The Case Against Intergenerational Accounting - The Accounting Campaign Against Social Security and Medicare


The Levy Economics Institute of Bard College Public Policy Brief No. 98, 2009


In recent years we have been subjected to a rising cacophony of nonsense about a looming financial crisis. No, we are not referring to the current, very real, meltdown of private financial markets. Rather, we are told, future unfunded entitlements will bankrupt our government as the baby boomers retire. Social Security and Medicare are the main source of what former Comptroller General David Walker has called the “super-subprime crisis.”

Social Security and Medicare have always had enemies, closely allied to private insurance companies who would like the business and to fund managers and others who would profit from privatization of the associated revenue streams. But recently, these enemies have been given a boost, and a claim to respectability, by the creation of “intergenerational accounting,” an economic method that purports to calculate the debt burden our generation will leave for future generations. This brief assesses intergenerational accounting and related aspects of what we call “The Accounting Campaign Against Social Security and Medicare.”

In intergenerational accounting, federal government revenue and expenditure streams are compared over very long periods—even over infinite time. “Deficit gaps” are then used to measure the financial burden of these commitments, and therefore the alleged solvency or insolvency of the government. Discounting the sum of the differences back to the present permits infinite sums to be translated into very large, but finite, numbers. The results, amounting to tens of trillions of dollars, are headline-grabbing and scarylooking. Evidently, this combination makes them irresistible. Even the Board of Trustees of the Social Security Administration (SSA) began to dabble in such arithmetic several years ago.

Now the Federal Accounting Standards Advisory Board (FASAB)(Notes 1) is proposing to subject the entire federal budget to such accounting. It has issued two “exposure drafts” titled “Comprehensive Long-Term Projections for the U.S. Government” (ED 1) and “Accounting for Social Insurance, Revised” (ED 2), and is soliciting comments on its recommendations. In this brief, we argue that these proposals are not only wrongheaded, but also dangerous. We examine the purpose of budgeting at the federal government level and explain why government should not be subject to the same sort of accounting and financial constraints that apply to private households or business firms. We conclude that intergenerational accounting should play no role in federal government budgeting, and that arguments based on this concept do not support a case for cutting Social Security or Medicare.

General Principles of Federal Budget Accounting

Even though some principles of accounting are universal, federal budget accounting has never followed, and should not follow, the exact procedures adopted by households or business firms. There are several reasons why this is true.

First, the government’s interest is the public interest. The government is there to provide for the general welfare, and there is no correlation between this interest and a position of surplus or deficit, nor of indebtedness, in the government’s books.

Second, the government is sovereign. This fact gives to government authority that households and firms do not have. In particular, government has the power to tax and to issue money. The power to tax means that government does not need to sell products, and the power to issue currency means that it can make purchases by dispensing IOUs. No private firm can require that markets buy its products or its debt. Indeed, taxation creates a demand for public spending in order to make available the currency required to pay the taxes. No private firm can generate demand for its output in this way.

Neither of these statements is controversial; both are matters of fact. Nor should they be construed to imply that government should raise taxes or spend without limit.

However, they do imply that federal budgeting is different from private budgeting and should be considered in its proper, public context.

While it is common to regard government tax revenue as income, this income is not comparable to that of firms or households. Government can choose to exact greater tax revenues by imposing new taxes or raising tax rates. No firm can do this; even firms with market power know that consumers will find substitutes if prices are raised too much. Moreover, firms, households, and even state and local governments require income or borrowings in order to spend. The federal government’s spending is not constrained by revenues or borrowing. This is, again, a fact, completely noncontroversial but very poorly understood.

The federal government spends by cutting checks—or, what is functionally the same thing, by directly crediting private bank accounts. This is a matter of typing numbers into a machine. That is all federal spending is. Unlike private firms, the federal government maintains no stock of cashon-hand and no credit balance at the bank. It doesn’t need to do so. There are surely limits of wisdom and prudence on federal spending, as well as numerous checks, balances, and self-imposed constraints, but there is no operational limit. The federal government can, and does, spend what it wants

Tax receipts debit bank accounts. So does borrowing from the public. These are operationally distinct from spending. There is no operational procedure through which federal government “uses” tax receipts or borrowings for its spending. If, perchance, one chooses to pay taxes in cash, the Treasury simply issues a receipt and shreds the cash. It has no need for the income in order to spend. This is why it is a mistake to look at federal tax receipts as an equivalent concept to income of households or firms.

As we discuss below, federal government spending has exceeded tax revenues, with only brief exceptions, since the founding of our country. There is no evidence, nor any economic theory, behind the proposition that federal government spending ever needs to match federal government tax receipts—over any period, short or long. The deficit per unit of time is the difference between taxing and spending over that time. To repeat, taxing, on the one hand, and spending, on the other, are operationally independent. Any reasonable observer should conclude that federal government spending is not, and need not be, dependent on, constrained by, or even related to tax revenues in the way that the spending of households or firms is related to their incomes.

The difference between microeconomic and macroeconomic accounting is also pertinent. An individual household or firm has a balance sheet that consists of assets and liabilities. The spending of that household or firm is constrained, in a fairly concrete sense, by its income and balance sheet—by its ability to sell assets or to borrow against them. It is meaningful to say that its ability to deficit-spend is constrained: a household must get the approval of a bank before spending can exceed income, and therefore borrowing is subject to banking norms. But if we take households or firms as a whole, the situation is different. The private sector’s ability to deficit-spend, to spend more than its income, depends on the willingness of another sector to spend less than its income. For one sector to run a deficit, another must run a surplus. This surplus is called saving—claims against the deficit sector. In principle, there is no reason why one sector cannot run perpetual deficits, so long as at least one other sector wants to run surpluses.

In the real world, we observe that the U.S. federal government tends to run persistent deficits. This is matched by a persistent tendency of the nongovernment sector to save. The nongovernment sector accumulates net claims on the government; the nongovernment sector’s “net saving” is equal (by identity) to the U.S. government’s deficits. At the same time, the nongovernment sector’s net accumulation of financial assets (or “net financial wealth”) equals, exactly, the government’s total net issue of debt—from the inception of the nation. Debt issued between private parties cancels out; but that between the government and the private sector remains, with the private sector’s net financial wealth consisting of the government’s net debt.

This identity does not change once we allow for a foreign sector, which is just a part of the nongovernment sector. Since the United States has in recent decades run persistent current account deficits, the foreign sector has been accumulating net financial claims in dollars—thus the role of dollar-denominated securities as reserve assets. Whether this is “good” or “bad” does not change the accounting identities. It is identically true that U.S. government deficits equal nongovernment surpluses, and U.S. government debt equals nongovernment net financial wealth. Yet these macroeconomic relations are not obvious when one looks to individual firms or households, or if one examines the U.S. government as if it had an independent balance sheet.

Do the FASAB Exposure Drafts Recognize the General Principles of Federal Budget Accounting?

The reporting proposed by the two FASAB exposure drafts does not appear to recognize the fundamental differences between public and private budgets. There are numerous problems in the drafts: Some of the most basic principles of accounting are neglected. Key terms are left ill defined or undefined. Projections are misused. Unjustified policy prescriptions are slipped into the drafts in the guise of accounting standards. And revenues are matched to spending for parts of the federal budget—notably, Social Security and Medicare—in ways that have no economic justification.

A Basic Principle: Liabilities and Assets

The FASAB drafts are intended to be “statements of financial condition” for “the government” and for “the nation.” These two concepts—government and nation—are not interchangeable. To use them interchangeably, as the exposure drafts do, is a source of confusion.

In our understanding, a statement of “financial condition” is, in general, a balance sheet. These are constructed with two columns: one for liabilities and the other for assets. This very basic principle is no different for the public sector, or for the nation as a whole, than it is for private sector accounting. The “nation’s financial condition”—a term used repeatedly in the exposure drafts—is a combination of the financial condition of the government and that of its citizens. Yet the proposed “federal financial reporting” contains no mention of the assets that correspond to the liabilities that would be reported when accounting for “the nation.” For example, it would treat the obligations of the Social Security system as a liability. That same Social Security benefit liability is, of course, an asset to the public. The Social Security wealth of the current population is just as real as the liabilities that support it. Yet nowhere is this Social Security wealth reported or even remarked on. Put another way, a transfer program from one group of citizens to another, via the government or otherwise, merely transfers resources. It does not increase or diminish them. This is an economic reality, and any financial statement for “the nation” should reflect it

The picture is further confused by treating the forecast difference between Social Security benefits and FICA tax revenues, projected over time and discounted to the present, as a “net liability” of the government—and, by implication, of “the nation” as well. In this way, intergenerational accounting purports to show an “unfunded burden” on the government, for the benefit of the future retired population. This overlooks the fact that the underlying citizens who support the government are the same people: today’s workers will become, eventually, tomorrow’s retirees. It is therefore hard to see why workers should object if the burden of payroll taxes does not, in present value terms, equal the value of Social Security benefits.

We shall return later to other issues. Here, our point is a matter of accounting: the asset of payroll tax revenues to the government is just a liability to the working population, just as the liability of future benefits is an asset to the public. In both cases, the books balance between the public and private sectors—taken together, “the nation.” And if the public’s books taken alone don’t balance, it merely means that the private sector’s books, taken alone, don’t balance either: the deficit of the one is the surplus of the other. There is nothing alarming about this. Just as the public debt can be eternal and need never be paid off, a net debt position for Social Security and Medicare can likewise be eternal, since the government’s net deficit is balanced by the nongovernment sector’s net surplus. Detailing the balance sheet in full for “the nation” would be good financial-reporting practice. And, in this case, it would usefully reduce the “scare” content of claims that focus on liabilities without acknowledging the corresponding assets.

Ill-defined Terms: What Is a “Budgetary Resource”?

The proposed reporting speaks of “budgetary resources.” The apparent concern is that the federal government operates within the “budgetary resources” available to it. Specifically, the exposure drafts are concerned that budgetary resources be sufficient to “sustain public services and meet obligations as they come due.” But there is no clear definition of what “budgetary resources” means.

If what is meant is tax revenue, the definition is totally inappropriate. As we have demonstrated above, the government does not need tax revenue sufficient to match spending in order to “sustain public services and meet obligations as they come due.” This is obvious: the government almost never has sufficient tax revenue for that purpose. (It has run significant surpluses for only seven very brief periods in the history of the nation, each of them producing a depression or a recession.) This is why we have a national debt to begin with. Yet the U.S. federal government has never, in 233 years of operation, lacked for “budgetary resources” sufficient to “sustain public services and meet obligations as they come due.” This is also obvious, insofar as the federal government has never defaulted on its obligations, including making all interest payments on its debt.

If, however, the term “budgetary resources” means tax revenues and public borrowings sufficient to “sustain public services and meet obligations as they come due,” this, too, is inappropriate. The standard in that case would appear to be intended to inform the public about the borrowing capacity of the government of the United States. Yet the procedures contain no information about and no guidance as to how to assess this question.

Can we imagine that the U.S. domestic sector will reach a point such that it will refuse to accumulate dollar claims on our government in the form of currency and interest-bearing government bonds? If so, that would only mean that government spending would be immediately inflationary, to the extent the government tried to force deficit spending, as did Germany in 1923. Would we reach the point where American businesses would refuse to sell their wares for U.S. currency?(Notes 2) If households had more currency than desired, would they refuse to substitute it for Treasuries? Would private banks refuse to exchange excess reserves for Treasuries? We think not. Nor would it be catastrophic if households or banks did refuse to substitute cash or reserves for Treasuries—the Fed and the Treasury could simply avoid selling them.

Low long-term interest rates tell us that the markets are not troubled by this possibility, and that the U.S. government is not now facing financial concerns due to any one of these conditions (nor even due to the current global crisis!). Nor is it possible for such concerns, should they arise in the markets, to become actual problems even with the growth of “entitlements” over the next three-quarters of a century. The drafts presume that financing of Treasury spending could at some time become problematic but do not explain, operationally, how problems could arise. It is not sufficient to show that on some set of assumptions projected tax revenue might fall short of projected spending. Rather, there must be some explanation as to why that should be a matter of concern or why and how borrowing might become difficult or constrained—particularly given that we now have two centuries’ experience of accumulated tax shortfalls with, predictably, no suggestion of government insolvency.(Notes 3)

On the assumption that what is termed “budgetary resources” by the FASAB includes public debt issue, the proposed procedure betrays a false supposition that there is some economic limit to the nominal value of the bonds that can be issued by the U.S. Treasury. The reality is, no such limit exists. Nor does the government have to issue securities, operationally, in order to spend. As an operating matter, the government spends first and issues securities later by transferring funds from interest-bearing reserve accounts at the Federal Reserve to interest-bearing Treasury securities. (The latter are also merely accounts at the Federal Reserve.) Treasuries are net issued (in the open market by the Fed and in the new-issue market by the Treasury) when financial institutions in the aggregate hold more reserves than desired. The function of Treasury sales is to substitute bonds for reserves; Treasury spending cannot be constrained by nongovernment unwillingness to lend. The exposure drafts appear to wish to resolve problems that do not, and perhaps cannot, exist. At the same time, they ignore some real problems, to which we now turn.

Misuse of Economic Projections and Assumptions

The exposure drafts provide no guidance on the choice of economic assumptions to be used in making projections. This is a serious shortcoming, particularly insofar as it has become a habit for the Social Security actuaries to violate generally accepted accounting practices when making economic projections relevant to the financial flows of the Social Security system. Specifically, past performance is characteristically ignored and future projections are systematically pessimistic with respect to past performance. This has led in recent years to repeated, systematic revisions of the financial projections for Social Security that are always in the direction of rolling back the projected dates when benefits exceed payroll taxes and the so-called OASDI Trust Fund is exhausted. This pattern has been systematic, so much so that it is reasonable to conclude that the actuaries have been systematically and persistently pessimistic. FASAB guidance on this point should specifically address two issues: the proper relationship of economic projections to generally accepted accounting principles, and the appropriate ways in which to factor into projections the effect of policy changes on economic performance. We turn next to this question.

As a matter of plain English definition, one cannot assess the “impact on the country of the government’s operations and investments” without assessing the economic effects of such operations and investments. If, for instance, a “stimulus bill” produces a higher rate of growth and lower rate of unemployment than would otherwise be the case, then that is surely an “impact on the country of the government’s operations and investments.” What else could it be?

The procedures in the exposure drafts explicitly propose to ignore those impacts. That is, irrespective of the government action, the economic projections used to assess that action will not be changed. The assumption will be made, however arbitrarily, that there is no effect of that action on the rate of economic growth, on the rate of employment and unemployment, on the mix between consumption and investment, or on any other pertinent real economic variable. The inference will therefore be drawn that the program necessarily involves costs—entirely nominal, associated with the debt—without real economic benefits associated with higher growth or lower unemployment. This procedure is prima facie absurd. It can serve only to confuse public debate and to obstruct, rather than advance, public purpose.

It might be acceptable, and even necessary, for an individual household or firm to ignore possible effects of its actions on the rest of the economy. But when dealing with an entire sector, and especially with the public sector, this cannot be the case. The actions of the government sector taken as a whole cannot be assessed in isolation from their consequences for the nongovernment sector and the performance of the economy. For example, the government sector might want to run a surplus, but it cannot achieve this unless the domestic nongovernment sector and/or the foreign sector will run a deficit. So long as the nongovernment sector (including the foreign sector) wishes to save, it is futile and counterproductive, as well as unnecessary and pointless, for the government sector to wish to save as well.

Government spending can be excessive. But the consequence of excess government spending is not a refusal (on the part of foreign creditors or anyone else) to hold the bonds associated with deficit spending. It is, rather, a possible devaluation of the dollar and a possible decline of the country’s real terms of trade. But this possibility—an appropriate concern up to a point and under certain conditions—is also ruled out by the FASAB’s proposed assumption of unchanged economic conditions. Unlike the nonissues discussed above, this is a real concern, and one that deserves actual attention. So, again, the proposed reporting fails to promote understanding of the nation’s financial condition.

Note also that, in recent months, even as the U.S. budget deficit has grown and as the possibility of a large fiscal package implies much larger future deficits, interest rates have fallen and the dollar has appreciated. Clearly, the low short-term interest rates are due to a Federal Reserve decision to lower them, and nothing else. The rise in the dollar, despite sharply lower interest rates, is due to the fact that the rest of the world has run to the world’s safest asset, U.S. Treasuries—driving down long-term U.S. interest rates. These trends should be described as votes of confidence in the U.S. dollar and in the strength of the U.S. Treasury. Of course, the foreign holding of U.S. debt results from the willingness of foreigners to sell their excess output to us and to accumulate dollar assets; it, too, is an attribute of their confidence in the dollar as a reserve asset. The two exposure drafts do not consider these matters, nor do they provide any guidance regarding how to consider the U.S dollar’s role as an international reserve asset.

In a world of financial interdependence it is of course essential that accounting standards applied to the federal government show an understanding of the basic position of the U.S. dollar and dollar-denominated assets in the world economy.

Backdoor Policymaking: What is the “Fiscal Gap,” and What Does “Fiscal Sustainability” Mean?

In the exposure drafts, the FASAB introduces the concept of a “fiscal gap” and states as a policy norm that it would be desirable to “maintain public debt at or below a target percentage of gross domestic product.” There is an apparent belief that this is an accepted, perhaps even noncontroversial, position. And we agree. To set a target for the debt-to-GDP ratio, which implies that public debt can grow alongside GDP, recognizes that public deficits, rather than balanced budgets, are normal.(Notes 4)

Yet no such policy objective exists in any statute of the U.S. government. Nor can such an objective be justified by reference to any known economic theory or operational constraint. There are times, including the present, when the level of debt in relation to GDP should rise to advance public purpose. There are times when it should fall. There are times when it will fall or rise irrespective of policy. To repeat, there is no justification in law or theory for attempting to legislate an accounting standard with a debt-to-GDP ratio as a target for economic policy.

Further, the exposure drafts fail to distinguish between total public debt, public debt held by the public, guaranteed agency debt, and implicit liabilities in the form of guarantees. The guidance in ED 1 refers to these concepts as “alternatives” but fails to take a position as to which alternatives are meaningful and which are not. As such, the measure of the so-called “fiscal gap” is essentially meaningless.

The concept of “receipts” in the calculation of the fiscal gap should also be clarified. It should, of course, include bonds issued as well as tax receipts; though we repeat that these, even taken together, do not present an operational constraint on spending.

Putting issues of bonds and also reserves and currency into the total for receipts, of course, would make clear that the concept of fiscal gap, as well as its measure, is meaningless, since tax receipts plus “receipts from borrowing” (broadly defined as new issues of currency, reserves, and bonds) are by accounting identity equal to total spending. But, while the accounting information will always show that federal government spending equals tax receipts plus new issues of currency, reserves, and Treasuries, the exact ratio between federal government spending and any one of the items on the other side of that equation is largely determined by spending and portfolio decisions of the nongovernment sectors (Notes 5). Those ratios are “sustainable” so long as the nongovernment sector seeks to sustain them. For example, if the nongovernment sector prefers to accumulate cash and reserves, there will be no need to issue Treasuries (“borrowing” will fall); if the nongovernment sector prefers a shift toward Treasuries and away from cash and reserves, then more Treasuries will be issued (“borrowing” will rise). If the nongovernment sector decides to reduce its saving, it will spend more, the economy will grow faster, tax receipts will rise, and the budget deficit will shrink. These outcomes are equally plausible and equally sustainable.

The exposure drafts define “fiscal sustainability” as a condition of policy under certain arbitrary economic assumptions such that “public debt does not rise continuously as a share of GDP.” The difficulty here is that the assumption of a stable inflation rate under hypothetical conditions of excessive fiscal expansion is untenable. If fiscal expansion is excessive, inflation, and therefore nominal GDP, rises, and the public debt will eventually cease to rise as a share of GDP. This effect is known (to economists) as the inflation tax. The inflation tax is an automatic stabilizer, which prevents excessive growth of real demand, which is necessarily limited by actual output. Inflation is unpleasant, but an unlimited debt-to-GDP ratio is not a consequence of it. The inflation tax therefore vitiates the problem of “fiscal sustainability” as defined in ED 1.

Dividing Up the Budget in Arbitrary Ways

The FASAB proposes a minimum level of disaggregation for the basic financial statement. For example, projected receipts and spending for major programs such as Medicare and Social Security would be shown separately from all other government programs. This proposal also reflects a substantial misunderstanding of the purposes of federal budgeting.

The purpose of a program budget is to discipline the program. It is to hold managers accountable and to discourage fraud. This is why specific amounts of funds are appropriated to specific programs. Without budgetary constraints (as well as oversight and other means of exercising control), it is likely that “mission creep” would lead to continual expansion of any particular program. Thus, it is certainly appropriate to hold programs accountable to ensure that they do what they are supposed to do.

However, there is little public or government interest in reporting long-range projections of the “fiscal balance” of particular portions of the budget. And while officials in any program should be held accountable after the fact, there is little public purpose and no economic interest served by reporting the resulting, after-the-fact fiscal balance of particular portions of the federal budget. For example, if Congress appropriates $100 billion for a transportation project, those in charge should provide an ex post accounting for all spending. They should explain the reasons for cost overruns and their careers should depend on acceptable performance. However, whether the total tax revenue received from any particular source (e.g., gasoline taxes) equals spending on transportation over some arbitrary period adds nothing to this.

We do understand the desire to provide an ex ante projection of total federal government spending and revenues for coming quarters, or even years. This facilitates analysis of the expected impact of fiscal operations on aggregate demand and thus on economic growth, inflation, and employment. There may also be some interest in disaggregating spending and taxing to match in an ex ante manner transportation-related spending and transportation-related tax receipts. This is not because fuel taxes actually “pay for” transportation spending, but because such a process can perhaps help to discipline the budgeting process by “allocating,” in an ex ante sense, expected revenues among program spending. However, the success of the transportation projects should not be measured by the ex post balance between total spending and total tax receipts related to transportation over the course of a fiscal year or any other arbitrarily chosen period. It might be very poor public policy to cancel a vital transportation project merely because projected fuel tax revenues fall short of expected program spending.

By extension, the long-term success of Social Security should not depend on, nor be assessed by, the government’s matching spending on that program against some portion of federal tax revenue. The economic effects of budget deficits are the same whether they result from Social Security spending that exceeds payroll tax revenue or from transportation spending that exceeds transportation taxes. If, over time, we should find that projected payroll tax receipts fall significantly short of desired Social Security spending, then it would no longer make sense to adopt a budgeting procedure that dedicates—in a purely planning sense—payroll tax receipts to the Social Security program. In other words, whether we are setting fuel taxes or payroll taxes, the tax rate should be administered in such a manner that it achieves the public interest, not with a view to matching spending in any particular federal program. Likewise, when deciding how much to spend on transportation or Social Security, the program budget should be set to achieve the public purpose, rather than constraining spending to projected receipts from a specific tax.

So far as transfer programs are concerned, and given that both assets and liabilities should be reported and that we are concerned ultimately with the financial condition of “the nation,” a few exercises will demonstrate that assets and liabilities necessarily balance. The government’s deficit is the private sector’s surplus and, vice versa, contingent liabilities of the government are contingent assets to the public.(Notes 6) Therefore, it would seem unnecessary to present many alternatives since all would show the same thing. The “basic financial statement” proposed in the exposure drafts defies understanding. Efforts to make it clear are therefore somewhat beside the point. (The public purpose might be better served by efforts to make it confusing.) We naturally oppose the inclusion of “scare” charts such as those included in the drafts.

Arbitrary, Capricious, and Misleading Time Horizons

The FASAB’s proposed time horizons are also problematic. They are so long that they will involve making assumptions that are, in the nature of things, impossible.

An example is the assumption of current Medicare forecasts that health care costs will continue to rise indefinitely and more rapidly than nominal GDP, so that the share of health care in GDP rises without limit. While the focus of the exposure drafts is on implications for the federal budget, the effect on the private sector would be worse. In the limit, there would be few or no resources left to produce food, shelter, industrial goods, or education, and the health care burden on households and firms would become intolerable. This cannot happen; therefore, it will not happen. Stein’s Law applies: when a trend cannot continue, it will stop.

No understanding of the issues is gained by a procedure that necessarily incorporates unrealistic assumptions of this type. Since the time horizons are arbitrary, the present value of future “liabilities” can be blown up to any size, simply by changing time horizons and discount rates. Most readers of the proposed budgetary documents are unlikely to be aware that the exercise is purely arithmetic in this sense.

For Social Security and other permanent programs, what matters for long-range projections are demographics, technology, and economic growth.(Notes 7) Financing is virtually irrelevant. If by 2083 every U.S. citizen is over age 67, no financing scheme will allow us to meet our commitment to let people retire at a decent living standard at age 67. This, however, is most unlikely. Indeed, all plausible projections of demographic trends show only gradual and moderately rising real burdens on those of normal working age in terms of numbers of dependents (aged plus young) per worker. The Old Age, Survivors, and Disability Insurance (OASDI) part of Social Security currently moves less than 4.5 percent of GDP to beneficiaries, and that rises to about 6.5 percent over the next 75 years. On the one hand, this is a significant increase, but on the other hand, similar shifts have occurred in the past without generating economic crisis or intolerable burdens. And it still leaves over 93 percent of GDP outside OASDI.

Moreover, in economic terms a rise in this burden is substantially less worrisome when considered in the context of a falling stock market, which reduces dividend income and capital gains available to the wealthier elderly. The current crisis drives home the necessity of having the Social Security leg of the retirement stool—a leg that promises to deliver benefits no matter how poorly the economy performs. While the promise is in financial terms because of the manner in which benefits are calculated, benefits will tend to rise in real terms as the economy’s productive capacity rises. As the population ages, there will necessarily be a rise in the real burden of supporting them.8 The other legs of the retirement stool (private pensions and individual savings) cannot guarantee that the real needs of elders will be met. First, this is because financial markets are subject to wild swings—so that many will retire at inopportune times (when assets are falling in value). Second, there is no mechanism operating in financial markets to ensure that asset values rise sufficiently faster than prices of consumer goods, shifting a larger share of the nation’s output to the retired. Indeed, it is precisely the ability of Social Security to increase the share of output going to beneficiaries (that is, to raise the real burden) that will be required as the nation ages. Finally, if all of our projections turn out to be incorrect, Social Security benefits can be changed (increased or reduced as necessary) as a matter of public policy rather than as a result of the performance of financial markets.

The growing “real burden” of providing for an aging population is captured by the projection that, while we have three workers today “supporting” each beneficiary, the number will fall to only two workers sometime around midcentury. Two questions follow from this. First, can we expect productivity to rise enough over the next half century to ensure that two workers will, indeed, produce as much as three workers do today? All reasonable projections—including those of the SSA’s Trustees—do assume this. Indeed, over the past half century, productivities of workers in manufacturing have doubled or tripled, depending on the industry—far more than what is necessary to guarantee that we will have enough output to raise the living standards of retirees, workers, and other dependents. Suppose (however unlikely the event) that productivity does not rise by the necessary amount. Is there any purely financial change we can make to the program—including privatization—that will avoid a “crisis”? The answer is clearly no. Getting more money into the hands of future retirees would just mean that they would bid more of tomorrow’s production away from workers and other dependents, leaving those groups worse off. To be sure, there would be policy actions that could attenuate the crisis by raising the ratio of workers to retirees (immigration of workers in 2050, for example), but financial expedients are not among them. If worse comes to the worst, so that we have fewer workers per beneficiary and no increase in productivity in 2050, then taxes will have to be raised or benefits cut—or some combination of the two—to apportion the pain of lower living standards. But it is best left to voters in 2050 to make such a decision.

In short, it serves no useful purpose to project financial shortfalls for Social Security and Medicare into a far distant future, and no purpose whatever to revise those programs today on the basis of such projections. We summarize the reasons:

First, Social Security spending need not be politically constrained by tax receipts from any particular source.

Second, so far as fiscal impacts on the economy go, what matters is the overall fiscal stance of the government, not the stance attributed to one part of the budget.

Third, the most important factors determining future real burdens are demographic and technological, not financial. Uncertainties about demographic and technological trends increase exponentially as the length of the projection period increases. Almost any projection of birth rates, family size, death rates, or labor productivity for 2085 would be equally plausible, and very slight changes to trend rates for any of these variables would make huge differences for projections of real burdens. For this reason, basing policy today on such projections is almost certainly swinging in the dark.

Fourth, if we do face problems in the distant future due to aging of the population, they are not financial problems. The federal government will always be able to make all benefit payments as they come due; the only question is whether the payments correspond to an appropriate share of total product at that time.

Conclusion: The Folly of Intergenerational Accounting

Many of the FASAB’s proposed procedures appear to rely on the notion of intergenerational accounting. This exercise attempts to assess financial burdens through time, especially with a view to claiming that financial decisions taken in one generation can impose burdens on another. But this argument is specious. It refuses to count as real assets the infrastructure and other national assets that the current generation will leave for future generations. And it does not understand that federal government debt never needs to be retired, any more than private sector net saving needs to be eliminated.

In real terms, there obviously are no intergenerational transfers, except for the knowledge, physical assets, and larger environment that the present leaves to the future. The real goods produced in 2050 will be distributed to those alive in 2050, regardless of the public debt in existence at that time. Then, just as now, the deficits of the state will fund the nominal savings of the nongovernment sectors. In short, intergenerational accounting is a deeply flawed, experimental, and unsound concept. It should not be included in any government accounting.

In general, and in conclusion, the FASAB’s exposure drafts have not made a persuasive argument about basic matters of accounting. The Board should work on getting these matters straight and stay very far away from the additional challenges of determining public policy. We mean no disrespect to members of the Board, for many others have been seduced by equating household balance sheets with those of the federal government. No household can forever spend beyond its income plus its ability to borrow. But that fact is simply irrelevant to a discussion of federal government spending. Federal spending can, and almost always does, exceed tax receipts. And that is almost always a good thing because it provides the wherewithal to allow the nongovernment sector to save in the form of highly desired, safe, dollar-denominated financial assets. Further, there is an important counterbalancing asset to the government’s liability: the accumulated financial, physical, and human capital of our nation that is available to be called upon should we ever need to mobilize capital to serve the public purpose.

The notion that there is some “unfunded liability” amounting to tens of trillions of dollars is hogwash. There cannot be any “underfunding.” The U.S. government always has the operational ability to make all payments as they come due, and could do so even if one concluded that government liabilities exceed private assets through some strange accounting mistake or trick.

We apologize for the blunt tone of these remarks, but these are important matters. We fear the FASAB has been led astray by intergenerational warriors, who must not be allowed to take control of our federal budgetary process. The danger is, of course, very real, for the application of “intergenerational accounting” to Social Security and Medicare can only mean the gutting of these vital programs, which are the mainstays of life security for America’s elderly—and for the working population that hopes to be elderly some day.


  1. According to its website, “The mission of the FASAB is to promulgate federal accounting standards after considering the financial and budgetary information needs of citizens, congressional oversight groups, executive agencies, and the needs of other users of federal financial information. Accounting and financial reporting standards are essential for public accountability and for an efficient and effective functioning of our democratic system of government. Thus, federal accounting standards and financial reporting play a major role in fulfilling the government’s duty to be publicly accountable and can be used to assess (1) the government’s accountability and its efficiency and effectiveness, and (2) the economic, political, and social consequences of the allocation and various uses of federal resources.

  2. Since the funds to pay taxes come only from public spending, this would also imply a refusal to pay taxes.

  3. Looking overseas, it might be interesting, for example, to know whether there is a point at which, despite continuing surpluses in China’s trade with the United States, the People’s Bank of China might become unwilling to add to its stock of U.S. Treasury bonds (and whether, if that were to happen, it would matter). There is no mention, let alone analysis, of the policies of the People’s Bank in these documents. Indeed, we note that all indications of the intention of the People’s Bank are to the contrary: China continues to pursue policy that will allow it to accumulate dollar reserves and bonds.

  4. Thus, if the public debt is 50 percent of a $16 trillion GDP and the nominal growth rate is 5 percent, it would be normal under the proposed guideline for deficits to equal $400 billion per year. Recognizing this would certainly represent progress compared to a desire to balance the budget. It is obvious, though, that this implicit recommendation conflicts with the main thrust of the exposure drafts.

  5. In fact, it is the Federal Reserve’s job to accommodate these decisions as part of interest rate targeting, through what it calls “offsetting operating factors.”

  6. In an open economy, foreigners can accumulate a portion of the government’s debt. This opens the possibility that the U.S. current account deficit could reverse to a surplus, with foreigners using their dollar claims to increase consumption of U.S. output. That would stimulate U.S. production and growth. How it would affect projections of U.S. government tax receipts and spending (and whether that matters) is not considered by the drafts. It could affect terms of trade and real living standards - again, matters that are not considered. It would not affect the government’s ability to make all promised payments as they come due.

  7. In an open economy, imports of goods and services are also relevant for the support of retirees. Even if the ratio of retirees to U.S. workers is rising, the real burden of providing for Social Security beneficiaries need not rise if foreigners want to sell their output to the United States in exchange for reserves

  8. The caveat discussed in the previous note again applies. We will not continue to mention this point in the discussion that follows.