retiree health benefits, ASOP No. 6, tried to keep emphasis on the projection, not the present value. My WSJ piece was a warning to those taking FAS 106 as a solution to a question FASB had never posed—What is a retiree health benefit worth? The 1992 WSJ piece identified three public concerns converging on that question: “the aging of America, the unending rise of health care costs and the consequences of misguided corporate paternalism.” In building economic models of the impact of the first two, actuaries gave corporate managers knowledge of the third, and those executives set about turning off the faucet. In the context of the time, this was inevitable.
Concerns with aging and health care costs remain, with a pandemic having twisted the viewing prism on each. Another challenge for actuaries—which, if met successfully, could open new practice areas of societal and environmental impact—is to refine and define how risk and uncertainty fit with discounted present values.
JEFF PETERTIL, MAAA, ASA, FCA, MBA, is an independent consulting actuary who served on the Academy Board of Directors in 1990–1993 and 2010–2013.