Fourteen
A Triumph of Governance: Social Security
The social security program was going broke. That means revenues from payroll taxes (the FICA deductions from paychecks) dedicated to Old Age and Survivors Insurance (OASI) were smaller than the amount being paid to the elderly. Originally, the system had a cushion of budget authority, to be called upon when benefits exceeded revenues; by 1982, however, that was nearly gone. Unlike the budget-deficit "crisis," social security's difficulties were concrete and calculable: by a specific date, some 31.6 million checks would be held up. The federal government was in danger of defaulting on what is, save for the national defense, its biggest commitment.
Two years of social security politics did not create hope for a settlement. Democrats felt that Republicans exaggerated the problem, wanting to cut social security, not out of concern for the elderly, but as part of their ideological attack on "big government." Republicans felt that Democrats were engaging in unmerciful demagoguery, beating up on the GOP while failing to acknowledge real problems. Each view had a measure of truth. Let us now quickly review the dimensions of the problem to set the stage for the last-minute rescue.
After the May 1981 debacle, and after the president was convinced not to include social security in his "September offensive," Congress, with Reagan's support, passed a bill to restore the minimum benefit, which had been eliminated in the 1981 reconciliation. OASI pensions are just part, though the largest part, of the insurance system funded by payroll taxes under the Federal Insurance Contributions Act. FICA also includes unemployment insurance and the disability and health (medicare) trust funds.[1] Because disability and health funds were still in the black, the Senate, in its version of the minimum benefit bill, provided authority for OASI to borrow from its companion funds for the next ten years.
But Barber Conable, Republican leader on Ways and Means, insisted that this borrowing authority must expire at the end of 1982. He hoped that such an imminent date would encourage other action in the lame-duck session after the 1982 election, when at least some portion of Congress no longer would have to worry about voter retribution.[2] the National Commission, on which Conable sat, would report in November, and, in the face of imminent default, Congress would have to act.
The National Commission first met in February 1982. Chairman Alan Greenspan, former CEA chair and future Federal Reserve chair, joked that the fifteen members might produce fifteen minority reports.[3] President Reagan's appointees included three Republican business types—Greenspan, Robert Beck, and Mary Foley Fuller—along with conservative Democrats Alexander Trowbridge (head of the National Association of Manufacturers) and former congressman Joe Waggoner of Louisiana. Howard Baker appointed Finance Chairman Bob Dole, social security subcommittee chairman William Armstrong (R-Colo.), and John Heinz (R-Pa.), chair of the Special Aging Committee. Like Baker, Bob Michel appointed his party's committee leaders—Conable and subcommittee ranking member Representative William Archer. Thus, Reagan had chosen five conservatives, while Congressional Republicans selected a set of committee leaders divided between those who staunchly opposed revenue increases, Armstrong and Archer, and those who might accept a mix of revenues and benefit cuts. Democrats appointed liberals, passing over almost all committee leaders. Robert Byrd appointed Senator Moynihan, the subcommittee ranking member, and Lane Kirkland, head of the AFL-CIO. Speaker O'Neill appointed Claude Pepper, former Representative Martha Keys, and Robert Ball. Ball, former commissioner of social security and O'Neill's real representative, had helped organize Save Our Security, a coalition of interest groups. He had been in the program since the beginning and was the party's expert. Ball's expertise was matched by that of Robert Myers, the National Commission's staff director; Myers, former chief actuary, at the time was deputy commissioner of the Social Security Administration. Myers was the Republicans' expert on the program.
The Commission thus knew enough and represented virtually everybody. Conservatives could outvote liberals, but that didn't mean much because no agreement rejected by Ball, Kirkland, Pepper, and Moynihan was about to pass the House.
Before the Commission could do much, social security became part of the larger budget fight. Efforts by a few participants to cut a deal under cover of the Gang-of-17 negotiations broke up in animosity. How much to cut COLAs, how much in taxes the president would allow in
return, and, especially, who would take the blame were the issues over which the Gang of 17 collapsed during Reagan and O'Neill's acrimonious meeting.
The Gang's collapse put Domenici out front, where, the reader may recall, he talked Reagan into the $40 billion social security "plug" in his First Resolution for FY83. That, too, was blown out of the air quickly. When the National Commission next met, the result was, in Time 's words, "a partisan shooting match, with cameras rolling. Senator Moynihan charged that the administration had 'terrorized' older people into thinking that they won't get their Social Security." Republican Senator Armstrong declared that "we have done everything to avoid making this a partisan issue"; then he blamed Democrats for the $40 billion cutback proposal.[4] After that dustup, commission meetings sank into somnolence; congressional Republicans stopped making dangerous proposals; and attention shifted to the election.
It is hard to disentangle the social security issue from the recession and from "fairness" in order to assess its independent effect on elections. Paul Light's analysis makes three important points. First, many Republicans who survived the election did so, particularly in the Senate, only by very slim margins. Therefore, they were nervous about any issue that could anger the public. Second, to defeat them at the polls, social security did not have to work directly, as if the Republicans had done something to harm it; social security had only to symbolize the entire Democratic argument that Reaganomics was unfair. Third, Republicans believed social security had cost them dearly in lost seats.[5] A well-positioned lobbyist summed up the political situation:
By the time the November election was over, Baker, Darman and that crowd were in charge of the issue. This is an overstatement but from conversations then it was clear that they would take their own grandmothers off social security to get a deal with the Democrats. They were in the trough of a depression, the President's popularity was at bottom, they had lost their House majority, and they wanted it done, behind them! The polls showed that if anything went wrong with social security the president would get blamed.
Yes. And clearly something could go wrong by July. Pragmatists in the White House believed that, as one told us, "the only way we Republicans can deal with social security is if the leadership of the Democrats is on board." That gave the Democrats an advantage. But it didn't necessarily mean that the president would go along with a deal.
The administration's best scenario was that a solution—that could be enacted in the lame-duck session—would emerge from the National Commission's meetings on November 11–13. But that would not happen:
the session already had enough to handle; the Commission was too divided, and probably too large, to forge an agreement; and the Speaker wanted his twenty-six new Democrats in place before reforms were considered.
The Commission did, however, define the scope of the problem, an action surely necessary to end the extensive disagreement over its seriousness, which depended on the economy's performance. In 1972 the Social Security Administration (SSA) had estimated 15 percent inflation and 12 percent real wage growth from 1973 to 1977. Those assumptions justified indexing benefits to inflation after a 20 percent increase. Unfortunately, inflation turned out to be 41 percent, and real wages rose only by 1 percent. In response, President Carter and Congress put together a rescue package that changed the calculations of some benefits (the formula had been more generous than intended) and scheduled payroll-tax increases at intervals until 1990. Assuming 28 percent inflation and 13 percent real wage growth during 1978–1982, the system would have been quite healthy; but unfortunately inflation was 60 percent and real wages had shrunk by 7 percent. The discrepancy was more than enough to swamp social security again.[6]
As a matter of policy and prudence, this experience might have caused participants to err on the side of pessimism. Yet liberals objected: to define the problem as drastic served the conservative agenda—radical reform as opposed to incremental change. And conservatives were embarrassed because pessimistic forecasts would contradict the administration's economic scenarios. The SSA responded in 1982 by creating a range of scenarios, of which alternative 3 was very pessimistic (shrinking real wages through 1984 and barely any wage growth for the rest of the decade), alternative 2A was the administration's FY83 economic forecast, and alternative 2B represented the actuaries' fairly pessimistic best guess (a weak recovery in 1982, fairly similar to the low-ball forecast on which Martin Feldstein would insist for the FY84 budget).
Alternative 2A predicted an $82 billion shortfall for OASI from 1983 to 1989. At that point, the OASI payroll tax would rise from 9.5 percent to 10.6 percent (half each from employers and employees), and the system would stop losing money. It was not good but not so terrible, certainly no reason to make big cuts. By November 1982, however, alternative 2A was not looking credible; it predicted only 8.9 percent unemployment for 1982. As Senator Moynihan, who had downplayed the crisis, later commented, "There was a point when we thought you might squeeze by, but when unemployment went to 11 percent, there was no point kidding yourself."[7] Greenspan, Staff Director Myers, and most Republican commission members had no stake in the administration's forecast; they were willing to go with something like alternative 2B, which
assumed a $184 billion OASI shortage through 1989, larger than the disability surplus.[8]
When the National Commission met, Greenspan suggested they look for savings in a range from $150 to $200 billion, and the members agreed. In these matters, involving huge sums, accuracy plus or minus $25 billion is the most one can hope for. The shortfall was around 14 percent of projected costs of the program for 1983 to 1989. The commission agreed also on the size of the system's long-range problem: 1.8 percent of taxable payroll. To comprehend that estimate requires an understanding of the social security trust fund.
The most important thing to know about the financing mechanism of social security is its pay-as-you-go system. What is paid in during a given fiscal year as contributions goes out that same year as benefits. That is all there is to it. Everything else is misleading—everything.
Social security looks like insurance: individual and employer make equal contributions; the worker collects after retirement. He or she pays into, and benefits are paid out of, a trust fund. The trust fund pays benefits each year out of that year's contributions plus any accumulated balance. If benefits exceed contributions, the balance declines. From 1975 through 1981, the OASDI balance fell from $44 billion to $25 billion.[9] If contributions outpace benefits, as during most of the system's history, the balance rises.
Benefits are shaped by eligibility rules, inflation rates, and demographics. Contributions are shaped by tax rates, taxable income definitions, wage and employment levels, and population. (Only wage income is taxable; in 1982 up to the first $32,400 could be taxed. The rationale for this cap is the insurance principle: higher contributions could be justified only by higher benefits. But as it stands, higher-income participants receive a smaller benefit relative to contributions than do lower income; thus, the system is progressive.) If people live longer, social security costs more. If fewer people are born, there are fewer wage earners to contribute.
Over the long run, population factors have the largest influence on the relation of contributions to benefits. In 1955, when the system was young, there were 8.6 workers per beneficiary. A 4 percent payroll tax was sufficient to cover benefits. By 1982, with 3.2 workers per beneficiary, about 12 percent was needed. The ratio was expected to remain stable as the baby-boom generation began to dominate the work force. After the year 2000, however, this group will start retiring, to be replaced by their children, the "baby-bust" generation. By 2030, under the alternative 2B assumptions, there would be only 2 workers per beneficiary. If benefits were to be maintained at the same level, contributions would have to equal 16.8 percent of the taxable payroll. Because existing law
set taxes at 12.4 percent, the system would run massive deficits. If the pessimistic alternative 3, which predicted even lower birthrates, were true, contributions of 23.9 percent of payroll would be needed. Nearly half of all benefits would have to come from the principal of the trust fund.[10] Dollar amounts involved would depend on inflation; over the period, deficits would be in the trillions. And, like all complex calculations, things could turn out a lot better—or worse.
Peter Peterson, former Commerce secretary and organizer of the "Bipartisan Budget Appeal," was among the many conservative critics who used long-term projections to argue for major changes. Peterson wanted a series of reforms that would lower benefits, particularly for people with higher incomes.[11] To his ideological right, the Heritage Foundation reported that "the only way to cure the fiscal dilemma of OASI is through fundamental redesign of the social security system … [converting it into] an actuarially sound individual annuity program."[12] Contributions would be invested in a fund tied to the stock market's performance or some other reflection of the private economy. Ronald Reagan wanted to make the program voluntary. The program's defenders felt that the whole notion of such long-term projections was a mite unreasonable. "The feeling that we must make final decisions now for the year 2035," said Henry Aaron of the Brookings Institution, "would be rather as if we had expected Calvin Coolidge to make binding policy for Americans of the '80s."[13]
Polls showed that the proportion of the public who had "only a little" or "no" confidence in the ability of the system to pay benefits when they retired rose from 49 percent in 1979 to 75 percent in 1982.[14] "My children are concerned with whether they ought to contribute," said Dan Rostenkowski. "They're entitled to doubt whether there's going to be solvency when they expect to retire."[15] The critics argued that these fears presaged "generational warfare." In fact, the young supported current benefits at least as strongly as did the recipients.
Though the critics' crocodile tears over lack of support for the program were unjustified, proponents of the program felt that the long-run political consequences of long-term deficit projections might not be so minimal. If they could reduce the size of the long-term "problem" without reducing benefits much, it was worth doing.
"Fixing the problem," according to virtually everybody, meant combining reduced benefits and an enlarged trust fund, so that during bad years the system could live off its "savings." The National Commission defined the target for a long-term solution, not as the shortage in any particular year, such as 2035, but rather as the average shortage over the entire period from 1982 to 2056. According to the alternative 2B projections, this would be 1.82 percent of taxable payroll.[16]
A difficulty with all this went unidentified: the "trust fund" is a mirage . To withdraw money from the fund, the government must cash in its assets. The assets, however, are government bonds. To pay off those bonds, the government must either sell more bonds or raise revenue through taxes—which is exactly what it would do if there were no trust fund at all. Merely building up the fund per se makes no difference.
The fund would matter only if it held assets in the private economy: stocks, bank accounts, buildings, whatever. Then cashing in the fund would reduce the need to tax or to borrow. But that possibility poses all sorts of problems: A fund of this sort, large enough to fund social security, would have to own a substantial part of the U.S. economy. Who would control the fund? Or, whom would the fund control? Few politicians have ever seriously suggested that the trust fund hold assets to make it meaningful because this would raise directly the thorny issue of relations between the public and private sectors. What would happen, for instance, if social security were made up of private annuities invested in the stock market and a "Black Monday" caused a great loss in value? Not only would government be damaged, it might well either feel obliged or be compelled to make up the loss. To the cost of employee-employer contributions, therefore, would be added the price of government emergency funding when the market declined. This is not to say that building up the fund has no effect; it just doesn't mean what it is supposed to mean. Each year's surplus, made up of real cash contributions, reduces the unified budget deficit. It thus would reduce federal borrowing, increase the national savings rate, and, by that logic, increase long-run productivity and growth.
But, as far as cumulating resources over time is concerned, the trust fund is empty. Think of it this way: when, in a given year, there is a shortfall (i.e., payments exceed contributions) the government makes up the difference by borrowing, taxing, or printing money. At some future time, some people think that shortfalls would be made up by cumulation of past social security surpluses. Not really true. Treasury bonds or other IOUs from one part of the government to the other might total in the trillions, but these pieces of paper are not claims on private sector assets. Consequently, the government must still make up a shortfall by borrowing, taxing, or printing money.
Social security is funded annually. Consequently, the actual shortfall in, say, 2035 could be ameliorated only by changing the planned benefits and contributions in 2035. Pretending for the moment that those could be forecast (i.e., that alternative 2B projections were accurate), what would a solution involve?
In 2035, according to 2B, OASDI would represent 6.05 percent of GNP. In 1982 it was 5.16 percent of GNP.[17] What, one wonders, was
the big deal? If the economy were worse, social security would be just one of many problems. True, the actual burden on workers would increase by more than the change in proportion of GNP. The percent of taxable payroll, an increase from 11.78 percent in 1982 to 17.02 percent in 2035, reflects the fact that when there are fewer workers their wages make up (or are assumed to make up) a smaller portion of GNP.[18] Anyone who wanted to deal with these more worrisome taxable payroll figures thus would have to adjust the ratio of beneficiaries to workers. More children or more immigrants (the latter being far easier to supply) would do the trick, but this was beyond the scope of social security legislation. Instead, as Pickle had suggested in 1981, Congress could raise the retirement age.
The whole argument about the long run was decidedly unreal. The problem was not so big; the solution of building up the trust fund was no solution at all; the most relevant factor, the supply of workers, was not even on the policy map. To complete the unreality, the rhetoric of panic largely ignored one area of really frightening numbers, much worse than for OASDI—medicare. Under alternative 2B, Hospital Insurance (HI) costs were projected to rise from 1.30 percent of GNP in 1982 to 3.97 percent in 2035—more than double the increase in OASDI. The increase was 9.1 percent of taxable payroll.[19] Of course, projections of medical costs were unreliable, involving many more factors than pensions do. And, since the short-term crisis for medicare wasn't expected until, say, 1990, that program was not on the table.
Why, one might ask, did the liberals not simply scoff at the long-term problem? Though we can only speculate, there is a reasonable answer: in order to say that the whole argument (about fund solvency and an average shortfall of 1.8 percent of taxable payroll) made no sense, liberals would have to admit that the insurance notion and the fund itself were mythical. That was too big a risk to take. All in all, some ostensible fix, even if it cut benefits from where they would have been thirty years down the road, was safer. Robert Ball explained that public confidence required "a plan that would result in the Board of Trustees saying officially that the program was in full actuarial balance both in the short and long-term."[20]
Although both sides agreed on the size of the problem, they were far from agreed on its solution. The fundamental, liberal-conservative, disagreement was the same as that for the rest of the budget: tax hikes versus spending cuts. Conservatives wanted all the latter, liberals all the former, and centrists a mix. On November 11, Senator John Heinz of Missouri, the most centrist of the Republican negotiators, suggested a 50-50 split. Republicans, however, were avoiding specific proposals. Dole figured that proposals to cut benefits were politically dangerous.[21]
On November 12 the Democrats proposed advancing the 1990 tax increase to 1984, which would add $132 billion; tax increases on self-employed workers would mean another $18 billion. (If you don't have an employer, there is no employer contribution. Rather than have no social security for the self-employed, the system would charge them a higher individual contribution—though not as much, even after the fix, as the regular employee and employer contributions combined.) Two other provisions fit the long tradition in social security financing of increasing short-term revenues by including in the program a new group, which would begin paying immediately but begin collecting only decades into the future. These provisions included (1) federal employees within social security (worth $21 billion) and (2) newly hired state and local employees (garnering another $13 billion).[22]
Coverage expansion, convenient as this sounded, was not so easy to achieve. State and local coverage might turn out to be unconstitutional. If made mandatory, social security contributions might be defined as a tax on state and local governments—explicitly forbidden. Federal employee coverage had been urged by social security administrators and resisted by employees since the 1940s. This group already had a pension system with larger contributions that provided larger benefits. If federal employees had to join social security, they would be contributing to both systems, paying more but not receiving more. Employee unions objected strenuously and had been successful for a generation.[23] The AFL-CIO, a strong proponent of social security in all other matters, had backed its federal unions.
"Extending coverage was a given," recalls one lobbyist, "except for the AFL-CIO, which did not want to admit to its federal sectors that it was going to cave." Labor could live with that so long as the rest of the deal was good enough. The liberals wanted Reagan to sign first before they went to O'Neill; then Reagan would look like he was initiating large tax increases. The offer was rejected.
On Saturday, November 13, liberals tried again, suggesting that state and local coverage be replaced by a three-month delay in the date of the COLA. "Delay" may sound temporary, but it was really a permanent cut because, for three months each year, benefits would be at a lower rate (without inflation adjustment) than they would otherwise have been. Over seven years, benefits would be about $23 billion lower.[24]
The Speaker approved this package, which included concessions from both Lane Kirkland (federal employees) and Claude Pepper (COLAs). Yet the conservatives weren't interested in what was still overwhelmingly a revenue solution.[25] Nor was Rostenkowski who, in an open letter to his House colleagues and the commissioners on November 12, asserted his committee's prerogative to draft a solution and expressed skepticism about a payroll-tax increase during a deep recession.[26]
Just as with the Gang of 17, National Commission members' opinions were secondary to whether those who were not there—Reagan, O'Neill, Rostenkowski—could agree. The principals were not talking, and the commissioners did not have the power to bargain out an agreement. The formal sessions of November 11–13 ended without issue. A last meeting on December 10, 1982, was quick and desultory.
What was needed was not formal negotiations but a way for the two sides to feel out, test, and influence each other. Commission member and nominal Democrat Alexander Trowbridge was a Washington veteran who knew all the players. After the November meetings, he began developing packages, talking to other commissioners, asking their positions, moving toward John Heinz's target of the 50-50 split between revenue increases and benefit cuts. Trowbridge had no commitments from anyone, but he created an impression of movement. That gave Stockman something to take to the president. Stockman asked Reagan if Trowbridge's latest plan—$80 billion in taxes; $50 billion in a sixmonth COLA delay; $32 billion in coverage expansions; and an increase in the retirement long term, to sixty-seven—would be acceptable.[27] Reagan and authorized Stockman to begin secret negotiations with the commission to get them to report such a package.
Stockman had to negotiate with the liberal leaders; then the White House and the Speaker would try to bring along the rest of the commission. In essence, the president's appointees were replaced by the core White House legislative team—James Baker, Richard Darman, Kenneth Duberstein, and Stockman. The negotiations had to be kept secret: because failure might produce red faces; because interested parties might not keep quiet; and because the Democrats generally distrusted Stockman. The White House team began with personal contact between Stockman and Moynihan, his old mentor, and between Darman and Ball, who knew each other from HEW in the early 1970s. Ball made clear that the Trowbridge plan would not do but did not close the door. The old friends, Stockman and Moynihan, were rebuilding trust. When the Senate reconvened, Moynihan enlisted Dole for "one more try" at negotiating a package on behalf of the commission.
On January 4 Dole, Greenspan, and Moynihan met to arrange the secret talks. By adding Conable, Ball, and the four White House representatives, they had included one member from each of the commission's factions—House and Senate majority and minority and the chairman. Members with strong interest-group or ideological positions—such as Pepper, Kirkland, and Armstrong—were excluded. Meetings began the next day.
Because no staff were allowed, Stockman and Ball were the key players. They knew the most. Everyone understood that President Reagan and Speaker O'Neill held the real votes. Each side used the argument
that its own extremists had to be placated, both to force compromise from the other camp and to emphasize that they were all reasonable, in this together.
The meetings could not stay secret if for no other reason than that participants couldn't judge what would play in Peoria without doing some checking beyond the closed doors. The fourteen-million member (according to 1982 records) American Association of Retired Persons (AARP), the AFL-CIO, and such congressional powers as Rostenkowski had to be sounded out. On January 8 Senator Armstrong demanded to be included; Dole had no way to refuse. Armstrong, finding too much support for tax increases, went public and threatened to rally interest groups (including plausibly, the AFL-CIO) against "a massive tax increase."[28] The other negotiators retaliated by truing to distract Armstrong in the meetings and by keeping him away from Ball and Stockman.
Agreement on cost was crucial, particularly in order to put a price on a COLA delay. Stockman, who wanted as big a package as possible in real dollars (to reduce the deficit), jawboned the Democrats into accepting lower inflation assumptions; this meant that any COLA delay would be estimated to raise less. Democrats, therefore, would have to accept other provisions or a longer delay. The Democrats went along—in part because they were not sure they could win a spitting match over assumptions and in part because, on this one, Stockman seemed right.[29]
After agreeing on economics, the National Commission had to reach its target (still between $150 and $200 billion over seven years) in the most politically acceptable way. That meant allocating as little pain as possible in the present, making no part of the package so big that a major player would feel obligated to defect. Whereas Trowbridge had four major provisions, Stockman and Ball moved toward a much more complex package. The parts were smaller; there were a few side payments, provisions benefiting one faction or another so it would complain less; and many provisions were beyond the ken of most nonexperts. Obscurity might reduce the political heat.
"The principle of equality was very important to the White House," Paul Light quotes one Democrat as explaining, "but if we had gone to a strict 50-50 split we would have needed $80 billion in benefit cuts. That was simply not acceptable to us."[30] A six-month COLA delay and a two-stage acceleration of the previously scheduled tax increases yielded $40 billion each, a 50-50 split. But Democrats did not want to yield more benefits. "The key," Light's source recalled, "was to go outside either pure taxes or benefits, and find some new area of agreement." Democrats found it by taxing benefits.
Social security benefits had never been taxed, partly on the grounds that people were just getting their own money back. People who received
social security alone would not be taxed much anyway; their incomes would be too low. Millions of retirees, however, did have other income. The negotiators agreed that
beginning with 1984, 50% of OASDI benefits should be considered as taxable income for income-tax purposes for persons with Adjusted Gross Income (before including therein any OASDI benefits) of $20,000 if single and $25,000 if married. The proceeds from this taxation, as estimated by the Treasury Department, would be credited to the OASDI trust funds.[31]
Because the tax money was to be fed back into the trust fund, taxing benefits really meant a backdoor cut in benefits for more affluent recipients.
Ball and the Speaker had long been willing to tax benefits. Stockman agreed, for reasons of different principle. The provision was worth $30 billion in the short term and more in the long run. Yet taxing benefits for higher, not lower, incomes threatened the premise that social security was a program for everybody. Instead, social security would become an income-transfer program, creating class divisions that might threaten its political support. In the short run, Republicans could object because the tax was a cut to mostly Republican constituencies. It was also vehemently opposed by AARP, largest of the aging lobbies. AARP is as much a giant buying club as a political group; its power would be diluted by the need not to alienate any portion of its fourteen million members (as of 1982). Yet fourteen million is many voters, voters who were better off and more active than most. In short, taxing benefits involved risks; at the least, neither side wanted to lead the way walking that plank. But the risks were worth taking because the provision provided $30 billion that either side could claim as a victory in the scorekeeping of revenue increases versus benefit cuts.[32]
Coverage expansions, COLA cuts, tax-rate increases, and taxation of benefits would yield $130 billion. Another $18 billion was found through increasing the tax on self-employed persons, who previously had paid 75 percent of the combined employer/employee rate. It seemed fair to have them pay the full rate; after all, they got full benefits. But, because corporations deducted payroll contributions from taxable income, thus reducing the corporate income tax, negotiators chose to allow self-employed individuals to do the same. As a result, the $18 billion in extra trust fund revenues would reduce general revenues by $12 billion, as the self-employed claimed larger deductions. The provision was a veiled transfer of $12 billion from general revenues to the trust fund.
The last large part of the package was another disguised general revenue transfer. Before 1957, soldiers had received social security credit for their service without paying into the fund. Current law required
contributions from the general fund each year to cover these "excess" benefits; however, someone came up with the bright idea of crediting all future payments as if they were received together right then and there. The "proper" amount of such payments was, one negotiator recalls, "absolutely arbitrary." It was raised to $18 billion during the last days in response to concerns, particularly from AARP, that the package was not big enough to assure the public that social security was safe.[33] One lobbyist called the "lump sum military wage credit" (the official name) "the tooth fairy." But this money under the pillow was crucial because it was immediate, "a quick one-time infusion to get through 1983–84, to jump-start" the plan. It meant the system was "fixed," in the first year, without cutting or raising taxes. Without it the system would have gone bust in July 1983.
Agreement on the final package was not reached until midday on Saturday, January 15. In addition to the six main proposals, it included a raft of smaller ones.[34] The package, totaling $168 billion for 1983–1989, provided only two-thirds of the long-term target. Although the five core commissioners were agreed, others objected to various parts of the package. In spite of the COLA increase, Claude Pepper went along. His support was essential not only because of his leadership of the elderly but also because, in the new Congress, as chairman of the Rules Committee he would be able to block action. Armstrong and Archer, however, resisted both the tax hikes and the general revenues. They hoped also to get business representatives to object. On the afternoon of January 15, however, President Reagan telephoned Mary Fuller, Robert Beck, and Alexander Trowbridge, strongly urging them to go along. "It was tremendous pressure," a negotiator recalled, and the three complied. After a polite minuet that ended with the Speaker, the president, and the commission simultaneously committed, the commission endorsed the package, 12 to 3. Armstrong, Archer, and ex-representative Waggoner dissented.
For those most interested in reducing the size of government, the package was a defeat. Taxation of benefits did not reduce intrusive government; only the COLA cut was acceptable. It was a victory for those most interested in maintaining benefits. For the needy even the COLA cut was offset by a change in rules for Supplementary Security Income (SSI). In terms of ideology, the Speaker came off better than the president. Yet Reagan was getting the issue off the table; and if benefit taxation were viewed as a benefit cut, then the real cut in benefits ($70 billion) was greater than the increase in taxes ($58 billion). Reagan was less purist, more willing to take the best he could get, than were some of his allies.
From the standpoint of Reagan's aides, the best thing about the deal was its completion. "Once we stopped being revolutionaries and started
being system conservers," a Reaganite recalled, "it was a tremendous accomplishment." "We even cut social security," said one proud Democratic politician. "We just didn't say we were doing it. Everyone agreed to say we weren't doing what we were doing." They were showing that they could govern, that reasonable legislators (themselves) could find solutions to difficult problems. But the fact that they felt unable to tell the public what they were doing did not bode well for the future.
All that remained was convincing Congress to go along. First, this meant House Ways and Means. A strong endorsement from the committee would help a lot; a weak or divided endorsement would hurt. As one committee source put it, "the deal was ours to screw up." A fast track was needed to have the agreement in place by July, preferably even before Easter recess. After that came budget time—no time for other problems. Public opinion was confused: mildly negative on the individual provisions, mildly positive on the package.[35] Interest groups, equally confused, were also defensive.
Unlike lobbies for the aging, federal employee groups had little to gain from the package; unlike business, they had few comparable interests that might induce moderation so as to preserve long-term relationships. Indeed, the federal employee lobbies, particularly the powerful postal unions, had few relations with Ways and Means; they would get no satisfaction there. Instead, the coalition of federal retirement lobbies, Fund to Assure an Independent Retirement (FAIR), announced a $3 million lobbying campaign whose goal was to win separate votes on employee coverage on either House or Senate floors. As the 1980 and 1982 reconciliations showed when COLA cuts were defeated, the CSR supporters had a lot of clout.
Rostenkowski at Ways and Means could rely on Pepper at the Rules Committee, backed by the leadership, to protect him from FAIR.[36] Rosty's real problem, one source recalls, was "a fundamental difference between conservatives and liberals on the long-term solution." Pickle wanted to add his own plan to the commission package, raising the retirement age to sixty-seven. Pepper, supported by the AFL-CIO whose industrial and craft workers strenuously objected, wanted a tax increase in the year 2015. "We knew if we went one way Pepper wouldn't allow it on the floor, but Pepper's version wouldn't pass on the floor."
Pickle's subcommittee could not give the package bipartisan backing because its Republicans were led by Archer, who had already dissented from the commission report. The subcommittee reported the bill by a 7 to 4 margin. In full committee, the package was attached to proposals from other subcommittees involving medicare, unemployment insurance, and SSI.
The Ways and Means health subcommittee proposed and full committee
accepted a major reform of medicare. The Health Care Financing Administration would establish a schedule of "diagnosis-related groups" (DRGs); each DRG would establish a set fee for treatment for each diagnosis, rather than continuing to pay the hospital for activities, (surgeon's fees, tests, number of days in the hospital), as in the traditional "fee-for-service" system. Setting government costs, and thus "prospective" hospital income in advance, would give hospitals incentives to control their costs; thus, the government would have a way—the payment schedule—to regulate its costs. How a few far-away federal administrators would triumph over many closed-by doctors in monitoring millions of transactions was not specified.
Clearly, prospective payment was intrusive government regulation. But, in TEFRA, Congress had ordered the Health Care Financing Administration to develop a plan; and the administration decided that this system was its best hope for cost cutting. Time described it as "the only element [of administration proposals] that does not automatically send temperatures soaring."[37] Liberals refer regulating hospitals to cutting benefits. And everybody involved knew that medicare's financing problems, though less immediate than those of social security, were in fact more severe. Therefore, the politician effectively conspired to shortcircuit objections from medical providers by attaching prospective payment to the social security rescue before the providers could mobilize opposition.
Prospective payment sailed through Ways and Means. The full Committee gave the GOP a victory by removing a plan by Pickle to allow general revenue subsidies if OASDI got into future trouble. The major issue was what to do about a long-term solution. The full committee faced the same Pickle-and-Pepper dilemma as had the subcommittee. Rostenkowski wanted Pickle's solution, but he had to get past Pepper. Rosty and his allies convinced the Speaker, and Pepper, to leave it up to a floor vote. Let each offer his plan, devoid of sweeteners, with Pepper having the advantage of going last. It was explained to us, by someone we'll call Peter Piper:
We had to do a massive selling job to convince both Pickle and Pepper to trust us. We knew how it would turn out, that Pickle would win. But both sides thought they could win. The Speaker thought the liberal side would win…. Otherwise we would never have gotten a vote…. We had 35 members, and the Republicans kept insisting on a vote in the committee. They kept saying, "We have 20 votes." Rosty kept saying, "No, you have more, you have me and some others, but if we adopt it in committee we'll never get a rule."[38]
On March 9, Pickle won a decisive victory. His amendment carried (228 to 202), supported by 152 Republicans and 76 Democrats. Pepper's
failed (132 to 296), despite impassioned support from both the Speaker and Pepper whose speech received a standing ovation. The retirement age would increase from sixty-five to sixty-six, in two-month steps, from 2004 to 2009, and to sixty-seven, again in two-month steps, from 2022 to 2027. All Ways and Means Republicans, and a bare majority of Democrats, supported Pickle. Many conservative Republicans, who intended nonetheless to vote against the full package, voted here for what they considered a change for the better.[39] This was not a tax.
After he lost on the long-term provisions, Pepper could no longer endorse the package. The Speaker, even as he supported Pepper's amendment, called on his troops to pass the bill in the end; and it did pass, 282 to 148. The center had held.
Next came the Senate. On the long-term issue, Finance adopted (13 to 4) Heinz's proposal to cut benefits by 5 percent in 2000 and to raise the retirement age, in small steps, to sixty-six by 2015. The Committee rejected (11 to 6) Russell Long's motion to have new federal employees not included in social security until creation of the new, supplementary civil-service pension system. Long was backing employee advocates who had argued that new employees would be paying into both social security and the old civil-service retirement, without being able to benefit from both. Dole maintained that any delay would only help unions in the fight for their real goal—repeal of the coverage extension. Finance adopted an assortment of small changes, enough to satisfy Armstrong, and the committee reported its bill out (18 to 1) on March 11.
The Senate moved quickly to consider the package, only to meet a roadblock thrown up on an extraneous issue: withholding income tax for interest and dividend earnings, a major part of the 1982 TEFRA revenue package. The banks (we will discuss this in a subsequent section) had whipped up a fire storm of demands that such withholding be repealed. John Melcher (D-Mont.) tried to amend the social security rescue to delay withholding by six months. Dole was furious, but Melcher wouldn't budge. At this point, the president, Republican leaders, and the AFL-CIO (particularly concerned about the unemployment benefits attached to the bill) ganged up to break the logjam. On March 22, Vice President Bush, presiding, ruled that Melcher's amendment, by reducing revenues, violated the revenue targets of the FY83 budget resolution and was therefore out of order. Perhaps because the issue had been redefined—but more likely due to pressure from all sides and to a suspicion that withholding could be repealed later—senators rejected Melcher's motion to waive the budget act, 54 to 43.[40]
As is its tradition on tax bills, the Senate passed a series of fairly insignificant amendments. One, however, was important. On March 22, Russell Long offered his amendment to delay coverage of new federal
employees until a new supplementary benefit plan was established. A frequent ally of the employees, Ted Stevens, whose subcommittee would have to draft that plan, objected; so did Dole. But the FAIR coalition represented six million angry citizens who did have a point; and some of them, the postal workers, were politically very active in every state. Long's amendment passed on a voice vote. The Senate passed its bill (88 to 9) on March 23.
"Despite the separate victories," Light reports, "there was one 'small" problem: neither bill could pass in the other chamber."[41] That is a bit extreme; both chambers wanted a bill, and, compared to no bill at all, the differences might not look so big. But there was a series of disagreements, ranging from the long-term solution to federal employee coverage, that could be compounded by institutional jealousies. Rostenkowski wanted to "win" the conference, and, had Dole felt the same way, things might have been more difficult.
The conference was also up against time pressure. Congress was due to recess for Easter on March 24, the end of the day the conference began. Bargaining went on for most of that day. The Senate gave in on the long-term retirement fix and, rather willingly, on federal employees. The toughest issue was a "fail-safe" for the short-term solution. The House rejected automatic COLA cuts. The Senate insisted that the House's provision—making the COLA the lower of wages or inflation if reserves fell below 20 percent after 1987—was too little and too late. In the final agreement of the conference, the House provision was set to be effective at the beginning of 1985, and the "trigger" was lowered to 15 percent. The COLA change, therefore, would be announced before the 1984 election—dangerous politics.
With heavy Democratic support, the House passed its conference agreement (243 to 102) late on March 24. The Senate's deliberations ran less smoothly. Some senators sensed an upset when both Armstrong and Long announced their opposition; the House's victories now eliminated the compromises that had won their support. But leaders of both parties rallied, arguing that they could not afford a failure. "I do not like the bill," said Minority Leader Byrd. "I wish the problem would just go away." But defeating the bill would not accomplish that. "This is not a perfect bill," Howard Baker declared, "but we are not a perfect body."[42] Resting on the safety of imperfection, the Senate passed the conference agreement, 58 to 14. At 2:00 a.m. on March 25, the Senate adjourned for Easter. If not altogether resurrected, at least social security had been saved.[43]
The rescue was a triumph of governance. The politicians had shown they could act to avert a crisis, that reasonable people of the center could prevail over polarized, divisive forces that cared more about spending
or fending off taxes than about saving the system. The social security issue seemed to have taught a lesson: imminent crisis would force responsibility.
The social security package, on balance, exemplified government. Yet it did not imply that the overall federal deficit problem, as distinguished from social security alone, could be solved. Let us review both the problem and the solution to see why.
Consider the short term. First, the package barely dented the budget deficit; of its $168 billion, over seven years, only about $120 billion, that is, $18 billion a year, reduced a federal deficit that was ten times larger. The balance were transfers from general revenues or account maneuvers: the "tooth fairy" military accounting credit ($18 billion); the partial self-employment tax increases that would come out of income tax receipts (but offset by tax credits); and federal employee coverage (it was not generally understood that new employees who would be contributing to social security would no longer be contributing to the Civil Service Retirement Fund, which actually had a higher—7 percent—contribution requirement). The package amounted to less than half of TEFRA. If we count federal employee coverage as a real step, the total still came to around 10 percent of social security liabilities. The federal deficit was about twice as large a part of federal policy commitments as the social security shortfall was of its overall commitments. The social security problem (and solution) was much smaller than the deficit problem .
Yet there was a problem, and they had solved it. We have argued, contrary to much scare talk, that the long-term problem actually was not so big. If the economy, with its demographic multiplier, really went into a tailspin (alternative 3), the system could not be financed; but under those circumstances there would be much trouble in other areas long before social security hit the wall. Given mediocre economic performance (alternative 2B), the problem was manageable—except that the standard notion of how to manage it, building up the fund, was meaningless because the fund's assets, government bonds, were no more than a promise to tax or borrow. Raising the retirement age addressed the real problem—the demographic balance between workers and retirees. Taxation of benefits meant that by 2030, with inflation, most benefits would be taxed—which meant, in fact, they would be reduced. Politicians had addressed the long-term "problem" of underfunding by changing the long-term promise, that is, by reducing future benefits.
We have said that budgeting is about commitments: what kind and how many you make and, in a time of deficits, what kind and how many are broken. The contrast between the short-term and long-term fixes of social security reflects a natural tendency to try to keep commitments on which people have relied for years and a greater willingness to change
commitments when there is a long time—here, a generation—for recipients to adjust to the change. The "tooth fairy" may have been fake; but it looks more respectable if we recognize it, not as a long-term commitment of general revenue, but as a device to maintain existing commitments that very few Americans wanted to abrogate. If budgeting is conceived only as a matter of income and outgo—the budgeters' perspective—then the "tooth fairy" is a scam. But it looks more like statesmanship if we see the task of budgeting as choosing commitments so that members of a society can live together.
The charge of statesmanship on social security we have leveled against our national politicians is disputed on two quite different grounds. First, social security now and for years to come contributes its annual surplus to reducing the deficit. Evidently, this is a worthy act. Why, then, should the social security program not only be exempt from attacks on the deficit but also have its surpluses credited against its future shortfalls? A good reason, which we have previously discussed, is that social security taxes do have an impact on the deficit because they reduce willingness to pay other taxes levied on the same individual. Unless the right ear pays income taxes and the left foot social security, so the person who pays has no glimmer of the total tax burden, social security is part of the deficit problem.
Furthermore, proponents of the "social security is not to blame" thesis talk as if they wanted to eliminate its pay-as-you-go feature in order to credit early surpluses, when the system starts and people pay in but do not take out, to later deficits, when more people receive than contribute. By doing this, however, the conflict between generations that has been muted by political statesmanship is once again highlighted. The difficulties involved in departing from pay-as-you-go are emphasized by the other major objection: the quick fix will lead to unacceptably large increases in social security taxes in the first half of the twentieth century.
The debate over social security has been enlivened by contributions from free-market advocates, especially Peter J. Ferrara, who wish, in essence, to privatize the system. Their arguments are cogent, even illuminating, although we do not find them compelling. It is true that as social security taxes rise and fewer contributors support a larger number of beneficiaries, due to the aging of the population and the maturing of the system, combined with its increasingly redistributive character, a significant minority of contributors will take out less than they put in.
It is also true that black people, because they do not live as long as whites, take out considerably less. We would stipulate further that calling social security a middle-class subsidy is a misnomer; after all, if everyone were given a chance to opt out, people of middle income might well be
better off financially investing in a private annuity. What neither complaint addresses is whether, given the usual interplay of politics and economics, there is a remedy that is not worse than these maladies; we doubt it.
It is clear to everyone that social security cannot be allowed to default. This means that current and (likely) increased future contributions will continue. Where, then, will the money for private annuities come from? Ferrara has the only idea—income tax credits. While in the past a rescue of social security has often been proposed that would pay beneficiaries out of general revenues, it is now proposed that people be enabled to get out of social security via this subsidy.[44] The deficit, to say the least, would have to decline substantially before this proposal even had a chance to be considered. Should it pass, moreover, a program of phasing in private annuities as a substitute for social security would depend upon the march of the markets. Deep downturns could hardly be tolerated. The temptation to interfere with markets would increase. Making retirement depend on avoiding deep market fluctuations does not seem to us a good way of promoting capitalism, which requires a willingness to face substantial loss.
It is likely that social security taxes will rise in the next century, as they have in this one. In thirty to fifty years, people may not mind so much paying a higher proportion of a much larger income. Or encouraging immigration may prove more attractive. For the time being, however, our politicians have maintained the viability of government's most popular program. This is not bad, especially absent evidently superior alternatives.
Sometimes commitments conflict: a commitment to provide certain benefits, for example, counters a commitment to balance budgets. In the social security case, however, commitment to maintain benefits and commitment to fiscal solvency went together: there could be no benefits without solvency. Thus, even the most radical of the program's supporters, Claude Pepper, would, if necessary, give a little on benefits. And solvency had no independent value of its own. Contrast that with the wider budget dispute, that saw "balance" as an independent value, and in which arguments claiming that benefits depended on balance were at best long-run and dubious.
The conflict of commitments was much less severe in social security than in the federal deficit. The short-term social security fix was an effort to maintain existing commitments. Any solution to the overall federal deficit required breaking many such commitments—whether by cutting defense effort, or slashing the poor, or raising taxes substantially, or changing the social security promise. Indeed, it would be nearly impossible
to cut social security in later budget-deficit packages because the argument that benefits were being cut to preserve them would no longer make sense. Politicians talked about "waste, fraud, and abuse" because that implied that it was possible to keep the promise to balance the budget without breaking other promises. It wasn't true then, and it isn't true now.