We should not be surprised if Congress does nothing to fix Social Security before 2034, when the trust fund runs dry. Although experts first identified the long-term solvency problem nearly three decades ago, and opinion surveys have repeatedly shown that fixing the problem is one of the public’s top priorities, legislators have never voted on a proposal to fix it — not in committee, not on the floor, not in the House, not in the Senate.

Congress failed to act in the late 1990s, when some officials proposed using the growing budget surplus either to strengthen traditional Social Security or to jump-start the creation of individual accounts. It failed again in 2005 when President Bush proposed partial privatization. And the principal reason for congressional inaction is clear: insolvency is a long-term problem without short-term consequences.

Everything will change in 2034. Suddenly, insolvency will become an urgent problem with enormous consequences. Absent congressional action, an estimated 83 million Social Security recipients — 18 million more than today — will face automatic benefit cuts of 21 percent. Another 8 million people filing for Social Security benefits that year will face similar reductions from what they would otherwise collect.

At the same time, the threat of insolvency will become an urgent political problem for legislators. If they cannot find ways to block or reduce these enormous cuts, legislators will face hostile voters at their next election.

Congress has five options. One is to rerun the 1977 or 1983 playbooks, choosing from the incremental solutions. Another is to adopt some variant of full or partial privatization, creating an advance-funded, “defined-benefit” program for younger workers and using debt to fund current and near retirees. A third is to use general funds to pay benefits, breaking the century-old restriction that requires Social Security benefits to be funded exclusively with Social Security taxes. A fourth is to adopt a short-term fix designed to get legislators past the next election, perhaps using general funds or debt. A fifth is to do nothing, forcing Social Security benefits to shrink immediately.

Incremental Solutions

Twice, legislators have enacted what they thought were long-term solutions to Social Security’s solvency problems. That the 1977 fix lasted six years, while the 1983 fix is on track to last for more than a half century, is more a consequence of happenstance than choice. And both reforms offer practical lessons for future policymakers.

In 1977, the challenge was surging inflation. The first three automatic cost-of-living adjustments, which commenced in 1975, increased annual benefits by 8.0, 6.4 and 5.9 percent. Unfortunately, Social Security’s revenue stream was not growing fast enough to support those increases. As insolvency loomed, legislators explicitly chose to protect all beneficiaries — current and future — by raising additional revenue. They increased the payroll tax rate by one-quarter phased in over 13 years, and increased the maximum taxable wage base by 68 percent, phased in over three years.

When the threat of insolvency returned in 1983, legislators chose a different path. They reduced Social Security benefits by raising the retirement age, delaying cost-of-living adjustments, and taxing first upper-income beneficiaries and later middle-income beneficiaries. They also increased tax revenues by accelerating already-scheduled rate increases and eliminating the preferential rate for selfemployed workers. Benefit cuts were seven times greater than tax increases.

Why did legislators protect beneficiaries in 1977 but favor taxpayers in 1983? One explanation focuses on agenda-setting. In the first instance, Democrats controlled the House, Senate and presidency. Exclusive control gave Democrats the ability to place their own preferred options on the decision agenda. In the second instance, Democrats controlled the House, while Republicans held the Senate and presidency. Shared control allowed a wider range of alternatives to reach the decision agenda.

A second explanation focuses on changing policy preferences among officeholders, particularly among Republicans. David Mayhew, a political scientist at Yale, has shown that the vast surge of liberal legislation that began under President Johnson continued unabated under Presidents Nixon and Ford, with support not only from these presidents but also from Republican legislators. These were the legislators who supported the 1977 solvency plan.

President Reagan’s 1980 election abruptly ended executive branch cooperation by placing a tax-cutter and Social Security critic at the head of the table. Around the same time, Republicans in the House and Senate began their decades-long shift to the right, as measured by floor votes. There was no way that a genuinely conservative president and a group of increasingly conservative Republican legislators would approve a solvency bill that relied exclusively on tax increases.

How policymakers decide to fix Social Security in 2034 will depend partly on whether one party controls all three institutions (House, Senate, presidency) or whether the parties share control. If there is unified Democratic control, say a 2030s version of President Obama, Speaker Nancy Pelosi and Senate Majority Leader Harry Reid — the agenda setters for the Affordable Care Act — we should expect different choices than if there is unified Republican control. Of course, if there is divided party control, we should expect choices somewhere between these two extremes, since the two parties would share control of the agenda.

The interesting question is what those two extremes would look like. Can we imagine that Democrats, if they control all three institutions, would repeat what their 1977 colleagues did — namely, approve a tax-only solvency bill, thus shielding current and future beneficiaries from cuts? Can we imagine that Republicans, if they control all three institutions, would stick to their signed pledges never to raise taxes, thus closing the solvency gap exclusively with benefit reductions?

Republicans have the tougher row to hoe. Although cutting Social Security benefits is not an impossible quest, as the 1983 reforms demonstrate, Congress imposed most of the 1983 cuts on temporally distant cohorts, not current retirees. Recall that legislators raised the full retirement age from 65 to 67, but implemented the change very gradually starting with people who were then 45 and ending with those who were then 23. The only cost that legislators imposed on current beneficiaries was a six-month delay in cost-of-living adjustments, a small price to pay for avoiding automatic benefit cuts that would have been larger.

A previous chapter outlined the political logic: impose modest cuts on current beneficiaries because they would swiftly punish legislators for anything too painful, and impose large cuts on future beneficiaries, because future reductions would not enrage voters as much as current reductions would.

But Republicans cannot use the 1983 playbook in 2034. At that point, policymakers will require something like 1.2 percent of GDP annually to forestall the 21 percent benefit cuts. They cannot raise such a colossal sum by adopting long-term solutions like raising the retirement age, because long-term solutions produce very little immediate revenue. Moreover, short-term solutions, like delaying costof- living increases or reducing benefits for affluent retirees, would produce only a fraction of the revenue required to keep benefits flowing. Put differently, benefit reductions can be part of a solvency plan, but alone they cannot prevent automatic benefit cuts.

Republicans have painted themselves into a corner. Most have signed pledges that they would never vote to increase taxes. Many have promised current retirees that they would do nothing to threaten their benefits. For three decades, they have been able to keep both promises. But in 2034, the two promises will become incompatible. If Republicans stick to their pledges — no new taxes — many of their constituents will be stuck with massive benefit cuts. Moreover, these cuts will be directly traceable to legislators’ actions or inactions.

Finally, Republican legislators will have to choose. Will they support their tax-averse donors and constituents, who might punish them if they violate their signed pledges? Or will they support Social Security beneficiaries — often large fractions of the active voters in their districts — who might punish them for allowing large benefit cuts to take place?

Decades of procrastination are therefore particularly damaging to the Republican antitax cause. Perhaps if Republicans had more wisely used their unified control of the House, Senate and presidency in 2005, they could have enacted gradual benefit reductions — say, slowly raising the retirement age and slowly shrinking cost-of-living adjustments. Those sorts of things might have closed the longterm solvency gap without having to raise taxes. In 2034, however, gradual benefit cuts are useless. Promising to be frugal in the future is no answer when the cupboard is already bare.

Democratic legislators face no such quandaries. Although many Democrats have promised to preserve all benefits for current and future beneficiaries, they have never pledged to do so without raising taxes. Indeed, many of them have sponsored or cosponsored bills that would both raise the maximum taxable wage base and increase the tax rate. To be sure, some legislators might get cold feet in 2034, perhaps pressured by their affluent constituents to pare back the proposed tax increases. But Democrats have no signed pledges constraining their actions. For them, the only question is how to allocate costs among various types of workers, employers and beneficiaries.

Decades of procrastination are also less damaging to the Democratic cause. If they had acted in the 1990s, say by gradually increasing both the tax rate and the wage base, they could have captured substantial revenue from the baby boomers, salted it away in the trust fund and thereby lessened the overall burden of supporting the boomers in retirement. Additional tax increases would still be needed to deal with lengthening life spans, but those tax increases would be much less painful than those that are required to fund the boomers’ retirement. In short, earlier actions by courageous politicians would have made later actions less expensive.

Despite the broad support among DemoThird Quarter 2022 79 tk cratic legislators for tax increases, they are unlikely to have enough votes to enact those increases without Republican support. Although the House is a majoritarian institution, the Senate requires a supermajority — 60 out of 100 senators — to break a filibuster. Sometimes Congress uses the annual budget resolution, where simple majorities suffice, to enact tax or spending bills, thus avoiding the filibuster. But the Congressional Budget Act, as amended in 1990, explicitly prohibits using budget resolutions to change anything about Social Security. It is the only federal program to have such statutory protection.

In short, changing Social Security is necessarily a bipartisan affair unless one party controls the House and the presidency and has a supermajority in the Senate. Such single-party dominance is rare. It has occurred for fewer than six months in the past four decades.

What is essentially a statutory requirement for bipartisanship profoundly shapes the politics of Social Security. Republicans could not have enacted partial privatization in 2005 without the support of at least five Democratic senators. Democrats cannot enact a tax-only solvency package in 2034 without the support of at least some Republicans.

To be sure, controlling the agenda still affects the outcome. Although Democrats may not be able to enact their most preferred plan with unified party control, at least they could begin with something closer to that ideal and then gradually modify the plan as they work to attract Republicans. That is likely to be more satisfying for Democrats than if Republicans have the first-move advantage.

This statutory requirement for bipartisanship essentially weeds out extreme proposals. For example, the 2019 Larson bill, which nearly 90 percent of House Democrats cosponsored, includes tax increases equal to 3.7 percent of taxable payrolls — almost 1 percentage point higher than necessary to close the long-term deficit — and proposes using the extra revenue to enhance various benefits.

Although I can imagine that some Republican legislators in 2034 might accept some tax increases in order to forestall benefit reductions, just as they did in 1977 and 1983, they are unlikely to accept a tax-only solution that is accompanied by even higher benefits. More likely, Republicans would expect some benefit reductions in exchange for their support.

One reason bipartisanship is common for most major enactments is that legislative leaders seek to enact durable laws. For less durable victories, executive orders often suffice. Notice how brand-new presidents often issue a flurry of executive orders that simply reverse orders promulgated by their predecessors. But we seldom see such reversals on Capitol Hill, not just because the filibuster limits what bare majorities can accomplish, but because legislators enact most major bills with oversized majorities. In fact, political scientists James Curry and Frances Lee show that legislative leaders working in today’s polarized environment are just as likely to build large bipartisan majorities as did leaders working in less polarized times.

Leaders are especially likely to seek large bipartisan majorities when they must impose painful costs. If Democratic legislators somehow found a way to pass the Larson bill with a purely partisan coalition — perhaps by eliminating the filibuster as they did in 2013 for most judicial nominations — their party would then own the tax increases. Republicans could attack them forever for increasing taxes so much.

By working to forge a bipartisan consensus, however, coalition leaders may not obtain the bill of their dreams. But they can enact legislation that binds both parties to the costs, thus protecting both sides from dagger-pointing.

Mitch McConnell, the top Republican leader in the Senate since 2007, has long claimed that reforming entitlement programs like Social Security is not politically feasible when a single party controls the House, Senate and White House. It is only when the parties share control of government that they can protect themselves from the partisan attacks that would otherwise follow from one party increasing taxes or cutting benefits.

Curry and Lee also demonstrate that picking off a few moderate legislators from the other party’s rank and file is more difficult today than it was before the parties became so polarized. One reason is that each party is more cohesive. Another is the near disappearance of moderates. Today’s party leaders, rather than negotiating with a small group of moderates from the other side, negotiate directly with the other party’s leaders. These negotiations, when successful, tend to produce very large majorities, often including the bulk of each party’s members.

Privatization

The idea of allowing workers to divert some portion of their payroll contributions to individual investment accounts once attracted bipartisan support in Congress. It no longer does. When George W. Bush made individual accounts the centerpiece of his 2000 presidential campaign, prompting his opponent, Al Gore, to oppose the proposal both during the debates and in a barrage of political advertisements, privatization became a partisan issue.

Today, Democratic legislators regularly reject such proposals. Meanwhile, Republican legislators advocate “market solutions” to the solvency problem, although carefully avoiding the now-toxic term “privatization.”

It is tough to know how committed Republican legislators are to enacting privatization (here used as a neutral descriptive term). Of course, some wealthy Republican donors are deeply committed to reinventing Social Security, so the incentives are strong for legislators to parrot support for the cause. But campaign talk is cheap talk. It does not bind legislators to work actively to advance privatization. It does not compel them to vote for privatization if given the opportunity.

As we saw in 2005, when President Bush campaigned tirelessly for individual accounts, even the most conservative legislators were fair-weather friends. They spoke glowingly of giving workers control over their payroll contributions, but they did little to advance the president’s plan. Even today, Republican legislators rarely introduce bills that would create the market-like solutions they regularly praise. Former representative Paul Ryan introduced the most recent privatization plan more than a decade ago.

One reason for the dearth of recent privatization proposals is that the costs of transitioning to a new system increase steeply as insolvency nears. Before President Bush proposed individual accounts in 2005, actuaries expected the trust fund would last until 2042. Program designers, therefore, had a long lead time to gradually restore solvency. If such a plan had passed, Social Security finances would have turned worse for a while, as younger workers diverted their payroll contributions to individual accounts and the government borrowed to pay benefits to current and near retirees. But eventually the system would have returned to balance, as benefit cuts for temporally distant cohorts kicked in.

As 2034 nears, however, the need to borrow to jump-start privatization will increase dramatically. First, the government would no longer have a large trust fund to help subsidize benefits. That loss alone would require borrowing an additional 1.2 percent of GDP annually just to keep benefits from declining. This new borrowing would be in addition to the massive borrowing required to cover the diversion of workers’ contributions to individual accounts.

Second, gradual changes in the benefit formula would no longer have time to work their magic. Third, the entire boomer generation would already be collecting retirement benefits. Rather than having young boomers transition to individual accounts, as they would have under the 2005 plan, traditional Social Security would have to support all boomers until their demise. In short, delay is very costly for the privatization cause.

Privatization still offers Republican legislators one tempting treat: no new taxes. For politicians who have signed the Norquist pledge, that treat could still prove irresistible. Yes, the government would have to borrow many trillions of dollars beginning in 2034. But massive borrowing did not dissuade Republicans from supporting President Bush’s 2001 tax cuts, the Bush-era wars in Iraq and Afghanistan, or President Trump’s 2017 tax cuts. If Republicans did not mind borrowing for war-making and tax-cutting, they might not mind borrowing to avoid the political nightmare of choosing between their tax-averse supporters and their benefit-dependent constituents.

Still, the devil is in the details. Some early privatization plans proposed that the government borrow for many years to support current and near-retirees, and then repay the debt as the costs of traditional Social Security declined. But as President Bush discovered in 2005, devising politically acceptable ways to reduce future benefits, and thereby repay the debt, is fiendishly difficult. What was difficult then will be even tougher in 2034 because the revenue needs will be even greater.

Also problematic is the notion that privatizers can somehow guarantee benefits for current and near retirees. AARP and other opponents will surely remind voters that Republicans have a troubled history when it comes to extending the government’s borrowing authority, sometimes demanding cuts in domestic programs as the price for raising the debt limit to avoid government default. The worry is not just that Republican legislators might use future votes on debt extensions to demand reductions in Social Security benefits. Republicans might also insist that the ballooning debt — ballooning because of privatization — necessitates deep cuts in health, education, food stamps and the like. Party reputations are sticky: it is tough to break free from them.

Finally, Republicans would face the problem of attracting Democratic support. Bipartisanship is just as much a requirement for Republicans’ privatization proposals as it is for Democratic plans to increase taxes. Here the problem is that Democratic legislators have few reasons to abandon traditional Social Security. Still, if Republicans control the agenda, maintain unity among themselves, block every counterproposal and wait until moments before the trust fund runs dry, they might attract the handful of Democrats they would need.

But maintaining Republican unity to dismantle one of the government’s most popular programs is far from assured. Republican legislators would face the same constituency pressures to save traditional Social Security as Democratic legislators would.

The fundamental problem for the privatizers is that traditional Social Security is popular. It enjoys broad public support among Democrats and Republicans, workers and retirees, the young and the old, the poor and the better-off. In order to replace it, the proponents of change need to make the replacement program popular, too. Only then would legislators feel comfortable voting to reinvent Social Security.

President Bush understood this, as he mounted his “60 stops in 60 days” national tour. But he was unable to move the needle because the opposition enjoyed too many advantages. Democrats had been the more trusted party on Social Security for decades. AARP had direct channels for reaching and mobilizing its 35 million members. Not much has changed since then, except that Republicans no longer have a national leader with a policy fervor for individual accounts.

General Funds

A third option is to use general funds to close the solvency gap. General funds are funds derived from all government revenue sources, excepting only those derived from special purpose taxes such as the payroll tax, which supports the Social Security and Medicare trust funds, or the gasoline tax, which supports the Highway Trust Fund. General funds include revenues derived from broad-based taxes, such as corporate or personal income taxes, as well as borrowed funds that are used to close the gap between overall revenues and overall spending.

The political attraction of using general funds is that it separates voting on benefits from voting on taxes. Legislators can claim credit for delivering benefits without being blamed for imposing costs. When spending general funds, all questions about costs are deferred to the future. Indeed, given the federal government’s growing propensity to borrow rather than tax — it has run large annual deficits since 2002 — questions about costs are often deferred indefinitely.

There are ample precedents for using general funds to subsidize programs that have their own dedicated taxes. Congress once supported highway construction exclusively with the gasoline tax. But since legislators last raised the gasoline tax in 1993, they now supplement this dedicated tax with general funds. Since inception, Congress has funded Medicare with a combination of general funds, a dedicated payroll tax and user premiums.

Policymakers have had good reasons for not using general funds to subsidize Social Security. President Roosevelt argued that a tight link between taxes and benefits served two important ends. It would protect Social Security from hostile actors — “No damn politician can ever scrap my Social Security program” — but it would also protect the program from unreasonable expansion. Legislators could not expand the program unless they were also willing to increase taxes.

This tight link has worked for nearly a century. The program’s detractors have never found a way to dismantle Social Security because workers earn their benefits by paying a dedicated tax. But neither have the program’s champions been able to expand benefits since 1972 because legislators have been unwilling to increase taxes.

Using general funds in 2034, however, would offer several attractions. First, it would protect Republican legislators from having to vote to raise the payroll tax, which would violate their pledge never to raise taxes. Second, it would help Democratic legislators move the program’s revenue base from the regressive payroll tax to the progressive income tax, thus making the program more equitable. The sticking point is that Social Security’s revenue needs are colossal — 1.2 percent of GDP.

Would legislators be willing to borrow that sum, year after year, essentially taking on longterm debt to pay Social Security’s monthly obligations?

Of course, the general funds alternative to massive borrowing is to increase income taxes. That alternative makes sense for many Democrats, who prefer progressive taxes. But it hardly helps Republican legislators, who oppose raising both income taxes and payroll taxes.

In short, general funds are no panacea. They might help coalition leaders close a portion of the solvency gap, but only in conjunction with other revenue-raising or benefit-cutting provisions.

Stalemate

A fourth possibility is stalemate. If Democrats and Republicans cannot find common ground as the clock ticks down, program administrators would have no choice but to reduce everyone’s benefits by 21 percent. That is the law. Experience suggests, however, that legislators would find a short-term fix to keep the benefits flowing. That is what they did in 1981, when they authorized the Social Security trust fund to borrow from other trust funds, thus pushing insolvency past the 1982 congressional elections and making 1983 the year for action. That is what legislators regularly do when they cannot agree about the rest of the federal budget. They adopt shortterm continuing resolutions that keep funds flowing, usually at the previous year’s level.

But borrowing from other trust funds would not last very long in 2034, given Social Security’s enormous need for immediate revenue. Moreover, experts forecast that the Medicare trust fund will be depleted long before then, leaving the retirement funds for military and civilian personnel as the only government trust funds with significant assets.

If stalemate continues, the next stopgap would be using general funds to pay Social Security benefits. Although using general funds for Social Security is currently illegal, Congress could suspend that prohibition. In short, policymakers could buy as much time as they need to enact long-term changes — if majorities, or in the case of the Senate, supermajorities, agree to extend the clock. Eventually, one side or the other will conclude that actual benefit cuts are necessary to spur serious negotiations, and refuse to approve additional extensions. At that point, the first three options are back on the table.

A fifth possibility is complete policymaking breakdown. Legislators prove unable to enact either a short-term fix or a long-term solution. As the law requires, program administrators would then cut everyone’s benefits by 21 percent. Next, legislators would take their cases to the people, arguing why their own plans — or their party’s plan — were superior, and why citizens should reelect them and their party teammates to fix Social Security. Few legislators would welcome an accountability election like this, where swarms of ill-tempered seniors are out for revenge. And that is why legislators are likely to fix Social Security moments before it plunges over the solvency cliff.

The Finale

No one can be certain what Congress will do in 2034. We don’t know what role Social Security will play in the presidential and congressional campaigns leading up to trust fund depletion, although the issue is likely to rise in importance as insolvency nears. We don’t know which party will control the House, the Senate or the presidency, and therefore whether a single party will control the agenda. We don’t know whether the president or top congressional leaders will care more about enacting a specific solution — say, partial privatization — or whether they will simply want to fix the solvency problem without endangering legislators’ careers.

All else equal, incremental solutions are more probable than reinventing the system. First, traditional Social Security enjoys enormous public support. Retirees like it, but so do workers. Democrats like it, but so do Republicans. The poor like it, but so do better-off folks. Only the uber-wealthy seem to detest it.

Although creating broad public support for some type of privatization is certainly possible, it is hard to imagine a worse time for doing so than during the weeks before Social Security plunges over the fiscal cliff. Its 83 million beneficiaries, as well as millions of soon-to-be beneficiaries, will be looking for reassurance, not fresh beginnings.

Second, many of the incremental solutions are quite popular. For example, most surveys show that large bipartisan majorities favor increasing or abolishing the maximum taxable wage base. Adopting either alternative would be quite lucrative. Abolishing the wage base would eliminate nearly three-quarters of the long-term actuarial deficit; merely restoring the wage base to where it was in the early 1980s would eliminate one-quarter. When Medicare was in crisis, legislators, desperate for revenue, first raised and later eliminated that program’s maximum taxable wage base. Desperate legislators in 2034 will be tempted to do the same.

To be sure, 6 percent of workers earn more than the wage base, and would therefore pay some additional taxes, but most costs would be concentrated on the richest 1 or 2 percent of workers. Although some of these affluent workers may be generous campaign donors, the ultimate political currency is votes, not dollars. The 83 million enraged Social Security beneficiaries can do more political harm than even the most irate campaign donors can.

Surveys also reveal public support for raising the tax rate, at least modestly, and perhaps substantially. Of course, most surveys have asked about raising the tax rate gradually, which is more appealing. Unfortunately, gradualism will not be an option in 2034, given the need for immediate revenue. So, it is hard to know where people would stand on substantial and immediate tax increases. Similarly, surveys show public support for gradually raising the retirement age, though not too high. Although such gradualism does little to create immediate revenue, it is the best protection against continuing increases in longevity.

Third, incremental solutions make it easier for people to calculate their own personal stakes than do novel solutions. Workers can imagine how their income would decline with specific tax increases. Retirees can imagine what life would be like with particular benefit reductions. Such calculations are more difficult when policymakers consider novel solutions such as privatization.

To be sure, policymakers sometimes enact novel solutions with uncertain effects. The Clean Air Act of 1990, which introduced a cap-and-trade program to curb the sulfur dioxide emissions that cause acid rain, is a good example. Stakeholders struggled to calculate how the new system would affect their own interests, while policymakers inched toward a compromise.

But policymakers rarely adopt novel fixes under great time pressure. Once Social Security’s fiscal cliff is within sight, incremental solutions are more straightforward, not just for stakeholders calculating their personal interests but also for legislators calculating their electoral interests.

Fourth, compromise is easier with incremental solutions than it is with novel solutions. For traditional Social Security, all the basic elements — retirement ages, replacement rates, tax rates, the maximum taxable wage base, cost-of-living adjustments — are measured on interval scales. When policymakers negotiate about one of those elements — say, doubling versus quadrupling the maximum taxable wage base — all the intermediate options are obvious. For many novel solutions, however, the negotiating space is poorly defined. It is hard to find a midpoint between nominal categories, such as creating or not creating individual accounts.

How serious is the Republican pledge never to support tax increases? Can we imagine that Republican legislators would be so loyal to their tax-averse supporters that they would reject every plan to protect their benefit- dependent constituents? History suggests that elected politicians sometimes suspend their professed principles in order to resolve thorny policy problems. Although President Reagan introduced Republicans to the joy of tax-cutting with his 1981 plan to reduce income taxes, the same president and the same Republican legislators raised income taxes in 1982 and 1984 when the initial cuts proved too costly.

In 1981, Reagan proposed reducing the statutory increases in Social Security’s payroll tax, scheduled for 1985 and 1990. Days later, he abandoned that proposal. And two years later, he signed a solvency bill that not only accelerated those statutory increases but also raised taxes on self-employed workers.

His successor, George H. W. Bush, made a similar switch. Although he accepted the presidential nomination with the memorable lines, “Read my lips: No new taxes,” he later signed the Omnibus Budget Reconciliation Act of 1990, which did, in fact, raise taxes.

Not even the Norquist pledge has kept Republican legislators from approving tax increases. Consider Republicans’ predicament after approving the massive 2001 tax cuts. The tax cuts themselves would expire after ten years because they were enacted with a budget resolution, where a simple majority suffices, rather than a tax bill, where a supermajority would be needed to avoid a filibuster. Later, Congress extended the tax cuts through 2012, but with an agreement that any further stalemate would generate not only automatic tax increases but also automatic spending cuts for all federal agencies.

On New Year’s Day 2013, the stakes were enormous. Inaction that day would reverse all the Bush-era tax cuts, require the sequestration of more than $100 billion from agency budgets, and thereby threaten the still-fragile economy. At that point, House Republicans caved. They voted 232 to 2 to adopt a procedure that would allow floor consideration of a bill that would make the tax cuts permanent for everyone except the wealthy. An hour later, 85 House Republicans joined Democrats to approve the bill itself.

The first vote guaranteed that the nation would not plunge over the fiscal cliff and, more importantly, that Republicans would not be blamed for the economic consequences. The second vote guaranteed that most Republican House members would not be punished for violating their pledge to oppose all tax increases. But notice that 36 percent of Republicans voted explicitly to allow tax increases for affluent taxpayers, and 99 percent of Republicans voted to allow such a bill to reach the floor.

New Year’s Day 2013 reveals the practical limits of the Norquist pledge. Yes, Republican legislators work hard to avoid raising taxes, in part to avoid electoral and donor repercussions. But when faced with more severe electoral repercussions, as would likely follow from sailing over the 2013 fiscal cliff, Republicans swallowed hard and accepted that politics is about trade-offs. No doubt, many legislators will come to the same conclusion just moments before skidding over the 2034 solvency cliff.

Here is where general funds (more accurately, debt) can play an important role. General funds would allow legislators to approve gradual changes in the other basic elements of a solvency package — tax rates, benefit cuts, retirement ages and the maximum taxable wage base. Without general funds, legislators would have to approve sudden changes in these elements, thus imposing large and immediate costs on workers, employers or beneficiaries. General funds could reduce the immediate costs, thus making them less noticeable, less traceable and less dangerous.

Waiting until 2034 has one political advantage. It transforms policymaking from a slow, deliberative process, where partisan legislators often take extreme positions to appeal to their most attentive supporters, into a fullblown crisis, where legislators, fearing the electoral consequences of inaction, welcome compromise solutions.

The tragedy of 9/11 brought legislators together for bipartisan lawmaking. So, too, did the Great Recession, when bipartisan majorities approved the $700 billion Troubled Asset Relief Program. Similarly, in the early months of the Covid-19 pandemic, bipartisan majorities approved four relief bills totaling $2.9 trillion. The prospect of Social Security sailing over the fiscal cliff will be just as scary as these three crises. At that point, party leaders will negotiate a solvency package, and relieved legislators will approve it.