[*PG1285]RECKLESS DISREGARD: THE BUSH ADMINISTRATION’S POLICY OF CUTTING TAXES IN THE FACE OF AN ENORMOUS FISCAL GAP
Abstract: The Bush administration’s policy of sharply cutting taxes while increasing government spending is both misguided and harmful. Presumably rationalized in private as a way of shrinking government over the long term without paying a current political price, this policy in fact increases the government’s distributional intervention by handing money to current voters at the expense of younger and future generations. Moreover, the ballooning fiscal gap may lead to an Argentina-style meltdown in the U.S. government’s position as a borrower in world capital markets, potentially yielding chronic inflation, unemployment, and bank and currency crises that may affect our economic productivity for an indefinite period.
The dominant feature of fiscal policy in the George W. Bush administration has been cutting taxes while vastly increasing current and projected future spending, in the face of an already enormous fiscal gap. Many commentators have noted the slide of the annual budget picture from surpluses on the order of $230 billion to deficits exceeding $400 billion.1 The trend in the annual numbers is trivial, however, compared to the long-term picture, which was bad even during the surplus years, but has grown significantly worse, in large part due to the Bush administration’s fiscal recklessness.2
[*PG1286] Serious disruption is likely to result from the stress that the Bush-enhanced fiscal gap places on the U.S. political system. There is a major possibility of an Argentina-style meltdown in the U.S. government’s position as a borrower in world capital markets, potentially yielding chronic inflation, unemployment, and bank and currency crises.3 The increased fiscal gap also will harm future generations and will cause many current workers, who otherwise could have counted on lesser (though still significant) reductions in Social Security and Medicare benefits, to enter retirement with inadequate resources.4
What could be the Bush administration’s rationale for expanding, rather than seeking to narrow, the fiscal gap? Apart from indifference or misunderstanding, along with short-term political expedience and an eagerness to reward friends, the idea seems to be that the tax cuts will create strong pressures on the spending side, leading to a smaller government down the road. (Never mind that the Bush administration has increased even non-defense spending, while also vastly expanding Medicare through the enactment of an unfunded new prescription drug entitlement.)5
Unfortunately, the Bush administration’s ideological fervor for tax cuts has caused it to overlook the fact that the likely consequence of its tax enactments will be a larger government.6 The contrary view of the Bush administration and conservative activists appears to rest on spending illusion, or confusing the amount of the nominal dollar flows between individuals and the government with the actual size of government.7 After establishing a more plausible interpretation of government size as a function of the government’s allocative and distributional effects on the society, it becomes clear that the tax cuts’ main effect over time is likely to be increased wealth redistribution from younger to older generations.8 This is no less a case of government activism than taking money from the rich and giving to the [*PG1287]poor, with the important difference that—even though future generations may be richer than we are—it is less defensible normatively.
In the end, the Bush administration’s tax policy, if assumed to be at least partly motivated by principle rather than merely short-term political convenience, brings to mind the words of George Santayana that “[f]anaticism consists in redoubling your efforts when you have forgotten your aim.”9 Tax-cutting fervor has been divorced completely from the underlying premises that might make it appealing, most obviously through the lack of any serious effort at spending discipline, but more fundamentally through the Bush administration’s courting of a fiscal crisis that is unlikely to advance its presumed ideological goals. It has created seemingly permanent deficits, but they will not be permanent because they cannot be. Even in the short run, the 2001–2003 tax acts have so enhanced instability in the fiscal system that political rent extraction by Washington politicians from a variety of interest groups can be expected to reach new heights.
The remainder of this Article proceeds as follows. Part I discusses how the long-term fiscal picture is best analyzed, where the United States now stands, and the impact of the 2001–2003 tax cuts.10 Part II examines the “size of government” concept and the likely effects of the tax cuts.11 Finally, Part III considers the significance of the fiscal gap.12
Annual budget deficits and surpluses are the main measures used to evaluate the overall long-term budget picture. Deficits have resonated periodically as an American political issue for more than two centuries.13 As recently as 1995, a balanced budget amendment to the U.S. Constitution passed the House of Representatives and fell only one vote short of Senate approval.14 Many of the amendment’s sup[*PG1288]porters, of course, subsequently became avid cheerleaders for bloating the deficit through the enactment and expansion of George W. Bush’s tax cut proposals.15
The political prominence of deficits partly reflects their symbolic significance—connoting to the likes of Ross Perot, for example, that the government must be wasteful, inefficient, and unbusinesslike.16 People also are prone to exaggerate the analogy between public and private debt, overlooking that the federal government does not face the same type of default risk as a private individual because it borrows in its own currency and can raise taxes.17 Nonetheless, concern about deficits reflects their perceived relationship with four underlying issues of genuine substance.18 These issues are as follows:
(a) Generational policy—There has long been concern about leaving future generations the burden of repaying public debt that was incurred to finance budget deficits.19 This concern would disappear to the extent of Ricardian offsets, or increased saving by altruistically minded members of current generations to offset the debt burden left for their heirs.20 Empirical evidence with strong theoretical backing, however, suggests that Ricardian offsets to the government’s fiscal transfers from future to current generations are extremely limited.21
(b) Macroeconomic issues—Deficits and surpluses are thought to affect the macroeconomy in various ways. For example, Keynesian fiscal policy traditionally emphasizes the counter-cyclical use of deficits to counter recession and surpluses to cool off inflation.22 Other macroeconomic concerns about deficits emphasize whether, as public dissaving, they reduce national saving, and whether they lead to higher interest rates that might make future government borrowing more costly and might raise the “hurdle rate” for new business investment.23
(c) Size of government—Conservative opposition to budget deficits, and greater liberal tolerance thereof, long reflected the belief that [*PG1289]debt financing encouraged higher government spending. The basic idea was that spending programs would be easier to enact if current taxpayers were not handed the bill immediately and if the assignment of the cost to anyone could be deferred. Over the last ten years, however, this has changed fundamentally. Many conservatives (and the Bush administration) instead have embraced the hypothesis that cutting taxes, and thereby creating large budget deficits, is the best way to restrain government spending over the long term. Meanwhile, liberals apparently agreeing with the conservatives that taxes rather than spending levels are the main political variable at the margin, have increasingly favored lower deficits and higher surpluses.
(d) Sustainability of current policy, or degree of specification of a feasible long-term policy—Over the long run, the present value of the government’s inflows must equal that of its outlays. This is the basic no-free-lunch principle, or “budget constraint” in economic parlance, which holds that everything ultimately must be paid for. Any fiscal policy that fails to satisfy this constraint is unsustainable and will have to change.24
For individuals, the government’s failure to announce a sustainable fiscal policy means that their long-term personal and business planning must go forward without the aid of a credible projection of its long-term policy course.25 Those who are reasonably attentive must decide how to deal with substantial uncertainty, while the inattentive may face severe shocks when fiscal policy finally does address the funding gap.26 An example would be entering retirement with inadequate savings by reason of Social Security and Medicare cutbacks that one failed to anticipate. Worse than any of this, however, is the possibility of a severe macroeconomic crisis if investors in world capital markets lose their confidence that the government is willing and able to place its fiscal policy on a sustainable course.
All four of the above issues are important. Unfortunately, however, the annual budget deficit or surplus is poorly designed to illuminate any of them.27 The core problem is that it is not an economically meaningful measure for two reasons. First, it is only a current-year measure and thus ignores the accrual of future expected cash flows [*PG1290]between the government and other persons.28 Second, it is sensitive to labeling because amounts denominated as loan principal are ignored.29 Suppose, for example, that the government officially re-described the payment of Social Security taxes followed by the receipt of Social Security benefits, as involving taxpayer loans followed by the repayment of principal plus interest (plus or minus a transfer to the extent the benefits did not equal the taxes in value). The result would be to change dramatically the reported budget deficits or surpluses for all years, even though nothing would have changed on the ground.
Seventy years ago, these shortcomings of the annual budget measure might not have mattered as much, despite their inviting “smoke and mirrors” manipulations, such as postponing outlays for a day so as to fall in the next fiscal year. With the enactment of Social Security and Medicare as long-term entitlement programs, however, the issue of accruing future commitments became too important to ignore. As experience has shown, consideration of all four of the above issues can be distorted badly by focusing on annual budget outcomes.
(a) Generational policy—Because Social Security and Medicare are the two primary tools in our fiscal policy for wealth transfer between generations, a focus on budget deficits can lead to a radical misunderstanding of the burdens that we are imposing on future generations. A classic example arose in the 1950s, when President Dwight D. Eisenhower, though apparently sincerely concerned about deficits for this reason, overlooked the fact that, through expansions of Social Security, he had “presided over a very large redistribution from [younger and future] generations to the elderly of his day.”30 Indeed, this “redistribution in the 1950s toward the old and away from the young and unborn was more significant than that of the [deficit-ridden] 1980s.”31
Suppose Congress in 2003 had provided strict pay-as-you-go payroll tax financing for the new Medicare prescription drug benefit that it enacted, thus avoiding any impact on the budget deficit for any year. The result would nonetheless have been a sizeable wealth transfer from younger to older generations. Current seniors still would have received the benefit without paying for it, while younger generations would have had to pay over time both for their own benefits and for those given to current seniors. The effect, therefore, would have been comparable to [*PG1291]that of debt-financed deficit spending on behalf of current seniors, but debt and deficit measures would have missed it completely.
(b) Macroeconomic issues—Counter-cyclical fiscal policy is probably the area in which deficits retain the greatest relevance as measures.32 Keynesian stimulus relies in large part on the idea that people are myopic or cash-constrained, and hence will adjust their current consumer spending, based on this year’s cash flows between themselves and the government, more than they “ought” to if acting rationally on a farsighted basis.33 Even here, however, the annual measure may be misleading insofar as people are not totally myopic, and often take some account, even if erratically and unpredictably, of changes to their long-term expectations regarding government policy.34 Furthermore, because more farsighted economic actors, including those with significant stakes in the capital markets, adjust their behavior as well, the fact that some people are myopic or cash-constrained may cease to have as great an effect on aggregate behavior as one otherwise might have expected.
In any event, however, the case for discretionary counter-cyclical policy, or relying on new enactments rather than automatic responses of tax revenues and certain outlays (such as unemployment benefits) to the business cycle, was widely discredited by the 1980s on political economy grounds.35 The main concerns were that stimulus would come too late (like a thermostat that detects cold weather in January but does not turn on until July), and that politicians would use it disingenuously as a one-way ratchet inviting deficit spending but never contraction.36 Discretionary stimulus then enjoyed a pseudo-revival under the Bush administration—albeit one that merely confirmed the latter political economy concern—with two distinct strains. The first was that “[n]o politician wishes to be cast in the title role of It’s the Economy, Stupid.”37 Thus, Congress in 2002, “remind[ing] us that policy makers may go where economists fear to tread,” actually passed a conventional stimulus bill with incentives for new investment and ex[*PG1292]tension of unemployment benefits.38 The second, and perhaps more significant, strain emerged in 2003, when the Bush administration, determined to cut taxes but lacking the budget surplus rationale of 2001, hit upon the pretext that more stimulus was needed, notwithstanding that its new proposals had been designed to meet totally different objectives. For example, they were directed neither to individuals with high marginal propensities to consume nor, at the business end, to new investment. The take-away lesson from 2003 was that, when there is public concern about recession or unemployment levels, the President (if he controls the policy agenda and lacks intellectual scruples) can make whatever tax cut proposal he likes, labeling it “stimulus” and holding over his foes the threat of blaming them for any problems in the economy if it does not pass.
(c) Size of government—Deficit reduction has never been more than tangentially related to restricting the size of government.39 Under some circumstances, it certainly is plausible that new spending programs will be harder to enact if, under an annual balanced budget requirement, they would have to be paid for immediately. Yet this is merely one scenario among many, as the last few years have shown through the increasing identification of budget deficits with lower taxes rather than higher spending. William G. Gale and Brennan Kelly suggest, for example, that budget-busting tax cuts tend to be associated with budget-busting spending increases as periods of fiscal lenity alternate with periods of fiscal discipline.40
More fundamentally, however, the effects of current policy choices on both tax and spending levels must be evaluated over the long term, not just for a single year. Thus, consider the inadequacy of the annual budget measure to constrain the enactment of new entitlements, such as Medicare in 1965 or Medicare prescription drugs in 2003, which start small but can be expected to grow exponentially, along with their political constituencies, over time.41
[*PG1293] (d) Sustainability of current policy, or degree of specification of a feasible long-term policy—A government that has the capacity to borrow (or to print money that people accept) does not need to balance its budget annually. Furthermore, its performance in this regard in a given year, or even the level of its accumulated public debt, may offer little information concerning whether its current or announced fiscal policy is on a sustainable course. This is especially true in an era in which programs like Social Security and Medicare, representing stated long-term commitments and purporting to use dedicated financing, occupy so large a place in the federal budget.42
This problem has been officially recognized up to a point, through the creation of trust fund reporting for Social Security and Medicare that includes annual projections of program solvency over a seventy-five year period.43 These projections, however, even when considered alongside the annual budget measure, fall considerably short of offering adequate information. There are three main problems with relying on the seventy-five year Social Security and Medicare projections. First, they are limited to taxes and spending under arbitrarily defined subsets of the fiscal system as a whole. They therefore fail to offer a comprehensive picture, with the consequence that they can be (and have been) manipulated by simply changing the revenues or outlays that are attributed officially to a particular program.44 Second, because the projections purport to measure whether the programs have been self-financing since their inception, they include historical [*PG1294]information about specified revenues relative to outlays (shown as positive trust fund balances) that have no direct bearing on whether current policy is sustainable into the future. Third, seventy-five years simply is not a long enough period over which to evaluate these programs. When long-term fiscal imbalances are as severe as they are under existing Social Security and Medicare, truncating the projection horizon at seventy-five years may result in understating the present value of the imbalance by two-thirds or more.45
Thus, annual deficits and surpluses fail to provide adequate information regarding fiscal policy. For the concerns relating to their macroeconomic and size-of-government effects, it is not clear what an improved measure would look like. For generational policy and under-specification, however, improved measures not only are possible, but have been developed recently.
To measure the relative lifetime treatment of members of different age cohorts, the tool of choice is generational accounting, which offers a measure of lifetime net taxes and lifetime net tax rates.46 Lifetime net taxes equal taxes paid minus transfers received, computed in present value terms from birth.47 Dividing this amount by one’s lifetime income, likewise computed in present value terms from birth, provides a measure of one’s lifetime net tax rate.48 Thus, suppose that on average the members of a given age cohort paid lifetime gross taxes of $3.5 million, received lifetime transfers of $1 million, and had lifetime income of $10 million. The average lifetime net tax for members of the cohort would be $2.5 million, and the lifetime net tax rate would be 25%.
If generational accounting computations were made by assuming that current policy will continue indefinitely, they would ignore the [*PG1295]principle that everything ultimately must be paid for. Yet current policy does not describe how the fiscal gap will be eliminated. Moreover, assuming a particular response would mean that one was no longer projecting current policy, but instead making a prediction about future policy. Generational accounting responds to this conundrum by treating the entire revenue shortfall as eliminated through a net tax increase on future generations.49
The operating assumption, therefore, is that everyone alive today, including newborns, will not have to pay for any of the fiscal gap, leaving those born next year or thereafter to bear it in full.50 This is concededly unrealistic and is meant to provide “an informative counterfactual, not a likely policy scenario.”51 It does, however, indicate the trend implied by current policy because delaying the necessary fiscal changes tends to leave the burden to be borne by future generations. To the extent that future generations are shown to have higher lifetime net tax rates than current generations, it is clear that higher burdens either are assigned to them expressly, or will have to be borne by them if the fiscal gap is not addressed while present generations remain on the scene.
The President’s Annual Budget briefly included generational accounting forecasts, but this practice ceased during the William J. Clinton administration, reportedly because senior officials found them too embarrassing given the political emphasis on current budget surpluses.52 Generational accounting proponents continued for a while to issue periodic forecasts, using Congressional Budget Office (“CBO”) projections but with certain revisions to the present course of tax policy and discretionary spending as well as to assumed discount and growth rates. A forecast shortly before the enactment of the Economic Growth and Tax Relief Reconciliation Act of 2001 (the “2001 Act”) showed lifetime net tax rates of 17.68% for current newborns and 35.81% for future generations, suggesting that the fiscal gap (the key source of the difference) would more than double lifetime net tax burdens, perpetually into the future, if not addressed while members of current genera[*PG1296]tions are still alive.53 No estimates have been issued since that time, but the disparity is probably much greater now because the fiscal gap (discussed below) has increased several-fold, reflecting both revised long-term forecasts and the impact of the 2001–2003 tax changes.
A measure of lifetime net tax rates may be a bit too abstract to offer a compelling snapshot view of big policy choices like the recent tax cuts or Medicare prescription drug benefit. For that purpose, a measure stated in dollars would be preferable. Although other candidates have been suggested, my own suggestion involves the use of what I call the “generational transfer” measure for policy changes.54 This is a present-value measure of how a given proposal would change generational distribution on a going-forward basis relative to prior policy.
To illustrate both the measure and its limitations, a current illustration of the generational transfer measure may help. Jagadeesh Gokhale and Kent Smetters projected that, as of 2002, current generations, defined as people then over the age of fourteen, would receive net benefits, or transfers in excess of taxes, worth $10.1 trillion from Social Security and $15.4 trillion from Medicare.55 Current generations also were projected to pay, over the rest of their lives, taxes (other than for Social Security and Medicare) with a present value of $32.6 trillion.56 Overall, then, these people would pay remaining lifetime net taxes of $7.1 trillion (the excess of the other taxes over the net Social Security and Medicare benefits).57
Considered in isolation, this figure has little meaning. To begin with, it carries no implication that current generations are net losers [*PG1297]from fiscal policy, even on a going-forward basis. After all, it excludes the value of in-kind services received from the government. Such value would not have to be high—indeed, twenty-two percent of the value of the $32.6 trillion in taxes would suffice—in order for current generations to be net winners from fiscal policy on a going-forward basis.58 Even that, however, would carry no implication that current generations are net winners on a lifetime basis. The basic point of Social Security and Medicare is to provide individuals with benefits after retirement in exchange for taxes paid during their working years. Thus, the fact that a group, such as current generations, that includes many older people should be net winners from the present moment until the end of their lives does not really illuminate how these people have been treated overall.
Suppose, however, that a given policy change, such as expanding the Medicare prescription drug benefit, would reduce the remaining lifetime net tax for current generations from $7.1 trillion to, say, $5 trillion. This would indicate that the policy involved a $2.1 trillion transfer to people over the age of fourteen, evidently from those who are younger or as yet unborn. In a sense, this repeats the assumption of generational accounting that future generations will pay in full for the change, but there is less risk of confusion regarding the point that is being made. Rather than seeming to project the actual course of future policy (as people often seem to think generational accounting does), we are simply measuring what this enactment would do absent other future enactments.
One further refinement would make the generational transfer measure more informative still. Suppose we disaggregated further into more distinct age groups, such as the following four groups: (a) future generations, defined as the unborn plus those who are currently fourteen or younger; (b) the young, defined as those from ages fourteen to forty; (c) the middle-aged, or those from ages forty to sixty; and (d) the old, or those age sixty and above. This potentially would be a most illuminating measure.
There are two problems with the generational transfer measure. First, increasing the fiscal gap would appear to be a win for all current voters. Emphasizing that the gain came at the expense of future generations, however, would help illustrate the core point about generational policy, which is that (leaving aside efficiency issues) it is a zero-sum game. It would have been interesting to hear the Bush admini[*PG1298]stration explain why transferring trillions of dollars from future to current generations through the 2001–2003 tax cuts and the Medicare prescription drug benefit was a desirable policy move.
Second, consider a tax cut that is financed by reducing in-kind government services to the people who would have paid the taxes. The generational transfer measure treats those people as benefiting from the full amount of the tax cut, rather than simply from the excess, if any, of the taxes over the value of the services they otherwise would have received. This oversight is unfortunate, but unavoidable with fiscal measures that operate against the background of valuation difficulties for government services. If this point makes the generational transfer measure undesirably confusing, then presumably all tax distribution tables should be eliminated as well.
In Do Deficits Matter?, I called for a measure of our fiscal policy’s current under-specification, projected forward over an unlimited time horizon, although I showed an unfortunate rhetorical tin ear by calling the proposed measure “tax lag” or the “economic accrual national debt.”59 One of the first major efforts by economists to make such a computation more felicitously called it the “fiscal gap,” defined as “the size of the long-run increase in taxes or reductions in non-interest expenditures (as a constant share of GDP) that would be required immediately” in order to keep current government debt constant as a percentage of gross domestic product (“GDP”).60 In this formulation, “the fiscal gap is stated as a flow, like the annual budget deficit or the amount you must pay each year on a bank loan. The fiscal gap can also be stated as a stock, like the national debt or the principal you owe on a bank loan.”61
Taking into account the 2001 tax changes, but assuming that they would remain in effect permanently rather than being “sunsetted” after 2010, and also making certain other reasonable adjustments to official CBO projections, the “flow” fiscal gap as of 2002 was estimated to stand at 11.07% of GDP.62 This implied a “stock” fiscal gap of about [*PG1299]$74 trillion if one discounted future cash flows at a 3% rate and assumed that real GDP would grow at an annual rate of 1.5%.63
The most recent estimates (as of early 2004), suggest a total fiscal gap of $73 trillion.64 This amount represents the sum of the following totals (rounded to the nearest trillion dollars):
(1) According to the latest report of the Social Security Trustees, an unfunded Social Security obligation through the infinite horizon of $10.4 trillion;65
(2) According to the latest report of the Medicare Trustees, an unfunded obligation through the infinite horizon of $21.8 trillion for Part A of Medicare (pertaining to hospitalization benefits);66
(3) According to the latest report of the Medicare Trustees, a need for infinite horizon general revenue contributions of $23.2 trillion to pay for Part B of Medicare (pertaining to outpatient care);67
(4) According to the latest report of the Medicare Trustees, a need for infinite horizon general revenue contributions of $16.6 trillion to pay for the new prescription drug benefit;68 and
(5) According to a recent estimate by economists Jagadeesh Gokhale and Kent Smetters, an infinite horizon fiscal imbalance, as of 2004, of $753 billion for the entire federal government apart from Social Security and Medicare.69
In one key sense, the $73 trillion forecast is wildly over-optimistic. It projects future tax revenues by assuming the continuation of the laws on the books, although in two important respects these laws do not describe actual, currently intended policies in a credible way. [*PG1300]First, in passing the Bush tax cuts, Congress provided that the entire 2001 Act, and many of the provisions in the Jobs and Growth Tax Relief Reconciliation Act of 2003 (the “2003 Act”), would expire (barring future legislation) within periods ranging from two to nine years.70 The main reason for these “sunsets”—which proponents of the two Acts insisted would not be permitted to take effect—was to lower the official ten-year estimates of the Acts’ revenue cost by more than fifty percent.71 Thus, looking at either what the proponents openly intended or the likely resolution of the politically unrealistic policy path suggested by the sunsets, the sunsets result in a significantly higher estimate of the fiscal gap.
Second, the $73 trillion forecast reflects assumptions that nothing will be done to rein in the rapid growth of the alternative minimum tax (the “AMT”), a parallel tax system to the regular income tax that uses lower rates but a broader base, ostensibly to ensure that people do not over-use tax preferences.72 Until 2000, less than one percent of taxpayers paid the AMT in any given year.73 Under current law, however, the CBO estimates that the number of taxpayers paying the AMT will grow from just over one million in 2001 to nearly thirty million in 2010—at which time the AMT will be paid by more than 60% of taxpayers earning between $50,000 and $100,000, more than 90% of those earning between $100,000 and $200,000, and more than 80% of those earning between $200,000 and $500,000.74 This level of AMT application seems likely to be politically intolerable given the compliance burdens of functioning under two parallel tax systems. A fix would be extremely costly, however. It was recently estimated that, for the years 2005 through 2014, repeal of the AMT would cost $660 billion if the 2001 Act is allowed to expire, and $1.09 trillion if the Act is extended.75 They also have estimated that, by 2008, repealing the AMT would be costlier than repealing the regular income tax while retaining the AMT.76
[*PG1301] The growth of the AMT has two main causes. First, an income exemption amount that it uses to reduce the number of individuals who are subject to it is not indexed to inflation, and thus continually declines in real value unless expressly increased. Second, by sharply cutting the regular tax without addressing the AMT, the Bush administration put many more people on the latter system because the AMT is payable when it exceeds regular tax liability. If people have to pay the greater of A and B, sufficiently cutting A inevitably shifts people to B.
It would have been bad enough if the Bush administration, when deciding how to structure its 2001 and 2003 tax cuts, merely had failed to prioritize adequately the need to address the AMT. This already would have been irresponsible, given the magnitude of the AMT problem and the difficulty of solving it straightforwardly given the overall fiscal gap. What is even worse, however, is that the AMT, like the sunsets, appears to have been a deliberate tool in the Bush administration’s characteristic, Iraq War-style willingness to engage in crude, shameless, and shortsighted deception. Specifically, the AMT aided enactment of the 2001 and 2003 tax cuts by gulling voters and taxpayers who did not read the fine print. Projected AMT revenues permitted the official estimates of the cost of the enactments to be hundreds of billions of dollars lower, without widespread understanding that, for example, by 2009 people earning between $100,000 and $500,000 would be losing more than 50% of their Bush tax cuts to the AMT.77
Obviously, any estimate of the U.S. fiscal gap over an infinite time horizon is extremely sensitive to economic and demographic assumptions and highly subject to significant change from year to year even if policy remains constant. This is no ground for dismissing its significance, however. As Alan Auerbach and Kevin Hassett have noted, risk aversion suggests that we should be more concerned, not less, about a risky fiscal gap than we would be about one with the same expected value but that was fixed.78
Moreover, the prediction of a huge fiscal gap is quite resilient. It does not, for example, vary sharply with the business cycle in the [*PG1302]manner of the annual budget deficit or surplus. The resilience of the fiscal gap to new information or revised assumptions reflects the fact that its two principal causes, population aging due to increasing life expectancy and growth in U.S. healthcare expenditure relative to GDP, are considered highly resilient.79
Might economic growth provide the cure? Anglo-American history, such as that of England after the Napoleonic Wars and the United States after the Civil War, shows that enormous public debts can simply be outgrown if the economy takes off sufficiently.80 The current U.S. fiscal gap, however, appears to be growth-proof because various expenditure programs, no less than taxes, are pegged to GDP over the long run. In particular, Social Security benefits are indexed to rising wage levels and thus to productivity growth, while the Medicare fiscal gap is actually worsened by GDP growth if healthcare expenditures rise in conjunction with, but faster than, GDP. Recent estimates have shown, therefore, that a greater than expected rise in GDP actually might make the fiscal gap larger rather than smaller.81 Thus, we are likely to have to deal with the fiscal gap in the reasonably near future regardless of economic growth.
The only plausible change to projected trends that could reduce the fiscal gap drastically is a decline in the rate at which healthcare expenditures grow relative to the economy. In theory, it is possible that technological advances could have this effect. Consider, for example, the use of a cheap pill to clear up heart problems without expensive surgery, or the prospect of deferring end-of-life expenditures, which often are quite high, by enabling people to live longer.
Historically, however, technological advances have continually increased healthcare’s share of the economy.82 This trend is probably demand-driven and likely to continue until incentives in the healthcare industry are changed significantly. Medicare, which virtually eliminates seniors’ incentive to be cost-conscious in seeking cov[*PG1303]ered healthcare, is only part of the problem.83 Even for people under the age of sixty-five, cost-consciousness is the exception rather than the rule. The income tax exclusion for employer-provided healthcare gives employees an incentive to over-insure, even if this results in wasteful expenditure that increases the cost of their coverage.84 Plus, perhaps less significantly, Medicaid reduces the cost-consciousness of beneficiaries in seeking healthcare. When the great majority of healthcare consumers have so little reason to be cost-conscious, not only will they use the technology at hand in an uneconomical way, but the firms developing new treatments will tailor the products that they choose to develop for this audience. Thus, we cannot anticipate a change in the rate of healthcare expenditure growth—other than one forced on consumers by a wrenching political change—to render the fiscal gap less onerous than it now appears.
Why would the Bush administration have sought enormous tax cuts in the face of an enormous fiscal gap that was well known to leading economists even amid the budget surpluses of the late 1990s? Without getting too ad hominem about it, one should note that a key reason for the Bush administration’s eagerness to slash taxes was ideological, relating to anti-tax and anti-government sentiment that in recent decades increasingly has swept conservative and Republican thought. What if the advocates of the tax cuts had it backwards, however, and enacting enormous tax cuts in the face of a large fiscal gap was not actually a step toward smaller government?
For many decades, a dominant theme in American conservative thought and politics has been battling “big government.” Although partly waged on the regulatory front, the main action for at least three decades has centered on the fiscal system and is well conveyed by President Ronald Reagan’s famous, oft-repeated charge that the Democrats liked nothing better than to “tax and spend.”85 Conserva[*PG1304]tive Republican advocacy of tax cuts, after premiering at the national level in Ronald Reagan’s 1980 presidential campaign and the ensuing Economic Recovery Tax Act of 1981, became a core ideological and policy aim with the promulgation of the “Contract with America” in the 1994 congressional campaign, and then bore fruit in the enactment of the 2001–2003 tax cuts.
Controversial as the tax cuts have been, their supporters and opponents alike generally agree that they are steps toward smaller government. Although merely reducing the government’s tax take for now as spending continues to rise, they put our long-term fiscal policy on so unsustainable a course as to portend a need for much tighter spending controls in the future—including cuts in Social Security and Medicare because so much money is spent on those programs.86
This may be deliberate, reflecting anti-government sentiment, even though the Bush administration has simultaneously supported vast spending increases including a $16.6 trillion Medicare prescription drug benefit.87 Although the Bush administration itself has been circumspect about the pressures on future spending created by its tax policy, well-connected conservative activists are more forthcoming. Tax-cutting advocate Grover Norquist, for example, states that his “‘goal is to cut government in half in twenty-five years,’” and thus “‘to get it down to the size where we can drown it in the bathtub.’”88
The English writer Saki once observed, “When one’s friends and enemies agree on any particular point they are usually wrong.”89 So it is this time. The point of agreement between supporters and opponents of President Bush’s tax policy, that the 2001–2003 tax cuts are steps toward smaller government, reflects a shared misunderstanding of what the notion “size of government” can be reasonably interpreted to mean. More specifically, it rests on spending illusion, or confusing the amount of the nominal dollar flows between individuals and the government with the actual size of government. Once we develop a more plausible interpretation of government size, we can see [*PG1305]that the 2001–2003 tax cuts are probably steps on the road to larger government because their main effect will be to increase wealth redistribution from younger to older generations.
The size of government is an empirical idea, presumably concerned with the magnitude of the government’s effect on some set of outcomes. For convenience, one can divide the outcomes of greatest relevance to tax policy into the following two categories: allocative outcomes, concerning how society’s resources are used, and distributional outcomes, concerning how the resources are divided between people.
An immediate problem in evaluating the size of government, in terms of effects on allocation and distribution, is that one must ask, compared to what? A no-government state of nature may have initial logical appeal as the counterfactual baseline. This, however, seems to require a thought experiment—asking what society would look like if government did not exist—that requires not just speculation but motivation. Even if one could specify empirically just what a no-government world would look like, one next would have to explain why this would be relevant to our assessments of the actual world in which we live. Libertarians imagining the state of nature may grant it privileged normative status in determining which distributional policies are permitted or just.90 They are all too prone, however, as Liam Murphy and Thomas Nagel tartly note, to “imagine life roughly as it is now, with jobs, banks, houses, and cars, and lacking only the most obvious government services such as Social Security, the National Endowment for the Arts, and the police.”91 They overlook that property is a “legally constructed social relation,” and that without government, “there would be no right to use, enjoy, destroy, or dispose of the things we own.”92 If one interprets the hypothetical state of nature in light of what actually would happen if the government suddenly and magically disappeared or had never existed to begin with, it is plausible to think that we would find ourselves in a Hobbesian world where [*PG1306]“everyone’s level of welfare would be very low and . . . roughly equal” (because opportunities to earn and save would be limited) and indeed where the world’s population would be far lower than we observe under the actual circumstances.93
The no-government world would be too remote from the actual one to help in measuring the size of government even if we more optimistically imagined a Lockean state of nature in which people generally accepted that “no one ought to harm another in his life, health, liberty, or possessions.”94 As Barbara Fried notes, there is no discernible basis for assuming that people’s talents would have similar, or even the same proportionate, value in such a world as in the actual one.95 Moreover, there is no reason to think the economy would look anything like our actual one. Thus, we really have no way of comparing distributional or allocative outcomes in that world to our actual one, even leaving aside the motivation for the comparison.
Does this mean we cannot make any sense or practical use of the “size of government” concept? If it did, then exposure of the libertarian error in ignoring the government’s role in establishing markets and property rights would mean that one could not distinguish between, for example, the government’s allocative and distributional role in the 1880 American economy and in the 1980 Soviet economy. One need not embrace the libertarian error, however, in order to accept that the government, having established the prerequisites for a stable social order, can vary the level of its further interventions, both allocative and distributional.
A measure of the size of government ought to be designed in light of the main issues that we want it to illuminate. In this instance, libertarianism is at one pole of the underlying dispute, and socialism at the other. At the libertarian pole, market outcomes, once their preconditions have been established, prevail without any further government intervention. At the socialism pole, government decisionmakers not only establish background institutions, but also exert ongoing control over production and distribution. In between the two poles, where all [*PG1307]countries with market economies find themselves, governments attempt in varying degrees, through government production, regulatory commands, and fiscal tools such as taxes, transfers, and subsidies, to exert some influence over allocative and distributional outcomes. The deeper underlying question in steering between the two poles goes to the relative merits and defects of market and government processes, and to the value or disvalue of redistribution (relative to the libertarian pole), such as from rich to poor or healthy to sick.
Against this background, a measure of government size has two related advantages. It can provide some sort of index or comparative measure for showing where we are—as compared, for example, to other countries or to where a given policy change would put us. In addition, given that people likely are making assumptions about government size anyway (as in Ronald Reagan’s “tax and spend” charge), a well-grounded measure can prevent us from being misled by alternatives that are cruder and more formalistic, and hence, more manipulable.
Because libertarianism is at one pole in the underlying policy debate, its assumptions (although erroneous) can be used reasonably in constructing the baseline against which government size is measured. In effect, in the course of measuring the government’s interventions beyond simply establishing the prerequisites for a functioning market economy, one may indulge in the fiction that “life roughly as it is now, with jobs, banks, houses, and cars” would exist in any event.96 In economics jargon, this is a partial equilibrium analysis, or one that considers only the relatively direct ramifications of government rules and programs. As such, its use can be illuminating, given near-universal acceptance of at least the minimal libertarian state, so long as one keeps in mind that its baseline is merely a convenient reference point, too arbitrarily constructed to have the normative status often claimed by libertarians.
Actually constructing a size of government measure is well beyond the scope of this Article. The use of a single metric would be questionable in any event, given the multi-dimensional character of the problem. Consider, for example, the question of how to integrate [*PG1308]the degree of redistribution with the scope of civil liberties and with the legal system’s method of coordinating competing property claims. Instead, this Part briefly considers the government’s effects on allocation and distribution through explicit taxes and spending as well as regulatory provisions that can be conceptualized as tax or spending equivalents. My aim is simply to establish enough of a background understanding to permit preliminary consideration of the size of government effects of the 2001–2003 tax cuts, as opposed to trying to bring the measurement process anywhere near to fruition.
Allocatively, one might start by measuring the goods and services that the government directly supplies. Cost rather than value might need to be the metric, on the ground that efforts to specify the latter would be too difficult and controversial.97 To this, however, one might want to add the allocative consequences of people’s responses to government rules that change the relative prices of different commodities. The word “changes,” of course, implies a preexisting set of prices, but here we might think in terms of the neutrality notion that is commonly used in public economics. A lump sum tax, for example, such as a uniform head tax, does not affect relative prices, whereas consumption and income taxes raise the price of market consumption relative to leisure. Income taxation additionally raises the price of future consumption relative to current consumption.
The government also can affect allocation through income effects, such as the changes in consumption that result from redistribution between people with different preferences. An example would be reducing national saving, wholly without regard to incentives to save, by transferring wealth from high savers to low savers. Should income effects be included when we evaluate the size of government, even though, once the redistribution (accounted for separately) has occurred, they simply reflect consumer preferences and market forces? The answer to this question depends on what we actually care about when we debate the proper size of government.
Consider two examples. In the first example, the government takes money from people who do not like ice cream and gives it to people who do. As a result, even though ice cream lovers are free to spend the transfers however they like, more ice cream is sold, simply through the normal operations of our market economy. In the second [*PG1309]example, tracking actual Social Security, the government takes money from young people and gives it to old people. This reduces national saving because older people, having less of an extended future to worry about, tend not to save as much. The well-founded (though not uncontroversial) view that Social Security historically has done just this is prominent in the debate concerning program reform.98 Logically, it is exactly parallel to the ice cream example, but there is a significant difference. We care more about saving than about ice cream consumption—in economic parlance, because it may have positive externalities. When it comes to ice cream, leaving aside the health effects of calories and cholesterol, there is really no reason for anyone but the consumer to care how much ice cream he or she has. Let the consumer’s tastes determine the answer. Increased national saving, however, is thought by many to benefit people other than the savers themselves—in particular, younger people and future generations, who may benefit if productive use of the savings enables them to live in a more affluent society.
Suppose that someone, therefore, complains that the government, by creating a huge, unfunded Social Security system in 1935, reduced national saving to today’s detriment. To say that this is not really part of the “size of government” because it merely reflected market forces at work given the underlying wealth transfer would seem strained at best, desperate at worst. The bottom line, unfortunately, can be no clearer than this: we care about wealth effects when we care about them.
Distributionally, one might start by netting people’s taxes against their transfers, and then comparing their net taxes to their net benefits from other government policies, such as the provision of public goods. A system of pure benefit taxation, if it could be adequately defined in terms of properly relating net taxes to net benefits, might then emerge as the no-redistribution baseline. This, in turn, should be done on a lifetime basis, comparing lifetime net taxes to lifetime net benefits. This calculation would resemble generational accounting, which measures people’s lifetime net tax rates as a percentage of their lifetime incomes, while ignoring public goods provision because the benefit therefrom is too hard to measure. Due to the public goods problem, benefit taxation has proven difficult even to [*PG1310]define, and grave empirical measurement difficulties would remain even if a given definition were accepted.99
Libertarians sometimes claim that a flat-rate tax—whether on income, consumption, or something else—is as close to a benefit tax as one can practically come, and thus constitutes “an indispensable part of the Lockean program of taxation.”100 Barbara Fried has refuted this argument, suggesting that it is merely political fallback for people who oppose the distributional consequences of a progressive rate structure but assume that regressive rates and uniform head taxes are politically out of bounds.101 Still, where we observe large differences, going in the same direction, both in lifetime net taxes and in lifetime net tax rates, it is reasonable to infer that redistribution is afoot, going from the high-tax to the low-tax individuals, unless there is reason to think that the tax-side difference is being systematically offset by a matching benefit-side difference.
We therefore have a very rough basis for evaluating the likely effects of the 2001–2003 tax cuts on the allocative and distributional size of government. Do they seem likely to increase or reduce the government’s allocative effect on the economy via direct provision of goods and services plus responses to its effects on relative prices? And do they seem likely to increase or reduce redistribution, gauged by differences in lifetime net taxes or tax rates, if we have no reason to think that any such differences are being offset on the benefit side?
One more preliminary step remains before considering the tax cuts enacted in 2001–2003. I suggested above that the Grover Norquist theory that massive tax cuts can be used to “‘cut government in half . . . [so] we can drown it in the bathtub’” relies on spending illusion, or confusion between the actual size of government and the gross amounts of the nominal dollar flows between the government and private persons that are denominated as “taxes” and “spending.”102 There certainly are situations in which Grover Norquist would be correct in [*PG1311]assuming that we can reduce the size of government by cutting taxes and spending. For example, reducing government spending on national defense and public education by $50 billion per year and cutting taxes by the same amount presumably makes the government smaller, at least allocatively. Things are not always this straightforward, however, especially not in the currently prevailing long-term budgetary setting.
For an initial example of spending illusion, suppose we use preferential income tax rules (such as special deductions or credits) to pursue the very same allocative policies that might otherwise have been pursued through a broader-based, more neutral income tax plus explicit government appropriations. Use of the special tax benefits caused both taxes and spending to be nominally lower than under the direct appropriations route. Yet the government cannot really be smaller if we stipulate (as is entirely plausible) that the alternative sets of rules have the same allocative and distributional effects.
Social Security and Medicare, under which people pay taxes during their working years and then get benefits after reaching age sixty-five, help to suggest further illustrations, which for convenience initially can involve debt financing. Suppose Jill and Bill live in a two-person, two-period society. In Period One, the government spends $10.00 supplying a public good that benefits Jill and Bill equally. It finances the expenditure by borrowing the entire $10.00 from Jill at the market interest rate of 10% per period. In Period Two, when the government owes Jill $11.00 (and does not supply any further public goods), it must choose between (a) levying a uniform head tax of $5.50 in order to raise the money it owes her and (b) reneging on the debt.
Suppose first that the government reneges. Using formal, conventional definitions of taxes and spending, this means that it will have levied taxes of zero in both periods while spending $10.00 in Period One and zero in Period Two. By contrast, if the government levies the head tax, Period Two taxes rise from zero to $11.00, and Period Two spending rises from zero to $1.00.103 Thus, from the standpoint of spending illusion, reneging appears to lead to smaller government. In fact, however, it would mean that over the two periods, the government engaged in substantial redistribution from Jill to Bill, rather than zero redistribution, while supplying the same public goods.
[*PG1312] Libertarians might not like the decision to renege in Period Two because it involves the expropriation of a contract right, and thus a government taking equivalent to taxation.104 The conclusion, however, that Period Two reneging would make the government bigger, rather than smaller, did not rely upon the fact that Jill may have been told in Period One that she had a contract right to repayment. Suppose the government chose instead to levy a $10.00 tax on Jill in Period One. Then, in Period Two, suppose it had to choose between (a) doing nothing and (b) levying a tax of $5.50 on both Bill and Jill and handing the $11.00 to Jill (or alternatively just taking $5.50 from Bill and handing it to Jill). The allocative and distributional consequences of (b) generally would be the same as those from honoring the bond in the first example. Thus, despite the absence of a Period One promise to Jill to even out the distribution in Period Two, option (b) would continue to imply a smaller government overall. From the standpoint of nominal cash flows, however, option (b) would appear to involve even bigger government than in the earlier example. Indeed, if the government went through the motions of levying a $5.50 tax on both individuals and then handing $11.00 back to Jill (rather than simply taking $5.50 from Bill and giving it to Jill), then Period Two would involve both “taxes” and “spending” of $11.00.
The key to this example in each of its variants—and to current real-world tax cuts that lead to future reductions in Social Security and Medicare spending—is that the Period Two decision comes in midstream. It is easy to accept that government would have been smaller in the example if taxes and spending had been zero in both periods. Once the government takes action, however, an immediate shutdown— even one constrained by honoring express contractual commitments in place—does not necessarily lead to a smaller government. It may instead lead to increased redistribution if the alternative in Period Two would have been to even things out on an all-periods basis.
In evaluating how the 2001–2003 tax cuts will affect the size of government, an initial problem is that they are only part of the story. Over time, under what economists call the long-term budget constraint, the present value of government inflows and outlays must be equal. Only resources on hand can be spent, and everything ultimately must be paid for (the no-free-lunch principle). Accordingly, reducing cash inflows through the tax cuts implies compensating changes, in the form of reduced outlays or offsetting future tax increases relative to the case of non-enactment. What ought to be evaluated, then, is the entire package, not the 2001–2003 tax cuts standing alone. The problem analytically is that the rest of the package has not yet been specified and will not be specified any time soon.
Fortunately, if that is the right word, our long-term fiscal picture is clear enough to permit highly educated guesses about the broad character of the offsets. An initial salient fact concerns the tension between the long-term budget constraint and the announced long-term fiscal policy path of the U.S. government. As noted above, the fiscal gap or fiscal imbalance (“FI”) may stand at about $73 trillion, whereas a “sustainable fiscal policy requires FI to be zero.”105
At some point, therefore, taxes will have to increase or spending will have to decline. Moreover, to the extent that spending declines, discretionary domestic spending is too small a component to bear the major brunt. Social Security and Medicare cuts will almost certainly have to do most of the heavy lifting on the spending side. Thus, it was recently estimated, prior to the 2003 tax cut, that, in principle, the FI could be eliminated by (a) raising federal income tax collections by 68.5%, (b) raising payroll tax collections by 94%, (c) cutting discretionary spending by 104.1% (which is not only mathematically impossible but also would imply no defense budget), or (d) cutting Social Security and Medicare outlays by 45.3%.106
Accordingly, in specifying the full package of changes associated with the 2001–2003 tax cuts, two main possibilities merit attention. [*PG1314]The first is that the tax cuts are offset by future tax increases. The second is that they are offset by future Social Security and Medicare cuts.107 Each of these two options can then be evaluated for its size of government effects.
The enactment of substantial future tax increases should not be discounted on the ground that the current political environment is so vehemently anti-tax. Things may look very different once the payment of Social Security and Medicare benefits is at risk. I have suggested elsewhere that, within the next fifteen years, the enactment of a consumption-style value-added tax (“VAT”) on top of the existing income tax, and the use of inflation as a deliberate policy tool for partly reneging on current obligations, are significant possibilities.108 This, of course, is just speculation, and we really do not know what the tax increases will be. We can confidently predict, however, two things about them if enacted. First, because the tax increases will not take effect until the future, they will result in the application of higher tax rates to future than to current economic activity. Second, by applying mainly to younger or future taxpayers by today’s perspective, they will result in the application of generally higher lifetime net taxes and tax rates to younger than older generations. The former of these two points matters allocatively, while the latter matters distributionally.
For two reasons, the application of higher tax rates to future than to current economic activity is likely to cause added economic distortion, and thus is likely to increase the real size of government by magnifying its impact on economic production. First, the application of higher tax rates to future than to current activity may induce taxpayers to shift taxable transactions from high-tax to low-tax years, especially as the transition nears and begins to take a more definite and predictable form. Second, even if economic activity cannot shift between years, the application of higher rates to some years and lower rates to other years tends to increase total economic distortion. It is a public economics truism that the deadweight loss from a tax generally rises more than proportionately with the rate of the tax, and indeed with the square of the rate.109 Thus, for example, “[d]oubling a tax [*PG1315]quadruples its excess burden, other things being the same.”110 This suggests that overall distortion will be higher if the rates are high in some years and low in others than if they are constant at the intermediate rate required for long-term revenue equivalence. That is, the reduction in distortion in low-rate years will be more than offset by the increase in distortion in high-rate years.
Accordingly, to the extent that the 2001–2003 tax cuts are offset by future tax increases, they seem likely to increase the size of government allocatively. Only if the newly enacted taxes were a great deal less distortive than those they replaced would this conclusion be likely to change. Distributionally, however, it seems even clearer that the package of current tax cuts plus future tax increases makes the size of government larger. Indeed, the distributional impact is really the big enchilada—persisting in the scenario in which Social Security and Medicare benefits are cut—that makes the big-government character of the 2001–2003 tax cuts almost indisputable.
Consider again that recent generational accounting forecasts suggest that future generations, if left to bear the entire FI, will face lifetime net tax rates more than double those being imposed on any current generations.111 One need not exaggerate the extent to which proportionate tax rates offer a definite non-redistributive baseline in order to see that this implies sizeable transfers from future to current generations. The imbalance is the product of having run Social Security and Medicare on a somewhat Ponzi-like basis, in which early generations received free benefits, rather than reflecting an increase in the in-kind services that future generations get from their greater net tax payments.112
The 2001–2003 changes exacerbated this redistribution from future to current generations. By lowering current generations’ already low lifetime net tax rates in exchange for raising such rates for future generations, the package of tax cuts now for tax increases later unmistakably increased the government’s inter-generational redistributive role. This effect is so significant that the package very likely made the government larger in distributional terms, even if one assumes that [*PG1316]the tax cuts reduced intra-generational redistribution, such as from rich to poor, through the fiscal system.
The prior Subpart’s conclusion may seem a bit too easy. It should be no surprise that cutting taxes now in exchange for raising them in the future fails to make the government smaller. Surely this was not the scenario that proponents of the tax cuts envisioned, except insofar as they were seeking purely current political advantage rather than pursuing a principled long-term policy of shrinking the government.
Part of the answer to any such challenge to this analysis is that the proponents should have envisioned this scenario. Future tax increases are only to be expected when one enacts massive tax cuts in the face of a huge fiscal imbalance plus predictable future political pressures to continue to provide Social Security and Medicare benefits with little prospect of reducing the imbalance by decreasing discretionary government spending. Still, to make the story complete, one also must consider the scenario in which government spending on Social Security and Medicare declines by reason of the increased FI created by the tax cuts. Although seniors and the American Association of Retired Persons bring enormous political power to defending Social Security and Medicare spending, relatively disguised or indirect cuts may soon be politically possible, and in the long run, they really cannot be avoided.113
Looking at the 2001–2003 tax cuts as part of a package with future Social Security and Medicare cuts does little to change the prior Subpart’s conclusion, however. Distributionally, it makes no difference at all (except insofar as different people bear the entitlements cuts as opposed to the tax increases). Either way, lifetime net tax rates are lowered where they are already relatively low and raised for younger people and the members of future generations for whom the rates would have been high in any event. A dollar is a dollar, whether one is losing it in the form of extra gross tax payments or in the form of reduced receipts through government transfers.
[*PG1317] What about allocatively? Here, if anywhere, is where the tax cuts might be expected to lead to smaller government. The allocative effects of Social Security and Medicare, however, taking as given their transfer content, are much smaller than they might initially appear. Under the influence of spending illusion, these two programs may look like big government because the cash flows that pass through them are so huge. In 2002, for example, government spending on the two programs exceeded $750 billion.114 Thus, it is natural to think of these programs as having huge allocative effects. This is misleading, however. In point of fact, Social Security and Medicare are fairly bland programs allocatively.
Social Security actually deals in cash, which people can spend as they like, albeit not until retirement. It is a negative tax, paid to people who previously paid positive taxes, and cutting it is thus very much like increasing taxes directly. Therefore, cutting Social Security to increase generational redistribution through the fiscal system should not be thought of as creating smaller government even allocatively, except in the sense that it reduces the effect of Social Security in requiring people to save a portion of their lifetime incomes (net of taxes and transfers) for retirement. This, of course, is hardly a benefit to people who “escape” being forced to save for retirement by having their savings simply taken away from them instead.
One subtlety about Social Security that makes analysis of its allocative effects more complicated is that “most contributors are likely to view the system’s . . . payroll tax as a pure tax” without counting as an offsetting negative tax the marginal benefits that they accrue by working and earning more.115 Thus, even an individual whose net lifetime Social Security tax was zero might be deterred from working by the burden of the Social Security tax as considered without regard to the accrual of benefits. One should keep in mind, however, that, even with future benefit cuts, the taxes needed to pay off already accrued benefits (or those that for any other reason are politically hard to eliminate) are likely to be very high, and that cutting taxes now probably means that they will need to be still higher in the future (and thus more distortionary for the reasons discussed in the prior [*PG1318]Subpart). Immediately reducing the accrual of future Social Security benefits, if that indeed affects the political likelihood that they will be paid, would have been considerably more constructive than allowing them to keep accruing unchecked and simply tightening the fiscal vise that all taxes and spending will face in the future.
Medicare, concededly, has allocative effects that Social Security lacks because of its in-kind benefits.116 Seniors get healthcare, albeit furnished by private providers rather than directly by the government, even when they would prefer the cash, and with rules poorly designed to give them an incentive to economize on healthcare that offers only modest or negligible benefits.117 Its incentive and income effects have surely been huge contributors to the enormous growth of healthcare expenditure in recent decades, both absolutely and relative to the economy.118 Moreover, the administering government agency (the Center for Medicare and Medicaid Services) prepares comprehensive price lists for covered medical services and monitors millions of payment requests. Thus, in an important respect, the government might be smaller allocatively if Medicare expenditures declined and the program’s impact on the economy thereby was lessened.
Yet even this effect can be overstated easily. For one thing, as empirical research has shown, healthcare is an area in which consumers’ price-sensitivity is relatively low.119 Being healthy is a high priority, and doctors often make treatment decisions for their patients (who may lack the expertise to use their own judgment) on diagnostic grounds that reflect only limited consideration of cost.120 Thus, although price sensitivity is great enough to suggest that Medicare could save billions of dollars if restructured to make patients and doctors more cost-conscious at the margin, the allocative effects nonetheless appear to be small, taking as given the income effects, relative to the dollars involved. As for the distributional effects, although Medicare may result in significant lifetime redistribution within an age cohort because some individuals get so much more than others, the package of current tax cuts plus future Medicare cuts (beyond those that would otherwise [*PG1319]have been necessary) is likely to increase the already huge redistribution from younger to older generations through the program.121
Even for Medicare, therefore, it is plausible that a package of the 2001–2003 tax cuts plus future benefit cuts would make the government larger on balance. The bottom line would depend on how one compared the reduction in the government’s allocative effects to the likely net increase in its redistributive effects. When one considers that tax increases and Social Security cuts are likely as well, however, and that these more clearly increase the size of government, the case that the tax cuts will probably, over time, make the government larger on balance becomes quite powerful. The overall package is one of much greater redistribution to older generations, accompanied by only a possibility of reduced allocative effects.
Looking beyond the tax cuts themselves to the overall budget policy of the Bush administration, the case for an increase in the size of government becomes almost irrefutable. This, after all, is an administration that in its first two years increased federal outlays by $222 billion, or from 18.4% to 19.4% of GDP.122 Only about 40% of this increase was for defense spending, suggesting that one could not attribute all (or even most) of it to the events of September 11, 2001.123 By its third year, the Bush administration was busy procuring a new $16.6 trillion Medicare prescription drug benefit.124
If nothing else, this new drug benefit effectively exploded any notion that the 2001–2003 tax cuts were aimed at shrinking the government’s allocative effects via Medicare, because simply recouping the expansion of government from this new benefit would be a challenge. Rather, the Bush administration appears to be seeking short-term political advantage and favors for particular groups, deceptively clothed in anti-government rhetoric. Then again, conservative ideologues of the Grover Norquist genre,125 even when opposed to the Bush administration’s spending increases, appear no less convinced than the Bush administration itself (if it is sincere) that tax cuts somehow have a greater impact on the size of government than [*PG1320]spending increases.126 They evidently do not, or choose not to, understand the long-term budget constraint, under which a dollar of added government outlays today, if it does not reduce expected future outlays (which more likely it would increase, given the effect on political expectations), implies added taxes with a present value of a dollar.
The above analysis suggests, however, that the 2001–2003 tax cuts likely would not have moved fiscal policy in a smaller-government direction even if the Bush administration had been interested in restraining spending.
One possible response to my suggestion that Republican tax-cutting increased the size of government involves arguing for a longer time frame. To be sure, people tell me, Republican tax-cutting has increased the already huge transfers from future generations to current seniors. They surmise, however, that the tax cuts are pointing us towards a new steady state in which the government will henceforth be smaller because it no longer will be able to afford big programs such as current Social Security and Medicare.127
If this scenario of transition to a stable new small-government steady state made any practical sense, there would indeed be a genuine tradeoff. The government would be transferring more wealth between individuals today, but in the future steady state, it might indeed be doing less. Thus, the issue would be which time frame mattered to you.
Unfortunately, this scenario, which few if any of its proponents have thought through with any thoroughness, is far-fetched at best. I call it the “mañana scenario,” as in, “We cannot take any responsible steps to rein in the entitlements programs today, and indeed, we are expanding them. But fear not, because we will do it all mañana.” I also call it the “come the revolution scenario,” as in, “Come the revo[*PG1321]lution, we will be able to do all of the tough things that we are simply too politically afraid to do today.”
Obviously, the questions are, “What mañana?” and “What revolution?” Why, or on what ground, do these people think that the politics of the future will be so different from the present that all of the hard choices, put off today will be feasible then? Have they really considered the political scenarios that the threat of an impending credit collapse by the U.S. government, which seems to be what they expect, would most likely involve?
Although I will postpone a fuller examination of these scenarios to the next Part, which examines the significance of the fiscal gap,128 a few main points that relate to the size of government are worth making here. First, plainly it is true that an enormous fiscal gap creates strong pressure not to increase outlays, and indeed it creates pressure to reduce them. Again, however, this is not necessarily the same as reducing the size of government. The effect at any time may be to increase generational redistribution, as the elderly in each period continue to push fiscal burdens forward, playing “hot potato” with younger voters, who generally are less politically organized. A scenario of continuously increasing lifetime net tax rates, which certainly cannot be ruled out here, is far indeed from the small-government, limited-redistribution Valhalla.
Second, the revenue pressure on the government may be so great—especially with seniors clamoring against benefit cuts—that it simply cannot be met through the type of straightforward, visible, widely distributed tax increases that are generally most efficient. For example, the use of inflation or even hyper-inflation to ease the fiscal crunch is likely to become very tempting. Stealth tax increases and ostensibly one-time takings from various groups also may become the order of the day. An example of the latter would be using inflation to devalue outstanding, non-indexed government bonds.
Third, the use of regulatory mandates to impose burdens on one group in exchange for benefits to another group will become ever more tempting as a substitute for the government spending that would have been financed (in many cases, more equitably and efficiently) by broad-based taxes. The fiscal restraint era of the late 1980s through the 1990s provided ample advance warning of this likely trend. One example was the Clinton administration’s 1994 healthcare plan, which would have relied on employer mandates to provide much of the financing [*PG1322]off budget.129 As it happened, this effort failed, in part because small business owners were politically well-positioned to resist the mandates. Other major new mandates of the era, however, such as the Americans with Disabilities Act, demonstrated how much could be done this way.130 A fiscal gap encourages the use of mandates even when they impose more targeted burdens, are more intrusive, and are less efficient than outlays financed through general revenues.
Finally, the sheer uncertainty that results from having a huge fiscal gap with no resolution in sight makes the government, in a sense, more intrusive. By cutting taxes and increasing spending, Congress has pointed a loaded gun at its own head (or rather at the heads of future Congresses), but we do not know when and how the gun will go off. Thus, anyone engaged in long-term planning, such as for retirement, must deal with considerable uncertainty about future government policy. For example, should you save more for your retirement because you simply cannot count on any specific component of the existing Social Security and Medicare commitments? Or should you instead save less because a big part of narrowing the fiscal gap will probably be to squeeze the people who had enough foresight to plan properly?
This is not just a matter of making it harder for people to forecast future government policy (and to insure adequately against the downside) when they make retirement plans. Having a huge fiscal gap that remains unaddressed is a surefire political formula for making all of the competing interest groups in Washington continually invest heavily in seeking to influence future government policy. Nothing is really safe, and no government commitment can be taken for granted for more than a few years. With even Social Security and Medicare likely to be on the chopping block, no group can afford to rely on political inertia to protect what it now has. This is an enviable setting for fundraising by politicians.
President Bush admittedly inherited the FI, which results mainly from advances in life expectancy and healthcare technology that make Social Security and Medicare ever more expensive, but the consistent central feature of his fiscal policy has been to increase it as [*PG1323]much as politically possible.131 Both in 2001 and 2003, the current administration and the Republican majority in Congress went out of their way to make the fiscal system even less stable, and unstable sooner, than if they merely had been larding an already immense FI. The mechanism of choice was the sunsets, which provided that the entire 2001 Act, and many of the provisions in the 2003 Act, would expire (barring further legislation) within periods ranging from two to nine years.132 Although the main reason for the sunsets was to lower the official ten-year estimates of the Acts’ revenue cost, deceptive bookkeeping and gaming of the budgetary rules designed to induce long-term fiscal responsibility were only part of the effect.133 A further effect, unacknowledged but not necessarily unforeseen, was to guarantee that Congress would have to continue considering major tax legislation throughout the foreseeable future. For example, would interested parties with millions of dollars at stake really do nothing as the scheduled disappearance of bonus depreciation in 2005, or of estate tax repeal in 2011, grew nearer? Indeed, would not lobbying over permanent extension of the expiring tax cuts be expected to commence immediately? Congress could not have done more than it did in 2001 and 2003 to ensure that the size of government, in the sense of the resources devoted to trying to influence its ongoing decisions, would be as high as possible for years to come.
Part I of this Article showed that the United States entered the twenty-first century with a huge fiscal gap (notwithstanding budget surpluses at the time), that was made significantly larger by the tax cuts of 2001–2003.134 Part II demonstrated that the tax cuts, in light of the redistribution from younger to older generations that will result from the need for compensating changes such as future tax increases and cuts in Social Security and Medicare, probably will not advance their presumed rationale of making the government effectively smaller.135 Disposing of this rationale, however, still leaves the ques[*PG1324]tion of how the tax cuts, through their effect on the fiscal gap, actually ought to be evaluated. This requires evaluating the long-term significance of the fiscal gap—an issue that, understandably, has received less attention to date than the necessary groundwork of defining and quantifying it.
The fiscal gap is merely an accounting measure concerning a set of statements made actually or implicitly by policymakers. One takes a set of rules on the books, fleshes them out by inferring the future policies implied by government programs that have not been legally specified beyond the current year, and then measures their degree of consistency with the long-term budget constraint. The detection of a fiscal gap means that the above statements about future policy cannot all be correct, and thus that they will have to change.
One line of thought that I have encountered, though not seen in writing, responds by asking, so what? The government’s statements about future policy change all the time, and its statements about policy in the distant future are not very meaningful yet. The mere fact that the statements will have to change does not mean that anything is seriously wrong. We should worry about fundamentals, not about whether some given set of policy statements are sustainable under expected economic and demographic conditions.
This critique might further be spelled out to include the following arguments:
These arguments may have some credence, yet the picture they draw is quite misleading. Although not entirely incorrect, they are seriously incomplete, and thus, their suggestion that the fiscal gap is not cause for alarm should be rejected. Because they have some substance, however, the rest of this Part addresses them in turn.136
Given the growth-proof character of the fiscal gap, reflecting that outlays as well as revenues are pegged to the size of the economy, I have elsewhere analogized our current fiscal policy to a hypothetical alimony agreement in a divorce between a high-earning corporate executive and his non-working spouse, in which the parties agree that each year he will pay her 30% of his salary, and she will get 50% of his salary.137 We know this is impossible. No matter how much or little he earns, the 30% that he is supposed to pay cannot equal the 50% that she is supposed to get. Moreover, growth in his salary does nothing to resolve this divergence, and, indeed, makes it larger as measured in absolute dollars. Yet it seems as if the parties ought both to do reasonably well, no matter where between 30% and 50% they settle things, so long as he continues to earn large amounts of money.
Nonetheless, it is also clear that the lack of a feasible agreement could be quite detrimental to both of them. For example, they might collectively spend more than 100% of the available funds. They might dissipate the wealth in costly litigation, or they might end up in a situation in which he, facing arrears, has lost his marginal incentive to earn.138 The lack of a feasible agreement, therefore, in a bad scenario, [*PG1326]might destroy the high earnings that seemingly were more fundamental to their welfare.
In the present setting of U.S. fiscal policy, this is no idle story. Bitter political conflict is likely in the future as seniors’ Social Security and Medicare benefits are endangered, and the choice between cutting benefits more, or raising taxes on younger people more, becomes politically unavoidable. Insofar as higher taxes are adopted, they may affect incentives to work and to save and may promote widespread tax cheating (pitted against rising enforcement efforts) along with the growth of the black market economy. People’s long-range planning, whether for career choice, business investment, or their own retirement, will be made more difficult by the lack of any predictable policy path. Politically, although the government grows more unpopular because it cannot meet its commitments and must raise taxes, rent-seeking activity aimed at influencing its decisions will become more prevalent with so much of its policy in play.
Additionally, Laurence Kotlikoff and Scott Burns identify two more definite reasons why deferring any response to the fiscal gap may end up significantly reducing economic growth.139 First, because of the tax hikes on workers needed to support contemporaneous seniors, capital shallowing, or a reduction in the amount of capital invested per worker, is likely to take place.140 This is an income effect, rather than an aspect of tax rates, reflecting that younger people tend to save more than seniors because they have greater future needs. It is no less real for being an income effect, however, and it suggests that “we’re likely to see a decline in capital per worker, labor productivity, and real wages (apart from any increases in the latter two variables associated with technological progress).”141
Second, and potentially even worse, however, is the scenario of fiscal distress experienced recently by such countries as Brazil and Argentina.142 Governments that are unable to meet their outstanding commitments in any politically tolerable way are prone to respond by printing money, thereby generating hyper-inflation.143 This in itself [*PG1327]can be economically destructive, for example, by increasing financial uncertainty and forcing the use of barter rather than money. Even before hyper-inflation emerges, however, nominal interest rates may skyrocket as soon as financial markets begin to anticipate it as a likely response. Moreover, real interest rates may skyrocket even if the government initially resists the temptation to print money, leading to recession (because fewer business investments can meet the hurdle rate from borrowing) and making the temptation all the greater. Thus, actual government default becomes ever more likely because cheap borrowing to float the annual budget deficit is no longer an option.
As Laurence Kotlikoff and Scott Burns note, Brazil has seen interest rates on government bonds rise to 27% even though the inflation rate is below 10%, reflecting widespread expectations that it will default.144 Argentina has seen a major economic meltdown, featuring bank failures, high inflation, an 80% currency devaluation, a one-third reduction in output, and a 25% unemployment rate.145 Solutions to this sort of fiscal crisis are elusive until the government places its policy on a sustainable course, both actually permitting it to meet its commitments and persuading the capital markets that it has things in order. This may require, as in Russia, coldbloodedly sacrificing the welfare of one generation of the elderly, whose benefits are renounced, in order to benefit all future generations.146 Laurence Kotlikoff and Scott Burns suggest, however, that the most likely outcome is a bad steady state (albeit mitigated by technological advances), featuring chronic inflation, high interest rates, and periodic bank and currency crises.147
This scenario presently may seem farfetched for the United States, but consider the following thought experiment. Because the fiscal gap is currently estimated at about $73 trillion, what would happen today if the U.S. Treasury tried to issue $73 trillion worth of bonds without offering any credible plan to finance the bonds by sufficiently raising taxes or to reduce government outlays?148 This undoubtedly would far exceed the U.S. Treasury’s credit-worthiness, not to mention market demand for the bonds, particularly if other countries with comparable or greater fiscal gaps simultaneously did the [*PG1328]same thing.149 It thus would be expected to lead to skyrocketing interest rates and serious questioning of the U.S. government’s solvency.
How does the actual current fiscal course compare to this? One trivial difference is that at least we will get to roll out our demand for credit a bit more gradually, and thus avoid a crunch caused purely by momentary illiquidity. A more important difference is that we will not need to sell bonds with a 2003 present value as great as $73 trillion to the extent that we change our fiscal policy by raising taxes and reducing outlays. At present, however, this appears to be so difficult politically that one cannot place much faith in it. We will be lucky, over the next few years, if policy changes do not make the fiscal gap significantly larger. Even if there is no policy change, by 2008, the FI is projected to rise by more than 20% to $54 trillion, due purely to the passage of time, and in particular, to the ever-nearing retirement of people in the baby boom generation.150
Leaving aside these points, however, the implications of a $73 trillion fiscal gap are not much changed by the fact that we are not immediately seeking to borrow that amount on the capital markets. The present value, in 2003 terms, of the amount that it appears we will have to borrow is not changed by waiting. Deferral only means that the amount we need grows at a market interest rate, here assumed not to include inflation and default premia. Thus, anyone who thinks that the U.S. government would not be able to borrow $73 trillion on reasonable terms today, and that policy changes narrowing the fiscal gap are too much to hope for until such time as the capital market reaction begins to be felt, should take very seriously the prospect that we are headed along the same path as Brazil and Argentina.
The Bush administration’s share of the blame for this state of affairs goes beyond the tax cuts and spending increases that it has pushed to enactment, along with those to be added in the near future if the sunset provisions are extended or repealed. It is to be blamed as well for relentlessly pushing its tax-cutting ideology without any concern for spending discipline, and thus for creating a political atmosphere in which urging fiscal responsibility is close to impossible. The Bush administration is likely to get the blame it deserves when people ten to twenty years from now look back in search of the causes of the problems they are facing.
[*PG1329] People also will remember the difference between this administration and, say, the Reagan administration, which, after its “riverboat gamble” tax cut in 1981, agreed to deficit-reducing tax increases in 1982 and 1984.151 The Reagan administration also can share the credit for the bipartisan 1983 Social Security changes, which improved that program’s fiscal posture and that of the government as a whole.152 The Reagan administration also avoided involving itself in anything akin to the Bush administration’s proposal to create an unfunded new prescription drug entitlement in Medicare. When a major new Medicare benefit, for catastrophic coverage of seniors’ medical expenses beyond the program’s limits, was (temporarily) added late during its time in office, the Reagan administration agreed that this should be done on a fully funded basis.153
People who anticipate relying on Social Security and Medicare for their retirement support will need little persuasion that the fiscal gap, by endangering their benefits, is actually a serious problem. Nor will people who are ideologically attached to the programs and their New Deal/Great Society provenance. What, however, if one is more of a skeptic about these programs’ merits and current design? What, however, if one is more of a skeptic about these programs’ merits and current design? Then one may regard the threat that the fiscal gap poses to them (reflecting their contribution to it) with more equanimity. Indeed, one might even hope, though with what political prospects is unclear, that the fiscal gap will prompt arguably desirable changes to the programs, such as raising retirement ages in the hope of inducing peo[*PG1330]ple to work longer, or restructuring Medicare to induce greater cost-consciousness by enrollees and their healthcare providers.154
There are very good reasons, however, for having programs that require people to save a portion of their lifetime incomes, net of taxes and transfers, for their retirements, and to take this retirement saving in the form of inflation-indexed life annuities and health insurance coverage.155 The central problem is that people tend to be myopic, and thus tend to violate the principles of lifetime consumption smoothing that rationality would seem to dictate. Accordingly, their welfare may be improved by forcing the consumption smoothing on them. Forced consumption smoothing also may have positive externalities if people who entered retirement with inadequate saving and health insurance coverage otherwise would be supported by need-based transfers.
However true it may be that Social Security and Medicare ought to change significantly, simply cutting retirement benefits, unless completed as part of a fair and sensible overall distribution of the costs of meeting the fiscal gap, risks harming seniors who have not adequately provided for their own retirement saving. Indeed, this raises the possibility that people who break even, or better, financially as a result of the 2001–2003 tax cuts (plus offsetting future tax increases and benefit cuts) will end up having lower lifetime welfare as a result.
Whether the Bush administration realizes it or not, in one sense, its tax policy may prove to have been highly progressive. By transferring wealth from future to current generations, it may be taking money from the relatively rich and giving it to the relatively poor. This depends, of course, on just how badly the fiscal gap hurts future economic performance, on whether things otherwise go well in the future (the fiscal gap is hardly the only threat we face), and on just how large the transfers are. For many decades, however, driven by technological advances, the United States has been growing steadily wealthier (periodic business cycle downturns aside). Worker productivity and real per capita GDP have been continually on the rise.156 Techno[*PG1331]logical advances are certain to continue, and they will tend, unless overwhelmed by other factors, to keep these trends operating.
Moreover, trends in real per capita GDP actually may understate the real improvement in material conditions that technology brings over time. To be sure, various disamenities that GDP ignores, such as pollution and over-crowded highways, may worsen. On the positive side of the ledger, however, GDP includes goods and services at their market prices without regard to consumer surplus, or the amount in excess of the price that one would have been willing to pay if necessary.
Consider electrification and the spread of car and air travel in the first half of the twentieth century, followed more recently by gadgets ranging from microwave ovens to personal computers to cellular phones to DVD players. Or consider that millions of people still live who were born at a time when penicillin had not yet been introduced, and pneumonia was known as the “old person’s friend.” The amount that we would be willing to pay for antibiotics to cure pneumonia, and that people for centuries would have been willing to pay, surely is far greater than its cost. Improvements in life expectancy and healthcare treatment options indicate that things really have been getting better, and may continue to do so, in a fundamentally important sense.
One might ask, therefore, what is wrong with the Bush administration’s fiscal policy. Taking money from richer people and giving it to poorer people is frequently considered a central aim of distribution policy. Should we applaud the generational consequences of increasing the fiscal gap, subject only to our degree of confidence that things will nonetheless actually be getting better overall?
Generational equity is a rich topic, which I cannot explore fully here. Suppose, however, that we consider a straight utilitarian framework, in which the aim is to maximize aggregate well-being over time, but we owe no specific duty to future generations other than weighing their welfare equally per capita with ours. Such a framework tends to favor progressive redistribution, subject to concern about its efficiency costs, on the ground that wealth generally has declining marginal utility. Accordingly, one might think that utilitarianism offers support for significant redistribution from future to current generations, assuming again that technology really does assure that things will continue to improve.
[*PG1332] For three reasons, however, the Bush administration’s generational policy via the fiscal gap is dubious from a utilitarian standpoint. First, GDP may understate the real improvement in material conditions over time due to consumer surplus. If money can buy increasingly better things as time passes, then people who live later might value a marginal dollar more than their predecessors, even though they are better off. Suppose that you could give a million dollars to either of two individuals with brain tumors: one living in 2004 who cannot be helped, or one living in 2054 who, at great expense, can be cured. From a utilitarian standpoint, it is hard to argue against giving the money to the latter individual, even if she is better-off. Thus, the point about GDP understating the consumer surplus from technological advances actually leans against redistribution from future generations to us. We should not deprive them of the opportunity to enjoy material benefits that are not even available to us in the first place.
Second, although living longer is a good thing, it indicates that one needs greater resources. People who buy life annuities (or support Social Security’s character as a life annuity) thereby honor the principle that sometimes good things, rather than bad ones, need to be “insured” against. Thus, even though future generations are presumably better-off than us insofar as they can expect to live longer, this is reason (independent of the consumer goods available at any point in time) for favoring them, rather than us, in distribution policy.
Third, utilitarian distribution policy is all about tradeoffs. Even redistribution from rich to poor within the same age cohort (living with the same technology) is not worthwhile on balance if the loss of welfare from the efficiency cost exceeds the gain in welfare from transferring resources to people who get greater marginal utility from them. Recall the scenario whereby the fiscal gap generates not only very high tax rates but also a capital market meltdown for the U.S. government and financial structure, potentially with long-lasting adverse macroeconomic effects. This suggests that, given how the political system is likely to respond to the fiscal gap, increasing it is a very costly and inefficient way of transferring resources from future to current generations. We simply may be imposing too great a burden on them for it to be worth the benefit to us even if our marginal utility of a dollar is greater than theirs (and even if the meltdown is so far down the road that those now living, or at least voting, need not worry about facing it themselves).
Accordingly, although it is difficult to specify the optimal inter-generational distribution policy, it is plausible that the fiscal gap ought not, from this standpoint, to have been widened. This conclu[*PG1333]sion would be strengthened by any sense of moral obligation toward future generations that discouraged over-burdening them or limiting their capacity to enjoy a better life than ours.
A final point in dismissal of the fiscal gap might note that its two main causes, increasing life expectancy and advances in medical technology, are actually good things. Would we really be better off if people died sooner and could not be helped as often by medical treatment, even though our long-term fiscal projections would then look much better? Is alarmism about the fiscal gap therefore misplaced?
It is easy to agree that, in that scenario, we might be better off on balance than we are today, but this does not change the fact that, in our actual situation, unlike that hypothetical one, our fiscal policy risks making things much worse than they need to be. Even if one would rather be lucky than wise, it is better still to be both.
Suppose a rational individual, who knew she would have to pay for her own retirement, faced an increase in life expectancy and better (but costlier) medical treatment options in the future. She would respond by saving more during her working years. Retirement saving is more valuable if you will need to draw on it for a longer period and can use it to greater effect than previously in improving your quality of life. If she failed to do this, although she still might be glad on balance about the changes, her retirement years might be filled with regret that she had failed to make the best use of her opportunities.
The same points hold for a country that collectivizes retirement financial risk by enacting social insurance programs for retirees. It similarly ought to respond to rising life expectancies and costly healthcare improvements by saving more collectively, whether through private or government saving. Americans, however, like people in many other countries, have resisted this course both in their personal lives and in the policies they have selected through the political process. The fiscal gap reflects this failure, and shows that we are doing unnecessarily badly in certain respects that we could control, even if in other respects we remain fortunate.