May 2004
Federal Reserve Bank of Dallas

The Federal Budget: Developments and Outlook

Since last summer, concern about the federal budget deficit has become more pronounced and widespread. In this presentation, I discuss recent budgetary developments and the budget outlook.

I begin by discussing the growth in the deficit over the last few years and the role of economic factors and policy changes in contributing to that growth. I then discuss the economic impact of budget deficits, particularly the reduction in government saving and the potential reduction in national saving. Finally, I discuss the budgetary outlook for the upcoming decade, explaining why the deficit is projected to shrink over that period and describing the economic uncertainties and policy choices that may alter that projection.

Recent Budget Developments
Sharp swing from surplus to deficit. In March, the Congressional Budget Office estimated the deficit for fiscal 2004, which began last October, at $477 billion, as shown in Figure 1. Of course, this is only an estimate, since the fiscal year doesn’t end until September. In fact, recent evidence suggests that the 2004 deficit will come in somewhat lower, due to the stronger economic recovery. Nevertheless, I use the CBO numbers throughout this presentation, because they’re the latest official estimates.

Figure 1

In dollar terms, $477 billion would be the largest deficit in all of U.S. history. That fact, among others, has led to concern about a budget crisis.

It’s more reasonable, though, to measure the deficit as a share of GDP. As shown in Figure 2, using that measure makes the picture somewhat less dramatic. At 4.2 percent of GDP, the estimated 2004 deficit is the eighth-largest since the World War II period. The deficit was larger in 1976, 1983 through 1986, 1991, and 1992.

Figure 2

Still, the swing from surplus to deficit has been stunning in both its size and speed. The budget was in surplus from 1998 through 2001, with the surplus peaking at 2.4 percent of GDP in fiscal 2000. Over the last 4 years, there has been an estimated budgetary shift of 6.6 percentage points of GDP.

Spending increases, revenue declines. The swing into deficit reflects both a rise in spending and a fall in revenue. Figure 3 plots the growth rate of three different spending categories and of revenue, relative to GDP, from fiscal 2000 to fiscal 2004.

Figure 3

The first two bars refer to discretionary spending—those programs whose funding levels are set annually by Congress in appropriation bills. About half of discretionary spending goes to defense and about half to nondefense programs. Since fiscal 2000, defense spending has increased 7 percentage points per year faster than GDP. Nondefense discretionary spending has risen 4 percentage points per year faster than GDP. The next bar refers to entitlements—those programs that automatically pay benefits to eligible recipients and that do not require annual appropriations. Entitlement spending has grown 2 percentage points per year faster than GDP. Finally, revenue has declined 7 percentage points per year, relative to GDP.

Deterioration in budget outlook. Figure 4 compares the recent path of the budget to the path projected in the Congressional Budget Office’s (CBO’s) January 2001 baseline. The current estimate for 2004, a deficit of 4.2 percent of GDP, contrasts sharply with the 2001 projection, a surplus of 3.3 percent of GDP. Going forward, the March 2004 baseline, discussed further below, is much less optimistic than the 2001 baseline.

Figure 4

What turned the 2004 surplus projected then into the deficit we now observe? Forty percent of the deterioration is due to economic factors that CBO did not predict, including the 2001 recession and the stock market slump.[1]

The January 2001 baseline was CBO’s prediction of what would happen to the deficit if the laws and policies then in place were not changed. But, Congress and the President actually made policy changes that enlarged the deficit, accounting for the other 60 percent of the deterioration. Those policy changes are split about equally between spending increases (27 percent) and tax cuts (33 percent).[2]

The spending increases largely reflect the growth of discretionary spending, both defense and nondefense, mentioned above. The 2001 baseline assumed that discretionary spending would remain at its 2001 level, adjusted for inflation. The higher spending level actually adopted this year therefore counts as a spending increase.

Figure 5 shows the recent increases in defense and nondefense discretionary spending and puts them in historical perspective. The pickup in defense spending largely occurred in response to the September 11, 2001 terrorist attacks and includes the military operations in Afghanistan and Iraq. As a share of GDP, though, defense spending remains well below the values observed during most of the last 40 years. Most categories of nondefense discretionary spending have also increased in the last few years.

Figure 5

No major policy changes have affected 2004 entitlement spending. Starting in 2006, however, entitlement spending will be boosted by a law adopted last December, which adds a prescription drug benefit to Medicare.

Tax cuts have come in three installments. A June 2001 law lowered income and estate and gift taxes; except for one minor provision later made permanent, this law is scheduled to expire in its entirety on December 31, 2010. A tax stimulus package followed in March 2002. The latest tax cut, adopted in May 2003, provided tax relief for dividends and capital gains through the end of 2008 and accelerated certain provisions of the 2001 tax cut.

Figure 6 describes the role of economic factors and the tax cuts in lowering federal revenue. In fiscal 2000, revenue was 20.8 percent of GDP, the second-highest value in all of U.S. history. Without any tax cuts, economic factors would have reduced the revenue share by 2.6 percentage points. The three tax cuts reduced the revenue share by another 2.4 percentage points. The combined result is an estimated revenue share of 15.8 percent, the lowest since 1950. (Developments since March suggest that the actual 2004 revenue share may be somewhat higher than this estimate.)

Figure 6

I should note that official estimates of the revenue loss from the tax cuts may be overstated. These estimates assume that tax changes do not alter macroeconomic aggregates, such as GDP and employment. (The estimates do attempt to include the effects of tax changes on microeconomic variables, such as capital gains realizations and fringe benefit payments.) Under some circumstances, a tax cut can boost real GDP, causing a revenue gain that partly offsets the direct revenue loss. Economists do not fully agree on the likely size of such a feedback, although there is a consensus that it would usually not be large enough to fully offset the revenue loss. The official estimates do not include a feedback of this kind.[3]

Before discussing the future budgetary outlook, it is useful to examine the economic impact of deficits.

Economic Impact of Deficit
Reduction in government saving. Government saving is the government’s net investment in capital minus its budget deficit. Deficits therefore represent negative government saving, unless they finance investment in government capital.

Running a deficit permits tax cuts or spending increases today. Servicing or repaying the resulting government debt, however, requires tax increases or spending cuts tomorrow.

Potential reduction in national saving. Government saving is of limited importance in its own right. It is merely one component of national saving, which is the sum of government saving and private saving. Because deficits are negative government saving (except when they finance government investment), they reduce national saving, unless they cause an offsetting one-for-one rise in private saving.

In some cases, private saving may be increased by policies that produce a deficit, such as tax cuts that enhance incentives to save. Moreover, if households recognize that deficits will result in future tax increases or spending cuts, they may save more to prepare for those future burdens. In many cases, though, it is likely that budget deficits reduce national saving to some extent.

A reduction in national saving has important economic consequences. It raises living standards today, as resources are consumed rather than saved. But, it lowers living standards tomorrow, compared to what they otherwise would have been, by reducing future national income.

The exact mechanism depends on whether the economy is open to international trade and investment.

In a closed economy, a reduction in national saving puts upward pressure on interest rates and reduces investment. With less investment, the capital stock is smaller than it otherwise would have been. With less capital available to aid in production, future output is lower than it otherwise would have been. Lower output translates into lower incomes throughout the economy, including lower wages.

In an open economy, a reduction in national saving is likely to increase the inflow (or reduce the outflow) of foreign capital. This change in capital flows must be financed by a larger trade deficit (or smaller trade surplus). In this case, the reduction in national saving need not cause a reduction in investment—foreign savers can finance the investments for which domestic savers fail to supply funds. If there is no reduction in the capital stock, there will be no reduction in the future output produced inside the United States. Nevertheless, the future incomes of Americans will still fall (relative to what they otherwise would have been), because more of the output produced inside the United States must be paid to the foreign savers who financed the investment and own the capital.

Current and past saving. As shown in Figure 7, private saving (which includes both personal and corporate saving) has generally fallen as a share of GDP throughout the last 40 years. National saving has also followed a general downward trend.[4]

Figure 7

The difference between the two series is government saving. During most of this period, national saving was lower than private saving, as government saving was negative. From 1998 through 2001, when the federal budget was in surplus, government saving was positive, so national saving exceeded private saving. In 2002 and 2003, when the federal budget moved back into deficit, government saving again turned negative.[5]

In 2003, private saving was 5.3 percent of GDP while government saving was negative 3.8 percent. National saving was 1.5 percent of GDP, the lowest value since 1934.

Although this figure shows how private saving and government saving add up to yield national saving, it does not establish the extent to which changes in government saving have caused changes in national saving. There is no way to conclusively determine what private saving would have been if government saving had been different.

Deficits are not all that matter. Even if deficits have a significant effect on national saving, tax and spending proposals should not be evaluated solely by how they affect the deficit. The allocation of government spending across different types of programs is also important; for example, transfer payments do not have the same effects as improvements to public infrastructure. Tax and spending changes can alter incentives to work, save, and invest, with important implications for efficiency and economic growth. They can also affect the distribution of disposable income, with important implications for fairness. Tax and spending changes can also influence the business cycle. Programs that make payments from one age group to another, such as Social Security and Medicare, can have profound impacts on private and national saving and on the fiscal burdens borne by different generations.

Budget Outlook During the Next Decade
Deficit shrinks under baseline. As shown in Figure 8, under CBO’s March 2004 baseline, the deficit, as a share of GDP, shrinks throughout the next decade, especially after 2010. By 2014, the budget is almost in balance.

Figure 8

Several factors combine to produce this result. Under the baseline, discretionary spending keeps up only with inflation, meaning that it steadily declines relative to GDP. At the same time, three factors cause baseline revenue to rise relative to GDP. First, the brackets and exemption amounts for the regular individual income tax are adjusted each year only for inflation, not for real economic growth. As people’s incomes rise faster than inflation, they move into higher brackets, a process called real bracket creep. Next, the brackets and exemption amounts for the individual alternative minimum tax (AMT) are not adjusted at all, even for inflation. As a result, AMT payments will sharply increase in upcoming years—by 2010, one person in four will be on the AMT rather than the regular income tax. Finally, the tax cuts adopted in 2001, 2002, and 2003 are smaller after 2004 and completely expire by December 31, 2010 (three months into fiscal 2011), as illustrated in Figure 9. The expiration of the tax cuts explains the sharp shrinkage of the baseline deficit after fiscal 2010.

Figure 9

A countervailing factor, also shown in Figure 9, is the growing cost of the Medicare drug law after it takes effect in 2006.6 Rising medical costs and the retirement of the first baby boomers also push up entitlement spending over the next decade. Nevertheless, as shown above, the deficit still shrinks during this period.

Deficit shrinks less under President’s budget. As mentioned above, the baseline assumes that no policy changes occur. To get a better picture of what may actually happen to the budget, it is necessary to consider the policy changes that Congress and the President may adopt. I now examine CBO’s estimate of the budgetary effects of the policy changes proposed by the President, in the fiscal 2005 budget that he released in February.

While the baseline lets discretionary spending keep up with inflation, the President proposes a more restrictive policy. CBO estimates that, under the President’s proposals, total discretionary spending would grow, in nominal terms, at an average rate of 1.1 percent per year from 2004 to 2009, significantly less than inflation. Defense spending would grow at a 1.4 percent average rate and nondefense discretionary spending at a 0.7 percent average rate. The slow growth rate for defense spending is facilitated by the fact that the 2004 defense budget includes Iraq and Afghanistan costs that are not expected to be present in 2009. The President proposes budget rules to restrict spending, similar to the rules set forth in the Budget Enforcement Act during the 1990s.

On the tax side, the President proposes making the 2001 tax law and most of the 2003 tax law permanent and also proposes some other smaller tax cuts. Unlike the rules set forth in the Budget Enforcement Act during the 1990s, the budget rules proposed by the President would not impose any restrictions on tax cuts.

As shown in Figure 10, the President’s proposals would result in lower revenue than the baseline, especially after 2010. Even so, the revenue share would still rise from its historic low in 2004, due to the other factors mentioned above—real bracket creep, the sharp rise in AMT payments, and the smaller size of the tax cuts after 2004.

Figure 10

Figure 11 charts the net impact of the President’s tax and spending proposals. The President’s budget would result in a slightly smaller deficit than the baseline during the next 6 years. After 2010, it would result in significantly larger deficits than the baseline, because the tax cuts would not expire. The deficit would still shrink, though, from an estimated 4.2 percent of GDP in 2004 to 1.6 percent of GDP in 2014.

Figure 11

Debt burden remains within historical range. Figure 12 shows the projected path of the federal debt. Under the baseline, the debt grows from 36 percent of annual GDP at the end of fiscal 2003 to 41 percent at the end of 2010. After the tax cuts expire, it declines, falling back to 36 percent at the end of 2014. Under the President’s budget, the debt grows to 40 percent of annual GDP by the end of 2006 and remains at roughly that level through 2014.

Figure 12

These debt burdens are within the range of recent experience—larger than those observed during the 1970s, but smaller than those observed during most of the 1990s. They are much different, though, from the debt burdens projected in the January 2001 baseline. Under that baseline, the entire federal debt would have been paid off by 2009.[7]

Other factors affecting the budget outlook. Of course, the two projections I’ve been discussing hardly cover the full range of possible budget outcomes.

To begin with, both projections rely on CBO’s economic assumptions, which (as CBO points out) are subject to great uncertainty. CBO assumes average annual real GDP growth of 2.9 percent from fiscal 2004 to 2014. For each reduction of 0.1 percentage point per year, sustained throughout the decade, the 2014 deficit would grow by about $50 billion and the 2014 debt would grow by about $240 billion. Interest rates, the stock market, and medical costs are also uncertain.

Furthermore, the policies ultimately adopted by Congress and the President may differ from either the current policies in the baseline or those in the President’s budget. In general, most of the likely policy changes point toward larger deficits.

Notably, neither the baseline nor the President’s budget includes permanent AMT relief, even though there is a political consensus that such relief should and will be provided. The costs of such relief grow over time and could approach 0.5 percent of GDP in 2010.

Also, there is likely to be pressure to increase defense and nondefense discretionary spending. Some have argued that the spending levels in the baseline (let alone those in the President’s budget) are not adequate to meet public needs. On May 5, the President himself requested $25 billion of additional funding for operations in Iraq that had not been included in his budget.

The new Medicare drug benefit has also been criticized by some as inadequate and there may be pressure to make it more generous. The President may also propose Social Security changes that would increase deficits during the next decade, although no such proposals are in his 2005 budget.

Conclusion
During the last 4 years, the budget has swung sharply from surplus to deficit, representing a sharp reduction in government saving. The swing into deficit reflects a combination of higher spending and lower revenue, caused by a combination of economic factors and policy changes. These developments have probably reduced national saving, relative to what it otherwise would have been. A reduction in national saving imposes significant economic costs—a sacrifice of future national income.

Despite their significant economic costs, neither the current deficit nor those projected for the next decade can be described as a crisis. The deficit and the debt are within their historical ranges, though towards the upper end of those ranges. Also, the deficit is projected to decline over the next decade, although that projection is subject to considerable uncertainty.

This does not mean, however, that there is no budget crisis. The short-term budget outlook is overshadowed by the looming Social Security-Medicare challenge, to which Chairman Greenspan and others have repeatedly called our attention. The projected long-run growth of these programs has profound implications for private, government, and national saving, as well as for the fiscal burdens facing future generations.

—Alan D. Viard

Notes CBO distinguishes between “economic” and “technical” changes. Economic changes are revisions to the variables in CBO’s economic forecast, such as GDP, employment, inflation, and interest rates. Technical changes are changes in all other factors (except policy changes) affecting revenue and spending, such as the stock market, medical costs, or income distribution. For simplicity, I combine these changes and refer to them as “economic.” I classify additional interest payments resulting from tax cuts as part of the tax cuts rather than as spending increases. Increases in refundable income tax credits paid in cash to households that do not owe income tax are classified as spending increases. Chapter 5 of the 2004 Economic Report of the President discusses these issues and describes recent efforts by CBO and the Joint Tax Committee to estimate the effects of tax changes on macroeconomic aggregates. During my service on the staff of the Council of Economic Advisers, I participated in the writing of this chapter. These saving measures are net of depreciation. Gross saving is much higher, but of little relevance. Savings that merely replace the depreciation of existing capital do not increase national wealth. As this recent experience indicates, movements in government saving (as computed by the Bureau of Economic Analysis) tend to closely follow movements in the federal budget surplus. Nevertheless, the two measures are not identical. BEA uses a measure of federal government borrowing slightly different from the official federal budget deficit. Also, government saving includes net government capital investment. Furthermore, government saving includes saving by state, local and tribal governments. Finally, the saving data refer to calendar years rather than federal fiscal years. These estimates, like the others in this presentation, are CBO estimates. The Office of Management and Budget estimates the costs of the drug law to be about one-third larger. Actually, the baseline projection recognized that it would be difficult to literally repay some of the debt before it matured. By 2009, though, the cumulative surpluses would have allowed the federal government to accumulate financial assets greater than its remaining debt, leaving the government with no net indebtedness.

About In Depth This article is based on a presentation by Alan D. Viard, Research Officer at the Federal Reserve Bank of Dallas. The views expressed are those of the authors and do not necessarily reflect the positions of the Federal Reserve Bank of Dallas or the Federal Reserve System.