SVG
Commentary
National Review

Our Democratic Debt

Christopher DeMuth: We're borrowing to fund personal benefits, not public goods.

christopher_demuth
christopher_demuth
Former Distinguished Fellow
U.S. President Barack Obama delivers his State of the Union speech on Capitol Hill on January 28, 2014 in Washington, DC.(Larry Downing-Pool/Getty Images)
Caption
U.S. President Barack Obama delivers his State of the Union speech on Capitol Hill on January 28, 2014 in Washington, DC.(Larry Downing-Pool/Getty Images)

The federal government’s total debt is approaching $18 trillion. Its operating deficit was more than $1 trillion in each of the years 2009–12 and $680 billion in 2013. These numbers are too immense and unfamiliar to be useful. (A trillion is not yet even a standard measure—it means a thousand times a billion in the United States and a million times a billion in much of Europe.) Better to convert them to portions of the economy and government, so that the current debt is 103 percent of U.S. gross domestic product and the 2013 deficit was 4 percent of GDP and 20 percent of federal spending. These ratios put the dollar figures in perspective. The GDP ratio shows the burden of the debt (a larger economy can afford to borrow more, just as a higher-income family can afford a larger home mortgage), and the spending ratio shows how much of our government we are declining to pay for with our taxes. And they facilitate comparisons over time (effectively adjusting for inflation) and across nations with larger and smaller economies. But ratios are still just numbers: They need interpretation to tell us what they mean for our personal circumstances and those of our government.

We are not getting much help from public officials and policy experts, whose interpretations tend to be abstract and amorphous. The consensus formulation, embraced by President Obama, Speaker of the House Boehner, and the Congressional Budget Office, among many others, is that our current debt and deficits are “unsustainable.” This suggests that they are tolerable for the time being but will need to be reduced by some degree sometime in the future. Such a judgment has the advantage of sounding responsible and admonitory while suiting the short time horizons of democratic politicians and their preoccupation with immediate electoral exigencies.

And they have a point: Why not kick this can down the road? Experts have been warning for decades that our debt and deficits are unsustainable, yet here we are today, out and about and being sustained by an economy and government that continue to chug along even though the debt is bigger than ever. The only debt crises most of us have noticed have been the periodic impasses between the president and Congress over the debt ceiling—the statutory mechanism for enforcing Congress’s second enumerated power (Article I, Section 8 of the Constitution), which is to “borrow money on the credit of the United States.” Recurring annual deficits have obliged the Treasury Department to ask Congress to raise the ceiling a dozen times since 2000, and Congress is always reluctant to recognize the gap between spending and tax revenues that its policies have created. The most recent stand-offs threatened, in July 2011, default on U.S. debt service payments or drastic cuts in government spending, and then produced a two-week government shutdown in October 2013. But Congress invariably resolves these crises by raising the ceiling to make room for additional borrowing, whereupon everyone sighs in relief and gets back to business.

A somewhat edgier formulation of our fiscal situation is that debt and deficits are “robbing our grandchildren.” This is Speaker Boehner’s position today, and it was President Obama’s position when, as a senator, he opposed President Bush’s proposed debt-ceiling increase in 2006—but Obama renounced it when campaigning for his own increase in 2011. It seems to be the position of the opposition party—an attempt to use moral suasion when practical forms of persuasion are unavailable. I think there is much truth in the expression but that it is too broad and rhetorical. When a city sells revenue bonds to finance highway or airport improvements, is it robbing from future generations? The “greatest generation” of Americans that fought and won World War II borrowed hundreds of billions of dollars to do so; was it robbing from the future or securing the future?

A third formulation is that high public debt impedes the private economy, slowing its growth. The economists Carmen Reinhart and Kenneth Rogoff, in a widely noted 2010 article in The American Economic Review, compared debt ratios and economic growth across nations and time periods. They found that debt of more than 90 percent of a nation’s GDP is associated with significantly lower levels of economic growth, and argued that debt ratios this high are an important cause of lower growth rather than merely the result of low growth and a smaller economy.

This approach has the advantage of making debt an issue of the here and now rather than mañana, but it yields no more than a rule of thumb. Government borrowing can undoubtedly promote rather than retard economic growth when it is invested in such things as improved highways and education. (Despite today’s heavy state and local debts and occasional bankruptcies, the municipal-bond market—much of it invested in physical infrastructure and secured by the proceeds—is flourishing.) And many things other than debt can influence economic growth: More efficient taxing, and more productive spending and regulation, can promote higher growth and thereby reduce the burden of debt. The mechanisms by which a given amount of debt affects economic growth involve many contingencies and theoretical arguments.

As a result, efforts to relate government debt to current economic performance quickly run into disagreements over technical economics, political philosophy, and policy tactics. If you ask the eminent economists Martin Feldstein and Laurence Kotlikoff how to improve economic growth today, they will tell you that reducing the debt is essential. If you ask the eminent economists Lawrence Summers and Paul Krugman, they will tell you that increasing the debt, and spending it on useful things such as infrastructure and job training, is essential. Some economists argue that the debt itself is a sideshow; the essential thing, they say, is to reform tax, spending, and regulatory policies to spur productivity and “grow our way out of the debt.”

I am in the camp of the deficit hawks such as Feldstein and Kotlikoff, but I do not think the reason we are failing to address the problem is that some other smart and influential people are deficit doves. Rather, I think that our political institutions and political leaders have accommodated themselves to deficit spending and growing debt and acquired a stake in their continuance. Disagreements over the consequences and immediacy of the problem are always resolved in favor of borrowing more to address the problems of the moment and deferring “debt consolidation” (through some combination of higher taxes, lower spending, and higher economic growth) to a later time. The American body politic has acquired deficit-attention disorder.

It won’t last, as we shall see, but for the time being it is getting worse. The pitched battles of the House tea-party Republicans in 2011–13 were probably the high-water mark of debt-ceiling brinkmanship deployed as a tactic for spending and deficit reduction. The pathetic final score: more than $3 trillion in immediate new debt, all of it promptly borrowed and spent; $1 trillion in promised future spending “sequesters,” spread over nine years through 2021 (and already being relaxed in this year’s budget), during which time the debt will grow by another $6 trillion; a tax increase limited to high-income taxpayers that will raise only about $600 billion through 2021; and a substantial increase in the amount of debt the CBO projects over the next ten- and 25-year periods. In the end, when Congress reopened the government following the October 2013 shutdown, it did so in a way that provided a fitting coda to the drama: It did not raise the debt ceiling to a new and higher level but rather suspended it, permitting the Treasury to borrow as needed without any further authorization. By thus excusing itself from its constitutional duty to provide for the shortfall between spending and taxing, Congress moved the debt out of the headlines and away from public attention through the 2014 elections. The suspension lasts only through March 2015—but by then another election will be in sight, and a precedent will have been set.

Three features of our current fiscal circumstances are almost never mentioned explicitly in political debate, official reports, and academic studies. Each one is simple and straightforward yet will come as a surprise to those whose understanding of the debt comes from newspapers and the blogosphere. Taken together, they provide a more useful account of our situation than generalizations about sustainability, robbing our grandchildren, and economic growth. They also point to the source of our deficit-attention disorder—and to a range of prognoses.

1. Today’s government debt is far larger than anything in our national experience.

It is sometimes said that today’s debt is our highest peacetime debt—that our debt has been higher only in times of war. This is incorrect. Our major wars have indeed required substantial borrowing, but the debts of the Revolutionary War, the Civil War, and World War I were in each case only about 30 percent of GDP. The federal debt passed 40 percent of GDP for the first time during the Great Depression and peaked at 119 percent in the final year of World War II, falling the next year to 103 percent, the same as today’s.

On average, the official federal debt during the years since the 2008 financial crisis is about what it was during the years of World War II—years of desperate national mobilization in the greatest military conflagration in human history, coming on the heels of a prolonged depression that had already left the government deeper in hock than ever in its history. But the official figures miss profound changes in government since the 1940s and ’50s that have made our actual debt far larger. The changes are of three kinds.

First, state- and local-government debt has grown substantially, from about 7 percent of GDP in the late 1940s to 18 per- cent ($3 trillion) today. Now as then, state and local debt includes much that is invested in new and improved physical infrastructure, such as roads and water and sewer systems, which is relatively unproblematic. But the recent municipal bankruptcies in Detroit and elsewhere, and the dire fiscal problems of states such as Illinois and California, show that many states and cities have gotten themselves badly overextended. (They have done so by finagling with balanced-budget requirements and capital- budgeting procedures—a cautionary example to those who advocate such measures at the national level.) In addition, the post-war development of public-sector unionization has generated generous pension commitments that are now terribly underfunded—by $4 trillion, or 24 percent of GDP, according to the best estimates. That more than doubles today’s actual state and local debt, bringing it to $7 trillion.

These are not federal debts, as they would be if we had a unitary government like that of France. And we have, for the most part, avoided bailing out bankrupt localities, and thereby nationalizing local debt, in the manner of the European Union’s bailout of Greece. But some state and local debt (no one knows how much) does sponge up to Washington through the operation of grant programs such as Medicaid, and these programs may be used selectively to assist failing localities such as Detroit. A cascade of state and city insolvencies similar to the corporate insolvencies of 2008 would generate a similar panic and intense pressure for explicit federal bailouts—or disguised bailouts, in the form of Federal Reserve pressure on major banks to purchase dubious state and local bonds. But even without bailouts, the size and uses of today’s state and local debt—more than half of it for income support for retirees—are already adding to the special risks of the federal debt, which I will explore in the coming sections.

Second, a major federal innovation in recent decades has been the promotion of private borrowing, especially for home mortgages and college tuition, through loans and loan guarantees on easy terms and at preferential interest rates. These programs have generated enormous but unrecorded federal debt (“implicit debt”), because government accounting fails to capture the prospective costs of loan defaults or loan forgiveness. When Fannie Mae and Freddie Mac, the government-sponsored mortgage underwriters, went insolvent in 2008, the federal rescue cost nearly $190 billion that had not been in the budget. The White House now counts Fannie and Freddie as profit centers, because they have since paid more than $200 billion in dividends on the government’s preferred stock. But this is absurd: As the CBO and others have pointed out, it completely ignores the risks of default on the firms’ nearly $4 trillion in loan guarantees, which fair accounting would record at several hundred billion dollars (the 2013 deficit alone would have been $780 billion rather than $680 billion). The federal government now owns 85 percent of the nation’s outstanding student loans, which total more than $1 trillion. Here, the feds do make some, but not much, allowance for default risks. With half of recent college graduates either unemployed or working in jobs that don’t require a college degree, the unrecorded debt is probably substantial, and indeed it is already appearing in the form of President Obama’s order to cap the loan repayments of lower-income graduates.

Third, by far the largest implicit federal debt is to the Medicare and Social Security programs. It is the difference between the programs’ future benefit payments and future tax revenues under current law, discounted to a present value. Some of that difference is explicit debt: Revenues from the programs’ payroll taxes have exceeded benefit payments in years past, and the excess was spent on other federal programs and booked as IOUs from the Treasury to Medicare and Social Security. This “interagency” debt (which includes several smaller programs as well) now amounts to $5 trillion of the nearly $18 trillion of official federal debt, and it is to be repaid as Medicare and Social Security spending outpaces revenues from payroll taxes. But that is a small fraction of the gap between future benefits and future revenues—which amounts to a present value of $50 trillion to $200 trillion, depending on such variables as the number of years projected and whether one assumes that Congress will repeal statutory expenditure reductions (such as cuts in reimbursement rates to Medicare physicians) in the future as it has in the past.

This implicit debt is almost entirely a post–World War II phenomenon. It is the result, first, of the major expansions of Social Security in the 1960s and ’70s and the enactment of Medicare in 1965; and, second, of demographic developments (the baby boom and subsequent decline in the birthrate and increase in longevity) that are now destroying the programs’ pay-as-you-go financial structure. The exact size of the implicit debt is not important; projections so far into the future involve tremendous uncertainties over such things as economic growth and tax revenues. The important point is that it dwarfs the official debt by any reckoning and gives teeth to the notion of “unsustainability.” We know enough about some key variables—such as the size of the baby boom and the younger generations, and trends in longevity and medical costs—to say with confidence that continuing the current Social Security and Medicare programs would require a level of resources that no one has any idea how to obtain.

The resources could come from increased borrowing, increased taxes, reduced spending on other programs, or default on outstanding debt—and each one looks self-defeating. The requisite new borrowing or debt defaults would push interest rates and payments to levels that would consume other government programs and eventually Medicare and Social Security themselves; the requisite spending reductions would consume basic government functions that affect the welfare of the elderly as much as Medicare and Social Security do; and the requisite tax increases would take us well into Laffer Curve territory where revenues begin to fall.

2. Our high debt reflects the normalization of annual deficits.

Until a half-century ago, spending beyond current tax revenues was limited to investments such as canals and railroads and to emergencies such as wars and natural disasters. Investment debt was paid down over the life of the investments, and emergency debt was paid down when the emergencies subsided. Andrew Jackson—paragon of populist fiscal rectitude, founder of the Democratic party, and spiritual godfather of the Tea Party—wrestled the remaining debts from the Revolutionary War and the War of 1812 all the way down to zero. Even John Maynard Keynes, who in the 1930s pioneered the idea of deficit spending to surmount economic depressions as well as other emergencies, assumed that the debts would be repaid once the economy rebounded.

But Keynesian balancing was never even tried. In the 1950s, the federal fisc oscillated between small deficits and small surpluses, more the result of economic fluctuations than of deliberate policy, and averaged to a small surplus for the decade. The 1960s saw the beginning of deficit spending as a routine practice rather than a temporary expedient: JFK’s experiments with tax-cutting stimulus, followed by LBJ’s Great Society and Vietnam “guns and butter” spending, yielded a stream of annual deficits that averaged 4 percent of spending for the decade. The deficits were tiny by today’s standards, and they were exceeded by economic growth that continued to chip away at the World War II debt as a share of GDP, but they established a new normal. As government officials, economists, and the general public became increasingly comfortable with deficits, borrowed money became a continuous, growing source of routine annual spending. The federal government has run a deficit in 40 of the 44 years since 1970, averaging 10 percent of federal spending in the 1970s, 18 percent in the 1980s, 10 percent in the 1990s, and 11 percent in the 2000s. The deficit rocketed to 36 percent of spending in 2009–12, then fell to 20 percent in 2013.

The current projection is that deficits will average 16 percent of spending (3.4 percent of GDP) over the coming decade, then rise sharply to pay the mounting unfunded Social Security and Medicare bills of the baby boomers (and the accompanying rise in interest payments on mounting debt). The Obama administration has suggested that it is comfortable with regular deficits at the current level. The demurrals of congressional Republicans are mainly rhetorical: If you take all the spending-reduction proposals championed by fiscal-reform leaders such as House Budget Committee chairman Paul Ryan, significant annual deficits will continue for at least a decade and probably for much longer.

Chart 1 outlines the saga of routine annual deficits and rising debt since 1970. It is important to note that the debt did not grow as a share of GDP in every deficit year, because economic growth sometimes outpaced the deficit. An economic boom—the “dot-com bubble” of the late 1990s—even produced a surprise surplus in 1998 after President Clinton had proposed a deficit budget. President Clinton and the Republican Congress both deserve credit for maintaining surpluses through 2001, but their four-year deficit hiatus, which ended with the dot-com collapse and then the 9/11 terrorist attack of 2001, is a reminder that fiscal policy is often hostage to events in the larger world. The strongest pattern, unfortunately, has been for extraordinary events to produce extraordinary deficits and unanticipated new borrowing. These have included the Arab oil embargo in the 1970s; Reagan’s defense build-up in the 1980s; the 9/11 attack, wars in Iraq and Afghanistan, and Hurricane Katrina in the 2000s; and several financial crises along the way, the most severe and costly occurring in 2008. The frequency of such “extraordinary” events is an important part of our debt predicament, one we will return to at the conclusion of this essay.

__3. As debt and deficits have grown, their function has changed from funding public investment to funding private consumption.__

The traditional purpose of government debt is the provision of public goods with long lifespans, such as waterworks or aircraft carriers. Borrowing spreads their costs among the current and future taxpayers who benefit from them. Needless to say, the investments do not always turn out well. President Jefferson’s Louisiana Purchase was a triumph; President Obama’s Solyndra was a bust; President Ford’s swine-flu-inoculation program is still studied as a policy debacle; I am dubious about high-speed rail from Cedar Rapids to Chicago. And such investments are often controversial not only in prospect but also in retrospect. Midway through the Reagan administration, Senator Daniel Patrick Moynihan, a firm critic of Reagan’s economic policies, proclaimed, “The 1980s will be remembered as the decade when America borrowed a trillion dollars and threw a party.” I would say that it was the decade when America borrowed a trillion dollars and terminated the Soviet empire, but some would disagree with that assessment even now. Politics is aspirational, and government projects are open to varying interpretations (except where a project produces its own revenue stream, like a toll highway, to eventually settle the matter). The essential point is that debt was traditionally intended to expand and improve the future—through a larger and better capital stock, greater security against foreign threats, recovery from disasters and emergencies, and higher levels of invention and productivity.

These are not the primary purposes of today’s government borrowing, which is being used primarily to pay for immediate private consumption. The routinization of deficits and the growth of debt since 1970 have been accompanied, and encouraged, by a profound transformation in the nature of federal spending, illustrated in Chart 2. I have divided federal spending into two broad categories, based on line items in historical tables from the Office of Management and Budget. The first, Public Goods, consists of defense and international operations, from diplomacy to foreign aid; highways, airports, and other physical infrastructure; national parks and other preserves; courts, law enforcement, and the regulatory agencies; the National Institutes of Health and other research-and-development efforts; Congress; and smaller “general government” items. The second, Payments to Individuals, consists of money payments such as Social Security, unemployment compensation, and various welfare programs, and of in-kind provisions such as Medicare and Medicaid, food stamps, and housing subsidies. I have some issues with OMB’s bean counters: I would count farm subsidies and the Export-Import Bank as Payments to Individuals (because they are uncompensated subsidies to designated persons and firms) and the veterans’ hospitals as Public Goods (because they are a form of deferred compensation to military personnel). But I want to stick to the government’s categories, and my adjustments would have only a small effect on the aggregate figures and the story they tell.

Chart 2 traces Public Goods and Payments to Individuals as percentages of federal-program spending (not including interest on the debt) from 1970 through 2014. During this period—when large annual deficits became routine and debt grew from 36 percent to 103 percent of GDP—Payments to Individuals soared from 36 percent to 75 percent of annual spending while Public Goods plunged from 64 percent to 25 percent. Tracking spending to GDP, also shown on the chart, reveals a similar pattern: Payments to Individuals grew from 6 percent to 15 percent of GDP while Public Goods fell from 11 percent to 5 percent. The two signal innovations of post-1960s government—continuous borrowing to support regular spending, and spending primarily on private consumption—were concurrent. (Incidentally, essentially all of the fall in Public Goods has been in national defense; domestic public goods have stayed about constant, with the exception of increased spending on law enforcement—reflecting the costly “homeland security” measures introduced after 9/11 and also the continuing growth of federal criminal law.)

Incorporating the growth of “tax expenditures”—implicit public spending consisting of tax credits or deductions for specified private expenditures—would amplify the trends displayed on Chart 2. Most of that growth has been for personal benefits, such as the tax exclusion of employer-provided health insurance, the mortgage-interest deduction, and the earned-income tax credit rather than for putative public goods, such as the tax breaks for energy-efficient cars and appliances and for corporate research and experimentation. (I will leave it to the metaphysicians to categorize the tax benefits for children and child care.)

One cannot allocate the deficit in any given year, or the growth of debt over time, between my two spending categories. Receipts from bond sales and receipts from tax levies are a common pool, used to pay for spending programs as the bills arrive. But the tandem growth of Payments to Individuals and of deficits and debt is remarkable. And there is little doubt that these payments have become not only the dominant form of spending but also the more politically entrenched. The government categorizes all Public Goods as “discretionary” and most Payments to Individuals as “entitlements” not subject to congressional appropriations. The 2013–21 spending sequesters, adopted in the debt-ceiling compromise of August 2011, are focused almost entirely on discretionary programs. The partial government shutdown of October 2013 did not shut down Social Security, Medicare, unemployment benefits, or food stamps.

Moreover, debt and consumption spending have been joint projects at several critical points. Social Security and Medicare are pay-as-you-go transfers from wage earners to retirees—so earlier recipients (those who are older), who paid program taxes for less than their full working careers, receive relatively greater benefits than younger recipients. That in turn has created pressure to increase current benefits faster than payroll taxes over time, most dramatically in the 2002 addition of a Medicare prescription-drug benefit, 75 percent of which was financed by new borrowing. It is this Ponzi-scheme dynamic that has propelled implicit debt to impossible heights. At the same time, the biggest growth in loan guarantees and subsidies has come from increased private borrowing for home mortgages and for college tuition, room, board, and expenses. These are often justified as investments in household and human capital, but they clearly involve large elements of consumption. The extension of loans on extremely lenient, non-investment-grade terms, and the continuous expansion of debt-forgiveness programs to those whose loans did not turn out well, are powerful inducements to use the loans for consumption.

It is often said that the American government has become an insurance state; more precisely, it has become a consumer state. The shift from public goods to private consumption is a pervasive feature of contemporary government, not limited to spending programs. Regulation, too, has shifted from promoting production (transportation, communications, energy) to promoting consumption (environmental quality, personal health and safety, consumer protection). In the aftermath of the 9/11 terrorist attack, when the public and media were asking what regular citizens could do to support the government response, the government’s answer was: Keep shopping and traveling. In the aftermath of the financial collapse of 2008, which revealed that major financial firms had gotten themselves impossibly overleveraged and engorged with incontinent loans, the government’s flagship response was: a new consumer-protection agency. And when the collapse precipitated a dramatic rise in unemployment and a fall in employment, President Obama’s responses were emphatically consumerist: a stimulus program that was mostly transfer payments rather than investments, and a sweeping takeover of the health-care sector intended to make its products and services less costly and more widely available. In the months before the national elections of 2012, the president involved himself with great fanfare in requiring universities and health insurers to furnish birth-control products free of charge to students—which was not even insurance, much less production, but pure consumption.

Let us stipulate that consumption beyond the basic necessities is a good and worthy thing. For many it is what production is for—work to live, not live to work—and no one is indifferent to it. The government of an affluent, egalitarian nation is going to concern itself with the level, quality, and distribution of consumption among its citizens. And, at a time of slowing income growth, it is going to take steps to maintain the purchasing power of the middle class. What is striking about our consumption state, and menacing, is that it is a debt-financed consumption state. Public investments often fail. Wars are lost at least half the time. But borrowing for consumption is not aspirational. It is intended to ease the present rather than enlarge the future; and, when pursued on the massive scale we have become accustomed to, it can only diminish the future.

The essence of our debt predicament is not so much its size as its source. It has grown to unprecedented levels not because of external crises or failed or over-ambitious public ventures, which would be bad enough. Rather, it has arisen from within, generated by our democracy itself. Through a long sequence of accommodations to immediate contingencies and opportunities, we have built a system in which the electorate expects, and political officials provide, a higher level of personal benefits than of tax collections to pay for the benefits. The difference is taxed to younger and future generations—to be paid in the future, by unknown subsets of them, through higher outright taxes, reduced benefits, debt defaults, or inflation. The situation is intractable because most of those being burdened are not voters (most are not even alive) and cannot be part of a constituency for reform.

It is true that non-voting future generations also bear part of the burden of debts for investments in infrastructure, deterring and prosecuting wars, and other long-lived public goods. But the political dynamics, not just the economic dynamics, are fundamentally different. For one thing, living persons bear most of the burden (the non-monetary burden) of wars and natural disasters. More important, investment and disaster management are self-limiting, while consumption is open-ended. One can make a plausible case for only so many highways, battleships, foreign interventions, and cancer-research centers. But there is no inherent limit to the demand for more and better consumption, and for redistribution from the future to the present to pay for it.

There are three basic scenarios for resolving America’s debt problem. First, far-sighted political leaders could put our entitlement and income-transfer programs on a path to solvency, through reforms that have already been thoroughly worked out in academic and think-tank research—such as voucherizing Medicaid, raising the Social Security retirement age and indexing its benefits to prices rather than wages, and pre-funding both programs for future generations. Second, high inflation could reduce the debt burden at the expense of bondholders by shrinking the real value of fixed-interest payments and bond redemptions (sometimes called slow-motion partial default). Third, resolution could await a crisis—a war or other costly emergency, or a sharp increase in interest rates that balloons the cost of new and replacement bonds—that forces a choice between outright default and abrupt, severe benefit cuts.

There are precedents for each scenario and no need to predict which one (or which combination) we will follow. The first path is obviously the best but will remain open for only another decade or so before the demographic arithmetic begins to dictate benefit reductions that are immediate and steep rather than prospective and gradual. The second option would export some of the immediate losses, because foreigners own nearly half of our publicly held debt, but it would also injure most domestic savers regardless of their Treasury holdings. The third solution is obviously the worst, because it destroys settled expectations, imposes windfall losses on many people of modest means, and, in the extreme, inflicts widespread misery and loss of social morale (as in Greece) or tempts outright confiscation of private assets (as in Argentina). Confiscation is not so alien from the American experience as one might suppose: It’s what FDR did in 1933 in seizing everyone’s gold coins at a fraction of their value.

Whatever the path, it will lead to a welfare state that is more constrained and limited and less generous than the one we are enjoying today. It is possible to have a generous welfare state financed by current taxation; Sweden is close to one, and it has even put its retirement pensions on a budget. But Sweden is a small, homogeneous, collectivist, consensus nation, and America is a vast, heterogeneous, individualistic, fractious nation. Our transition from debt-supported to tax-supported consumption is going to be much more contentious. And, because we are not going to share current wealth so agreeably as the Swedes do, our welfare state is going to get smaller than it is today.

It is important to note, finally, that even in the best of all possible worlds—the first scenario, in which we begin the transition now and give everyone plenty of time to adjust—the transition is going to be a period of considerable national risk. To be sure, risk always involves positive as well as negative possibilities. The official projections of our future debt burden assume that economic growth in the coming decades will be lower than in the past. That is because, among other things, work-force participation will be lower in an older society, and because many of the 20th century’s big growth-promoting developments (such as in transportation, urbanization, and the entry of large numbers of women into the work force) have pretty much run their course. But there could be happy surprises: Dramatic discoveries in medicine, energy, or information technology could propel economic growth far beyond expectations, diminish the real burden of our existing debt, subdue the dynamics of population aging, and give debt-financed consumption a renewed franchise. Deficit doves who think we can grow our way out of our debt are implicitly counting on such big positive surprises.

That would be wonderful and is devoutly to be wished. However, a central responsibility of government—one might say its first responsibility—is to be prepared for big negative surprises. The purpose of maintaining low public debt is not to adhere to abstract notions of fiscal rectitude. Rather it is to maintain the government’s capacity to respond forcefully to severe emergencies that citizens cannot manage on their own. Such emergencies—foreign aggression, epidemics, natural disasters, financial collapses, and economic depressions— have occurred frequently throughout American history. Actuarially, at least one is likely to occur in the decades that it will take to reduce our debt to a manageable level of, say, 30 percent of GDP without risking serious social dislocations. If a severe shock should come our way soon, the government’s response will be seriously weakened by its inability to borrow massive resources immediately. And the very fragility of our financial circumstances will invite emergencies of the man-made kind.

The theory of social insurance is that government is able to reduce the lifetime risks facing individual citizens by spreading those risks among large populations and compensating for lack of individual foresight. Adding the expedient of massive borrowing to the equation has produced a different, untoward result. Whatever success our debt-financed welfare state has had in reducing risks among the citizenry, it has certainly centralized a great deal of risk in the government itself, thereby hobbling its ability to provide the most important form of social insurance of all.