[*PG1347]THE SOCIAL INSURANCE CRISIS AND
THE PROBLEM OF COLLECTIVE SAVING:
A COMMENTARY ON
SHAVIRO’S
RECKLESS DISREGARD
Abstract: Long-range Social Security and Medicare spending projections vastly exceed projected program revenues. If left unchecked, the resulting fiscal imbalance (estimated at $40 to $70 trillion in present value terms) would fall primarily on future generations. To avoid generational inequity, and perhaps fiscal meltdown, Professor Daniel N. Shaviro and others propose immediate fiscal austerity. This reply Commentary argues that near-term austerity is unlikely to play a significant role in overcoming the fiscal imbalance, which can be thought of as a balloon payment due in the mid-twenty-first century. Significant near-term fiscal austerity would eliminate the public debt and replace it with a public surplus. Political economy theory and U.S. public debt history suggest that this path is infeasible. This Commentary also stresses the importance of disaggregating the “Social Security and Medicare” problems. Contrary to popular belief, Medicare is by far the larger problem, and the Medicare imbalance is driven by projected spending increases, outpacing overall economic growth indefinitely. These observations suggest that a focus on Medicare cost control, rather than revenue enhancement, is called for.
Despite a projected social insurance fiscal crisis resulting from increased longevity and rising healthcare costs, the Bush administration has continued to advocate tax cuts and recently has sponsored and signed an unfunded prescription drug entitlement, exacerbating the current deficit situation. In Reckless Disregard: The Bush administration’s Policy of Cutting Taxes in the Face of an Enormous Fiscal Gap, Professor Daniel N. Shaviro argues that rather than decreasing the size of the government through limiting discretionary spending as the Bush administration apparently hopes, the growing deficits and fiscal gap [*PG1348](the present value of projected expenditures over revenues) will result in an enormous intergenerational transfer that actually represents larger government and unfairly burdens future generations with the cost of providing the baby boom generation with a secure and healthy retirement.1 Professor Shaviro argues that the Bush administration and Congress should reverse course and exercise fiscal restraint.2
This reply Commentary focuses on two issues that, although tangential to Professor Shaviro’s analysis, ultimately relate to both his fairness and size of government arguments. First, commentators routinely speak of a “Social Security and Medicare” problem in the singular, but it is important in evaluating potential solutions and in considering Professor Shaviro’s “size of government” argument to disaggregate the programs.3 Contrary to popular belief, Medicare is by far the larger problem, and the Medicare fiscal gap is driven largely by projected spending increases, outpacing increases in gross domestic product (“GDP”) for many years to come.4 This suggests that Medicare may be an appropriate “big government” target and that cost control should be the focus of our attack.5
Second, it is questionable whether near-term fiscal austerity is likely to have much bearing on the ultimate solutions to our social insurance problems or on how the burden of those solutions is divided among present and future generations.6 To reduce the burden on future generations in a meaningful way through near-term fiscal austerity would require a combination of tax increases and spending cuts that would replace the public debt with a near-term public surplus.7 The United States’ history of public (dis)saving and political economy theory suggest that this strategy is infeasible, at least within the current budgetary framework.8 We cannot “save” our way out of these problems, and thus, we should begin curtailing future benefits today.9
Professor Shaviro eloquently described the problem facing us, and this section will reiterate only a few key points. As a result of increased longevity, escalating healthcare costs, and the impending retirement of the baby boomers, our pay-as-you-go social insurance programs will not be sustainable under current funding and spending rules. The Congressional Budget Office (the “CBO”) has projected combined Social Security, Medicare, and Medicaid spending to rise from 8% of GDP in 2004 to 9% by 2012, 17% by 2040, and 25% by 2075.10 These obligations do not show up in snapshot deficit and public debt figures; in fact, Social Security and Medicare currently are producing surpluses which mask the size of the “rest of government” deficits.11 Recent analyses, however, put the “fiscal gap”—the current public debt plus the present value of projected non-interest expenditures minus projected revenues under current law—at $44 to $74 trillion.12 This shortfall is almost entirely attributable to the social insurance funding gap and dwarfs the actual public debt, which at the end of 2003 was about $4 trillion.13
This path is not sustainable and ultimately must result in a combination of tax increases and spending cuts. In addition, unless soon reversed, the fiscal imbalance will result in an unprecedented transfer of wealth from future to current generations. One can think of the situation as pay-as-you-go financing with an enormous balloon payment (on the order of $40 to $70 trillion) accruing interest and due in, say, 2050. [*PG1350]The question is whether those working in 2050 and beyond can and should bear this entire burden, or whether and to what extent those working and those retired today should bear a share through increased taxes, reduced benefits, or both. (Of course, previous generations also contributed to this problem, but it is too late to do anything about that.)
If the U.S. government stays on the present course, spending soon will overwhelm funding for both Social Security and Medicare, but the problems under the two systems actually have quite different dimensions. Whether viewed in terms of its share of the fiscal gap or its growing share of total federal expenditures, Medicare is by far the larger problem, and Medicare is a much longer-term problem than Social Security.
In a 2002 study, Jagadeesh Gokhale and Kent Smetters estimated that 83% of the long-term fiscal imbalance was attributable to Medicare and 16% to Social Security.14 Remaining federal programs were basically in balance. There are two problems, however, with this analysis.15 First, although Medicare Part B (Supplementary Medical Insurance) is funded through participant premiums and transfers from general funds rather than through payroll taxes, Gokhale and Smetters include only the premium receipts in calculating the Part B gap.16 This approach could be viewed as overstating the overall Medicare imbalance, but this is not too serious a problem because the fiscal gap attributable to Medicare Part A alone is almost three times the size of the Social Security gap.17 Second, and more important, dividing the imbalance between Social Security and Medicare could be considered arbitrary, at least on the funding side. Most Americans do not distinguish between payroll taxes dedicated to Social Security and those dedicated to Medicare.
Thus, it may be more appropriate to focus solely on expenditures, although the outcome is much the same. The CBO recently projected that spending for Social Security, absent changes, will rise [*PG1351]from 4.2% of GDP in 2003 to 6.2% of GDP in 2050.18 At that point, however, few of the baby boomers will still be living, and Social Security spending through the rest of the century would be basically flat.
Because of the tie to actual healthcare expenditures, Medicare cost projections are much more uncertain. Under their “middle-cost scenario,” however, the CBO projects Medicare spending to increase from 2.4% of GDP today to 8.3% in 2050.19 Medicare cost increases, moreover, are expected to continue to outpace GDP growth for the remainder of the century, leading Ronald Lee and Ryan Edwards to conclude that retirement programs, including Social Security, will account for only about one-eighth of the growth in federal spending over the twenty-first century, with healthcare for the elderly accounting for the bulk of the rest.20
Public perception of the relative magnitude of the Social Security and Medicare problems appears to be exactly the opposite of reality. Friends and colleagues whom I have queried uniformly believe that Social Security represents the larger problem, and that was my view before embarking on this project. It is interesting to consider briefly why this misperception exists. New York Times columnist Paul Krugman essentially blames a vast right wing conspiracy that is attempting to scare Americans into accepting privatization of Social Security by lumping the programs together and making the fiscal situation look more desperate than it is.21 But, if so, why have defenders of Social Security not made more of the fact that this discrete program is relatively solvent? Perhaps liberals believe it will be easier to save Medicare by tying its fate to a more politically popular Social Security program.
Whatever the reason for the common misperception, recognizing that Medicare represents the lion’s share of the social insurance problem in this country has important implications. Unlike Social Security, the Medicare problem is driven in large part by real growth in expenditures, is relatively distant, and, as a result, is more uncertain.22 As [*PG1352]Professor Shaviro observes, Social Security payments are pure cash transfers that have little allocative effect on the economy.23 Medicare, however, is another story. Again, as Professor Shaviro notes, “Its incentive and income effects have surely been huge contributors to the enormous growth of healthcare expenditure, both absolutely and relative to the economy, in recent decades.”24
That growth indeed has been enormous. Total national health expenditures (“NHE”) have grown from 5.1% of GDP to 14.1% between 1960 and 2001.25 The differential growth rate has slowed in recent years, but even in the last decade, average annual NHE growth exceeded GDP growth by 1.5%.26 In addition, Medicare spending has grown more rapidly than private healthcare spending, with Medicare costs increasing from 4.1% of total NHE in 1967 (the year of Medicare’s inception) to 17% in 2001.27 Of course, a large part of differential Medicare growth can be attributed to demographics, but even on a cost-per-enrollee basis, Medicare cost increases continue to outpace both GDP and overall NHE growth rates.28 The 1% excess growth rate projected by the Medicare Trustees would actually be a significant improvement over recent history: between 1980 and 2003, Medicare spending growth outpaced GDP growth by 2.3 percentage points.29
The history and projections suggest two conclusions. First, efforts to manage the Medicare fiscal gap should be focused on the expenditure side of the equation. Second, and related, Medicare is an appropriate “big government” target. Professor Shaviro argues that, rather than decreasing the size of government by limiting spending as the Bush administration seems to hope, the intergenerational redistribu[*PG1353]tion that inevitably will arise under current policies actually represents larger government.30 Frankly, however, this is not the kind of “big government” issue that Republicans salivate over. A public Medicare system comprising 8% or more of GDP is.
The other notable aspect of the Medicare situation is that the Medicare balloon payment facing future generations largely is attributable to funding and spending imbalances that will occur after current generations are gone.31 As a result of the persistent differential between the projected growth in Medicare costs and GDP, only about 40% of the present value difference between projected expenditures and revenues is attributable to past and current generations.32 Forty percent is a large burden to pass on to future generations, but even if one is skeptical of our ability to rein in Medicare costs, generational fairness does not require living generations to account for the entire Medicare fiscal gap.
Part II of this Commentary demonstrates that the U.S. social insurance crisis is largely a Medicare crisis and suggests that expenditure control is the most promising avenue of attack, but it does not explain why closing the gap through increased funding is not equally plausible.33 This Part illustrates why significant pre-funding of future social insurance obligations through near-term fiscal austerity is unrealistic.
Returning to the balloon payment analogy, one can envision projected fiscal obligations as funded through a pay-as-you-go system plus a $40 to $70 trillion balloon payment accruing interest and due mid-twenty-first century.34 Short of an economic miracle, it is extremely difficult to imagine mid-twenty-first century taxpayers paying off this balloon on their own, and even if they could, the result would be generationally unfair. If Congress were to institute a significant program of fiscal austerity today, we conceivably could spread that burden over several generations, thus reducing the risk of a fiscal meltdown and increasing generational fairness. Our public (dis)saving history and the political economy of public saving, however, suggest that near-term fiscal austerity is unlikely to play a significant role in alleviating the burden on subsequent generations.
At the end of 2003, the public debt stood at just under $4 trillion.35 Professor Shaviro cites two recent studies placing the “stock” fiscal gap—the present value of future tax increases or spending cuts needed to maintain the public debt at the current fraction of GDP—at $44 to $74 trillion, ten to almost twenty times the current debt, and notes that these estimates fail to reflect the latest round of tax cuts and the new Medicare prescription drug benefit.36 No one suggests, of course, that we catch up all at once. In 2002, Alan Auerbach, William Gale, and Peter Orszag projected that a permanent combination of tax increases and spending cuts amounting to 11% of GDP was necessary to close the fiscal gap.37 This was the level of austerity needed, according to that projection, to maintain the public debt at the current fraction of GDP over the long term, and maintaining a stable ratio of debt to GDP over the long term is required to keep the cost of servicing that debt from overwhelming the non-interest portion of the budget.38
Moreover, every year of delay in implementing austerity increases the ultimate pain. If Congress does not act until 2012, the required combination of perpetual tax increases and spending cuts necessary to eliminate the fiscal gap in Auerbach, Gale, and Orszag’s most plausible scenario would rise from 11% to 12.6% of GDP.39 Although not expressly stated, Professor Shaviro presumably believes that we should reverse course immediately and institute a fiscal austerity program on this order.
The situation has worsened since Auerbach, Gale, and Orszag published their 2002 study. The passage of an unfunded Medicare prescription drug benefit must have increased the fiscal gap substantially. Ignoring this increase, however, and using the figures discussed above, imagine if Congress were to enact a combination of tax increases and spending cuts amounting to 11% of GDP today and to maintain this program in perpetuity. Because the large increases in fiscal demands [*PG1355]are not projected to occur for twenty or thirty years, we would rapidly build a public surplus. Unless this surplus were somehow sequestered from the general public debt calculations, by 2007 the public debt would be eliminated entirely. By 2014 our current $4 trillion public debt would be replaced with a $10 trillion public surplus.40 Ultimately, this collective savings bubble would be drawn down as the baby boomers retire, but is the creation of this savings bubble feasible within the confines of our current budget and social insurance structures? Our national public (dis)saving experience says no.
This nation has carried a public debt consistently since its inception.41 The United States never has been in a surplus situation, although we came close to eliminating the debt in the 1830s.42 Over the last forty years public debt has swelled dramatically in nominal terms, from $250 billion to almost $4 trillion, but debt as a fraction of GDP has been relatively steady, ranging from about 25% to 50% of GDP and averaging about 36%.43 Moreover, annual budget surpluses have been rare, occurring only five times during this forty-year period.44 As an empirical observation, it would seem that the federal government is inexperienced, to the say the least, at creating public nest eggs.
One might think, however, that the Social Security and Medicare trust funds have been a mechanism for increasing public saving. In recent years, payroll tax receipts have far outpaced Social Security and Medicare drawdowns, resulting in significant increases in the trust funds. At the end of 2003, the combined trust fund balances exceeded $1.8 trillion.45 The trust funds, of course, are no more than [*PG1356]inter-governmental accounting entries. The trust fund balances are “invested” in government securities, but the T-bills, notes, or other securities representing the trust fund balances are not set aside. They simply represent an “I.O.U.” from the federal government’s general revenue to the social insurance systems. The projections suggest that if we remain on our current path, the trust funds will be exhausted in twenty-two years (Medicare) to thirty-eight years (Social Security).46 The narrow question, however, is whether the existence of these trust funds and recent large balances have had any effect on public saving.
Apparently, they have not. In 1985, the combined trust fund balances stood at $74 billion, less than 2% of GDP.47 The total public debt at the end of that year was $1.5 trillion, 36% of GDP.48 At the end of 2003, the trust fund balances were $1.8 trillion, representing about 17% of GDP.49 Total public debt was $3.9 trillion at the end of last year, again about 36% of GDP.50 The 15% of GDP increase in the trust fund balances appears to have been completely offset by other tax cuts or spending increases. Put another way, if the government were to borrow enough money to fund fully (and to segregate) the trust fund balances, public debt would jump to 53% of GDP.51 Eyeball empiricism suggests that the massive increases in the Social Security and Medicare trust funds have had no impact on the overall level of public debt.
If actual dollars need to be set aside (instead of simply passed between generations in a pay-as-you-go system) to promote generational fairness or to avoid a social insurance crisis in the future, it is not clear [*PG1357]that a collective approach to saving is superior to individual saving, or that a collective approach is even feasible. Individuals may suffer from myopia, but they have strong incentives to ensure their own well-being.52 Individuals who accurately assess their future needs can save for the future with relative security. It is possible that their savings will be eroded by inflation or confiscatory wealth taxation, but generally, they get to spend in retirement what they set aside today, as well as the earnings on those funds. Thus, they have every incentive to plan and to save.
Collective public saving, however, is more problematic. First, myopia does not disappear in the public setting. Perhaps the Bush administration’s apparent indifference to the coming fiscal crisis reflects public myopia. The Bush administration may be irrationally optimistic about the prospects for economic growth (despite the fact that much of the budget grows with GDP), or it may believe that technical innovation ultimately will reduce the costs of healthcare.
In fact, individual myopia likely is magnified in the public sphere. Professor Shaviro and others urge abandoning a myopic fixation with current deficits, debt levels, and ten-year budget outlooks, but I have yet to see a public official campaigning on a net present value budget analysis platform, and the salience of deficit and debt figures cannot be ignored.53 If anything, the evidence suggests worsening public myopia as sunset provisions are added to tax legislation to improve the optics of cuts while attempting to tie the hands of future Congresses.54 Few people would manage their own household budgets in this fashion. People may be individually myopic, but they do not simply ignore budgetary issues more than ten years out.
Second, collective public saving could be viewed as a collective action problem. Why should I agree to pay more taxes today and help build a collective nest egg, when other generations may shirk their obligations or divert the surplus we have built to Mars exploration or foreign wars? The shirking problem is particularly acute because the fiscal demands do not peak for many years.55 The Bush administration’s policies may reflect shirking instead of, or in addition to, myo[*PG1358]pia, and shirking would not be irrational here, because there is no way to bind future generations to a course of fiscal restraint.
Third, in crass political terms, the problem with collective saving is that future generations do not vote. Our pay-as-you-go system has worked effectively for as long as it has because the political muscle of the current beneficiaries was fairly well balanced against that of current earners and payers. Even if a considerable fraction of voters is foresighted and altruistic, this is unlikely to offset the effect of (rationally) selfish voters. Simply put, can we imagine the U.S. Congress sitting on a public surplus of $10 trillion in order to smooth out the social insurance cost bubble? I think it is extremely unlikely that surpluses would survive to meet the social insurance funding needs of future generations, and I suspect most taxpayers and voters would share my cynicism.56
Building a public surplus is probably not politically feasible. In fact, it would be a tremendous victory for fiscal austerity if we were to reduce the public debt to 25% of GDP by 2014. But this would entail a combination of tax increases and spending cuts of less than $2 trillion over the ten-year period against the CBO budget baseline, or perhaps $4 trillion against a more realistic baseline projection including certain 2001 and 2003 tax act extensions.57 Austerity of this magnitude would appear to be lost in the round of calculations of the impending fiscal gap of $44 to $74 trillion.58 It seems unlikely, then, that near-term fiscal austerity will have much to do with how we deal with our social insurance crisis.
It is disappointing, but not surprising, that the current administration continues to push tax cuts and spending increases in the face of looming Social Security and Medicare insolvency. This is sympto[*PG1359]matic of the difficulty of collective saving, and I would not expect to see significant fiscal austerity under a second Bush or even a Kerry administration. So, what can, should, or will be done?
The most likely scenario is that little will be done over the next ten to twenty years and that we will maintain public debt at about 35% of GDP, while the projected fiscal imbalance continues to grow. This scenario is troubling, but perhaps not as dire or unfair as Professor Shaviro suggests. Ultimately, benefits will have to be cut and taxes raised, but this scenario puts the most pressure on Medicare spending, which is the biggest part of the problem. Inevitable cuts in the baby boomers’ benefits may cause some hardship, but their voting strength will limit the hardship, and the benefit cuts will not be generationally unfair. The biggest problems in this scenario are that the working generation mid-century is likely to face very high tax rates and public debt will be pushed to untested limits.
A more responsible approach would be to begin curtailing future social insurance benefits now. Federal Reserve Chairman Alan Greenspan recently made such a proposal, and such action might be politically feasible if the cutbacks do not affect current retirees.59 Curtailing future benefits today would tend to shift some of the burden borne by mid-twenty-first century workers in the “do nothing” scenario to the baby boomers, thus increasing generational fairness.
The most responsible, and generationally fair, approach would be to begin curtailing future benefits today and to shore up funding. As discussed above, I am extremely skeptical of our ability to pre-fund future obligations within the current Social Security and Medicare structures.60 It is conceivable that the political economy problems could be overcome through substantial reform that results in effective sequestration of the trust funds from the “rest of government” accounts or even privatization, but it is unlikely. It is more promising, therefore, to focus efforts on the spending and benefit side of the equation.
Before closing, let me respond briefly to the principal point emphasized by Professor Shaviro in the present Article. He argues that the intergenerational redistribution that is inevitable if we do not change course should be viewed as “big government,” just like direct [*PG1360]government spending, and that, as a result, the present course leads to larger, not smaller, government.61 I agree in theory, but I am not persuaded fully by the argument for several reasons.
First, it is unclear how to weigh the distributional effects of intergenerational transfers against the allocational effects of direct government involvement in programs like Medicare. I am sure, however, how the Bush administration would weigh them. They would consider the projected Medicare excess growth rates (relative to GDP growth) a classic “big government” issue.
Second, in adding up the various allocational and distributional effects, we should recognize that near-term fiscal restraint also has “big government” implications if restraint results in a public surplus. Suppose Congress does begin building a public surplus that reduces intergenerational redistribution. Professor Shaviro believes this will forestall an increase in the size of government, but we also must consider the size-of-government effects of the public surplus administration apparatus.62 Currently, the Social Security and Medicare trust funds are “invested” in government securities. The fund managers are bookkeepers, and management costs are low. In a world of public surpluses, funds would have to be invested in non-governmental securities, which would require an elaborate investment management apparatus.
Collective public surplus management may be more efficient than private investment management, but this efficiency advantage does not address sufficiently the “big government” concern. If we envision the same level and distribution of collective savings taking place within a purely public system, a mixed public/private system (for example, mandatory individual accounts), or a purely private system, the allocative effects may be the same, but the governmental thumbprint and big government charge will loom largest in the first scenario. My intuition is that the size of the government is more that just the sum of its allocational and distributional effects. One also must consider the degree of interference with individual autonomy.
Finally, it is worth noting that discretionary spending already is coming under pressure as a result of growing deficits and public debt.63 Clearly, discretionary spending cuts cannot cure the fiscal im[*PG1361]balance, but if the Republican Party is to retain the mantle of the party of fiscal responsibility, it must regain its focus on direct government outlays in general, and discretionary spending in particular.
Professor Shaviro’s article adds to a long list of significant contributions he has made to our understanding of Social Security, Medicare, and the tax systems. I certainly share Professor Shaviro’s concern with the outlook for U.S. social insurance and the burden we appear to be creating for future generations, and I agree with much of his analysis. Our differences lie in emphasis as much as substance.
In this Commentary, I have attempted to demonstrate the importance of disaggregating the Social Security and Medicare problems. Disaggregation is complicated by the fact that payroll taxes are somewhat arbitrarily divided between the systems, but clearly, Medicare represents the larger problem. This matters because Medicare’s projected excess growth rates cry out for an expenditure-side solution, and the significant allocational effects of Medicare suggest that it is a legitimate “big government” target.
I also have argued that near-term fiscal austerity, although desirable for a host of reasons, is unlikely to have much impact on the resolution of our long-range social insurance crisis. I reach this conclusion simply because the current public debt is such a small part of the long-term fiscal gap that we would have to create a sizeable near-term public surplus to have any real impact. History and political economy theory suggest that this is unlikely to happen. Of course, we should not dig a deeper hole for our descendents to fill, but it seems more profitable to focus on gradually reducing social insurance benefits and controlling costs than attempting to improve social insurance funding.
Professor Shaviro’s argument that deficits are unlikely to reduce the overall size of government, if we take into account the redistributional effects of massive intergenerational transfers, is provocative and certainly could be correct, but the calculus is complex and largely ignores the role of individual autonomy. Even ignoring non-economic effects, however, one cannot predict which path—near-term fiscal austerity or starving the beast—results in “bigger government.” Thus, when thinking about deficits, debt, and the growing fiscal gap, I fall back to consideration of what is efficient, fair, and achievable. That analysis is difficult enough.