Ephemerides (August 2003)
August 21, 2003
At the end of last month, a judge in the notoriously plaintiff-friendly Southern District of Illinois handed down a decision declaring that cash balance pension plans are illegal. In certain circles, Cooper v. IBM Personal Pension Plan instantly became a case to rank with Brown v. Board of Education in the struggle for civil rights. The rights at issue in this instance are those of older workers, who are supposedly being victimized by an insidious novelty in methods of calculating pension benefits. According to such noisy voices as the Pension Rights Center, Wall Street Journal reporter Ellen Schultz and Business Week's Nanette Byrnes, companies are switching from traditional defined benefit pension plans to cash balance plans because they want to save money and strip loyal, long-term employees of the pensions that they had the right to expect. Furthermore, these critics have long insisted, cash balance formulas violate the Age Discrimination in Employment Act. Miss Schultz once alleged that cash balance conspirators had insinuated language into the preamble to a proposed IRS regulation in order to save their brainchild from immediate legal demise.
Readers who pay little attention to the specifics of a somewhat arcane subject may be surprised to learn that the object of all this venom is nothing more than an alternative way of looking at pension benefits, one that focuses on their present value rather than the anticipated future stream of payments. The traditional type of pension formula derives a life annuity from such inputs as years of participation in the plan, compensation over a specified period and a proxy for Social Security benefits. A cash balance formula credits participants with a lump sum amount each year, which can be converted into a life annuity at retirement if desired. A typical, fairly simple traditional formula looks like this: a life annuity beginning at age 65 equal to 1.5 percent of average annual compensation over the preceding five years, multiplied by the number of years of participation in the plan. A simple cash balance formula is on these lines: an annual credit equal to six percent of compensation, plus interest on accumulated credits at 100 basis points over the average rate for one-year Treasury bills over the preceding year. (There is a third variation, called a pension equity formula, that operates in much the same manner as a traditional formula but derives a lump sum instead of a life annuity. If the lump sum is payable at normal retirement age, the pattern of benefit accrual is like that in a traditional plan; if immediately, like that of a cash balance plan. The plan at issue in Cooper used a PEP formula before changing to a straight cash balance plan, but not enough facts are disclosed in the opinion to make clear how it worked.)
Under our sample traditional formula, a participant with 20 years of credited service and average pay over the past five years of $40,000 earns a life annuity equal to 30 percent of $40,000, or $12,000 per year. The accrued annuity is the same, whether he completes his 20 years at age 45 or 65, but the value of the benefit varies sharply with age. A guy who enters the plan at age 25 and quits to take another job at age 45 can look forward to $12,000 a year - but not until 20 years in the future. His similarly situated co-worker who started at 45 and is retiring at 65 can start drawing his annuity right away. As a result, the younger man's accrued benefit is worth less than a third of the elder's.
The value of benefits under a cash balance plan is, by contrast, insensitive to age at separation from service. A 30-year-old and a 60-year old ceteris paribus accrue benefits worth the same amount each year (though formulas can be, and frequently are, skewed to give higher accrual rates to older or longer-service participants). The nub of the "age discrimination" objection to cash balance plans is that failing to take age into account in computing benefits is tantamount to discrimination on the basis of age. Put another way, the judge who decided Cooper believes that the ADEA makes it mandatory for pension plan sponsors to incur higher costs for older workers than for younger (for, of course, value to the participant is merely the obverse of cost to the employer).
That may sound like a mildly implausible proposition, and we know quite definitely that Congress never had it in mind. The 1988 ADEA amendments that created the current version of the statute came at the end of a long controversy that was resolved by a backroom deal between Bill Brock, then the Secretary of Labor, and Democratic Congressional leaders. (Congressional Republicans heard about it after the fact.) The controverted issue was whether the longstanding practice of stopping pension benefit accrual under a traditional pension plan at the plan's normal retirement age (typically age 62 or 65) violated the ADEA. The original text of the statute, enacted in 1967, explicitly excluded bona fide employee benefit plans from its scope. Proponents of the ADEA declared that their purpose was to prohibit employers from making hiring or firing decisions on the basis of age, not to compel them to offer older workers benefits like health insurance and traditional pensions that grow more costly with age.
The administrative agencies with jurisdiction over the ADEA (initially the Department of Labor, later the Equal Employment Opportunity Commission) didn't like the employee benefit plan exclusion and tried to undermine it by issuing interpretations holding that an employee benefit plan was not "bona fide" unless it provided older workers with either the same benefits as younger workers or benefits that cost the same amount. Whether that interpretation comported with the law was a much disputed question. In 1988 a case presenting it squarely was working its way toward the Supreme Court. Since each side had much to lose from an adverse result (the Court's eventual decision was against the DoL/EEOC view, Public Employees Retirement System of Ohio v. Betts, 492 U.S. 158 (1989)), they reached a "compromise" (meaning that Secretary Brock caved in on all significant points). The Age Discrimination in Employment Act Amendments of 1988 generally adopted the "equal cost or equal benefits" rule. For pension plans, however, a special, purportedly less expansive, provision was inserted into ERISA and the Internal Revenue Code. It was headed (in its Code version - ERISA lacks section headings) "Continued accrual beyond normal retirement age" and provided that a defined benefit pension plan would not be in compliance with ERISA or with Code qualification requirements "if, under the plan, an employee's benefit accrual is ceased, or the rate of an employee's benefit accrual is reduced, because of the attainment of any age", with exceptions that are not pertinent here (ERISA, §204(b)(1)(H); I.R.C., §411(b)(1)(H)).
Given this background, the most stringent reasonable interpretation of the 1988 amendments is that they extended the "equal cost or equal benefits" rule to pension plans. The minimalist interpretation is that they do no more than require accruals to continue after normal retirement age (a position that finds more support in the statute's background than its text). Nowhere is there a hint that anyone imagined that Congress was requiring more of pension plans than of other types of employee benefit, that, for them alone, equal cost was not sufficient but equal benefits were mandatory in all cases. To extract that meaning from the statute, the judge in Cooper had to engage in hocus-pocus. He took the phrase "rate of an employee's benefit accrual" and read it as if it were "rate of increase of an employee's accrued benefit". In common parlance, those would be synonymous phrases, but "accrued benefit" is a term of art, defined as a benefit in the form of an annuity (ERISA, §3(23), I.R.C., §411(a)(7)). For technical reasons, there are circumstances under which a plan that defines participants' accrued benefits in lump sum form needs to convert them into actuarially equivalent annuities. Where that necessity exists, the statute painstakingly uses the term "accrued benefit". The section added by the 1988 amendments is not one of those places, but that did not bother the judge. He decided that the different terminology was attributable to considerations of good grammar and changed the words to reflect what the draftsmen must have really meant. Then, since benefits of equal present value translate into larger life annuities at a future age for younger participants, he concluded that, under a cash balance plan, "the rate of an employee's benefit accrual is reduced, because of the attainment of any age".
The chances that this feeble exercise in faux-literalism will survive on appeal are about the same as Arianna Huffington's of becoming the next governor of California, so the district court decision is not substantively important. It is, however, a morale booster for the demonizers of cash balance plans, who are likely to become yet more vociferous over the next few months and to start demanding Congressional action after they lose in the courts. Traditionally timid Republicans - phobic of being labeled "anti-worker", much less "anti-older worker" - will be tempted to accede.
The principal reason for resisting attempts to "reform" cash balance plans is not that they are an ideal compensation technique in every instance in which they are used. For some companies they are a good idea; for some not so good. Their most attractive feature at the present moment is probably the fact that the value of liabilities isn't affected by changes in interest rates. For traditional pension plans, the decline in interest rates during the past few years has been a financial catastrophe (as I have discussed elsewhere). For cash balance plans it has been unimportant. The human resources impact is more ambiguous. A program under which workers in their middle years can get their hands on a large chunk of cash by quitting (and only by quitting - pension plans, unlike the familiar 401(k) plans, can't lend money to participants or make hardship distributions) may be well-received, but it isn't the ideal strategy for minimizing expensive turnover.
There is no inherent reason why a cash balance plan should cost more or less than one with a traditional formula. Though Ellen Schultz, for one, seems convinced that the primary motive behind cash balance conversions is to save money, neither empirical evidence nor logical analysis supports that view. Benefits under any type of formula will be as generous or niggardly as the plan sponsor wants to make them. The pattern of accrual does not determine the ultimate cost.
Left to their own choices, companies currently (that is, as of 2000, the last year for which data are available) prefer traditional pension formulas to cash balance formulas by a three-to-one margin (in terms of the number of participants), but the gap is closing rapidly, having been 25 to one as recently as 1996-97 (DoL statistics). Groups like the Pension Rights Center would like to see the government step in and overrule the free market, telling companies that they must use traditional formulas or go out of the pension business. At the notion that lots of employers might make the latter choice, they scoff, insisting that pension plans are so valuable to their sponsors (allegedly as sources of paper profits to swell earnings-per-share) that they will be retained regardless of the extent to which the government dictates their design.
Let's leave aside for the moment whether that last contention is correct. Even if banning cash balance plans doesn't trigger an exodus of plan sponsors, there are two decisive arguments against it: First, by compelling employers to adopt plan designs that were less than optimal for their circumstances, it would waste economic resources. The effect would be the same as if every company had to pay government-mandated bonuses to a select group of lucky workers. The economy wouldn't collapse, but it would be smaller than otherwise would have been the case.
Second, the good done to older participants by the forced preservation of traditional formulas would be balanced by the harm done to younger ones. In effect, the government would be trying to transfer wealth from one group to another, a ploy that is as futile as it is unfair. The size of any worker's total compensation package is set by market forces. When the government tries to make one man's bigger and another's smaller by fiat, it simply forces the the parties involved to reach the some outcome more circuitously. In this case, keeping an unwanted traditional formula increases the pay of older participants above the market level, while reducing that of the younger cohort. Employers will predictably react by giving the disfavored group additional pay in other forms and taking steps to offset the compensation mandate for the favored group. Among the ways to accomplish the latter are (i) eliminate future pension benefits, substituting individual account plans, which have always operated in the same way as cash balance plans, except that participants have real, rather than hypothetical accounts and bear the risk of gain and loss; (ii) find methods of getting rid of older workers without violating age discrimination laws (far from impossible to accomplish); (iii) cut the value of benefits that are primarily of value to older workers, such as health insurance coverage.
Cash balance plans are not the most important issue of our era, but what happens to them will be a small indication of whether contemporary Republicans are, as some commentators assert, nothing by "Big Government Conservatives" and, more broadly, whether our ruling classes are getting their impulse to micromanage society and the economy under control.
August 13, 2003
The Wall Street Journal's Gerald Seib tells us, in today's Capital Journal column [link probably for subscribers only], about the Democratic strategy for scoring political points from the California recall election: "Nationalize the California recall circus by portraying it as part of a pattern in which Republicans try to use strong-arm legal tactics to seize political power they failed to win at the ballot box." What's the rest of the "pattern"?
First of all, Democratic leaders will cite the 2000 presidential election, in which, they contend, Republicans refused to allow a recount that would have shown Al Gore winning and instead forced a favorable but incomplete result through the courts until the Supreme Court gave them the election. The same tendency is playing out in Texas, Democrats argue, where Republicans are trying to jettison a court-drawn map of congressional districts by pushing through the state legislature a new map much more friendly to them, even changing state Senate rules to get it done.
And though most Democratic leaders are less eager to cite this precedent, some will argue that the impeachment of President Clinton showed that Republicans refused to accept the outcome of the two presidential elections he won.
These will make interesting talking points. The Democrats will essentially be arguing that resort to democratic procedures is "strong-arm legal tactics to seize political power".
In California, of course, the Republicans are trying to overturn the results of an election - by holding another election. Recalls may or may not be a great feature of the state constitution, but they can't be called defiance of the will of the electorate. If Governor Davis were doing a vaguely passable job in his office, the recall effort would have gone nowhere, just like a score of others (including a Democratic effort to oust Governor Ronald Reagan). As for legal tactics, the only ones visible so far have been the unsuccessful Democratic lawsuits aimed at stopping, delaying or eviscerating the October 7th election. Who is it who doesn't trust the ballot box?
The clash in Texas is between Republicans who want the elected state legislature to draw Congressional district boundaries and Democrats who want to keep a set of districts imposed by unelected judges. What's more, everybody agrees that the GOP map would more accurately reflect the political preferences of Texas voters. Under the present one, a majority of the seats are safely Democratic, even though Republicans control both houses of the legislature and every statewide elected office. Again, one can see which party is trying to get through legal tactics (in this case, illegal tactics, since Texas law - enacted back when there were almost no Republicans in the state - requires lawmakers to attend legislative sessions, and 11 Democratic state senators are hiding out in New Mexico rather than comply) what it can't gain through elections.
Florida's Presidential vote count is a more complicated and convoluted case, but, if the media re-recounts of the ballots mean anything, they show that almost any set of vote counting rules, consistently applied, give George W. Bush more votes than Al Gore. (CNN, "Florida Recount Study: Bush Still Wins"). The Democrats' complaint reduces to this: The U.S. Supreme Court' refused to allow the Florida Supreme Court to fashion inconsistent rules that would have maximized the chances of creating a Gore majority. That, once more, doesn't sound like the position of a party committed to popular rule.
Finally, there's impeachment. The Big He is already trumpeting that Gray Davis is another victim of the Vast Right-Wing Conspiracy. Well, I suppose that impeachment can be attributed to Republicans' "refus[al] to accept the outcome of the two presidential elections [Clinton] won", rather than to their disgust with his personal character, but how does it qualify as "strong-arm tactics"? The Constitution prescribes a method for removing the President from office; the bodies with the power to remove are themselves elected; and the defendant has ample opportunity to defend himself. When a party loses control of Congress through the ballot box, it must accept the right of the winners to exercise their Constitutional powers. At least, it ought to do so; otherwise, what's the point of holding elections?
My expectation is that this Democratic strategy will fade away. If it doesn't, the 2004 election will be a kind of seminar in political theory, but it won't be the Democrats who will be arguing in favor of democratic government.