At the request of the Senate Committee on Appropriations, the Congressional Budget Office (CBO), with contributions from the Joint Committee on Taxation (JCT), has prepared this analysis of the President's budgetary proposals for fiscal year 2004. CBO estimates that under the President's proposals, the deficit in 2003 and 2004 would rise to $287 billion and $338 billion, respectively (see Tables 1 and 2). For 2003, revenues would remain nearly unchanged from 2002, while outlays would increase by 6.6 percent under the President's plan. The following year, revenues would grow by 2.7 percent, while outlays would climb by 4.8 percent. As a share of the economy, revenues would dip below 17 percent in 2004 and outlays would reach nearly 20 percent, thereby producing a total budget deficit equal to 3 percent of gross domestic product (GDP).

Under the President's plan, over the 2004-2013 period, revenues would grow at an average annual rate of 6.1 percent, while the growth in outlays would slow to an average annual rate of 4.9 percent. Over those 10 years, under the President's policies deficits would persist but slowly decline, totaling roughly $1.8 trillion. However, annual deficits would be small as a percentage of the economy--less than 2 percent in most years.

In a departure from the practice of recent years, the Administration has submitted year-by-year estimates of its budgetary proposals for a five-year period instead of a 10-year period. Since the mid-1990s, lawmakers generally have used the 10-year period as the basis for making baseline budget projections and for measuring the costs of legislative proposals. But citing the uncertainty of making budget projections and estimates, especially in later years, the Administration has not provided annual estimates for fiscal years after 2008. CBO has documented the uncertainty involved in budget projections and estimates,⁽¹⁾ but in preparing this report, it has continued recent practice and has provided year-by-year estimates of the President's proposals for the 2009-2013 period.⁽²⁾

Overall, CBO's estimates of the President's budgetary proposals are similar to those of the Administration. For the 2004-2008 period, CBO estimates a cumulative deficit of $1.2 trillion under the President's policies; the Administration estimates $1.1 trillion.

Constructed according to rules specified in law and intended to serve as a neutral benchmark, baseline projections estimate what the future path of spending and revenues would be if current laws and economic assumptions remained unchanged. In conjunction with its annual analysis of the President's budget, CBO has updated its 10-year baseline projections that it published in January.⁽³⁾ CBO's revised baseline reflects the projected effects of increased spending resulting from the omnibus appropriation act for 2003 (Public Law 108-7), which was enacted in February; technical revisions that reduce estimates of federal revenues in the near term; other information that has become available since January; and associated increases in debt-service costs. The economic assumptions that underlie this baseline are unchanged from those for the previous projections.

CBO's revised baseline, which follows a pattern that is similar to its January projections, shows a deficit of $200 billion for 2004. Baseline deficits drop steadily thereafter and yield to small but growing surpluses after 2007. Under current laws and policies, over the 2004-2008 period, deficits would total about $360 billion--averaging 0.6 percent of GDP over that period. Steadily mounting surpluses in later years would produce a cumulative surplus of almost $900 billion for the 10-year period from 2004 to 2013. That projected surplus relies heavily on the assumed expiration at the end of 2010 of the tax cuts enacted in the Economic Growth and Tax Relief Reconciliation Act of 2001 (EGTRRA); that assumption, which is required by law, contributes about $600 billion to the projection of the cumulative surplus.

CBO estimates that the President's budget would increase deficits (or eliminate surpluses) relative to CBO's baseline in all years of the 10-year period. Those differences (including associated debt-service costs) sum to about $800 billion for the first five years and about $2.7 trillion for all 10 years. Nevertheless, CBO estimates that under the President's budget, deficits would decline in most years. As a percentage of GDP, the deficit under the President's policies is projected to fall to 0.6 percent in 2013. Under such a scenario, debt held by the public would remain roughly near its current share of the economy throughout the period (though nearly twice the level in CBO's baseline by the end of 2013).

Excluding debt service, about two-thirds of the increase in deficits under the President's budget (relative to the baseline) would be caused by reductions in revenues. The President proposes tax policies that would lower receipts by about $1.5 trillion between 2004 and 2013.⁽⁴⁾ About 40 percent of that drop in revenues would occur from 2011 to 2013 as a result of the President's proposal to permanently extend provisions of EGTRRA that expire at the end of 2010. Another 15 percent of the total decrease in revenues would occur in 2004 and 2005, largely from proposals to enact new tax cuts and to accelerate certain tax cuts that are scheduled to go into effect in later years. Nonetheless, cumulative revenues under the President's budget would represent 18.3 percent of total GDP for the 10-year projection period--about the historical average for federal revenues since World War II.⁽⁵⁾

CBO estimates that on the spending side, the President's budget would increase outlays by $725 billion (excluding debt service) for the 2004-2013 period relative to CBO's baseline. More than 85 percent of that total would come from the President's proposals to change various mandatory spending programs, the largest of which is his proposal to reform Medicare--estimated by the Administration to increase outlays by about $400 billion over the 10-year period. (CBO cannot estimate the cost of that proposal because the details are not yet available.) The President's proposals for programs funded by discretionary appropriations, as extrapolated by CBO beyond 2008, would increase outlays by $104 billion over the next 10 years relative to CBO's baseline. Defense outlays would rise by $211 billion and nondefense outlays would drop by $108 billion under the President's budget. Total spending under the President's budget would average 19.6 percent of GDP for the 2004-2013 period, CBO estimates--about the same share as in 2002.

In this report, CBO has estimated the President's budgetary proposals using traditional conventions and practices that do not include the proposals' possible macroeconomic effects. However, the Administration's proposals could affect the economy, which in turn would influence their budgetary impacts. To help better inform the Congress about those economic effects, CBO also prepared a macroeconomic analysis of the Administration's proposals. Presented in the last section of this report, that analysis uses various models and assumptions to indicate the range of potential economic and budgetary impacts of the President's proposals.

CBO's baseline projections and its reestimate of the President's budgetary proposals are subject to uncertainty. Neither of those estimates include the costs of the military conflict with Iraq and its aftermath, which could add tens of billions of dollars in spending this year and could have large effects on the budget in future years (see Box 1). Nor do those estimates include other possible demands on the budget, such as additional spending that may be necessary to respond to terrorist attacks or other contingencies. Furthermore, changes in economic growth from projected levels or changes in other economic factors also would affect the budget, especially federal revenues.
 

Box 1. Estimating the Costs of War with Iraq Last September, the Congressional Budget Office (CBO) was asked to gauge the costs of a war with Iraq. In its response, CBO explained that estimates of the total cost of a military conflict with Iraq and its aftermath are highly uncertain.1 They depend on many factors, including the strategy used, the duration of the conflict, the number of casualties, the equipment lost, and the need for reconstructing Iraq's infrastructure. In that previous analysis, CBO examined two possible force levels among the many that might be used to prosecute such a war. It now appears that the example emphasizing U.S. ground forces (as opposed to emphasizing air forces) is much closer in size and composition to the contingent that the U.S. military is employing for the war; in fact, the number of U.S. ground forces ordered to the Persian Gulf area now exceeds the levels that CBO assumed in its September 2002 estimate by one and one-third Army divisions and one Marine brigade. CBO has updated its cost estimate for the "heavy ground force" accordingly. CBO now estimates that the incremental costs of deploying a heavy ground force to the Persian Gulf (that is, the costs incurred beyond the amounts budgeted for routine operations) could be about $14 billion; after that, the incremental costs of prosecuting the war in Iraq could reach just over $10 billion during the first month of combat and subsequently fall to about $8 billion a month--although CBO cannot estimate how long the war might last. After hostilities end, the costs to return that force to home bases could be approximately $9 billion, CBO estimates. Further, the incremental cost of an occupation following combat operations could vary from about $1 billion to $4 billion a month. CBO provided no estimate of the potential costs for reconstruction or for foreign aid that the United States might choose to extend after the conflict has ended. Regardless of the composition of the force used, multiple unknowns exist about how the conflict with Iraq will unfold. If the Iraqi leadership or selected elements of its military capitulates quickly, ground combat could be short, as in Operation Desert Storm. If urban fighting is protracted or Iraq uses chemical or biological weapons against regional military or transportation facilities, the war might last longer. Given such uncertainty, CBO's estimates of the monthly costs of operations exclude expenditures for decontaminating areas or equipment affected by chemical or biological weapons. The war with Iraq could lead to substantial costs in later years, but CBO did not include such costs either because their magnitude cannot be assessed even roughly or because they depend on highly uncertain decisions about future policies. For example, the United States might leave troops or equipment in Iraq, which could require the construction of new military bases. Also, sustaining the occupation over time could require either increases in overall levels of active-duty and reserve forces or major changes in current policies on basing and deployment. Furthermore, the United States might provide Iraq with funds for humanitarian assistance and reconstruction, and it might provide substantial aid to allies and other friendly nations in the region. 1.  See Congressional Budget Office, Letter to the Honorable Kent Conrad and John M. Spratt, Jr., regarding estimated costs of a potential conflict with Iraq, September 30, 2002.

 

Changes to CBO's Baseline

Both CBO and the Administration construct baseline budget projections according to rules set forth in law, primarily the Balanced Budget and Emergency Deficit Control Act of 1985 and the Congressional Budget and Impoundment Control Act of 1974. In general, those laws instruct CBO and the Office of Management and Budget to project federal spending and revenues under current laws and policies. As a result, baselines are not intended to be predictions of future outcomes; rather, they serve as neutral benchmarks that lawmakers can use to gauge the effects of spending or revenue proposals, such as those in the President's budget.

Compared with its January projection, CBO's updated estimate of the deficit for 2003 under current law has grown by $47 billion (see Table 3). Almost two-thirds of that change stems from lower projected revenues, reflecting weakness in collections to date. For the 2004-2013 period, CBO has reduced its projection of the cumulative surplus by $446 billion, nearly three-quarters of which derives from enactment of the omnibus appropriation act in February.

Overview of CBO's Baseline Outlook
CBO estimates that in the absence of additional spending or tax legislation, the deficit will grow from $158 billion in 2002 to $246 billion in 2003 (see Table 4). Although that amount would be one of the largest deficits recorded in dollar terms, at 2.3 percent of GDP, it would be well below the share of the economy that deficits accounted for in the 1980s through the mid-1990s. As a share of GDP, deficits peaked at 6 percent in 1983. If current laws and policies remained unchanged, CBO projects, deficits would decline after 2003 and switch to surpluses in 2008. Over the 2004-2008 period, the cumulative deficit would total $362 billion--more than double CBO's previous projection. For the full 10-year projection period, CBO estimates a cumulative surplus of $891 billion.

The surpluses that are projected to emerge in 2008 mount steadily and accelerate after 2010, when the EGTRRA tax cuts are scheduled to expire. Because of that assumed expiration and because projections are most uncertain in the later years of the projection period, the 10-year figure should be interpreted cautiously: surpluses projected for the last three years of the period total $1.1 trillion, whereas the preceding seven years show a cumulative deficit.

At the end of 2002, debt held by the public totaled $3.5 trillion, or 34 percent of GDP (see Table 5). Under CBO's baseline projections, such debt declines steadily after 2007, dropping to $3 trillion (17 percent of GDP) by the end of 2013. However, just past the 10-year baseline period loom significant strains on the budget that will intensify as the baby-boom generation ages and that may require significant increases in federal borrowing.

The Omnibus Appropriation Act
In CBO's baseline, the Consolidated Appropriations Resolution for 2003 (also known as the omnibus appropriation act) is projected to increase the deficit by $14 billion in 2003 and to reduce the cumulative surplus by $330 billion over the 2004-2013 period. Spending projected as a result of that legislation is estimated to add $82 billion in debt-service costs over the 10 years.

When CBO prepared its January projections, only two of the 13 regular appropriation acts--those for defense and military construction--had been enacted for 2003. Programs and activities funded in the other 11 acts were operating under a temporary continuing resolution. However, the President and the Republican leadership had apparently agreed that regular appropriations for 2003 should total about $751 billion in budget authority, so CBO adjusted its baseline to that level.⁽⁶⁾ The omnibus appropriation act, which was enacted on February 20, 2003 (for the fiscal year that began on October 1, 2002), consolidated the 11 outstanding appropriation bills into one and boosted total discretionary budget authority for 2003 to $766 billion.

The $15 billion increase in budget authority relative to CBO's January projections will add $9 billion to discretionary outlays in 2003, CBO estimates. About two-thirds of that increase is for defense programs. As specified in the Deficit Control Act, CBO extrapolated the 2003 level of appropriations through 2013, which results in a cumulative increase in defense outlays of $121 billion and an increase in nondefense outlays of $78 billion over the projection period.

In addition to providing funding for discretionary programs, the omnibus legislation also boosted mandatory spending. Increased agricultural assistance will add $3 billion to outlays in 2003. Higher payments to physicians for services that they provide to Medicare beneficiaries will add almost $1 billion in outlays this year. The rates paid to those physicians were scheduled to drop by 4.4 percent on March 1, 2003, but based on a provision in the omnibus appropriation act, the Administration replaced the decrease with an increase of 1.6 percent. For 2004 through 2013, CBO estimates that the change in rates for payments to physicians will boost Medicare spending by $53 billion.

Technical Changes to the Baseline
Other changes in CBO's estimates have increased the projected deficit for 2003 by $33 billion and reduced the cumulative surplus over the 2004-2013 period by $116 billion. Most of those technical revisions to the baseline occur over the next three years and are concentrated on the revenue side of the budget.

The near-term outlook for revenues has dimmed a bit since CBO published its January projections. In light of recent data on withheld taxes, CBO has lowered its estimates of revenues by $30 billion in 2003 and by more than $60 billion over the 2004-2008 period. The largest changes, in 2003 and 2004, amount to about 1.5 percent of total projected revenues in those years.

On the basis of new information from the President's budget, from year-to-date data on spending and receipts, and from other sources, CBO has also made technical reestimates of outlays. Because of faster-than-expected defense spending on operations and maintenance--which funds such activities as maintaining a presence in Afghanistan, fighting the global war on terrorism, and conducting military operations in Iraq--CBO now anticipates discretionary outlays to be $4 billion higher in 2003. CBO has also increased its estimate of Medicare outlays by $3 billion, mostly because of higher-than-anticipated spending recorded since September.

Offsetting some of the additional spending for this year is a net reduction in the estimated subsidy cost for credit programs.⁽⁷⁾ The budget includes dozens of programs that either guarantee loans made by private financial institutions or provide direct loans to individuals or businesses. Accurately projecting loan repayments, defaults, and changes in interest rates over the life of credit programs is difficult, and errors are inevitable. In every year since 1994, federal agencies have reestimated the cost of the credit subsidy for loans and guarantees that were made in previous years. Although the net budgetary impact of those changes is to reduce outlays by more than $1 billion for 2003, some agencies have reported sizable reestimates to the Office of Management and Budget. For example, the Export-Import Bank plans a negative adjustment of more than $3 billion, while the Department of Education's revision will boost outlays by almost $2 billion.

The largest technical change that CBO made in its estimates of outlays over the 2004-2013 period (other than a change in debt-service costs) was a $32 billion increase for Medicaid. CBO raised its projection because of such factors as higher spending on managed care, the enrollment of more children because of states' outreach efforts and the creation of the State Children's Health Insurance Program (SCHIP), and the approval of additional waivers that allow Medicaid programs to provide prescription drug benefits to low-income Medicare beneficiaries. In CBO's baseline, those increases are partly offset by lower spending to reflect efforts by states to address their difficult budgetary conditions by further restricting eligibility for Medicaid.

In addition, CBO upped its estimate of outlays for discretionary programs by $11 billion over the 10-year period, largely on the basis of information reported in the President's budget. That amount includes a mix of small increases and decreases in spending that raise net outlays by about $1 billion per year.

Partially offsetting those increases are revised estimates for Medicare, which reduce projected outlays by $10 billion over the 2004-2013 period. On the basis of updated information, CBO reduced its projected rate of increase in per capita spending for hospice services and for services furnished by therapists, health centers, and hospital-based laboratories.

Under CBO's baseline, as a result of the technical revisions that decrease projections of revenues and increase estimates of outlays, the Treasury will need to borrow more than it otherwise would have over the 2004-2013 period. By CBO's estimate, such additional borrowing would raise net interest payments by $39 billion over the decade.
 

Differences from the Administration's Current-Services Baseline

Both CBO and the Administration estimate that if current laws and policies remained in place, the budget would show a deficit for several years. The Administration projects a deficit of $158 billion in 2004, turning into a small surplus in 2006; CBO projects the emergence of a surplus in 2008. For the five-year period from 2004 through 2008, CBO's projection of the cumulative deficit exceeds that of the Administration by $248 billion (see Table 6).

Table 6.
Comparison of CBO's March 2003 Baseline and the Administration's February 2003 Current-Services Baseline
(In billions of dollars)
				2003	2004	2005	2006	2007	2008	Total, 2004- 2008
CBO's March 2003 Baseline
Revenues
	On-budget			1,360	1,466	1,617	1,741	1,853	1,963	8,640
	Off-budget			532	558	588	619	651	685	3,101
		Total		1,891	2,024	2,205	2,360	2,504	2,647	11,741
Outlays
	Discretionary			805	837	854	868	886	911	4,356
	Mandatory			1,177	1,223	1,277	1,332	1,403	1,484	6,720
	Net interest			155	164	197	217	224	226	1,027
		Total		2,137	2,224	2,328	2,417	2,513	2,621	12,103
			On-budget	1,768	1,839	1,935	2,010	2,093	2,187	10,063
			Off-budget	369	385	393	407	420	434	2,040
Deficit (-) or Surplus
	On-budget			-408	-373	-317	-269	-240	-224	-1,423
	Off-budget			163	173	195	212	231	250	1,061
		Total		-246	-200	-123	-57	-9	27	-362
Administration's February 2003 Current-Services Baseline
Revenues
	On-budget			1,335	1,475	1,646	1,738	1,825	1,919	8,603
	Off-budget			532	556	590	615	644	673	3,078
		Total		1,867	2,031	2,235	2,352	2,469	2,593	11,681
Outlays
	Discretionary			785	795	813	825	843	862	4,138
	Mandatory			1,185	1,221	1,269	1,318	1,387	1,465	6,660
	Net interest			161	173	193	205	211	214	996
		Total		2,131	2,189	2,276	2,348	2,440	2,541	11,794
			On-budget	1,760	1,805	1,883	1,944	2,024	2,112	9,768
			Off-budget	371	384	393	403	416	430	2,026
Deficit (-) or Surplus
	On-budget			-425	-330	-237	-207	-199	-192	-1,166
	Off-budget			160	172	197	211	228	243	1,052
		Total		-264	-158	-40	5	29	51	-114
Difference (CBO minus Administration)
Revenues
	On-budget			24	-9	-29	3	29	44	38
	Off-budget			*	2	-2	4	7	11	23
		Total		24	-7	-30	7	35	55	60
Outlays
	Discretionary			20	42	40	43	44	49	218
	Mandatory			-8	2	8	14	16	19	60
	Net interest			-6	-10	3	12	13	11	31
		Total		6	35	52	69	73	79	309
			On-budget	8	34	51	65	69	75	295
			Off-budget	-2	1	1	4	4	4	14
Deficit or Surplus
	On-budget			16	-42	-80	-62	-41	-31	-257
	Off-budget			2	1	-2	*	3	7	9
		Total		18	-42	-82	-62	-38	-25	-248
Sources: Congressional Budget Office; Office of Management and Budget.
Note: * = between zero and $500 million.

Differences in Projections of Revenues
In projecting revenues, CBO's baseline over the period from 2004 through 2008 is very similar to the Administration's--higher by about 0.5 percent. That relatively small difference obscures some larger deviations in specific years. CBO's revenue baseline is higher than the Administration's by $24 billion in 2003 then falls below the Administration's by $30 billion by 2005. Thereafter, CBO's baseline projection gradually moves higher than the Administration's, with the difference reaching $55 billion in 2008.

Differing economic projections explain most of the differences in the estimates of revenues. For 2003 and 2004, CBO forecasts a lower level of taxable income than the Administration does. Thereafter, CBO projects a higher level of income--resulting from higher estimates of corporate profits and nonwage personal income--thereby leading to the higher projection of revenues over the entire 2004-2008 period.

Offsetting some of that difference attributable to differing economic projections are technical estimating differences between CBO and the Administration--that is, differences in the estimated amount of revenue generated by a given macroeconomic projection. For 2003, CBO projects a total of $34 billion in higher receipts from such technical factors. Much of that difference stems from the Administration's decision to reduce its estimate of revenues by $25 billion (without allocating it to any specific revenue source) to reflect uncertainty. For 2005, CBO projects $32 billion less in revenues than the Administration does because of technical estimating differences about such factors as the effects of the expiration of the tax cuts for businesses enacted last year in the Job Creation and Worker Assistance Act of 2002 and assumptions about the permanence of the recent weakness in individual income tax receipts. For 2006 through 2008, the technical differences are much smaller.

Differences in Projections of Outlays
On the spending side of the budget, CBO's baseline for outlays is $6 billion higher for 2003 than the Administration's. CBO's March baseline includes the additional funding provided in the omnibus appropriation legislation, which was enacted after the Administration completed its projections. In addition, CBO anticipates higher defense outlays than does the Administration. For mandatory spending, however, CBO's baseline is lower than the Administration's by $8 billion primarily because of different estimates of outlays for Medicaid, refundable tax credits, and student loans. Because CBO projects lower enrollment in Medicaid, its estimates of spending for that program continue to be below the Administration's throughout the projection period.

Overall, for the 2004-2008 period, CBO's estimate of total outlays exceeds the Administration's by $309 billion; discretionary spending accounts for about 70 percent of that difference. CBO's projections of discretionary spending are higher than the Administration's largely because CBO included the spending from the omnibus appropriation legislation, used a higher rate of inflation to project budget authority for spending not related to federal pay, and assumed a faster rate of spending for defense appropriations.

The remaining 30 percent of the difference in projected outlays over the five-year period stems mostly from divergent estimates for Social Security, Medicare, and net interest. Because CBO projects a higher consumer price index (CPI), automatic increases in benefits to Social Security recipients are higher in CBO's baseline than in the Administration's. CBO also estimates that real (inflation-adjusted) benefits will grow more quickly and that retroactive disability income payments will be greater over the period. CBO's estimates for Medicare include the effect of the Administration's decision to boost the rates paid to participating physicians, while the Administration's estimates, which were prepared before that decision, do not. In addition, CBO anticipates higher Medicare spending in 2003 and more rapid growth in that spending over the 2004-2008 period. Although CBO's estimates of net interest are lower than the Administration's in the near term (because of lower projections of interest rates and a different assumption about the mix of securities issued by the Treasury), they surpass the Administration's starting in 2005 (as CBO's projections of interest rates are then above those of the Administration).
 

The President's Budgetary Policies

Overall, CBO's and the Administration's estimates of the President's budget are similar (see Table 7). Both anticipate that deficits will peak in 2004: CBO projects a deficit of $338 billion that year and the Administration, one of $307 billion. For the 2004-2008 period, CBO projects a cumulative deficit of $1.2 trillion; the Administration estimates a deficit of $1.1 trillion. Beyond 2008, under the President's proposals, the deficit would decline in most years, reaching a low of $102 billion in 2013, CBO estimates. The Administration did not provide such estimates beyond 2008.

Table 7.
Sources of Differences Between CBO's and the Administration's Estimates of the President's Budget
(In billions of dollars)
					2003	2004	2005	2006	2007	2008	Total, 2004-2008
Administration's Estimate
Deficit Under the President's Budget					-304	-307	-208	-201	-178	-190	-1,084
Sources of Differences Between CBO and the Administration
Revenues
	Baseline differences				24	-7	-30	7	35	55	60
	Policy differences				-4	-8	-5	3	*	-2	-13
				Total Revenue Differences	20	-15	-35	10	35	52	47
Outlays
	Discretionary				13	17	-1	-3	-3	-4	7
	Mandatory
		Baseline differences			-8	2	8	14	17	19	60
		Policy differences			3	7	13	4	4	3	30
			Subtotal, mandatory		-5	9	21	18	21	21	90
	Net interest				-6	-10	6	12	12	11	31
				Total Outlay Differences	3	16	26	27	30	28	128
All Differences					18	-31	-62	-17	6	24	-80
CBO's Estimate
Deficit Under the President's Budget					-287	-338	-270	-218	-173	-166	-1,164
Memorandum:
Economic Differences
	Revenues				-10	-13	2	26	46	60	121
	Outlays				*	-1	10	23	29	31	93
				Total	-9	-12	-9	2	17	29	28
Technical Differences
	Revenues				30	-2	-37	-16	-11	-8	-73
	Outlays				3	17	16	4	*	-2	35
				Total	27	-18	-53	-20	-11	-5	-108
Sources: Congressional Budget Office; Joint Committee on Taxation.
Note: * = between -$500 million and $500 million.

Policy Proposals Affecting Revenues
The President's budget proposes several changes to tax law that would significantly reduce revenues over the next decade. His proposals include an economic growth package, the extension of a number of expiring tax provisions, a variety of new tax incentives, a few simplifications of the tax code, and miscellaneous changes in the administration of taxes and other items. Many of the proposals to spur growth and extensions of expiring provisions relate to features of the Economic Growth and Tax Relief Reconciliation Act of 2001.

CBO and the Joint Committee on Taxation estimate that the proposals would reduce revenues by $35 billion and increase outlays by $4 billion--through their effects on refundable credits--in 2003 (see Table 8). For the 2004-2013 period, CBO and the JCT anticipate that the proposals would reduce revenues by $1.5 trillion and increase outlays by $96 billion. As a share of projected gross domestic product, the revenue reductions would average 1.0 percent over the 10-year period, with the largest reductions occurring in the final three years. A few of the proposed changes would increase revenues, contributing $3 billion over 10 years.

Proposals accelerating and making permanent the changes in EGTRRA account for about 55 percent of the revenue reductions in the package. A proposal to eliminate the double taxation of dividends constitutes an additional 27 percent. The most significant proposals are these:

Extend EGTRRA's Expiring Provisions. Currently, all provisions of EGTRRA still in effect on December 31, 2010, are set to expire the following day. The President's proposal would permanently extend all of those provisions, which include reductions in the marginal income tax rate, the child tax credit, relief from the so-called marriage penalty, education incentives, the repeal of the estate tax and associated modifications of gift and other taxes, retirement income provisions, and other incentives. The total reduction in revenues during the 10-year period would be $602 billion, and the increase in outlays would be $22 billion. In all cases save one, the reductions in revenues would occur after 2010. In the case of estate taxes, some revenue effects would occur shortly following the provision's passage, as taxpayers altered their estate planning in expectation of the permanent repeal of the taxes.

Exclude Dividends from Double Taxation. Currently, income from corporate activity is subject to being taxed twice, once under the corporate income tax and then again when taxpayers receive dividends or realize capital gains on their corporate stock. Under the President's proposal, taxpayers would be able to exclude from their individual income tax liability dividends on which corporate taxes had already been paid. Additionally, shareholders would receive an increase in their cost basis for tax purposes for amounts of corporate earnings not distributed as dividends but on which corporate taxes had been paid (thereby reducing capital gains liability upon realization). The proposal, which would become effective for corporate distributions beginning January 1, 2003, is estimated to reduce revenues by nearly $8 billion in 2003 and by $388 billion over the 2004-2013 period.

Accelerate Individual Income Tax Cuts Scheduled Under EGTRRA. Currently under EGTRRA, an expansion of the 10 percent tax bracket is scheduled to take place in 2008, a reduction in tax rates is scheduled for 2006, an expansion of the 15 percent bracket and an increase in the standard deduction for joint filers (the provisions addressing the marriage penalty) are set to phase in from 2005 to 2009, and an increase in the child tax credit is slated for 2010. The President proposes to make all of those features effective for tax year 2003 (and includes an advance payment, or "rebate," of the higher child tax credit). The JCT estimates that those provisions would reduce revenues by $25 billion in 2003 and $211 billion over the 2004-2013 period. They would also increase outlays for refundable credits by $23 billion over the next decade.

Permanently Extend the Research and Experimentation Tax Credit. Corporations can take a tax credit of 20 percent on certain research expenditures above a base amount. The credit is currently scheduled to expire on June 30, 2004, but the President proposes to make it permanent. The cost of doing so is estimated to be $56 billion between 2004 and 2013.

Increase the Amount of the Alternative Minimum Tax Exemption. The alternative minimum tax (AMT) is a parallel income tax system with fewer exemptions, deductions, and rates than the regular income tax; taxpayers pay the greater of the regular tax or the AMT. Without changes in the AMT, many taxpayers would not receive the full benefits of the EGTRRA tax cut. Hence, EGTRRA provided for an increase in the AMT exemption but only through tax year 2004. The President proposes to increase the exemption under the AMT in 2003 and 2004 and to extend it through 2005. After that, the AMT would revert to its pre-EGTRRA form. The resulting loss of revenue is estimated to be $1 billion in 2003, $36 billion between 2004 and 2006, and nothing thereafter.

Increase Expensing Provisions for Small Businesses. Businesses are currently permitted to expense (take the whole cost as a deduction in the first year instead of depreciating it over several years) up to $25,000 of investment in certain equipment. The benefit is phased out at investment levels exceeding $200,000. As part of his economic growth package, the President proposes to raise the amount permitted to $75,000, allow expensing for certain computer software (for which it is currently disallowed), and raise the investment level at which the benefit begins to phase out to $325,000. The proposal would be effective retroactively to the beginning of calendar year 2003. The cost is estimated to be about $1 billion in 2003 and $27 billion from 2004 to 2013.

Allow an Above-the-Line Deduction for Long-Term Care Insurance. The costs of long-term health insurance are currently treated largely as other medical expenses are. Taxpayers can take a deduction from taxable income if they itemize deductions and have total medical expenditures exceeding 7.5 percent of their adjusted gross income (AGI). The proposal would permit a deduction of premiums for long-term health care insurance (up to current annual limits) regardless of whether taxpayers itemized and without any percentage floor. The provision would be phased in through 2007. The cost from 2004 through 2013 would be $18 billion.

Allow Nonitemizers to Deduct Charitable Contributions. Taxpayers who itemize can currently reduce their taxable income by the amount of their charitable contributions. The President proposes to allow a deduction for nonitemizers (those who take the standard deduction) of up to $250 for individuals and $500 for joint filers for charitable contributions exceeding those amounts. The provision would become effective at the beginning of tax year 2003 and be indexed thereafter. The cost would be less than $1 billion in the first year and $15 billion over the 2004-2013 period.

Provide a Tax Credit for Developers of Affordable Single-Family Housing. The President proposes to create a new tax credit analogous to the existing low-income housing tax credit (LIHTC) for single-family homes. The LIHTC applies to low-income rental units; the single-family housing tax credit would apply to new or rehabilitated homes intended for eligible lower-income families. Like the LIHTC, the credit would be allocated to states and localities to be awarded to projects. Recapture rules would be implemented in the event that homes were resold to ineligible purchasers. Credit allocations would begin in calendar year 2004. The 2004-2013 cost would be nearly $15 billion.

Provide a Refundable Tax Credit for Health Insurance. The President proposes to create a refundable income tax credit for the cost of health insurance. The credit would be worth up to $1,000 per adult and $500 per child (for up to two children). It could cover a maximum of 90 percent of the cost of insurance for individual taxpayers with a modified adjusted gross income of $15,000 and lesser amounts for individuals with higher income, phasing out completely at a modified AGI of $30,000. It would become effective at the beginning of calendar year 2004. In total, the proposal would reduce revenues over the 2004-2013 period by $13 billion and increase outlays by $51 billion.

Expand Tax-Free Savings Plans. A variety of individual retirement accounts (IRAs) currently exist that can be used not only for retirement but for other purposes (such as education). The President proposes to unify many of those accounts into two tax-free savings vehicles--retirement savings accounts (RSAs) and lifetime savings accounts (LSAs)--and to expand their applicability.

For RSAs, individuals could contribute up to $7,500 annually, and no income limits would apply. Contributions would be taxable, but all earnings on the accounts would accumulate tax free. Withdrawals without penalty could occur after age 58 or because of death or disability. Accounts currently held in Roth IRAs would become RSAs. Additionally, traditional IRAs and nondeductible IRAs could be converted into RSAs in the same way as they currently can be converted to Roth IRAs.

Individuals could also contribute up to $7,500 annually to lifetime savings accounts with the same tax treatment as RSAs and, again, without limits based on income. However, withdrawals from LSAs could be taken for any purpose and at any age. Balances currently held in Archer medical savings accounts, Coverdell education savings accounts, and qualified state tuition plans could be converted into balances in LSAs.

Over the 2004-2013 period, the net revenue loss due to the expansion of tax-free savings plans would be nearly $7 billion. However, there would be a net revenue gain of almost $2 billion in 2003 and $10 billion from 2004 through 2008. Revenue gains would occur from 2003 through 2007 because many of the current vehicles receiving favorable tax treatment collect contributions on a pretax basis. Contributions to the new vehicles, however, would be made on an after-tax basis. As a result, the proposal would increase federal revenues at the time the contributions were made (but reduce revenues when withdrawals went untaxed later on).

Extend Nonrefundable Personal Tax Credits Against the AMT. Except under a temporary provision, individuals cannot take certain personal credits, such as the dependent care credit and HOPE Scholarship and lifetime learning credits, against their liability under the alternative minimum tax. The temporary provision, which permitted taxpayers to take the full amount of these credits against the AMT, was scheduled to expire in 2001. That provision has been extended through tax year 2003. The President proposes extending the exemption another two years through tax year 2005. The 2004-2006 revenue loss would be $1 billion, and there would be no losses beyond 2006.

Other Proposals. The President also proposes a large number of additional tax changes, including a variety of additional incentives for charitable giving and health care; incentives related to education, energy, and the environment; additional simplification of the tax code; changes in tax administration; the extension of additional expiring provisions; and reform of unemployment compensation. Altogether, those provisions would reduce revenues by $66 billion over the 2004-2013 period.

Policy Proposals Affecting Discretionary Spending
The President's budget would boost discretionary budget authority for fiscal year 2004 to $787 billion, CBO estimates, a 2.7 percent increase over the $766 billion enacted for 2003 (see Table 9). That increase would be smaller than the 4.2 percent jump in discretionary budget authority between 2002 and the current level for 2003. (The increase for 2003 may ultimately exceed 4.2 percent if the Congress provides additional funding for the war with Iraq and other needs.)

Table 9.
Comparison of Discretionary Budget Authority Enacted for 2003 and the President's Request for 2004, by Budget Function
(In billions of dollars)
								Increase or Decrease (-)
Budget Function			2003 Enacted		2004 Request			Billions of Dollars		Percent
Defense Discretionary			392.1		400.1			7.9		2.0
Nondefense Discretionary
	International affairs		25.4		28.7			3.2		12.8
	General science, space, and technology		23.0		23.5			0.4		1.8
	Energy		3.2		3.7			0.5		15.2
	Natural resources and environment		29.2		27.9			-1.3		-4.4
	Agriculture		5.7		5.3			-0.4		-7.6
	Commerce and housing credita		0.2		-0.5			-0.6		n.a.
	Transportation		22.6		22.7			0.1		0.4
	Community and regional development		11.7		14.2			2.5		21.1
	Education, training, employment, and social services		72.9		77.5			4.6		6.3
	Health		49.5		49.6			0.2		0.3
	Medicare (Administrative costs)		3.8		3.7			-0.1		-1.6
	Income security		44.0		45.8			1.8		4.1
	Social Security (Administrative costs)		3.8		4.3			0.4		11.7
	Veterans benefits and services		26.5		28.2			1.6		6.1
	Administration of justice		36.3		34.2			-2.1		-5.8
	General government		15.7		17.8			2.1		13.2
		Total Nondefense	373.7		386.6			12.9		3.5
Total Discretionary			765.8		786.6			20.8		2.7
Memorandum:
Department of Homeland Security			21.3		27.1			5.8		27.5
Transportation Obligation Limitations			41.3		39.6			-1.7		-4.1
Source: Congressional Budget Office.
Note: n.a. = not applicable.
a. Includes certain receipts (such as those from loan guarantees made by the Federal Housing Administration's Mutual Mortgage Insurance Program) and other collections (such as those from the Securities and Exchange Commission) that are recorded as negative budget authority and outlays.

The President submitted his budget before the omnibus appropriation act was enacted. In the budget, the Administration assumed that appropriations for 2003 would total $749 billion, nearly $17 billion less than the level contained in the act. Starting from that base of $749 billion, the request for 2004 sought an increase of 4.4 percent in discretionary budget authority.⁽⁸⁾ From 2004 through 2008, the President would increase discretionary budget authority at an average annual rate of 4.7 percent for defense activities and 2.3 percent a year for nondefense programs. In CBO's baseline over that same period, which assumes that discretionary spending grows at specified rates of inflation, budget authority for both defense and nondefense programs rises at an average annual rate of 2.6 percent.

If no further legislation is enacted that affects spending in 2003, CBO anticipates that discretionary outlays will total $805 billion this year. Under the President's budget, discretionary outlays would rise to $836 billion next year and to $922 billion by 2008 (see Table 10).

National Defense. The President's budget for 2004 would continue the upward trend in defense spending that began in the mid-1990s but at a slower pace than in recent years. The proposed budget would add $8 billion in discretionary budget authority for defense programs--an increase of 2 percent over the amount currently appropriated for 2003.⁽⁹⁾ By comparison, increases in budget authority averaged about $30 billion a year over the past three years. CBO estimates that the $8 billion increase--along with spending from budget authority previously provided--would boost defense outlays for 2004 by $14 billion over CBO's estimated level for 2003.

The 2004 request would increase funding for pay raises and other benefits for service members (by almost $4 billion), the development of new weapon systems (by $4 billion), and defense programs within the Department of Energy and various other agencies (by $2 billion). The Administration also proposes to reduce funding from the levels appropriated for 2003 for operations and maintenance and revolving funds (by almost $1 billion) and for military construction and family housing (by $1 billion). The 2004 request for the military personnel and operations and maintenance accounts does not include explicit funding for continuing the U.S. military presence in Afghanistan and prosecuting the war on terrorism and does not account for military operations in Iraq. (Nor does the funding appropriated for 2003 for defense explicitly include much of the money needed to conduct those operations in this fiscal year.) According to public statements by officials of the Department of Defense, the Administration will instead rely on supplemental appropriations to provide funding for those missions. After accounting for those activities, the increases in funding for defense for 2003 and 2004 may substantially exceed the levels witnessed in recent years.

For 2005 through 2008, the President's budget envisions an average annual rate of growth of 4.7 percent in budget authority for national defense, although that growth does not include funding for continued antiterrorism activities or for dealing with the aftermath of the war with Iraq.

Nondefense Programs. The President is proposing for 2004 a 3.5 percent increase in appropriations for nondefense discretionary activities above the level enacted for 2003, CBO estimates, including funds for the new Department of Homeland Security (see Box 2). With those funds excluded, the growth rate for nondefense budget authority for 2004 would drop to 2.2 percent.
 

Box 2. Requested Funding for Homeland Security For 2004, the President has requested about $35 billion in net discretionary budget authority for homeland security.1 About 55 percent of that amount ($19 billion) would go to the new Department of Homeland Security and the balance ($16 billion) would go to other departments and agencies that also have responsibilities for homeland security.2 In total, the President requested about $27 billion in net discretionary budget authority for the Department of Homeland Security, but only about $19 billion of that amount would provide funding for activities that fall within the Office of Management and Budget's (OMB's) definition of homeland security. The $19 billion would fund activities such as those of the Transportation Security Agency ($2.3 billion) and border enforcement and protection activities previously performed by the Customs Service and the Immigration and Naturalization Service ($7 billion). It also includes about $3.5 billion for the Department of Homeland Security's Office of Domestic Preparedness to provide state and local governments with grants and training to improve the ability of first responders (police, firefighters, and other emergency personnel) to address terrorist attacks. (The remaining $8 billion of the $27 billion requested for the Department of Homeland Security would go to activities such as maritime safety and immigration services. Such activities are not included in the $35 billion total for homeland security because they are outside of OMB's definition.) Of the $16 billion for homeland security activities performed by other departments and agencies, almost $7 billion would go to the Department of Defense, another $4 billion would go to the Department of Health and Human Services, and $2 billion would go to the Department of Justice. The Congressional Budget Office (CBO) cannot compare the Administration's total request for homeland security for 2004 with amounts appropriated for 2003 because the Administration has not finished reviewing the enacted spending levels to identify which funding falls within its definition of homeland security. When compared with the $29 billion in funding enacted for fiscal year 2002, however, the $35 billion request represents a 20 percent increase over the two-year period. The President is proposing a number of new programs for homeland security. The largest is Project BioShield, which would, among other things, create incentives to increase research for new vaccines. The President is requesting permanent, indefinite funding authority to enable the government to purchase vaccines as soon as they are demonstrated to be safe and effective. The Administration estimates that this proposal would require about $890 million in mandatory budget authority in 2004 and would cost about $3 billion over the 2004-2008 period, but the President's budget did not provide enough details about this proposal for CBO to provide an independent estimate. The Administration also proposes to increase funding for a number of existing programs. In particular, the President would increase funding for the Information Analysis and Infrastructure Protection Directorate of the Department of Homeland Security by about $650 million to allow the organization to assess the vulnerability of critical infrastructure, such as power plants, dams, and bridges. In certain instances, the President's request for 2004 represents a decrease from 2003 levels. For example, although the Administration currently estimates that about $9 billion in funding was enacted in 2003 for the Department of Defense's homeland security activities, the President proposes to reduce that amount to about $7 billion in 2004, because significant purchases of force protection equipment in 2003 would not be repeated in 2004. 1.  That figure, which reflects estimates by the Office of Management and Budget (OMB), includes about $3 billion in offsetting fees for the Transportation Security Administration and the Department of State. In addition, according to OMB's estimates, about $3 billion in mandatory spending would go toward homeland security, much of that offset by receipts. Total gross budget authority in 2004 for homeland security would thus be $41 billion. 2.  The Administration's definition of homeland security activities is not limited to those of the Department of Homeland Security. For a complete discussion of that definition, see Office of Management and Budget, Annual Report to Congress on Combating Terrorism (June 2002), available at www.whitehouse.gov/omb/legislative/combating_terrorism06-2002.pdf.

Among the budget functions that would receive the largest increases are community and regional development, which would receive a boost in funding of over 21 percent to increase grants to first responders--which include firefighters and state and local law enforcement personnel--and to cover payments for disaster relief (activities that both now fall within the jurisdiction of the new Department of Homeland Security). In addition, international affairs would receive an increase of almost 13 percent in 2004. The President proposes to use that money to create the Millennium Challenge Account (which is designed to provide assistance to countries that follow sound economic and social policies), increase military and economic assistance to certain states in the Middle East and South Asia, and pay for reconstruction programs in Afghanistan. Education, training, employment and social services would receive more than a 6 percent increase, with much of that going for increases in elementary, secondary, and postsecondary educational activities.⁽¹⁰⁾

By contrast, the President seeks to reduce funding for some budget functions below what has been enacted for 2003. Included in that group is the administration of justice, which would receive a cut of 5.8 percent, accomplished in part by reducing funding for the Department of Justice's grants to states (by $1.8 billion) and reducing election reform grants to states (by $1.5 billion). Natural resources and the environment would receive 4.4 percent less than in 2003 and agriculture would receive 7.6 percent less.

Policy Proposals Affecting Mandatory Spending
The President's proposals would add $621 billion to mandatory spending over the 2004-2013 period, CBO estimates. Proposals involving Medicare and Medicaid would account for 75 percent of that increase (see Table 8).

Medicare. The President's budget proposes an allowance of $400 billion for an initiative to modernize Medicare that would restructure aspects of the program and provide coverage for outpatient prescription drugs. The Administration estimates that the initiative would cost a total of $400 billion through 2013; however, the budget does not provide sufficient details for CBO to make its own estimate.

Medicaid and the State Children's Health Insurance Program. The President's budget contains a proposal that would allow states to voluntarily convert their federal funding for Medicaid and the State Children's Health Insurance Program into block grants. The grants, called State Health Care Partnership Allotments, would be based on spending levels in 2002 and would grow each year thereafter. States that participated would enjoy much broader flexibility in providing health benefits than current law allows, particularly for beneficiaries who currently are covered at the states' discretion. (States that did not participate would be unaffected by the proposal.) The Administration anticipates that states accounting for half of total Medicaid and SCHIP spending would choose the block grant option.

Again, the President's budget did not provide enough details for CBO to provide an independent estimate of federal outlays for that proposal. Key features of the proposal that have not been specified include the exact method that would be used to calculate the base amount for the block grants, the rates at which they would grow in later years, and the degree of additional flexibility that would be given to participating states. Therefore, in preparing this report, CBO incorporated the Administration's estimate of Medicaid and SCHIP spending for states assumed to choose the block grants. Because the budget does not display projections of Medicaid or SCHIP spending for the 2009-2013 period, CBO projected spending for those years by taking the Administration's projections for 2008 and inflating them using the annual growth rates for Medicaid and SCHIP incorporated into CBO's baseline.

CBO used the Administration's estimate of total spending for Medicaid and SCHIP in evaluating the proposal; however, underlying differences in baseline spending projections between CBO and the Administration lead to very different estimates of the proposal. CBO estimates that, relative to what spending would be if current laws and policies remained unchanged, the proposal would increase the federal government's outlays for Medicaid and SCHIP by $38 billion over the 2004-2008 period and by $73 billion over the 2004-2013 period. By contrast, the Administration estimates that the proposal would cost the federal government $9 billion over the 2004-2008 period and save $0.1 billion over the 2004-2013 period. CBO expects lower spending under current law than does the Administration; thus, the shift to block grants at the amounts estimated in the budget by the Administration (and used by CBO) would result in a larger increase in spending relative to CBO's baseline projections.

In addition, several other much smaller proposals affecting Medicaid and SCHIP would increase outlays by about $1.5 billion from 2004 to 2008 and decrease total outlays by about $1 billion from 2004 to 2013, CBO estimates.

Refundable Tax Credits. The Administration's tax proposals would add an estimated $96 billion to outlays over the 2004-2013 period because a number of the proposals involve refundable tax credits (see the discussion of the proposals affecting revenues for further description of the proposed changes). In particular, the President proposes to accelerate an expansion of the child tax credit and make it permanent, to extend the expansion of the earned income tax credit enacted in 2001, and to introduce two new refundable tax credits (one for health insurance and another for education). Accelerating the child tax credit and other tax relief so that they applied in 2003 would increase outlays by $4 billion in that year and $23 billion from 2004 through 2010, JCT estimates. Permanently extending EGTRRA would increase spending on those two credits by about $22 billion from 2011 through 2013. The health insurance credit would add $23 billion to outlays over the 2005-2008 period and $51 billion through 2013.

Postal Service. Under the President's budget, changes would be made to the way the U.S. Postal Service finances retirement benefits for many of its current and former employees.⁽¹¹⁾ The Office of Personnel Management projects that under current law, the Postal Service will eventually overfund its pension obligations for its workers by as much as $71 billion. Under the proposal, the Postal Service's payments to the retirement fund would decline by about $3 billion to $5 billion a year.

The budgetary impacts would flow from two aspects of the proposed change: the loss of receipts to the Civil Service Retirement System trust fund (which is on-budget) and the response of the Postal Service (whose net cash flow is classified as off-budget) to a sizable reduction in one of its major expenses. CBO estimates that the total budgetary effect of the proposal (that is, combining both on-budget and off-budget impacts) would be a cost of nearly $38 billion over the 2004-2013 period, as the result of lower postage rates and additional spending by the Postal Service for operations, capital investments, or both.

Customs User Fees. Under current law, customs user fees expire on September 30, 2003. The President has proposed extending those fees, which CBO estimates would increase offsetting receipts by $18 billion over the 2004-2013 period.

Other Initiatives. The President has proposed that states, rather than the federal government, pay the administrative costs of running their unemployment compensation programs. Under that proposal, states would be expected to fund those activities on their own, probably through their employment taxes. (Receipts and outlays from state accounts for employment taxes are included in the federal budget.) CBO estimates that the proposal would add about $17 billion to mandatory spending over the 2004-2013 period. At the same time, discretionary appropriations for those activities would be reduced by similar amounts.

The President has also requested $3.6 billion for 2003 to enable states to create personal reemployment accounts. Under that proposal, states could provide individuals who were likely to exhaust their regular unemployment benefits with bonuses of up to $3,000 to be used toward the costs of job training or overcoming other barriers to employment. If individuals were reemployed within a certain period of time without spending the entire benefit, they could keep the remainder. CBO estimates that the bulk of the requested funds would be spent in 2004.

The President's budget proposes to open a portion of the coastal plain of the Arctic National Wildlife Refuge to oil and gas leasing and development. By CBO's estimate, leasing sales from such a program would generate receipts (net of payments to Alaska) totaling $2 billion over the 2006-2008 period.

The President's budget includes four legislative proposals that would affect offsetting receipts from licenses awarded by the Federal Communications Commission (FCC) for use of the electromagnetic spectrum. The proposals would impose new fees on licenses used for analog television broadcasts and on licenses awarded by methods other than auctions, allow certain agencies to spend some auction receipts without further appropriations, and extend the FCC's authority to conduct auctions beyond 2007. Overall, CBO estimates that implementing those proposals could increase net outlays by $5 billion over the next five years (largely because some auctions would be delayed) but would reduce outlays by more than $2 billion over the 10 years from 2004 to 2013.

Differences Between CBO's and the Administration's Estimates
The differences between the Administration's estimates and the JCT and CBO's estimates of the proposals in the President's budget affecting revenues are relatively small through 2008 compared with the total costs of the proposals, although the differences increase in later years. According to the JCT and CBO's estimates, the proposals would reduce revenues by $13 billion more than the Administration projects for the 2004-2008 period (see Table 11). The JCT and CBO estimate greater reductions in revenues than the Administration does for several provisions, most notably for the increase in expensing for small businesses ($7 billion less in revenues); the dividend exclusion ($6 billion less); and the acceleration of the EGTRRA tax cuts ($5 billion less). The JCT and CBO also estimate a smaller increase in revenues from the expansion of tax-free savings accounts ($4 billion less). In the other direction, the JCT and CBO expect smaller net reductions in revenues from the two provisions affecting the AMT ($17 billion more) and the research and experimentation tax credit ($4 billion more).

Table 11.
Comparison of CBO's and the Administration's Estimates of the Effect of the President's Budgetary Proposals
(In billions of dollars)
					CBO				Administration				Difference (CBO minus Administration)
					Total, 2004- 2008		Total, 2004- 2013		Total, 2004- 2008		Total, 2004- 2013		Total, 2004- 2008		Total, 2004- 2013
Total Baseline Deficit (-) or Surplus as Projected in March 2003 by CBO					-362		891		-114		n.a.		-248		n.a.
Effect of the President's Revenue Proposals
	Extend expiring EGTRRA provisions				-5		-602		-6		-498		1		-103
	Provide dividend exclusion				-147		-388		-140		-360		-6		-28
	Accelerate individual income tax cuts				-190		-211		-185		-214		-5		3
	Extend experimentation credit				-19		-56		-23		-68		4		12
	Increase AMT exemption				-36		-36		-26		-26		-10		-10
	Increase expensing for small businesses				-15		-27		-8		-15		-7		-13
	Provide deduction for long-term care insurance				-4		-18		-7		-28		2		10
	Provide charitable contribution deduction for nonitemizers				-7		-15		-6		-13		-1		-2
	Provide tax credit for affordable single-family housing				-2		-15		-2		-16		*		1
	Provide refundable health insurance credit				-5		-13		-3		-2		-2		-12
	Expand tax-free savings				10		-7		15		2		-4		-9
	Extend AMT treatment of nonrefundable personal credits				-1		-1		-18		-18		17		-17
	Other proposalsa				-32		-66		-32		-52		-1		-14
				Total Revenue Effect	-454		-1,455		-441		-1,307		-13		-148
Effect of the President's Outlay Proposals
	Discretionary spending
		Defense			44		211		111		n.a.		-67		n.a.
		Nondefense			-37		-108		108		n.a.		-145		n.a.
			Subtotal, discretionary		7		104		218		n.a.		-211		n.a.
	Mandatory spending
		Medicareb			130		400		130		400		0		0
		Medicaid and SCHIPc			40		72		10		-3		30		75
		Health care tax credit			23		51		31		88		-7		-37
		Earned income and child tax credits			17		45		18		50		-1		-4
		Postal Service			15		38		9		31		6		7
		Unemployment insurance			2		17		2		17		*		*
		Reemployment benefits			4		4		2		2		2		2
		Customs fees			-8		-18		-8		-19		*		1
		ANWR (Net of payments to Alaska)			*		*		-1		-2		1		1
		Spectrum auctions			5		-2		5		-4		1		2
		Other			9		15		11		8		-1		8
			Subtotal, mandatory		239		621		209		568		30		54
	Net interest				103		530		102		n.a.		*		n.a.
				Total Outlay Effect	348		1,255		529		n.a.		-181		n.a.
Total Impact on the Surplus					-802		-2,710		-970		n.a.		168		n.a.
Total Deficit Under the President's Proposals					-1,164		-1,820		-1,084		n.a.		-80		n.a.
Memorandum:
Economic Growth Packaged
	Effect on revenues				-388		-663		-359		-615		-28		-48
	Effect on outlays				22		27		20		27		1		*
Sources: Congressional Budget Office; Joint Committee on Taxation; Office of Management and Budget.
Note: * = between -$500 million and $500 million; n.a. = not applicable; EGTRRA = Economic Growth and Tax Relief Reconciliation Act of 2001; AMT = alternative minimum tax; SCHIP = State Children's Health Insurance Program; ANWR = Arctic National Wildlife Refuge.
a. Includes interaction effect from enacting all provisions together.
b. CBO did not have enough detail to make an independent estimate of the allowance for modernizing Medicare. Instead, it used the estimate contained in the President's budget.
c. CBO did not have enough detail to make an independent estimate of the proposal to allow states to convert their funding for Medicaid and SCHIP into a block grant. Instead, it calculated the cost of the proposal as the difference between the Administration's estimate of total spending for Medicaid and SCHIP (for states assumed to choose the grants) and CBO's baseline estimate.
d. Includes seven provisions affecting revenues: acceleration of the 10 percent individual income tax bracket expansion, acceleration of the reduction in individual income tax rates, acceleration of marriage-penalty relief, acceleration of the increase in the child tax credit, elimination of double taxation of corporate earnings, increase in expensing for small businesses, and provision of alternative minimum tax relief to individuals. Also includes two provisions affecting outlays: personal reemployment accounts and the refundable portion of the child tax credit.

For the 2004-2013 period, the JCT and CBO estimate revenue losses that exceed the Administration's estimate by $148 billion. The largest differences are from the proposals to extend the EGTRRA tax cuts ($103 billion) and to provide a dividend exclusion ($28 billion).

On the outlay side, a number of significant differences exist between CBO's and the Administration's estimates of the President's proposals. The largest differences occur in estimates of discretionary spending; however, the variation almost entirely reflects underlying differences in baselines rather than different assumptions about the effects of the President's request. CBO's baseline for discretionary spending is higher than the Administration's because CBO incorporated the effects of the omnibus appropriation act (which was enacted after the Administration had released its budget) and because of other, technical factors. As a result, although the Administration estimated that its policies would raise discretionary outlays by $218 billion between 2004 and 2008 compared with its own baseline, when measured against CBO's baseline such spending is only $7 billion higher over those five years.

For mandatory outlays, CBO estimates that the President's proposals would increase spending by $239 billion over the 2004-2008 period--or by roughly $30 billion more than the Administration estimated for the proposals. Most of that difference results from the proposal to allow states to convert their funding for Medicaid and SCHIP into block grants. CBO's estimate of the impact of that proposal is $29 billion higher than the Administration's because CBO measured the cost against a lower baseline estimate of spending.

Another significant estimating difference between CBO and the Administration involves the President's proposal to reduce the Postal Service's payments to the Civil Service Retirement System. The Administration assumes that the Postal Service would initially use all of the realized savings to pay off its debt (which has no net budgetary impact), while CBO assumes that most of the funds would be used for capital projects and other operating costs or to postpone postal rate increases. Over the 2003-2008 period, the difference would amount to $8 billion in outlays. For the proposal to create personal reemployment accounts, CBO's and the Administration's estimates of total outlays for those accounts are the same ($3.6 billion) but CBO expects that the accounts would take longer to set up than does the Administration; consequently, CBO anticipates that all of the outlays would occur in 2004 and 2005, while the Administration expects significant outlays in 2003.

Other major differences involve the effects of certain tax proposals on outlays. Because the JCT and CBO assume lower participation than the Administration does for the refundable health tax credits, CBO expects the proposal to increase outlays by $37 billion less over the 2004-2013 period than the Administration does. In addition, the JCT and CBO expect the refundable child tax credit to increase outlays by $4 billion less than the Administration does. Finally, the Administration anticipates that holding lease sales for the right to develop oil and gas resources in the Arctic National Wildlife Refuge would generate gross receipts from bonus bids totaling $2.6 billion over the next five years. In contrast, CBO estimates that receipts from such sales would total over $4 billion (half of which would go to the state of Alaska).
 

CBO's and the Administration's Economic Assumptions

Because the Administration's economic forecast assumes larger tax bases for 2003 and 2004, it generates higher estimates of revenues for this year and the next; however, the opposite is true in subsequent years, when CBO's economic projections generate higher estimates of revenues. For the early years of the 10-year projection period, the Administration's forecast of wages and salaries plus profits--the income categories that have the largest effect on revenue projections--is greater than CBO's, but that difference is reversed during 2005. That pattern is largely the result of the difference between the Administration's and CBO's forecasts for the GDP price index. The Administration's forecast has consistently faster growth of real GDP than CBO's. However, because the Administration's forecast for growth of the GDP price index is more than 0.2 percentage points lower than CBO's, the Administration's projection of nominal GDP begins to fall significantly below CBO's during 2004 (see Table 12).

Table 12.
Comparison of CBO's, the Administration's, and Private-Sector Economic Projections for Calendar Years 2003 Through 2008
			Estimate 2002		Forecast		Projected Annual Average, 2005-2008
					2003	2004
Nominal GDP (Billions of dollars)
	CBO		10,443		10,880	11,465	14,154a
	Administration		10,442		10,884	11,447	13,919a
	March Blue Chip		10,446		10,948	11,499	n.a.
Nominal GDP (Percentage change)
	CBO		3.6		4.2	5.4	5.4
	Administration		3.6		4.2	5.2	5.0
	March Blue Chip		3.6		4.3	5.5	5.4b
Real GDP (Percentage change)
	CBO		2.4		2.5	3.6	3.2
	Administration		2.4		2.9	3.6	3.3
	March Blue Chip		2.5		2.6	3.6	3.2b
GDP Price Index (Percentage change)
	CBO		1.1		1.6	1.7	2.1
	Administration		1.1		1.3	1.5	1.7
	March Blue Chip		1.1		1.6	1.8	2.2b
Consumer Price Index (Percentage change)
	CBO		1.6		2.3	2.2	2.5
	Administration		1.6		2.2	2.1	2.2
	March Blue Chip		1.6		2.3	2.3	2.6b
Unemployment Rate (Percent)
	CBO		5.8		5.9	5.7	5.3
	Administration		5.8		5.7	5.5	5.1
	March Blue Chip		5.8		5.9	5.6	5.2b
Ten-Year Treasury Note Rate (Percent)
	CBO		4.6		4.4	5.2	5.8
	Administration		4.6		4.2	5.0	5.5
	March Blue Chip		4.6		4.2	5.1	5.7b
Tax Basesd (Percentage of GDP)
	Corporate book profits
		CBO	6.2		6.8	7.3	9.2
		Administration	6.3		7.1	7.2	8.4
	Wages and salaries
		CBO	48.1		48.1	48.1	48.0
		Administration	48.1		48.5	48.7	48.7
Tax Basesd (Billions of dollars)
	Corporate book profits
		CBO	653		739	842	1,267a
		Administration	659		771	830	1,120a
	Wages and salaries
		CBO	5,025		5,237	5,518	6,782a
		Administration	5,021		5,275	5,575	6,757a
Sources: Congressional Budget Office; Office of Management and Budget; Aspen Publishers, Inc., Blue Chip Economic Indicators (March 10, 2003); Department of Commerce, Bureau of Economic Analysis; Federal Reserve Board; Department of Labor, Bureau of Labor Statistics.
Notes: Percentage changes are year over year.
n.a. = not applicable.
Since the publication of an interim version of this report earlier this month, this table has been updated to include figures from the March Blue Chip survey.
a. Level in 2008.
b. Based on the 2005-2009 period.
c. The consumer price index for all urban consumers.
d. The Blue Chip survey does not include projections of tax bases.

That pattern is reinforced by differences in the projected relationship of the major tax bases to GDP. The Administration assumes that the total share of income going to wages and salaries plus profits is higher than CBO does through 2005 and slightly lower thereafter.

However, there are two aspects of the Administration's projections that partially offset the pattern in the latter years. The expectations for interest rates and unemployment are significantly lower than CBO's, particularly after 2004. The Administration's projection of the unemployment rate averages 0.2 percentage points below CBO's from 2003 through 2008; its projection of three-month Treasury bill rates averages 70 basis points below CBO's projection for 2005 through 2008. Those differences reduce the projected cost of servicing the national debt and the costs associated with unemployment.
 

The Potential Macroeconomic Effects of the President's Budgetary Proposals

The overall macroeconomic effect of the proposals in the President's budget is not obvious. For example, some provisions in the proposals would lower marginal federal tax rates on labor and capital income. By themselves, those provisions would tend to increase labor supply, investment in productive capital (such as factories and machines), and the economy's output. However, the proposals also would promote the consumption of goods and services by both the government and the private sector, which would tend to reduce investment. CBO's analysis suggests the proposals, on net, would probably increase labor supply but decrease investment and the stock of capital.

Largely because of those two opposing effects, the net effect on economic output could be either positive or negative--with the difference depending not only on how the private sector would respond to the proposals themselves, but also on how the proposals would influence what budgetary policies people might expect in the future. Importantly, regardless of its direction, the net effect on output through long-term changes to the supply side of the economy--including fundamental "inputs" such as labor supply or the stock of capital--would probably be small. Under most assumptions, the proposals' supply-side effects would raise or lower the level of output by less than a percentage point, on average, from 2004 to 2013.

That modest effect on the economy is not surprising. Taken altogether, the proposals would provide a relatively small impetus in an economy the size of the United States'. Excluding any economic effects, CBO estimates that in 2004 the President's proposals would reduce revenues by $117 billion, or 1.0 percent of gross domestic product, and would raise spending (including interest costs) by $21 billion, or 0.2 percent of GDP. From 2004 to 2008, the proposals would reduce revenues by $454 billion, or 0 .7 percent of cumulative GDP, and increase spending by $348 billion, or 0.5 percent of GDP.

The economic impacts should not, of course, be evaluated on a dollar basis alone. For example, as noted above, the proposals would alter marginal tax rates on capital and labor. Over the long term, the effects of budgetary policies depend on the degree to which they alter incentives to acquire skills, work, save, innovate, and undertake investments. Indeed, a subset of the President's proposals are intended to increase those incentives. Those proposals would not operate in isolation, however. The remainder of the revenue proposals and those that would increase spending embody few such incentives. They likely would tend to reduce growth in the long run by increasing government and private consumption, at the expense of saving and investment.

Taking account of the budget's potential effects on the economy could change the estimated budgetary cost of the President's proposals. But as with the macroeconomic effects, the direction of the influence could be positive or negative and is unlikely to be dramatic (see Figure 1). CBO estimates that the supply-side economic effects of the budgetary proposals could add as much as 10 percent to their cumulative cost or subtract as much as 15 percent over the period from 2004 to 2008, and add as much as 15 percent or subtract as much as 17 percent over the period from 2009 to 2013. The estimated cumulative deficit from 2004 to 2008 varies from as much as $1,242 billion to as little as $1,042 billion when supply-side effects are included, compared with an estimated $1,164 billion under baseline assumptions; the estimated cumulative deficit from 2009 to 2013 varies from as much as $942 billion to as little as $335 billion when supply-side effects are included, compared with an estimated $656 billion under baseline assumptions (see Figure 2).
 

Figure 1.

Various Models' Estimates of the Budgetary Savings or Cost (-) from Supply-Side Effects

(As a percentage of the conventional estimate of the President's proposals)*

Source: Congressional Budget Office.

Notes: The figure depicts the variation of the models' estimates from CBO's conventional estimate of the budgetary impact of the President's proposal assuming no macroeconomic feedbacks.

The models used (which are described in detail later) are these: (A) "textbook" growth model, (B) closed-economy life-cycle model with lower government consumption after 2013, (C) closed-economy life-cycle model with higher taxes after 2013, (D) open-economy life-cycle model with lower government consumption after 2013, (E) open-economy life-cycle model with higher taxes after 2013, (F) infinite-horizon model with lower government consumption after 2013, (G) infinite-horizon model with higher taxes after 2013, (H) Macroeconomic Advisers' model, and (I) Global Insight's model.

A negative number means that the macroeconomic feedbacks are estimated to increase the budgetary cost; a positive number, that they are estimated to reduce it (or provide savings).

* Assumes no macroeconomic feedbacks.

** Because the model is designed primarily to capture business-cycle developments, which are hard to predict beyond a few years out, CBO did not compute an estimate for the 2009-2013 period.

 

Figure 2.

Various Models' Estimates of the Deficit Under the President's Budgetary Proposals

(In billions of dollars)

Source: Congressional Budget Office.

Notes: The figure depicts supply-side effects only.

The models used (which are described in detail later) are these: (A) "textbook" growth model, (B) closed-economy life-cycle model with lower government consumption after 2013, (C) closed-economy life-cycle model with higher taxes after 2013, (D) open-economy life-cycle model with lower government consumption after 2013, (E) open-economy life-cycle model with higher taxes after 2013, (F) infinite-horizon model with lower government consumption after 2013, (G) infinite-horizon model with higher taxes after 2013, (H) Macroeconomic Advisers' model, and (I) Global Insight's model.

* Because the model is designed primarily to capture business-cycle developments, which are hard to predict beyond a few years out, CBO did not compute an estimate for the 2009-2013 period.

In addition to benefiting from supply-side growth from the accumulation of greater technologies, skills, labor supply, and capital over the long term, the U.S. economy may grow faster in the near term through "demand-side," or cyclical, growth--greater utilization of the existing labor force, factories, and economic capacity. From a demand-side perspective, budgetary policies that raise consumption (and other purchases) may increase economic growth temporarily, especially if the economy is operating below its potential. Including such demand-side effects would make the overall macroeconomic impacts somewhat larger, raising economic output by as much as 1.4 percent on average from 2004 to 2008.⁽¹²⁾ However, the direction of the budgetary effect is ambiguous, largely because the rise in GDP is estimated to be accompanied by a rise in interest rates.

While the demand-side effects would in some cases be somewhat larger than the supply-side effects over the next five years, such effects are temporary, and they should be viewed cautiously even when the economy is operating below its potential. First, the economy is likely to experience a cyclical recovery in the absence of budgetary policies that boost aggregate demand. Recoveries typically stem primarily from economic adjustments in the private sector. Moreover, the Federal Reserve may adjust monetary policy to aid recovery. Second, changes in spending and taxes can help boost the economy out of recession only if they are correctly timed--they must be enacted at a point of subpar economic growth and in a fashion timely enough to lead (and not follow) the recovery. Past experience in the United States and elsewhere suggests deliberate attempts to employ budgetary policies to aid cyclical recoveries have had little systematic success.

One key determinant of the net macroeconomic impact of a proposed policy change is how it would affect people's expectations of what taxes and other government policies they might face in the future (see Box 3). For example, to the extent people expect that proposals to lower taxes now will lead to higher taxes in the future, they are more likely to increase saving, and perhaps work more, today. But such effects on expectations are very hard to determine. Tax cuts could make people believe that taxes are more likely to rise in the future to finance the interest payments, or that spending is more likely to be cut. Alternatively, people might not worry much about future policy changes.
 

Box 3. How Would the President's Proposals Be Paid For? According to the Congressional Budget Office's (CBO's) projections, the President's budgetary proposals imply a deficit in every year over the next decade and would keep the ratio of debt to gross domestic product (GDP) over that period close to its current level of 34 percent. However, if spending and tax policies remained unchanged, as assumed under CBO's baseline, the ratio of debt to GDP would fall to 17 percent. That higher level of debt under the President's budget would imply higher interest payments and thus would add to the budget's financing requirements after the end of the projection period in 2013. For some time, that added need could be met by running higher deficits. However, the federal government could not follow such an approach indefinitely. At some point in the future under the President's proposals, either taxes would have to be higher than they otherwise would have been, or spending would have to be lower. Some analysts might argue that the President's proposals be compared with an alternative standard that includes other policy changes, rather than be compared against CBO's current-policy baseline, which assumes no policy changes. For example, compared with an alternative that included fewer tax cuts, less encouragement of investment, and more government spending, the President's proposals could look more favorable to growth. However, CBO has no basis on which to construct such an alternative for comparison, and all of its analyses of legislative proposals are made relative to baseline assumptions.

How Fiscal Policy Affects the Economy
The aggregate production of goods and services changes over time in two distinct ways. First, the economy's underlying potential to generate output rises with increases in the quantity and quality of the labor force, the size of the stock of productive capital, and the level of technological know-how. Economists refer to those three determinants of potential output as "supply-side" variables because they determine the quantity of goods and services that the economy is capable of supplying. Supply-side changes have a lasting effect on the economy.

Second, actual economic output cycles around its potential level, as unemployment rises and falls and the stock of capital is used more or less intensively. Those movements are referred to as demand-side, or cyclical, variations because they occur as the total demand for goods and services moves above and below the level of potential output. Unlike movements in the supply side of the economy, cyclical changes are temporary--built-in corrective forces tend to move the economy back toward the potential level determined by the supply side.

When the economy is below its potential level of output, policies that increase aggregate demand can increase output without running the risk of accelerating inflation. The President's budget would add to demand both by cutting taxes and increasing some transfer payments--which would increase the disposable income people had available to spend--and by increasing the government's own spending on goods and services.

The demand-side effects of budgetary policies depend critically on the way the Federal Reserve responds to them in its monetary policies. That response in turn depends on the state of the business cycle. For example, during a recession, the Federal Reserve would be unlikely to increase interest rates to offset budgetary policies that increased aggregate demand, but if the economy was robust, the Federal Reserve might do so.

But business cycles cannot be projected with any degree of reliability beyond a few years, and the same would be true of the Federal Reserve's actions. Consequently, CBO's analysis of the demand-side effects of the President's budgetary proposals is restricted to five years. In contrast, CBO evaluates supply-side effects over a conventional 10-year window.

In the United States, both supply-side and demand-side economic developments depend on the choices of millions of individuals about things such as what and how much to buy, how much to save and what assets to hold, and where and how much to work. While government spending and tax policies can influence those choices, and therefore the economy, the impact of budgetary changes on the economy is limited. Although the government plays a crucial role in establishing the legal and institutional framework within which the economy operates, once that general framework is in place, personal circumstances and preferences play a much larger role in people's behavior than do marginal changes in government policies.

The following sections review how the policies in the President's budget might affect the economy, first examining supply-side effects that would change potential output and then turning to demand-side effects.

The Quantity and Quality of Labor. The overall quantity and quality of labor is an important determinant of potential output. Most simply, an increase in the overall number of hours worked in the economy raises potential output. In addition, increases in educational attainment, the amount of training provided, workers' level of experience, or their degree of effort on the job raise the quality of each hour worked, increasing output. Some analysts might assert that the policies in the President's budget would affect the quality of labor. However, the ways in which budgetary policies affect labor quality are not well understood. For that reason, CBO's analysis concentrated on the effect of the budget proposals on the hours of labor supplied.

The President's budget would affect the hours of labor in two main ways. First, a number of provisions, such as accelerating the increase in the child tax credit and exempting most dividend income from taxation, would increase after-tax income without changing marginal tax rates. Such increases tend to reduce the number of hours worked because people can maintain the same standard of living with less work. Second, provisions such as the acceleration and extension of EGTRRA's reductions in marginal tax rates would increase the after-tax compensation for each additional hour of work in addition to raising after-tax income. Evaluating the effect of such rate reductions on labor supply is complicated by the fact that they have opposing effects--people earn more for each extra hour they work, which tends to encourage work, but can earn the same after-tax income in fewer hours, which tends to discourage work. Most studies, however, find that, on net, reductions in marginal tax rates increase labor supply, primarily by drawing secondary earners into the labor force.

To estimate the effect of lower marginal tax rates, CBO estimated the changes in the effective marginal tax rate on labor income--the rate at which the average additional dollar of compensation for labor is taxed (see Table 13). The percentage-point changes are smaller than the change in statutory income tax rates under the President's proposals because some of the compensation that people receive for working--such as employer-provided health benefits--is not taxed.

Table 13.
CBO's Estimates of the Effective Marginal Federal Tax Rates on Labor
(In percent)
Calendar Year	Tax Rates Under Current Law		Tax Rates Under President's Budget		Percentage-Point Difference		Percentage Change
2003	30.0		28.2		-1.8		-5.9
2004	29.7		28.4		-1.3		-4.3
2005	29.7		28.5		-1.1		-3.8
2006	29.2		29.2		-0.1		-0.3
2007	29.5		29.5		0		0
2008	29.7		29.7		0		0
2009	29.7		29.7		0		0
2010	30.2		30.2		0		0
2011	32.0		30.5		-1.5		-4.6
2012	32.0		30.5		-1.5		-4.6
2013	32.4		31.0		-1.3		-4.1
Source: Congressional Budget Office.
Note: Includes federal individual income taxes and payroll taxes.

Provisions in the budget proposals that would affect the level of the capital stock could also change compensation per hour of work by affecting productivity. If the proposals led to lower investment, that would imply a smaller stock of productive capital and therefore lower wages. A positive effect on investment would have the opposite effect. CBO incorporated those secondary influences on labor supply into its analysis.

CBO estimates that, overall, the President's budget would increase the number of hours worked somewhat--that is, the positive effect of lower marginal tax rates would outweigh the negative effect of increased after-tax income.

The budget's policies could also affect labor supply by changing people's expectations of future policies. The budget's proposals would increase the federal budget deficit, which could lead people to expect that some time in the future, taxes would have to be increased or transfer payments (such as unemployment compensation or Social Security) or government services would have to be cut to finance the federal government's increased interest payments. If people expect to face higher tax rates on labor in the future, they may try to work more before the rates go up and work less when the rates are higher. Even if they expect simply to have to pay more taxes (whether or not the marginal tax rate on labor goes up) or receive less transfer payments or government services, they may try to work and save more now in order to have more resources to compensate for the larger burden in the future. It is difficult to gauge, however, the degree to which people make decisions with so much foresight, the time horizon they consider in making plans, and the future policy changes they might expect. To deal with that uncertainty, in its analysis CBO used various assumptions about people's degree of foresight and expectations of future policies.

The Size and Composition of the Capital Stock. The President's budgetary policies would affect the size of the capital stock--the nation's stock of productive equipment such as factories and information systems--primarily through their impacts on government and private consumption and, therefore, on investment. The policies would directly increase government consumption relative to the level in CBO's baseline. That increased government consumption would tend to reduce investment in productive capital by reducing the resources available.

Some of the effect of higher government consumption on investment would probably be offset by an increase in the amount of foreign capital that was invested in the United States. However, most of the returns to those investments would accrue to foreigners and therefore would not be available to U.S. residents. For that reason, the additional foreign investment would not necessarily increase the resources available to Americans in the long run.

The President's budgetary policies would also influence private consumption in a number of ways. For one, the budget would increase disposable income through reduced taxes and increased transfer payments (such as a Medicare prescription drug benefit). That would tend to boost consumption, because people would probably spend some of that extra disposable income.

However, some tax proposals in the President's budget would tend to reduce consumption by increasing the after-tax rate of return on savings. Accelerating and making permanent EGTRRA's reductions in marginal tax rates, reducing the share corporate income subject to double-taxation, and expanding tax-free savings accounts would all reduce the marginal tax rates on income from savings. (For a detailed analysis of the President's proposals concerning double taxation and savings accounts, see Box 4.) Overall, those changes would increase the after-tax return on savings.
 

Box 4. The Potential Economic Effects of the President's Proposals to Reduce Double Taxation of Corporate Income and Expand Tax-Free Savings Accounts Two provisions in the President's budget--the proposal to reduce double taxation of corporate income by exempting from taxation most dividend income and some capital gains on corporate stock and the proposal to expand the availability of tax-free savings accounts--have unusually complex economic effects. Reduce Double Taxation of Corporate Income Under current law, some corporate income is taxed twice, once under the corporate tax and again when individuals receive taxable income in the form of dividends or capital gains. The President proposes to reduce significantly that double taxation of corporate income by eliminating individuals' tax liability for income that has already been taxed at the corporate level. The Congressional Budget Office (CBO) estimates that the proposal would eventually shelter some 90 percent of dividends and 40 percent of capital gains on corporate shares, although some of that sheltering would be redundant because only about half of dividends and one-quarter of those gains are now taxed.1 Because gains are effectively taxed at a lower rate than dividends and the proposal would shelter a smaller share of gains than of dividends, the dividend exclusion would account for more than 90 percent of the value of the reduction in revenues.2 The proposal and its economic effects are complex. First, eliminating taxes on most dividends and some capital gains would reduce the overall taxation of capital income. In general, that might be expected to lower the cost of funds for businesses because they could pay investors less before taxes to yield the same after-tax return. But the extent of the reduction in the cost of capital is unclear: some analysts hold to a theory of corporate finance which implies that the reduction in the cost of capital would reflect only the less than 10 percent of the tax saving stemming from the reduction in taxes on capital gains, while others hold that the reduction would reflect both the reduction of taxes on gains and the reduction of taxes on dividends.3 CBO has adopted a middle estimate of the implications of the President's proposal for the cost of capital for firms, largely because the proposal accords a saving incentive to a specific sector. In an open economy, such a targeted incentive would have results in between those predicted by either theory, even if the theory predicting a greater fall in the cost of capital was otherwise correct (as CBO normally assumes).4 Second, the proposal would tend to increase shareholders' consumption by raising the value of their corporate stock. The interaction of the current schedule of accelerated depreciation and the proposed cut in taxes would reduce the distinction between new and old corporate capital, raising the value of the existing stock.5 More important, share values would rise to the extent that the tax savings were not immediately offset by lower pretax returns stemming from more investment. (To the extent that the tax proposal encouraged extra investment, the size of the capital stock would rise, decreasing the pretax rate of return to capital and offsetting the tax savings to shareholders.) Corresponding to the disagreement about the size of the drop in the cost of capital, opinions differ about how much share values would rise. The theory of corporate finance that predicts a relatively large increase in share values predicts a relatively small decrease in the cost of capital, and vice versa. Because increased share values lead to more consumption, the President's proposal would help increase aggregate demand in the short run. However, the more it would help demand by raising consumption, the more it would hurt supply in the long run by lowering saving and investment. As with the cost of capital, CBO adopted a middle estimate for the increase in share values. Third, the proposal would lessen the disadvantage that the corporate sector now faces in the competition for capital. Currently, while some income from the corporate sector is taxed twice, the imputed income from owner-occupied housing is not taxed at all, and income from small businesses is taxed only once (at the personal level). That disparity in tax treatment leads to less investment in the corporate sector than is optimal for economic output. Lowering taxes on the corporate sector would allow that sector to attract additional capital from the other two sectors. In general, such a shift would improve efficiency, although it might conflict with other goals, such as supporting owner occupancy of homes or unincorporated businesses. Fourth, the proposal would tend to make equity financing more attractive to firms relative to debt financing, and it would make paying dividends more attractive relative to retaining earnings. Currently, interest payments are deductible from corporate income, so they are taxed only at the personal level. However, if a firm finances investment through equity, some of the returns are taxed at both the corporate and personal levels. So under the proposal, the difference between the effective tax on interest and equity returns would narrow. Also, because most investors currently face a lower tax rate on capital gains than on ordinary income and capital gains taxes are deferred until the gains are realized, firms are encouraged to retain earnings and build up the value of their stock rather than pay out dividends. Under the President's proposal, that incentive would no longer apply. The proposed reduction in the double taxation of corporate income would also interact with some of the President's other proposals and with current law. For instance, the President's proposal to expand tax-free savings accounts would increase the share of personal assets held in tax-free accounts--duplicating some of the effect that the proposal to reduce the double taxation of capital income would have on the cost of capital and on the allocation of capital among economic sectors. However, the expanded accounts would partly undo the impact that the proposal concerning double taxation would have in bolstering equity financing, because interest income (as opposed to dividends or gains) earned on assets in the accounts would not be taxed at either the personal or corporate level. That effect would be strengthened by the fact that the combination of the proposals would increase the share of interest-bearing assets in tax-free accounts--there would be little incentive to hold equities in such accounts if their returns were already largely sheltered from taxes. In addition, corporate income taxes are currently temporarily low, both because firms have relatively low earnings as a result of the sluggish economy and because the temporary investment incentives in the Job Creation and Worker Assistance Act of 2002 reduce the taxes corporations pay on earnings. Because the President's proposal would eliminates taxes at the individual level only on income that was already taxed at the corporate level, the low corporate taxes would limit the initial impact of the proposal on firms' cost of capital. And, in general, the lower combined tax on corporate income would reduce the tax value of accelerated depreciation and the deductibility of corporate interest. CBO incorporated the effects of the proposal to reduce double taxation of corporate income into its analysis in two ways. For the macroeconometric models that CBO used, it estimated the effect of the proposal on the cost of capital in different economic sectors and on share values. CBO then incorporated those estimates into the models, and the models' equations determined the ultimate effect on the economy. For the supply-side models, CBO estimated the overall effect on the average cost of capital and incorporated that estimate into the models. Those models have no mechanism to estimate the effect of the reallocation of capital. To incorporate that effect, CBO reviewed outside estimates of the effect of that reallocation on output, determined a middle-range estimate, and added that amount to the models' underlying estimates of the effect on output. That procedure added an average of 0.1 percent to the estimated effect on gross domestic product over the 2004-2013 period in those models. Expand Tax-Free Savings Accounts The President's budget includes a proposal to create retirement savings accounts (RSAs) and lifetime savings accounts (LSAs) to consolidate the current system of tax-free savings accounts for retirement and other purposes (such as education). The RSAs would replace the three-tiered system of traditional, Roth, and nondeductible individual retirement accounts (IRAs). Taxpayers could use the LSAs to consolidate other savings plans, including the Archer medical savings accounts, Coverdell education savings accounts, and qualified state tuition plans. The proposal would also up the contribution limits, eliminate some of the eligibility restrictions based on income, and liberalize some of the distribution rules. If the President's other proposals were also enacted, the proposal for savings accounts would not have any appreciable effect on the economy on average through 2013, CBO estimates.6 Most taxpayers would simply save the same amount in one of the new accounts as they would have saved in one of their current tax-free accounts. Moreover, people who currently have assets in taxable accounts could reduce their tax liability by selling those assets and putting the cash from the sale into the tax-preferred accounts--an action that would have no effect on private saving. Most new saving would be done in small amounts by taxpayers with few taxable assets to shift. However, the effects beyond 2013 could be larger. CBO estimates that after the first few years, the proposals for new tax-free accounts would have a slight positive effect on saving that would increase after 2013. 1.  Dividends and capital gains are not taxed if they accrue to tax-free accounts or nontaxable entities such as pension funds and nonprofit institutions. In addition, some gains are not taxed because the owner of the asset dies before the gains are realized. In that case, taxes are levied only on increases in value after the owner's death--the so-called step-up in basis at death. 2.  The effective tax rate on capital gains is relatively low in part because investors can defer the realization of the gains, because about half of all gains go untaxed on account of step-up in basis at death, and because some gains accrue to assets held in tax-free accounts. 3.  George R. Zodrow, "On the 'Traditional' and 'New' Views of Dividend Taxation," National Tax Journal, vol. 44, no. 4, part 2 (December 1991), pp. 497-509. 4.  Clemens Fuest and Bernd Huber, The Optimal Taxation of Dividends in a Small Open Economy, Working Paper No. 348 (Munich: CESifo 2000), available at www.cesifo.de. 5.  Alan J. Auerbach and Laurence J. Kotlikoff, Dynamic Fiscal Policy (New York: Cambridge University Press, 1987), pp. 134-136. 6.  The assumption that all of the proposals in the budget are enacted is important because their effects interact. For example, as described above, the proposal to reduce double taxation of corporate income would lessen the incentive to invest equities in tax-free accounts because the returns to those equities would already be largely tax-sheltered. Therefore, fewer people might take advantage of the accounts.

CBO estimated the average effective marginal tax rate on capital income--the rate at which the average additional dollar of capital income is taxed--with and without the budget's policies to estimate the changes in the rate of return on savings (see Table 14). Those changes in effective tax rates are smaller than the changes in statutory income tax rates under the President's proposals because some capital income (such as that which flows into tax-free savings accounts or pension funds) is not taxed.

Table 14.
CBO's Estimates of the Effective Marginal Federal Tax Rates on Capital
(In percent)
Calendar Year	Tax Rates Under Current Law		Tax Rates Under President's Budget		Percentage-Point Difference		Percentage Change
2003	13.8		12.6		-1.2		-8.5
2004	13.7		12.6		-1.1		-8.1
2005	13.7		12.6		-1.1		-8.2
2006	13.5		12.5		-0.9		-6.9
2007	13.5		12.5		-0.9		-7.0
2008	13.5		12.5		-1.0		-7.1
2009	13.5		12.5		-1.0		-7.1
2010	13.5		12.5		-1.0		-7.2
2011	14.1		12.6		-1.5		-10.5
2012	14.1		12.6		-1.5		-10.5
2013	14.1		12.6		-1.5		-10.6
Source: Congressional Budget Office.
Note: Includes federal individual and corporate income taxes.

The proposed reductions in taxes on capital income would raise the return on savings and affect consumption in two opposing ways, just as lowering the marginal tax rate on labor income has opposing effects on labor supply. The increase in the rate of return on savings would raise savers' wealth by increasing their current and future after-tax returns--which would tend to increase current consumption--but also increase the gain in future consumption for every dollar saved, which would tend to increase saving and reduce current consumption.

Perhaps partly for that reason, analysis based on empirical data tends to estimate that changes to the return on savings have a relatively small effect on consumption, which could be positive or negative. However, some models of behavior predict a large negative effect on consumption.

CBO attempted to span that range of estimates: some economic models used by the agency in its analysis assume that the rate of return on savings has little or no effect on consumption, while others assume that increasing the rate of return on saving reduces consumption--and increases saving--significantly.

Finally, as described in the previous section, the increased deficits under the President's budget might lead some people to anticipate changes in policy in the future. If people expected higher taxes, lower transfer payments, or less government services in the future, they might tend to reduce consumption in order to build up savings to compensate for those anticipated policies. CBO used a range of assumptions about those expectations in its estimates.

The President's proposal to make permanent the repeal of the estate and gift tax after 2010 is particularly difficult to analyze. To begin with, there is no clear consensus regarding the motive for leaving bequests, or even whether they are typically the result of a deliberate savings plan. If they are not, repealing the estate tax would not encourage saving. Moreover, those who believe that estate taxes affect consumption and saving disagree about the direction of the effect. A lower estate tax makes it cheaper for people to leave money to their heirs, which could encourage them to save more to leave larger bequests. In contrast, with a lower estate tax, people can leave the same after-tax bequest with less saving, which might induce them to save less. Also, other things being equal, a lower estate tax increases the after-tax size of bequests, which could lead potential recipients to increase their consumption and reduce their saving. Finally, although a great deal of attention has been focused on the role of estate taxes in sectors such as agriculture or activities such as entrepreneurial ventures, the implications for the economy as a whole are less clear.

Because so little is understood about how repealing the estate tax would affect consumption, most of CBO's estimates assumed that in their consumption and saving, people would respond in the same way as they have on average to past spending or tax changes that affected the budget deficit. That assumption implies that people would spend some of their increased after-tax income, increasing aggregate consumption. In one model, however, CBO assumed that people would respond in the same way they would to a change in lump-sum taxes, which have no effect on marginal incentives. That assumption implies that all of the increase in after-tax income would be saved, so consumption would not rise.

Most of CBO's estimates indicate that the President's budget would increase the sum of private and government consumption on net, which would tend to imply somewhat less investment and a smaller capital stock.⁽¹³⁾ Only under the most dramatic assumption about foresight--in which people are assumed to care just as much for future generations as they do for themselves--did CBO estimate the President's budgetary proposals would lead to a bigger capital stock. In effect, if people have a sufficiently long time horizon, they may recognize and counter the deleterious effects of policy on capital formation and, thus, future standards of living.

The President's budget could also affect potential output by changing the mix of capital over time. The proposal with the greatest potential to change the composition of the capital stock is the one to reduce double taxation of corporate income. Some corporate income is taxed twice: once at the corporate level by the corporate income tax and again at the personal level by the individual income tax. That tax treatment creates a distortion in the allocation of capital, discouraging investment in the corporate sector relative to the housing and noncorporate business sectors. As a result, less capital is held in the corporate sector than is efficient. The taxation of dividends also encourages firms to finance investment with debt rather then equity (because interest payments on debt are deducted from tax at the corporate level and so only taxed once), which may also lead to economic inefficiencies. Reducing the tax on dividends would lessen those inefficiences, thereby increasing overall economic output.

Entrepreneurship and Technological Progress. Budgetary policies might conceivably affect the economy by influencing the rate of technological progress. That avenue is potentially important because new and improved processes and products are the source of most of the long-term growth in productivity. Unfortunately, however, economists have little basis for estimating how budgetary policies influence technological innovation. Because so little is understood about the sources of technological progress, CBO has not incorporated into its analysis any effects of the budget on technological progress.

Demand-Side, or Cyclical, Effects. Government policies also affect the economy by adding to or subtracting from the total demand for goods and services in the economy. Increases in demand can cause firms to temporarily gear up production and hire more workers to meet the demand. That type of effect can be especially beneficial if the economy is operating below its potential, which, according to CBO's estimates, it currently is. In that case, if an adjustment to fiscal policy is well-timed, it can help move the economy back to equilibrium more quickly than it would have moved otherwise. Of course, if the adjustment is ill-timed, there are no such benefits, and there could be economic costs.⁽¹⁴⁾

Demand-side effects, however, can only temporarily raise or lower output above what it would have been otherwise--with or without demand-side effects, built-in economic forces tend to move output toward its potential level. Moreover, policies that increase demand by raising government or private consumption tend to lower output in the long run because they tend to eventually decrease investment and the size of the capital stock.

Description of Models and Results
CBO estimated the economic effects of the President's budget using several different models of the aggregate economy. Those models constitute simplified representations of the economy but differ substantially in the ways that they are constructed and the estimates that they produce. The models fall into two broad types. Three of the models that CBO used in its analysis--a "textbook" growth model, a life-cycle growth model, and an infinite-horizon growth model--estimate only supply-side effects. Two commercial macroeconometric models also used by CBO emphasize business-cycle aspects of the economy and are designed primarily to analyze demand-side effects, although they include some supply-side effects as well.

Ten-Year Analysis of Supply-Side Effects. CBO analyzed the supply-side effects of the President's budget on the economy through 2013 using three models: a textbook growth model, a life-cycle growth model, and an infinite-horizon growth model (see Box 5). The textbook growth model is not forward-looking--it assumes that people do not explicitly incorporate expected future policies into their current plans. The life-cycle model is so called because it assumes that people make life-long plans for working and saving but do not care about events after their death. By contrast, the infinite-horizon model assumes that people care about the welfare of their descendants as much as they care about their own. That assumption means people behave as if they will live forever.
 

Box 5. The Models That the Congressional Budget Office Used to Analyze the Economic Effects of the President's Budget Over the Next Decade The Congressional Budget Office (CBO) used three models to estimate the effects of the President's budget from 2003 to 2013: a textbook growth model, a life-cycle growth model, and an infinite-horizon growth model. The textbook growth model, which CBO uses to produce projections of the economy's potential output for the agency's 10-year economic baseline, is an enhanced version of the model developed by Robert Solow, a pioneer of growth-accounting theory.1 It assumes that output is determined by labor supply, the capital stock, and total factor productivity (which represents the state of technological know-how). The textbook growth model is not forward-looking--people do not respond to expected future changes in government policy. The textbook growth model incorporates no effects from demand-side, or cyclical, variations in the economy; the model assumes the economy is always at its potential level. The estimates using the textbook growth model incorporate effects of marginal tax rates on labor supply, which CBO estimated in a side calculation. Those effects increase labor supply relative to the level in CBO's baseline. By contrast, the capital stock is lower than the baseline level because of increased government and private consumption, which crowds out investment. The decrease in the capital stock is limited by two factors, for which the model includes assumptions based on past relationships. First, the increase in private consumption is dampened because people are assumed to increase their private saving by 40 cents for every dollar that the deficit rises. Second, for every dollar that national saving (private plus government saving) falls, the amount of foreign capital invested in the United States is assumed to rise by 40 cents. In the textbook growth model, changes in marginal tax rates on capital have no direct effect on spending by the private sector. The life-cycle growth model and the infinite-horizon growth model differ in fundamental ways from the other models that CBO used in this analysis. The two models incorporate simulated people who make decisions about how much to work and save in order to make themselves as well off as possible over their lifetime. Their behavior is calibrated so that macroeconomic variables such as the total amount of labor supplied and the size of the capital stock match the levels occurring in the U.S. economy. In the life-cycle and infinite-horizon growth models, people's consumption changes by a relatively large amount in response to changes in their after-tax rate of return on saving. Like the textbook growth model, those models do not allow for any demand-side effects. The people in the life-cycle and infinite-horizon models are assumed to be forward-looking--that is, they know all future changes in policy and alter their behavior accordingly. In terms of the degree to which people incorporate future events into their current behavior, this "perfect foresight" is at the other end of the range of possible assumptions from the assumption used in the growth model. Most people in the real world fall somewhere between those two extremes. However, in using those two assumptions, CBO has attempted to span a range of possible responses to the policies in the President's budget. Because people's behavior in the life-cycle and infinite-horizon growth models depends in part on future policies, using those models requires making assumptions about budgetary policies beyond 2013, the end of the projection period. Policies that increase deficits must be offset at some point in the future by taxes that are higher or spending that is lower than it would have been in the absence of the increased deficits. The assumptions about how and when to offset the bill that comes due have a large influence on the estimated economic effects over the 2003-2013 period. That influence stems from the fact that people anticipate the offsetting policies and plan accordingly. In its analysis, CBO used two different assumptions about how the budget would be stabilized after 2013: that taxes would be raised by a lump sum for everyone and that government consumption, which the models assume does not enhance people's well-being, would be cut.2 In general, if people believe that some time after 2013 their taxes will rise, they will work more and consume less in order to build up savings in preparation. Therefore, the effects on economic output before 2013 tend to be relatively more positive under that assumption. If, however, people expect government consumption to fall in the future, rather than taxes to rise, they do not need to work and save more in preparation (under the assumption that such consumption does not enhance people's well-being). So the effects on output over the first 10 years tend to be relatively more negative under that assumption. (If, on the other hand, government consumption was valued by people as highly as they valued their own consumption, the predicted economic effects from assuming a future fall in spending would be the same as those from assuming a lump-sum increase in taxes. The actual impact of government consumption on people's well-being probably falls somewhere between those two extremes.) The life-cycle and infinite-horizon growth models differ in what they assume about how far ahead people look in making their plans. The life-cycle model is calibrated so that the probability of death at a given age matches current U.S. mortality rates, and, as the name of the model suggests, people are assumed to take account of the impact of future economic or policy changes only on themselves and not on their children. In the infinite-horizon model, however, people behave as though the well-being of their descendants is as important to them as their own. That leads them to behave as if they expect to live forever. While that assumption cannot be ruled out, there is some evidence against it.3 The difference in the models' time horizons has an important effect on the resulting estimates. In both models, people expect the increase in deficits under the President's budget to be offset at some point in the future. However, a person in the life-cycle model, especially an older one, knows that he may die before an offsetting policy change occurs. Consequently, that person is less willing to work harder or save more during the 10-year projection period in order to compensate for any future tax increases. By contrast, people in the infinite-horizon model are certain that they (or, equivalently, their descendants, whom they care about as much as they do themselves) will be alive when the offsetting policy change is made. That certainty implies that the expectation of a future increase in taxes will have a greater effect on their current work and saving than it does in the life-cycle model. For that reason, the infinite-horizon model using the assumption of future tax increases produces the most positive estimates of the effect of the budget on the economy. In its analysis using the life-cycle model, CBO used two different assumptions about how open the economy is to flows of capital to and from other countries. One assumption is that the economy is completely closed--no capital can flow into or out of the country. The other assumption is that the economy is completely open and cannot affect the world interest rate--capital flows freely into and out of the country to keep the domestic interest rate equal to a constant world rate. The U.S. economy effectively behaves somewhere between those two extremes, because while it is relatively open to investment, it is so large that its economy can influence world interest rates. The estimated impact on U.S. incomes assuming an open economy tends to be more negative, or less positive, than that assuming a closed economy because of the premise that interest rates cannot rise. In a closed economy, policies that reduce the capital stock tend to increase interest rates, which gives people a greater incentive to save rather than consume and offsets some of the reduction in the capital stock and output. 1.  For a detailed description of the textbook growth model, see Congressional Budget Office, CBO's Method for Estimating Potential Output: An Update (August 2001). 2.  CBO also estimated economic effects assuming that marginal income tax rates, rather than lump-sum taxes, would be raised after 2013. Those results are not presented because they lie between those under the assumptions of lump-sum tax increases and cuts in government consumption. 3.  See Joseph G. Altonji, Fumio Hayashi, and Laurence Kotlikoff, "Risk Sharing Between and Within Families," Econometrica, vol. 64, no. 2 (March 1996), pp. 261-294; Paul Evans, "Consumers Are Not Ricardian: Evidence from Nineteen Countries," Economic Inquiry, vol. 31, no. 4 (October 1993), pp. 534-548; and T.D. Stanley, "New Wine in Old Bottles: A Meta-Analysis of Ricardian Equivalence," Southern Economic Journal, vol. 64, no. 3 (January 1998), pp. 713-727.

The life-cycle and infinite-horizon growth models produced estimates using three different assumptions for how the increased deficits under the budget will eventually be financed (those models require such an assumption about financing because they are forward-looking). The life-cycle model also produced estimates using two different assumptions about how open the economy is to inflows of capital from abroad.

The textbook growth model projection, which makes no assumption about future financing, estimates that the budget will decrease GDP by about 0.2 percent, on average, over the 2004-2008 period and by 0.7 percent over the 2009-2013 period (see Table 15). That model does not assume any direct effect of lower marginal (as opposed to average) tax rates and a higher pretax interest rate on private consumption, but it does incorporate CBO's calculation of the effect of marginal tax rates on labor supply.

Table 15.
CBO's Estimates, from Supply-Side Models, of the Effect of the President's Budgetary Proposals on Real Gross Domestic Product
(Average percentage change from CBO's baseline)
		2004-2008		2009-2013
Without Forward-Looking Behavior
Textbook Growth Model		-0.2		-0.7
With Forward-Looking Behavior
Closed-Economy Life-Cycle Growth Model
	Lower government consumption after 2013	-0.3		-1.5
	Higher taxes after 2013	0.5		0.3
Open-Economy Life-Cycle Growth Model
	Lower government consumption after 2013	-0.6		-0.5
	Higher taxes after 2013	0.3		0.6
Infinite-Horizon Growth Model
	Lower government consumption after 2013	0.2		-0.6
	Higher taxes after 2013	0.9		1.4
Memorandum: Effect on Real Gross National Product
Open-Economy Life-Cycle Growth Model
	Lower government consumption after 2013	-0.8		-2.0
	Higher taxes after 2013	0.3		0
Source: Congressional Budget Office.
Notes: The "textbook" growth model is an enhanced version of a model developed by Robert Solow, a pioneer of growth-accounting theory. For a detailed description of the model, see Congressional Budget Office, CBO's Method for Estimating Potential Output: An Update (August 2001). The life-cycle growth model, developed by CBO, is described in Shinichi Nishiyama and Kent Smetters, "Consumption Taxes and Economic Efficiency in a Stochastic OLG Economy," Technical Paper 2002-6 (December 2002). The infinite-horizon growth model is an enhanced version of a model first developed by Frank Ramsey; see Robert J. Barro and Xavier-I-Martin, Economic Growth (New York: McGraw-Hill, 1995). The three models reflect a wide range of assumptions about the extent to which people are forward-looking in their behavior: in the textbook model, their foresight is the least, while in the infinite-horizon model, it is perfect and extends infinitely to include a full consideration of effects on descendants.
In models with forward-looking behavior, CBO had to make assumptions about how the President's budget would be financed after 2013. CBO chose two alternatives--cutting government consumption or raising taxes.
Real gross domestic product (GDP) is GDP adjusted for inflation.

The estimates produced by the life-cycle and infinite-horizon models depend critically on how the President's budgetary policies affect people's expectations of budgetary policies beyond 2013. The life-cycle growth model projects that if people think the President's budgetary proposals would be financed by eventual decreases in government consumption, economic output would decrease by between 0.3 and 0.6 percent over the 2004-2008 period compared with CBO's baseline and by between 0.5 and 1.5 percent over the 2009-2013 period. However, the life-cycle model projects that if people think the proposals would be financed through a future lump-sum tax increase--an equal dollar tax increase levied on everyone--the proposals would raise output by between 0.3 and 0.5 percent over the first five years and by between 0.3 and 0.6 percent during the second. (Estimates assuming a future increase in marginal tax rates, not shown for brevity, fall between those assuming a future cut in government consumption and those assuming a future increase in lump-sum taxes.⁽¹⁵⁾) Estimates assuming an eventual increase in taxes tend to be more positive because people, as represented in the model, work and save more inside the 10-year projection period in preparation for the tax increase but not for a cut in government spending, which the model assumes people do not value. (Assuming that government consumption was valued as highly as personal consumption would lead to an estimate similar to the one assuming a lump-sum tax increase.)

The estimated economic effects of the budget also depend on the extent to which the economy is open or closed. Assuming an open economy--one in which international capital flows freely to keep U.S. interest rates equal to fixed world rates--tends to lead to larger estimates of GDP on average over the 2004-2013 period. However, that result occurs partly because investment is boosted by inflows of foreign capital, and most of the profits from the investments financed by those inflows go to foreigners rather than U.S. residents. The income of U.S. residents (represented by gross national product (GNP) in Table 15) is actually lower under the assumption of an open economy, despite the higher domestic output. (In a closed economy, GDP and GNP are identical, so the effect on GNP assuming an open economy can be compared directly with the effect on GDP assuming a closed economy.)

The proposals would have the most positive effect on output if people behaved as assumed in the infinite-horizon model and expected the proposals would be financed with a lump-sum tax increase. In that case, the proposals would raise output by 0.9 percent over the first five years and 1.4 percent over the second. As with the life-cycle model, assuming that people expect future cuts in government spending leads to more negative effects on output--an increase in GDP of 0.2 percent during the first five years and a decrease of 0.6 percent during the second. The infinite-horizon model tends to predict more positive effects than the life-cycle model if people expect a future tax increase because, as they are represented in the infinite-horizon model, people know that they (or their descendants, whom they care about as much as themselves) are going to bear the burden of any future increase in taxes.

The economic changes from fiscal policy would in turn affect the budget through 2013 (see Table 16). Under different assumptions, the economic effects of the President's proposals could increase their cost by as much as 10 percent or decrease their cost by as much as 15 percent over the 2004-2008 period and could increase their cost by as much as 15 percent or decrease their cost by as much as 17 percent over the 2009-2013 period.

Table 16.
CBO's Estimates, from Supply-Side Models, of the Cumulative Budgetary Impact of the President's Proposals
(In billions of dollars)
		2004-2008		2009-2013
Conventional Estimate of the President's Proposalsa		-802		-1,908
Budgetary Cost of the President's Proposals with Macroeconomic Feedbacks
Textbook Growth Model		-847		-2,126
Closed-Economy Life-Cycle Growth Model
	Lower government consumption after 2013	-846		-2,194
	Higher taxes after 2013	-745		-1,817
Open-Economy Life-Cycle Growth Model
	Lower government consumption after 2013	-880		-2,013
	Higher taxes after 2013	-753		-1,760
Infinite-Horizon Growth Model
	Lower government consumption after 2013	-775		-1,989
	Higher taxes after 2013	-680		-1,587
Budgetary Savings or Cost (-) from Macroeconomic Feedbacks as a Percentage of the Conventional Estimateb
Textbook Growth Model		-6		-11
Closed-Economy Life-Cycle Growth Model
	Lower government consumption after 2013	-6		-15
	Higher taxes after 2013	7		5
Open-Economy Life-Cycle Growth Model
	Lower government consumption after 2013	-10		-5
	Higher taxes after 2013	6		8
Infinite-Horizon Growth Model
	Lower government consumption after 2013	3		-4
	Higher taxes after 2013	15		17
Source: Congressional Budget Office.
Notes: The "textbook" growth model is an enhanced version of the model developed by Robert Solow, a pioneer of growth-accounting theory. For a detailed description of the model, see Congressional Budget Office, CBO's Method for Estimating Potential Output: An Update (August 2001). The life-cycle growth model, developed by CBO, is described in Shinichi Nishiyama and Kent Smetters, "Consumption Taxes and Economic Efficiency in a Stochastic OLG Economy," Technical Paper 2002-6 (December 2002). The infinite-horizon growth model is an enhanced version of a model first developed by Frank Ramsey; see Robert J. Barro and Xavier-I-Martin, Economic Growth (New York: McGraw-Hill, 1995). The three models reflect a wide range of assumptions about the extent to which people are forward-looking in their behavior: in the textbook model, their foresight is the least, while in the infinite-horizon model, it is perfect and extends infinitely to include a full consideration of effects on descendants.
In models with forward-looking behavior, CBO had to make assumptions about how the President's budget would be financed after 2013. CBO chose two alternatives--cutting government consumption or raising taxes.
a. CBO's estimate of the budgetary impact assuming no macroeconomic feedbacks.
b. A negative number means that the macroeconomic feedbacks are estimated to increase the budgetary cost; a positive number, that they are estimated to reduce it (or provide savings).