Federal Tax Policy in the New Millennium

William G. Gale

Mr. Chairman and Members of the Committee:

Thank you for inviting me to testify at this hearing.

My testimony provides perspectives on the emerging federal budget surpluses and examines the case for and against using the surplus for tax cuts versus other goals. My principal conclusions are as follows:

  • The federal surplus is a major achievement, but it is only the first step toward long-term fiscal sustainability. The short-term surpluses are an accounting illusion, and the long-term forecast shows a significant fiscal deficit, due primarily to social security and medicare.
  • Although previously announced fiscal surpluses were located entirely in the social security trust fund, there is a growing possibility that the non-social security portion of the budget will show significant surpluses in the near future.
  • It is well recognized that tax cuts should not be financed from build-ups in the social security trust funds. Debate now centers on whether to finance tax cuts from any emerging surpluses in the non-social security portion of the budget. However, the same logic that warns against using the social security trust fund to finance tax cuts—namely, that government accounting is misleading—also implies that non-social security surpluses that finance Medicare and government pensions should not be available to finance tax cuts either.
  • Although aggregate tax revenues are quite high, there is no evidence that the vast proportion of families are being overburdened by taxes. A number of measures show that at most points in the income distribution, taxes are at their lowest point in over 20 years. The increase in revenues is coming largely from increased income among the very highest-income households.

I. The issue

Just as the federal budget deficit was the centerpiece of fiscal policy discussions in the 1980s and early 1990s, emerging federal budget surpluses will dominate debate about the appropriate level of taxes over the next several years. Several schools of thought exist concerning how to best use the surplus. In his State of the Union address in January 1998, President Clinton proposed to “save social security first.” This approach is supported by a significant portion of the population and by many leaders in both parties.

A second view is that the surpluses should finance tax cuts. Last summer, then-House Speaker Newt Gingrich (R-GA) called for up to $1 trillion in tax cuts over the next 10 years, Rep. John Kasich (R-OH) called for $700 billion in cuts over 10 years, and Rep. Bill Archer (R-TX) circulated a list of tax cut ideas totalling $3 trillion over 5 years. These proposals, however, would cut deeply into the accumulation of social security reserves.

Senator Pete Domenici (R-NM) has proposed that only surpluses in the non-social security portion of the budget be allocated for tax cuts. This may be seen as a way to reconcile the two views above.

Another approach is to expand government programs. The surplus has generated proposals to increased funding for highways, education, foreign aid, wage subsidies for low-income workers, subsidies for inner city investment, environmental clean-up, civil rights, and numerous other items.

II. Understanding the budget surplus

A. The 10-Year Forecast

Improvement in recent projections
The turnaround in the official U.S. budget totals has been nothing short of astounding. From 1981 to 1995, federal deficits averaged $193 billion per year, or 4.0 percent of GDP. Adjusted for inflation, federal debt held by the public nearly tripled. Compared to the size of the overall economy, federal debt nearly doubled. In 1995, the federal deficit stood at $164 billion and deficits stretched “as far as the eye can see.”

Over the next several years, as shown in Figure 1, the budget forecasts improved dramatically. In January 1998, the CBO announced that the budget was essentially in balance and that surpluses totalling $660 billion were projected over the next decade.

By July, 1998, CBO revised the 10-year surplus estimate to $1.55 trillion (Table 1). The surplus is allocated between the off-budget (mainly social security) and on-budget (the rest of government) portions of the budget. The 10-year surplus in the off-budget portion is estimated to be $1.52 trillion. This reflects the amount by which payroll tax payments and interest earned by the social security trust fund on the Treasury bonds it holds exceeds social security benefit payments and administrative costs. Over the same period, the rest of the budget is in surplus by only $29 billion. In the next few years, the entire surplus resides in the off-budget portion of the unified budget. The remainder of the budget is currently in deficit, and is not expected to a show a significant surplus until 2006.

Table 1
CBO Baseline Budget Projections, July 1998*

in billions
of GDP
1999 80 –  37 117 0.9 –  0.4 1.3
2000 79 –  46 125 0.9 –  0.5 1.4
2001 86 –  45 131 0.9 –  0.5 1.4
2002 139 1 138 1.4 0.0 1.4
2003 136 –  10 146 1.3 –  0.1 1.4
2004 154 0 154 1.4 0.0 1.4
2005 170 5 165 1.5 0.0 1.5
2006 217 44 173 1.8 0.4 1.5
2007 236 55 181 1.9 0.4 1.5
2008 251 64 186 1.9 0.5 1.4
Total 1546 29 1516 —- —- —-
Average —- —- —- 1.40 –  0.02 1.42
*  Projections assume compliance with discretionary spending caps.

Sources of recent improvements

The dramatic improvement in the short-term status of the budget can be attributed to several sources. First, and perhaps most relevant to tax policy, George Bush’s willingness to abandon a poorly-conceived “no new taxes” pledge in 1990, and Bill Clinton’s 1993 tax increases greatly improved the fiscal status. These events should serve as indicators to policy-makers that taking a principled, long-term view in favor of fiscal soundness can pay off in the long-term even when there is vitriolic opposition to these policies in some quarters in the near term.


Second, aided by reduced deficit forecasts, sound monetary policy, and other fortuitous events (such as very low energy prices), the economy has expanded significantly and continuously. This has generated numerous channels through which tax revenues have increased. Lower unemployment has caused the number of tax filers to rise significantly faster than the population at large, and has caused significant increases in wage, which in recent years, accounted for about two-thirds of the increase in tax revenues. A booming stock market has raised revenue from capital gains taxes. In 1995 and 1996, capital gains revenues rose by 18 percent and 40 percent in 1996, reflecting changes in asset values (Figure 2). The cut in capital gains taxes in the 1997 tax act has had an ambiguous impact on capital gains revenues.

Third, despite the tax increases in 1990 and 1993, which only raised income taxes for the top 2-3 percent of households, higher-income households have done particularly well in the last 10-15 years, with proportional income gains that far outpace lower income groups. The rise in income at the high end of the distribution, coupled with the progressive tax changes in 1990 and 1993, has generated a larger increase in revenues than was anticipated.

Fourth, from 1990 to 1997, the taxable share of GDP has increased from 77.2 percent to 79.3 percent. Corporate profits, which are taxed at high rates compared to most personal income, rose from 5.5 percent to 9.6 percent. These increases are due to the strong economy, a slowdown in non-taxable fringe benefits (mainly health insurance) and lower corporate interest payments.

Lastly, spending restraint has made an important difference. Real discretionary spending fell by $66 billion (in 1992 dollars) between 1990 and 1998 (Table 2). Within that category of spending, defense fell by $95 billion, international fell by $5 billion, and domestic spending rose by $33 billion.

Table 2
Real Discretionary Outlays

Fiscal Year Total National
International Domestic
1990 540.9 325.5 20.7 194.8
1991 549.9 331.4 20.2 198.4
1992 534.0 302.6 19.2 212.3
1993 528.3 288.0 20.8 219.5
1994 518.2 272.5 19.4 226.4
1995 507.2 257.8 20.2 231.1
1996 480.0 240.5 16.3 223.2
1997 481.8 240.3 16.5 225.0
1998 474.3 230.1 15.9 228.3

Source: U.S. OMB, 1998

Sources of uncertainty
Any estimate of the future budget status hinges on a number of important uncertainties. In this case, several issues are paramount. The forecasts assume there will be no recession over the next 10 years. However, if a recession of average depth and duration occurred, it would reduce annual surpluses by about $100 billion to $150 billion for a few years. On the other hand, the forecasts are also based on a 2.1 percent real growth rate, which may be low relative to historical patterns.

The projections also assume that a recent surge in revenue will be permanent. Aggregate federal tax revenues have grown by 7.2 percent per year from 1992 to 1998, compared to only 5.2 percent for the economy as a whole. The precise source of the increase in revenues has yet to be determined. To the extent that it derives from increased wage income, it may well be permanent. However, to the extent that it derives from increased capital gains taxes due to a booming stock market, it is perhaps more uncertain.

A recent “guess-stimate” by Republican Senate staffers places the 1999-2008 surplus at $522 billion in the non-social security part of the budget, and the 2000-2009 surplus at almost $700 billion. These figures are speculative, but are based on the continuing strength of the economy, which raises the possibility that the revenue surge will be durable, unexpectedly low spending on Medicare, and other changes (Hager 1999).

These projections assume that spending is not increased above currently projected levels. In particular, the projections assume that Congress and the White House will accept extremely tight discretionary spending caps from 2000 to 2002, and then will accept a freeze in the nominal level of such expenditures from 2002 to 2009, despite inflation and population growth putting pressure on nominal spending. It is noteworthy that virtually all of the reductions in real discretionary spending that have taken place since 1990 have occurred in defense spending (table 2), where at least some downsizing was inevitable following the collapse of the Soviet Union. But large additional reductions there are likely to prove difficult. If so, then a major portion of future cuts will have to come from domestic spending, which has proven very difficult to cut in real terms in the past. Failure to adhere to these very strict limits could reduce the overall surpluses by about $200-$300 billion.

B. The long-term situation

The long-term fiscal imbalance is substantial, regardless of the most recent adjustments to the 10-year forecast. Although long-term forecasts face even greater uncertainty issues, Auerbach (1997, p. 338) notes that “there is relatively little disagreement among those who have made long-term projections that the underlying imbalance is large; the main issue is how large.”

Over the next several decades, the baby boomers’ retirement will place large demands on social security and medicare. According to the 1998 Social Security Trustees’ Report, the OASDI program is expected, under intermediate assumptions, to face negative cash flow starting in 2021 and to deplete the entire reserve fund by 2032. The overall deficit over the next 75 years is estimated to be 2.19 percent of taxable payroll over that period. Medicare is in even worse shape in the near term. Under intermediate assumptions, the Medicare trust fund will exhaust its resources in 2008. Over the next 75 years, the projected accumulated deficit in the HI (part A) program is about 2.10 of taxable payroll. It is important to note that the 75-year cutoff is arbitrary, and that both social security and medicare, under current projections, would face large shortfalls after the 75 year period is complete.

Medicaid is also slated to rise and there is general agreement that in the long-term, the problems created by Medicare and Medicaid together are more extreme than those created by social security. For example, the Congressional Budget Office estimates that social security benefits will rise from about 4 percent of GDP in 1997 to 7 percent in 2050. Medicare and Medicaid together rise from about 4 percent in 1997 to 10 percent in 2050.

Auerbach (1997) quantifies the long-term fiscal imbalance, assuming that the long-run debt GDP ratio converges to its current value. He finds that it would require a permanent rise in the primary surplus of 5.3 percent of GDP, accomplished through tax increases or spending cuts, to achieve long-run (permanent) fiscal balance, if action had been taken in 1997. This would be the equivalent of a 54 percent real, permanent reduction in medicare and medicaid, or a 58 percent cut in other expenditures, or a 47 percent rise in individual and corporate income taxes, These effects do not account for the debt service savings, but they also do not include any behavioral effects induced by added taxes. If action were delayed for 5 (20) years, the primary surplus would have to rise by 5.7 (7.1) percent.

If social security were reformed to achieve fiscal balance through 2070, the additional rise in primary surplus would be 4.5 percent of GDP. If social security were reformed on a permanent basis, the primary surplus would have to rise by an additional 4.1 percent of GDP to obtain fiscal balance. These figures show decisively that bringing social security into balance has significant quantitative effects, but is at most one-quarter of the long-term problem.

Even with changes made to ensure a balanced budget in 2002 and with permanent social security reform, Auerbach finds that about half of the original problem remains—the primary surplus would have to rise by 2.7 of GDP. But these spending cuts or tax increases would have to come on top of those needed to obtain permanent social security reform. Moreover, the base case already assumes cuts in spending (before any added cuts). All government spending except Medicare, Medicaid and social security fall from 11.3 percent of GDP in 1996 to 9.4 percent in 2007 and drop to 8.9 percent slowly thereafter.

In updated figures calculated in the spring of 1998, Auerbach finds that the base case is improved considerably by recent budget updates, but there is still a large fiscal imbalance. The required permanent rise in the primary surplus is 2.8 percent of GDP. However, this estimate depends crucially on cutting government spending other than social security, Medicare and Medicaid. If all other government spending were held constant, the required, immediate, permanent rise in the primary surplus would be 4.9 percent of GDP. At current levels, that would translate into a permanent $500 billion reduction in spending or increase in taxes. This suggests that even pessimistic long-term forecasts are making optimistic assumptions about cuts in government spending.

III. Should the surplus be used to finance tax cuts?

Deciding to use the surplus for tax cuts or anything else hinges on a number of factors that are discussed below.

A. Implications of the long-term situation

The long-term fiscal situation provides no justification for a large scale tax cut. The key point is that the emerging federal surpluses do not represent surpluses in an economic sense. Rather, they are largely an artifact of the peculiarities of federal government accounting. Specifically, the government keeps its books on a cash-flow basis. As a result, the official measure of the government deficit is a flawed and somewhat arbitrary measure of the burdens that fiscal policy places on future generations. Thus, even though the government has more money coming in than going out over the next 10 years, this should not be confused with having an economic surplus.

The government’s situation is similar in nature to that of a college student who perceives a surplus because summer earnings exceed living expenses, but neglects to factor in September’s large tuition bill. Likewise, if the government kept its books like a business would, it would show a shortfall under current circumstances, not a surplus.

To address this problem, we currently run surpluses in anticipation of increased future liabilities. The surplus has taken the form of increasing the social security trust fund, which has led to large off-budget surpluses and to distracting debates about whether the trust fund is “real.” The relevant point, however, is that surpluses in any form reduce government debt, allowing more funds to finance private investment and economic growth, and making it easier to meet future social security and medicare obligations. Tax cuts would raise public debt and reduce national saving, and so would work in the wrong direction.

B. How much is available over the next 10 years for tax cuts?

Despite the pessimistic long-term outlook, there is serious interest in using the short-term surpluses to finance tax cuts, but concern that doing so would exacerbate the long-term financing problems in social security.

A recent proposal, by Senator Domenici, would allow tax cuts, but only out of the non-social security portion of the surplus. This proposal is certainly more fiscally responsible than the tax-cutting frenzy envisioned by House Republicans last summer.

And, on the face of it, the Domenici proposal would avoid using the social security trust fund build-up to finance current tax cuts. But the proposal contains several other problems.

First, even if the trust fund is left alone to accumulate, social security still faces a long-term shortfall. Any resolution of this problem, even if it involves privatization, must in some way cut benefits or raise taxes. Thus, using only non-social security surpluses for tax cuts does not resolve the long-term financing problem in social security.

Second, the reason we should not use social security trust fund assets for tax cuts is that government budget accounting seriously misrepresents the long-term costs of the social security. But social security is only the tip of the iceberg when it comes to misleading government accounting. As noted above, the accounting for medicare creates similar problems, and faces severe long-term problems.

Less well known is the fact that the on-budget portion of the budget currently has net receipts of about $35 billion per year from the build-up of reserves in government pension programs. These items represent reserves in defined benefit pension plans. These reserves are owed to current workers when they retire. Thus, the case for treating these build-ups like social security trust fund build-ups is quite strong. Indeed, many states accumulate pension reserves in separate accounts that are excluded from their operating budget. At $35 billion per year, this accounts for about $350 billion over 10 years that is part of the surplus but, by the logic of omitting social security, should also be omitted from funds eligible to finance tax cuts.

Third, because of the way surpluses are added up over 10 years, the existence of a surplus of $700 billion, does not imply that taxes can be cut by $700 billion. This occurs because of feedback effects—lowering the surplus raises the debt, interest rates and interest payments and therefore imposes higher interest costs and a “fiscal penalty” just as raising the surplus or reducing the deficit reduced interest costs and created a “fiscal dividend.” These effects could be large. The CBO estimates that eliminating all surpluses over a 10-year period would cost about 19 percent of the surplus total in increased interest payments and about another 11.6 percent in “fiscal penalty.” Thus, only about 70 percent of the total of the 10-year surplus was available for tax cuts or spending increases.

Fourth, the proposal would require a change in the budget rules, which require that tax cuts be offset by other tax increases or mandatory spending cuts. Cutting the surplus, though, has the same economic effects as increasing a budget deficit, namely:

  • it will raise tax burdens placed on future generations,
  • it will raise interest rates,
  • it will raise government debt,
  • for both reasons, it will raise government interest payments,
  • it will reduce national saving, and
  • it will reduce tax revenues.

It is hard to see why, in the presence of these effects, and in the presence of the sizable long-term deficits outlined above, the budget rules should be waived for tax cuts out of the surplus.

These may seem like unfair criticisms; that is, it may seem like the goalposts have been moved back, now that we have reached a balanced budget in the short term. In a way, they have, but there is a good reason why. U.S. fiscal policy and the economy has benefitted from a demographic holiday during the last 15-20 years. If should not be all that surprising that we are able to generate budget surpluses while the baby boomers are in their peak tax-paying years. However, if the fiscal problems cannot be resolved by the time the boomers retire, the problems will be massive. The current surplus offers a unique opportunity to deal with these long-term problems and so, in my view, should directed toward them.

For all of these reasons, then, it is not appropriate to treat any non-social security surplus as “fair game” for tax cuts.

C. Are Americans overtaxed?

It is often claimed that Americans are overtaxed and so deserve a tax break. Note that this claim is not particularly related to whether there a surplus exists. Moreover, the assertion is subjective to a large extent. One’s views on whether Americans are overtaxed depend in part on how one values government spending. Nevertheless, there are some relevant facts to consider.

1. Aggregate Tax Revenues:

Historical Comparisons
Federal taxes were 20.2 percent of GDP in 1998, their highest levels since World War II. Table 3 and Figure 3 report historical data on federal tax revenues as a proportion of GDP. From 1950 to 1995, federal revenues fluctuated between 17 percent and 19 percent of GDP. There was a general decline in the importance of the corporate income tax and excise taxes and a commensurate increase in payroll taxes. Individual income taxes averaged 8.25 percent of GDP, with a range of 7.6-9.4 percent.

Table 3
Tax Receipts

(as a percentage of GDP)

Income Tax
Income Tax


1950-54 7.3 5.1 1.8 3.2 17.3
1955-59 7.6 4.4 2.2 2.8 17.1
1960-64 7.8 3.8 3.1 2.9 17.1
1965-69 7.8 3.8 3.7 2.6 18.0
1970-74 8.2 2.8 4.7 2.3 18.0
1975-79 8.1 2.6 5.4 1.9 18.0
1980-84 8.8 1.7 6.1 2.0 18.6
1985-89 8.2 1.7 6.6 1.6 18.2
1990-94 8.0 1.8 6.7 1.6