After slow growth during the first half of 2003, the economy now seems poised to expand at a faster pace. Since the first of the year, economic output has grown at an average annual rate of about 2 percent, reflecting not only tensions attributable to the war in Iraq but also a host of other factors, including the slow growth of foreign demand for U.S. goods, fiscal constraints on state and local spending, and businesses' concerns about the durability of the economy's recovery from the 2001 recession. Signs of a pickup in consumer and business spending in the second quarter, the rapid growth of federal purchases, enactment of tax cuts for firms, and a slightly more accommodative monetary policy have improved the economic outlook for the remainder of 2003 and for 2004. The Congressional Budget Office anticipates a rebound in demand in 2003 and real (inflation-adjusted) growth of gross domestic product that approaches 4 percent in calendar year 2004 (see Table 2-1 and Figure 2-1).

Figure 2-1.

The Economic Forecast and Projections

Sources: Congressional Budget Office; Department of Commerce, Bureau of Economic Analysis; Department of Labor, Bureau of Labor Statistics; Federal Reserve Board.

Note: All data are annual values; percentage changes are year over year.

a. The change in the consumer price index for all urban consumers, applying the current methodology to historical price data (CPI-U-RS).

Yet the outlook for 2003 and 2004 remains uncertain. On the one hand, foreign economic growth and foreign demand for U.S.-produced goods may be greater or smaller than CBO estimates, and the degree to which state and local governments are likely to curtail spending growth cannot be foreseen. In addition, the residual effects of certain economic developments in recent years--the fall in equity markets, the large reduction in households' net wealth, the drop in the personal saving rate, businesses' productive capacity that remains underused, and the U.S. economy's increased dependence on inflows of foreign capital--may continue to dampen growth to a greater extent than CBO has assumed. On the other hand, a number of favorable economic fundamentals, including low inflation and rapid productivity growth, may set the stage for another long period of strong growth. Beyond that, the forecast is of course subject to the uncertainty that surrounds the economic effects of the war on terrorism, developments in Iraq, and events elsewhere in the world.

Beyond 2004, real GDP growth will average 3.3 percent between 2005 and 2008 and 2.7 percent between 2009 and 2013, CBO projects. Growth of real GDP slows under CBO's projections as the output gap--the difference between GDP and potential GDP--closes. (Potential GDP is the highest sustainable level of GDP consistent with a constant rate of inflation.) Once the gap has been eliminated, projected real GDP grows at the same rate as potential GDP. CBO's baseline projections also reflect the macroeconomic effects of provisions in the Jobs and Growth Tax Relief Reconciliation Act of 2003.
 

CBO's Two-Year Outlook

Through the end of 2004, the economy will continue to recover from the recession of 2001, CBO forecasts. Because the unemployment rate will remain high and businesses' utilization of their productive capacity will still be low, the main determinant of GDP growth in the near term will be the speed with which the demand for goods, services, and structures grows and puts those underemployed resources to use.

Demand has not grown consistently in recent quarters. Late in 2002 and early in 2003, uncertainty about the underlying strength of the economy and the prospects of war in Iraq strained equity markets, raised oil prices, and depressed consumer confidence. Spending on consumer goods and services remained sluggish while businesses curtailed investment. For half a year--the last quarter of 2002 and the first of 2003--real GDP grew about half as fast as the trend growth rate of potential GDP. But sentiment among consumers and in the financial and energy markets improved in the second quarter of 2003. In addition, federal defense spending picked up sharply. As a result, the growth of real GDP rose at an annual rate of 2.4 percent. That growth would have been faster if firms had not met some of the increase in demand by drawing down their inventories. However, GDP in the second quarter could be revised upward because the data for trade and retail sales in June suggest that the economy was stronger than had previously been thought.

CBO's two-year forecast anticipates that demand will grow more rapidly in coming quarters than it did in the first half of 2003. In the near term, the growth of consumption will remain modest because households are likely to save much of the additional disposable resulting from the accelerated tax cuts in JGTRRA in order to rebuild their wealth. Businesses are likely to begin to restock, rather than draw down, their inventories and to increase their spending on producers' durable equipment and structures--so-called fixed investment. Consequently, real GDP is likely to grow by 3.8 percent in calendar year 2004, up from 2.2 percent in 2003. (See Table 2-2 for fourth-quarter-to-fourth-quarter percentage changes.) CBO's forecast incorporates the assumption that the federal government's spending will bolster the growth of demand over the next few quarters, but under CBO's baseline projections (which are described in Chapter 1), federal spending growth is expected to slow in 2004.

Table 2-2.
CBO's Economic Forecast for 2003 and 2004
		Actual 2002		Forecast
				2003	2004
Calendar Year Average
Real GDP (Percentage change)		2.4		2.2	3.8
Unemployment Rate (Percent)		5.8		6.2	6.2
Three-Month Treasury Bill Rate (Percent)		1.6		1.0	1.7
Ten-Year Treasury Note Rate (Percent)		4.6		4.0	4.6
Fourth Quarter to Fourth Quarter (Percentage change)
Nominal GDP		4.3		4.0	5.8
Real GDP		2.9		2.6	4.1
GDP Price Index		1.3		1.4	1.7
Consumer Price Indexa
	Overall	2.2		2.0	2.2
	Excluding food and energy	2.1		1.5	2.3
Sources: Congressional Budget Office; Department of Commerce, Bureau of Economic Analysis; Department of Labor, Bureau of Labor Statistics; Federal Reserve Board.
a. The consumer price index for all urban consumers.

CBO does not anticipate a quick drop in the unemployment rate from its current level of 6.2 percent. Typically, the unemployment rate falls when the growth of real GDP exceeds its potential rate. But even though the growth of GDP that CBO is forecasting exceeds its estimate of potential GDP, CBO expects that unemployment will average 6.2 percent in both 2003 and 2004. In part, that sustained high rate reflects caution on the part of employers; if they follow recent patterns, they are not likely to resume hiring immediately as demand begins to grow. In part, the rate also reflects the likelihood that the labor force will grow more quickly than it has in recent quarters, because people who became discouraged in their job searches in the past few years may now reenter the labor market and be tallied among the unemployed.

Financial Conditions and Monetary Policy

Financial conditions have generally improved since the beginning of this year, and CBO forecasts that monetary policy will continue to support the growth of demand through the end of 2003. An index of financial market conditions, combining the stance of monetary policy with a quantitative assessment of the channels through which that policy operates, can indicate roughly the extent to which financial markets support economic growth (see Figure 2-2).⁽¹⁾ Despite the Federal Reserve's efforts to boost the economy, the index, primarily because of a slumping stock market, indicated that financial and monetary conditions were not conducive to growth in 2001 and 2002. But as the U.S. dollar gradually depreciated and the stock market rebounded during the second quarter of this year, the index improved markedly.
 

Figure 2-2.

An Index of Monetary and Financial Conditions

(Percentage points of GDP growth)

Sources: Congressional Budget Office; Macroeconomic Advisers, LLC.

Notes: This index estimates how supportive financial markets are of the rate of growth of real GDP. It draws on statistical relationships between real GDP and financial variables such as interest rates, exchange rates, and equity values. When the index is positive, overall conditions in the financial markets are conducive to the growth of real GDP. When it is negative, overall financial market conditions are a drag on growth.

The last data point is the second quarter of 2003.

In May, the Federal Reserve said it believed that the risk of lower inflation was minor but exceeded that of higher inflation. In turn, participants in financial markets concluded that the federal funds and other short-term interest rates would drop further and would stay low longer than they had originally anticipated--possibly well into 2004. The yield on 10-year Treasury notes (which embodies expectations about future short-term rates) consequently fell from 3.9 percent at the end of April to 3.1 percent in mid-June, its lowest rate since the late 1950s.

In late June, however, when the Federal Reserve cut a quarter of a percentage point from its target for the federal funds rate (the rate that financial institutions charge each other for overnight loans of monetary reserves)--which left the rate at 1 percent--the yield on 10-year Treasury notes actually increased. That rise apparently reflected not only disappointment among financial market participants expecting a larger rate cut but also other factors, including increased signs of a pickup in real growth and perhaps new concerns about the prospect of large federal funding requirements. By early August, the rate on 10-year Treasury notes exceeded 4 percent.

CBO forecasts that short-term interest rates will remain at their current low levels through 2003 but will rise once the growth of demand begins to pick up (see Table 2-2). The interest rate on three-month Treasury bills will remain near 1.0 percent during the remainder of 2003, CBO estimates, and then increase in early 2004. The rate on 10-year Treasury notes, after averaging 3.6 percent in the second quarter, is expected to average about 4.1 percent in the second half of 2003, climbing to approximately 5.0 percent by the end of 2004.

Fiscal Conditions

Actions by federal policymakers from early 2001 to the present have mitigated the recent economic downturn and bolstered the recovery by cutting taxes and increasing spending. Cumulatively, those policies (not including shifts in the timing of payments) reduced the surplus in fiscal year 2001 by about $50 billion and increased the deficit in 2002 by approximately $180 billion. In 2003 and 2004, they are projected to increase the deficit by about $360 billion and about $520 billion, respectively. Policy actions by many state and local governments, which have faced budget shortfalls in recent years, are not likely to substantially offset the effects of federal policies (see Box 2-1).
 

Box 2-1. Are State and Local Fiscal Actions Offsetting Federal Fiscal Actions? Many state and local governments, faced with severe shortfalls of revenues, are taking action to avoid budgetary imbalances. Those measures on their own would tend to reduce aggregate demand--which is the opposite of the effect of recent federal fiscal actions. However, the federal government's fiscal actions are much larger than the measures that the state and local governments are likely to employ. As a result, the government sector as a whole is contributing to the growth of demand in the short term. The strong economy and soaring stock market of the late 1990s significantly boosted the growth of government revenues, enabling states and localities to cut taxes, undertake new programs, and hike spending for existing activities at the same time they were building their reserves to high levels.1 But the recession of 2001 and the stock market slump that began in 2000 severely reduced the growth of government revenues, which raised the prospect of substantial state budget deficits if remedial actions were not taken.2 Most states have constitutional or statutory requirements for balancing their general fund budgets (though most requirements also allow for the use of some temporary sources of financing); therefore, they had to act to address those prospective deficits. Initially, states responded to weakening revenues by using their reserves and tapping other sources of temporary funding. Increasingly, though, they have cut the growth of spending and raised taxes. For example, general fund spending by the states, which amounts to about 5 percent of gross domestic product (GDP), rose by 8.3 percent in fiscal year 2001, but spending growth then plummeted to a rate of 1.3 percent in 2002 and an estimated 0.3 percent in 2003.3 The reduction in the growth of states' general fund spending, however, is a narrow and thus potentially misleading indicator of the macroeconomic impact of the state and local sector. A more comprehensive measure of state and local spending can be drawn from the national income and product accounts (NIPAs), and that measure indicates much less of a slowdown. The NIPAs include all state spending--not just that from general funds (which constitutes about one-half of total state expenditures); the measure also takes into account spending by local governments, which exceeds total expenditures by the states.4 Total spending by the state and local sector as recorded by the NIPAs (which amounts to about two-thirds that of the federal sector) grew by 8.5 percent in fiscal year 2001, 5.9 percent in 2002, and 5.3 percent in 2003. (See Appendix B for a discussion of the NIPAs.) Yet looking exclusively at the slowdown in the growth of the NIPA measure of state and local spending still overstates the sector's macroeconomic impact. The decline in the growth of state and local revenues that can be attributed to the weak economy, although much smaller than at the federal level, has helped cushion disposable income. That buffer helps to offset some of the economic restraint coming from the slower growth of spending by state and local governments.5 CBO expects further budgetary actions by state and local governments that will tend to reduce the growth of aggregate demand, but they are unlikely to significantly offset the federal government's actions (including about $20 billion of aid to states over fiscal years 2003 and 2004 from JGTRRA), which are working to increase demand. The historical record suggests that the offset to federal fiscal policy will be relatively small: the budget balances of the state and local government sector fluctuate much less than those of the federal government relative to potential (cyclically adjusted) GDP (see the figure below). At the same time, slower-than-expected growth of the economy could reduce the growth of state and local government revenues below current projections, and that could result in larger tax increases and more spending restraint by state and local governments than are now expected.   Government Surpluses or Deficits (Percentage of potential GDP) Sources:  Congressional Budget Office; Department of Commerce, Bureau of Economic Analysis. Note:  The last data point is CBO's forecast for 2003. Accounting for Fiscal Pressures on State and Local Governments The fiscal difficulties of states and localities seem to be due to a combination of a slowdown in the growth of revenues and a lag in the slowdown in the growth of spending. Throughout most of the 1990s, total state and local spending rose relative to potential GDP, but revenues grew even faster, producing growing surpluses (see the figure below). In the late 1990s, spending grew faster than revenues, but the years of state and local surpluses did not end until the recession began in 2001. With the recession, the growth of revenues slowed dramatically, but the growth of spending did not slow quickly enough to avoid the onset of deficits.   State and Local Taxes and Spending (Percentage of potential GDP) Sources:  Congressional Budget Office; Department of Commerce, Bureau of Economic Analysis. Note:  The last data point is the second quarter of 2003. State Budget Shortfalls and Recent Actions Taken by States Thirty-nine states projected budget shortfalls for 2003, the fiscal year that ended in June for most states.6 (A shortfall is defined by the National Conference of State Legislatures as the difference between a state's projection of general fund revenue collections and projected general fund spending during a fiscal year. General fund revenues are essentially those raised by states through taxes and fees and exclude rainy-day funds, federal funds, and bond funds.) California, New York, and Texas, which account for 30 percent of states' general fund spending, accounted for almost 50 percent of the budget shortfall reported by states during fiscal year 2003. States have taken significant budgetary action since 2001 to respond to fiscal pressures. In addition to slowing the growth of general fund spending, they have drawn down their reserves from nearly $49 billion at the end of fiscal year 2000 to less than $7 billion at the end of fiscal year 2003. States have also imposed net tax increases since 2002 totaling almost $25 billion (on an annual basis), which is equivalent to about 5 percent of states' general fund revenues.7 In 2001, the net effect of changes in taxes and fees among all the states was a reduction in revenues of about $6 billion, with eight states increasing taxes and 28 states decreasing them. That was the seventh consecutive year in which states had cut taxes on a net basis. In fiscal year 2002, 15 states reduced taxes while 14 states increased them; the net effect for that year was an increase of about $0.3 billion. Three states--Minnesota, New Jersey, and North Carolina--accounted for the bulk of those increases. In 2003, 24 states enacted tax and fee increases totaling more than $8 billion. Three categories accounted for most of the net rise in taxes: cigarette and tobacco taxes (19 states had increases totaling almost $3 billion), sales taxes (11 states had increases totaling almost $1.5 billion), and corporate income taxes (eight states had increases totaling about $1 billion). Thus far in fiscal year 2004, 44 states have enacted tax hikes (mostly on sales) or various fee increases totaling over $15 billion (or about 3 percent of general fund revenues). At the same time, general fund spending is budgeted to grow by about 1 percent above 2003 levels. 1.  See Ronald K. Snell, Corina Eckl, and Graham Williams, State Spending in the 1990s (Washington, D.C.: National Conference of State Legislatures, July 14, 2003). 2.  The federal tax cuts contained in the Economic Growth and Tax Relief Reconciliation Act of 2001 (EGTRRA), the Job Creation and Worker Assistance Act of 2002 (JCWAA), and the Jobs and Growth Tax Relief Reconciliation Act of 2003 (JGTRRA) had adverse effects on state and local revenue collections, because many states and localities in some way tie their income taxes to federal tax calculations. To some extent, however, state and local governments have taken action to mitigate that adverse effect by "decoupling" their income tax calculations for individuals and businesses from some of the new federal tax provisions. 3.  See National Association of State Budget Officers and National Governors Association, Fiscal Survey of the States (Washington, D.C.: National Association of State Budget Officers and National Governors Association, June 2003). Most state fiscal years run from July through June. 4.  Although the NIPAs usually report data for state and local governments combined, a recent publication separates data for states and localities, reporting them on an annual basis for the 1959-2001 period. It also discusses several important differences between state budget and NIPA data. See "Receipts and Expenditures of State Governments and of Local Governments, 1959-2001," Survey of Current Business (June 2003), pp. 36-53. 5.  Estimates of the cyclical sensitivity of state and local budgets are presented in Brian Knight, Andrea Kusko, and Laura Rubin, "Problems and Prospects for State and Local Governments," State Tax Notes, August 11, 2003, pp. 427-439. 6.  See National Conference of State Legislatures, State Budget and Tax Actions 2003 (Washington, D.C.: National Conference of State Legislatures, July 23, 2003). According to the most recent survey conducted by the National Conference of State Legislatures, the shortfalls projected during 2003 exceeded 10 percent of general fund spending in 20 states, yet balances available in all general funds and "rainy-day" reserves totaled 3 percent of general fund spending that year. 7.  See National Association of State Budget Officers and National Governors Association, Fiscal Survey of the States.

Two pieces of tax legislation--the Economic Growth and Tax Relief Reconciliation Act of 2001 and the Jobs and Growth Tax Relief Reconciliation Act of 2003--have boosted disposable personal income. JGTRRA accelerates into 2003 reductions in personal income tax rates that were previously scheduled under EGTRRA but would have taken effect later. Beyond that, JGTRRA includes provisions that expand marriage penalty relief, temporarily increase exemptions under the alternative minimum tax, and temporarily expand the child tax credit (through 2004). JGTRRA also cuts tax rates on dividends and capital gains (through 2008). CBO estimates that the law's provisions lower personal income tax payments by a total of $39 billion in fiscal year 2003 (or by approximately 0.4 percent of projected GDP for that year)--almost entirely in the third quarter. The tax reduction will add about 8 percentage points (measured at an annual rate) to the growth of disposable income in the third quarter, CBO estimates.

JGTRRA also includes provisions to spur businesses to invest. The law expands incentives in the Job Creation and Worker Assistance Act of 2002 (JCWAA) to bolster business fixed investment by temporarily increasing, for tax purposes, the fraction of new investment spending that firms can "expense" (basically, deduct from their taxable income immediately rather than over time). JGTRRA allows firms, through the end of 2004, to expense 50 percent of the value of new investment in the first year after purchase; it also increases through 2005 the limit on the expensing of new depreciable assets by small businesses.

Composition of Demand Growth

The growth of demand stems from the actions of private decisionmakers--in the household and corporate sectors, both at home and abroad--and from decisions about government spending.

The Household Sector. Households' spending on consumption and housing, which expanded during the recent recession and has continued to grow since then, is likely to lag behind the growth of GDP in the near term. Real consumer spending grew at an average annual rate of 2.7 percent in the first two quarters of this year, a bit below its average of 2.9 percent during the second half of last year, and real residential investment (primarily new homes and improvements to existing homes) rose by more than 8 percent. Consumer spending is expected to continue growing at an average annual rate of between 2.5 percent and 3.0 percent; residential investment is forecast to grow slightly through 2004.

Several factors are likely to hold down the growth of household spending in the remainder of 2003 and in 2004. First, the effects of the sales incentives offered by automobile manufacturers appear to be waning. Second, the recent increase in long-term interest rates is expected to slow residential investment and curtail the boom in cash-out mortgage refinancing. Third, households' loss of wealth over the past three years--particularly wealth lost through the decline of the stock market, which has now only partly recovered--may continue to moderate the pace of household spending. Fourth, the slow recovery projected for labor markets is likely to restrain the growth of income. Finally, although the household sector as a whole appears to be in generally good financial health, some people are evidently feeling the strain of a long period of economic weakness.

Employment and Income. Weak labor markets have held down the growth of consumers' incomes since the onset of the recession. The outlook for the remainder of 2003 and for 2004 includes moderate gains in employment as the growth of demand picks up, but the recovery of labor markets is still likely to lag behind that of the economy.

The record of employment growth over the past two years has been even worse than in the "jobless recovery" of the 1991-1993 period (see Figure 2-3). (Another measure of employment looks slightly more hopeful; see Box 2-2.) From the business-cycle peak in March 2001 to December 2002, total nonfarm payroll employment fell by 1.8 percent (or 2.3 million jobs). It declined by another 0.3 percent (or 328,000 jobs) during the first seven months of 2003.⁽²⁾ That weakness reflects the lackluster nature of the recovery of demand in the economy and an extraordinary increase in productivity.
 

Figure 2-3.

Private Payroll Employment During Recoveries from Recessions

(Index, trough month = 100)

Sources: Congressional Budget Office; Department of Labor, Bureau of Labor Statistics.

Note: The trough of the last recession was November 2001, as designated by the National Bureau of Economic Research.

a. Average of seven recoveries during the 1949-1990 period, excluding the 1980 recovery.

 

Box 2-2. Uncertainty About Recent Employment Growth Assessing the current state of labor markets is more difficult than usual because the two employment surveys used in such analyses present conflicting stories about the recent growth of employment. (The two measures--the establishment, or payroll, survey and the household survey--are both published by the Bureau of Labor Statistics [BLS].) The household survey, which is based on interviews with individuals in their homes, implies that employment has recovered modestly since the recession reached its trough in November 2001. In contrast, the establishment survey, which is based on payroll data reported by firms, indicates that there are about 1 million fewer jobs in mid-2003 than there were at the trough. That disparity between the two stories holds up even after adjusting for some of the obvious differences in the surveys (the population adjustment that BLS made in the household survey in January 2003 and whether they count self-employed people or multiple jobholders). The two surveys usually provide slightly different pictures of employment growth during recoveries, but the difference is larger than usual this time. Moreover, this is the first instance in which one survey indicates employment growth while the other suggests contraction. The establishment survey better reflects the state of labor markets, the Congressional Budget Office believes, not only because other indicators also imply rather weak labor-market conditions but because large revisions or misreporting appears less likely for the establishment than for the household data. Data on tax withholding conform better to the establishment survey's results than to the household survey's; in addition, both the share of employed people who are working part-time for economic reasons and the still low labor force participation rate indicate weaker labor markets than those existing at the trough. Three other measures suggest the same conclusion: during the first half of the year, the unemployment rate rose, both initial and continuing claims for unemployment insurance remained elevated, and the help-wanted index fell.

Productivity growth normally increases during the economy's recovery from a recession, as businesses use existing labor and capital more intensively while remaining cautious about permanently adding workers in the face of excess productive capacity. As a result, employment--although growing--does not keep pace with output, and output per worker increases. In the past year, however, productivity has grown even more rapidly than in many past recoveries. Private payroll employment thus continued to decline as the growth of demand remained subdued. Consequently, productivity growth accounted for more than 100 percent of the rise in GDP over the period. According to CBO's forecast, productivity's share of GDP growth will fall to about two-thirds over the near term, an outcome consistent with increasing employment.

Yet the unemployment rate is expected to remain high, averaging 6.2 percent over the forecast period. In recent months, the rate rose from 6.0 percent last December to 6.4 percent in June and then fell slightly, to 6.2 percent, in July. Unemployment has been lower since the recession than it might have been because an unusually large number of people left the labor force, possibly because they were discouraged by the poor prospects for employment.

The number of people working (or actively looking for work) as a percentage of the working-age population dropped from 67.3 percent early in 2000 to 66.4 percent in mid-2003. Although such a decline is common in business-cycle downturns, the recent fall has been both slightly greater and more prolonged than usual. As the job market improves, some people who have stayed out of the labor force are likely to begin to seek work, and labor force growth will thus almost keep pace with the growth of employment in 2004, CBO estimates. By the end of 2004, employment growth will have picked up enough to begin to bring the unemployment rate down below 6 percent.

CBO anticipates that the growth of personal income will rise during the balance of 2003 and in 2004. Over the first half of this year, real disposable personal income grew at an average annual rate of 2.3 percent, although real wage and salary income grew by only 1 percent. CBO's forecast incorporates the assumption that wage and salary income will begin to rise and will grow at the same rate as nominal GDP. It also incorporates the sharp increase in disposable income that the tax cuts are expected to bring in the second half of 2003.

The Housing Market. In the near term, housing will no longer be a strong contributor to growth, CBO forecasts. By the end of June, interest rates for 15- and 30-year mortgages had fallen from December's levels by more than 80 basis points--to 4.6 percent and 5.2 percent, respectively--and were the lowest since Freddie Mac began recording them in 1971. Largely because of that drop in rates, the pace of new-home building was rapid during the first half of 2003. But the increase in real residential investment fell short, despite the dip in mortgage rates, of the contribution housing has made to previous recoveries (see Figure 2-4). By late July, mortgage rates had retraced all of their first-half declines. The housing market is thus likely to cool in the near term, although the level of housing starts remained high through July.
 

Figure 2-4.

Residential Investment

(Percentage of potential GDP)

Sources: Congressional Budget Office; Department of Commerce, Bureau of Economic Analysis.

Note: The last data point is the second quarter of 2003.

In the first half of 2003, as in 2002, many consumers chose to refinance mortgages to take advantage of exceptionally low interest rates. (Such refinancing does not alter the overall wealth of the household sector because the gains of people who refinance come at the expense of losses for those holding the original mortgages.) Freddie Mac estimates that homeowners cashed out about $50 billion of home equity from their mortgage refinancing in the first half of the year, following record cash-outs of $96 billion in 2002. Consumers tend to use about one-half of their cash-outs to finance additional spending on home improvements or consumer goods. They use the remainder apparently to make portfolio adjustments by, for example, paying down nonmortgage debt or increasing other investments.⁽³⁾

Households' Net Wealth and Saving. The net wealth of households (their financial plus tangible assets minus their debts) has fallen sharply since its peak in 2000, but it changed relatively little from the third quarter of 2002 through the first quarter of 2003 (the latest available data), holding at roughly $39 trillion. That plateau reflects the largely offsetting influences of the continued appreciation of households' real estate holdings, a decline in the stock market, and a sharp rise in debt, primarily home mortgage debt. Consequently, the ratio of households' net worth to their disposable personal income (the net wealth ratio) has varied little for the past few quarters (see Figure 2-5). The significant rebound in stock prices in the second quarter suggests that households' net wealth may have risen in recent months.
 

Figure 2-5.

Households' Net Wealth

(Ratio to disposable personal income)

Sources: Congressional Budget Office; Federal Reserve Board; Department of Commerce, Bureau of Economic Analysis.

Note: The last data point is the first quarter of 2003.

Given the drop in their net wealth since 2000, households have raised their rate of saving--but by less than recent experience suggests they might have (see Figure 2-6). Between the fourth quarter of 1999 and the first quarter of 2003, when the net wealth ratio fell from 6.3 to 4.9, the personal saving rate rose from 1.9 percent to 3.6 percent. The last time the net wealth ratio was 4.9 was in the third quarter of 1995, when the personal saving rate was 5.3 percent. That suggests that households may wish to raise their saving rate in the near term.
 

Figure 2-6.

Net Wealth and the Personal Saving Rate

Sources: Congressional Budget Office; Department of Commerce, Bureau of Economic Analysis; Federal Reserve Board.

Note: The graph displays the relationship between the personal saving rate and the ratio of household net wealth to disposable personal income from 1952 to the first quarter of 2003. Each point represents a saving rate/net wealth ratio pair for a quarter within that time span. The points arrange themselves into two distinct groups: the dark squares designate quarters prior to 1994 and the circles, quarters in 1994 and after. The figure suggests that the relationship between the saving rate and the net wealth ratio shifted downward after 1993.

Some analysts have worried (on the basis of comparisons with earlier quarters, also shown in Figure 2-6) that households might want to raise their saving rate even more--which would imply correspondingly weaker growth of consumption. For example, at the end of 1967, when the net wealth ratio was similar to the level in the first quarter of this year, the saving rate averaged 9.5 percent. However, as Figure 2-6 suggests, the relationship between the net wealth ratio and the personal saving rate appears to have shifted down since 1993, implying less cause for concern.

The Financial Health of the Household Sector. The household sector appears financially sound overall. Households' debt has risen sharply during the recovery, growing by 10 percent in 2002 and by another 10 percent in the first quarter of 2003. Home mortgage debt accounted for much of that first-quarter expansion; it rose by 12 percent compared with an increase of 4.3 percent in debt for consumer credit. The refinancing boom has allowed homeowners to repay more-costly consumer debt and lower their monthly mortgage payments. That boom, plus modest growth of consumer spending and lower interest rates, has helped stabilize the household sector's debt-service burden over the past year. Another indicator of the sector's financial health, the delinquency rate on a broad range of consumer loans at commercial banks, is now below a two-decade average of about 2.3 percent (see Figure 2-7). Moreover, the delinquency rate on conventional mortgages, which make up the vast bulk of outstanding mortgage loans, has steadied, skirting mid-1990 levels of a little over 3 percent of all loans during the first half of 2003.
 

Figure 2-7.

Delinquency Rate on Consumer Loans at Commercial Banks

(Percentage of loans)

Sources: Congressional Budget Office; American Bankers Association.

Notes: The data cover eight kinds of closed-end (nonrevolving) loans: personal, direct auto, indirect auto, mobile home, recreational vehicle, marine-related, property improvement, and home-equity and second mortgages.

The last data point is the first quarter of 2003.

Nevertheless, certain groups of borrowers are experiencing financial distress. The delinquency rates on bank credit card debt as well as on mortgage loans guaranteed by the Veterans Administration (VA) and Federal Housing Administration (FHA) remain at relatively high levels. However, VA and FHA loans are generally more prone to default than are conventional mortgages.

The Government's Purchases of Goods and Services. Recent federal spending policies contributed to the growth of demand in the first half of this year and are expected to raise demand further during the second half. Real federal purchases, measured on a national income and product account basis, will expand by about 10 percent this year, CBO estimates, driven largely by real defense spending. But CBO's budget projections are required to assume that appropriations after the current budget year will increase from their current levels by only the rate of inflation. Thus, CBO's economic forecast incorporates the assumption that real growth of federal spending for goods and services will slow in the future.

The pace of state and local governments' real purchases has slowed this year but has not substantially offset the strong growth of real federal purchases (see Box 2-1). With state and local revenues low relative to their trends in the late 1990s, the aftereffects of slumping equity markets and the recent recession have led state and local governments to curtail the growth of their purchases of goods and services to help correct fiscal shortfalls. As a result, real state and local spending (measured on a NIPA basis) is expected to be roughly flat in both 2003 and 2004 after expanding at an annual rate of almost 3 percent in 2002.

Net Exports. The U.S. trade deficit rose from 4.1 percent of GDP a year ago to a record 4.7 percent in the second quarter of 2003. According to CBO's estimates, the gap will continue to widen through the end of 2004. The drop in the value of the dollar against the currencies of the United States' major trading partners since early last year is still modest on a trade-weighted basis, but with the decline expected to continue through the end of 2004, the dollar's depreciation will eventually help narrow the trade deficit. In the interim, relatively weak growth abroad has continued to depress the growth of U.S. exports. If, as CBO expects, foreign economic growth picks up and the dollar continues to decline through the end of 2004, the real trade balance will improve in 2005 and beyond.

Foreign Economic Conditions. Overall economic conditions in the rest of the world are weak. With growth in the European Union (EU) and Japan lagging behind U.S. growth, and with developing economies dependent on export markets in more advanced countries, the United States is expected to continue as the primary locomotive of world economic expansion in the near term.

The economies of many major European countries continue to struggle toward recovery. Real GDP has contracted in several of them, including Germany, Italy, the Netherlands, and Switzerland. Germany, the largest economy in the euro zone (the EU countries that have adopted the euro), has technically entered its second recession in two years and, despite some signs of a nascent rebound, is not expected to make a strong recovery in the near term. Economic growth in the other two major European economies, France and the United Kingdom, remains only slightly positive.

Fiscal and monetary policies in the euro zone thus far have not played a significant role in increasing demand. The protracted economic stagnation in Germany and France has helped push government deficits above the limit (3 percent of GDP) set under the fiscal rules adopted by the euro zone nations, leaving little room for expansive policies. The European Central Bank cut its target interest rate by 75 basis points (0.75 percentage points) to 2 percent during the first half of the year; however, the appreciation of the euro against the dollar since March 2002 has made products priced in euros more expensive, offsetting some of the interest rate reduction's effect on overall demand in the euro area.

In Asia, Japan's economy in the second quarter grew faster than had been expected, expanding by 2.3 percent, but its near-term economic outlook continues to be dominated by the depressing effects of entrenched price deflation, "nonperforming" loans, and a large public debt. Among emerging nations in Asia, growth held up relatively better, although it lost some momentum partly as a result of the spread of SARS (severe acute respiratory syndrome) and the economic slump in the industrialized countries. Economic growth in China, which also lost steam in the second quarter as a result of SARS, is expected to resume its near 8 percent average annual rise in the second half of this year. Robust foreign demand for China's manufactured exports, which have become even more competitive in the international market as the yuan has depreciated with the dollar, has helped bolster China's continuing economic strength.

In South America, the economic crises in Brazil and Argentina are now past, but significant obstacles to sustained growth remain. In North America, a sluggish U.S. economy and an appreciating Canadian dollar have slowed the growth of real GDP in Canada to less than 3 percent. In Mexico, real GDP has declined for two consecutive quarters, in large measure reflecting a slump in U.S. demand for Mexican exports and direct competition from Chinese imports.

The Dollar's Exchange Rate. CBO expects the U.S. dollar to continue to gradually depreciate during the second half of 2003 and in 2004. The dollar fell in the first half of this year against a broad basket of currencies, extending a downward trend that began in March 2002. It declined most sharply against the currencies of many industrialized trading partners--with the exception of Japan--and more modestly against the currencies of other Asian nations (see Figure 2-8). That more limited depreciation may have been due to Asian governments' interventions in the foreign exchange markets.
 

Figure 2-8.

The Dollar's Exchange Rate Relative to Selected Currencies

(Index, January 2002 = 1)

Sources: Congressional Budget Office; Federal Reserve Board.

Notes: The yuan and yen are currencies of the People's Republic of China and Japan, respectively.

The last data point is July 2003.

The Corporate Sector. Businesses increased their investment spending only weakly in the first half of 2003 (see Figure 2-9). Real spending on producers' durable equipment rose--but at an average annual rate of only a little over 1 percent. Spending on information technology accounted for much of that modest advance, growing robustly in both the first and second quarters. Firms' spending on structures was largely flat but did include a lively pickup in drilling activity in the second quarter in response to higher energy prices. By mid-2003, total real business fixed investment was only slightly above its average level for 2002 as a whole.
 

Figure 2-9.

Business Fixed Investment

(Billions of 1996 dollars)

Sources: Congressional Budget Office; Department of Commerce, Bureau of Economic Analysis.

Note: The last data point is the second quarter of 2003.

Businesses drew down their inventories during the first half of 2003 after some modest rebuilding of stocks in 2002 (see Figure 2-10). They responded to the first quarter's weak demand largely by filling increased orders out of their inventories rather than stepping up production in the second quarter. As a result, the ratio of inventories to sales for the first half of the year turned down slightly.
 

Figure 2-10.

Change in Businesses' Inventories

(Billions of 1996 dollars)

Sources: Congressional Budget Office; Department of Commerce, Bureau of Economic Analysis.

Note: The last data point is the second quarter of 2003.

An anticipated rise in the growth of demand and the expiration of JGTRRA's partial-expensing provisions will encourage businesses to replace existing capital and expand productive capacity in 2004, CBO forecasts. JGTRRA's partial expensing is likely to increase business fixed investment by 0.2 percent of GDP in 2003 and 0.5 percent of GDP in 2004. (The effect is greater in 2004 in part because firms are likely to shift investment from 2005 into the second half of 2004 to take advantage of the provisions before they expire.) Firms' purchases of computer equipment are likely to account for much of that near-term investment, although purchases of other kinds of durable equipment as well as spending on structures will also expand in 2004. In addition, if demand strengthens in the second half of 2003 and in 2004, businesses are likely to restock their shelves after the inventory drawdown during the first half of 2003.

One factor that supports CBO's expectations of increased near-term spending by businesses is the improved financial conditions in the corporate sector so far this year. Corporate profits have bounced back from the low levels they reached during the recession, and CBO foresees rapid growth through the end of 2004. Also encouraging businesses to expand their spending is the recent rebound in equity markets and a reduction in the yields on corporate bonds, which have lowered the cost of capital and allowed firms to improve their balance sheets.

Demand for Products. A key requirement for a persistent upturn in business fixed investment is an expectation that the growth of demand will quicken. Such a speedup in the pace of growth appears to have begun. Final sales of domestic product (GDP minus inventory investment) grew at an average annual rate of 2.7 percent during the first half of this year, after expanding by 2.3 percent during the second half of 2002.

Profits. The rise in corporate profits in the first quarter of 2003 is likely to continue through 2004; CBO estimates that the GDP share of economic profits (profits from current production, adjusted for changes in the value of inventories and for depreciation of capital) will rise from 7.5 percent in the first quarter of this year to over 8 percent for 2004 as a whole. Rapid gains in labor productivity have helped push up corporate profits by enabling firms to produce more without adding new capital or hiring new workers. Declining prices for some goods and competition from lower-cost imports have worked against improved profitability, but the recent decline in the value of the dollar should further boost profits in the near term.

Some analysts are concerned that businesses' obligations to their workers, in the form of defined-benefit pension plans and health benefits, will severely undercut their profits in coming years.⁽⁴⁾ In March, the Pension Benefit Guaranty Corporation estimated that defined-benefit pension plans were underfunded by about $300 billion.⁽⁵⁾ Losses stemming from a decline in the value of a pension plan's assets must be recognized on accounting statements over a five-year period.⁽⁶⁾ If corporations were to make up those losses over five years, annual pension fund contributions would increase by at most $60 billion--which is a significant sum when compared with total profits in the corporate sector of $900 billion. That $300 billion estimate is uncertain, however, because the degree to which a pension plan is underfunded depends on a number of factors, including the value of its assets and interest rates. Although such obligations will lower profits, they will have no significant direct effect on either the national saving rate or the overall level of investment because the replacement pension contributions will be invested in financial instruments.

Corporations' rapidly rising health care costs are less likely to hurt their profits. Health benefit costs rose by 9.3 percent between March 2002 and March 2003, whereas wages and salaries grew by only 3.0 percent. But the increase in such costs probably has a smaller effect on profits than underfunded defined-benefit pensions do because the rise in health costs is partly offset by lower growth of wages. Moreover, the recent rapid growth of health benefits is likely to moderate.

The Cost of Capital. The cost of capital fell over the first half of 2003 as equity markets rebounded following the end of major combat operations in Iraq and investors began to view the corporate sector as less prone to the risk of defaults. After stagnating at almost six-year lows before and during the hostilities in Iraq, the Standard & Poor's 500 Composite Index had topped its late-2002 levels by the end of July. The resulting hike in share prices has lifted the market's valuation of the corporate sector and helped reduce the cost of financing new investment through stock offerings.

Similarly, recent improvements in the outlook for corporations have lowered the cost of firms' debt financing (see Figure 2-11). Yields on Baa (low-investment-grade) corporate bonds, for example, were about 40 basis points lower in July than they were in January. The yields on more-risky (below-investment-grade) bonds fell even further in the first half of the year. Even after the upturn that began in mid-June in the rates on 10-year Treasury notes, the resulting spreads (differences) between the yields on the notes and on highly rated investment-grade corporate debt--a measure of the perceived riskiness of the bonds--declined markedly from their first-quarter values, reflecting the end of major hostilities in Iraq and the waning of the effects of corporate accounting scandals. Spreads on low-investment-grade and speculative-grade debt, which represented almost half of outstanding corporate debt in the first quarter of 2003, also fell from their war-related peaks.⁽⁷⁾
 

Figure 2-11.

Interest Rates on Corporate Debt

(Percent)

Sources: Congressional Budget Office; Federal Reserve Board; Standard & Poor's Global Fixed Income Research.

Notes: A BB+ rated bond is a below-investment-grade bond of the highest quality.

The last data point is July 2003.

Businesses have used those improved conditions in the bond markets to bolster their balance sheets. By issuing long-term debt and using the proceeds to pay down short-term debt (in the form of commercial paper and bank loans), businesses have increased the average maturity of their liabilities and reduced the risks to their liquidity from relying on shorter-term funding. That increase in the overall maturity of firms' debt structures and the decline in rates have in turn reduced businesses' repayment obligations and net interest payments in the near term.

The conditions that businesses face when borrowing from banks have eased slightly over the first half of 2003 as compared with the previous two years. In the case of commercial and industrial (C&I) loans to large and medium-sized borrowers, fewer banks in the first two quarters tightened their lending standards or boosted the spreads that they require relative to their cost of funds. In addition, the demand for loans among enterprises of all sizes--although still weak--steadied somewhat in the first half of 2003. At the same time, however, the level of C&I loans failed to turn up from the decline that began in early 2001.

The Inflation Outlook

CBO forecasts a mild uptick, relative to the first half of this year, in the core rate of inflation in the consumer price index for all urban consumers (CPI-U).⁽⁸⁾ Inflation is likely to be higher in the near term because of rising prices for imports and faster growth of unit labor costs. Import prices (excluding those for petroleum and computers) were almost flat during 2002, but they are likely to climb by more than 5 percent both this year and next. Similarly, unit labor costs fell by more than 1.5 percent over the four quarters of 2002 as a result of the unusual increase in productivity. But if productivity growth returns to a more usual pattern, unit labor costs this year and in 2004 are likely to grow by more than a percentage point. Another factor pointing to an increase in inflation is that year-over-year percentage changes in other indexes (specifically, the spot-price indexes for nonfood, nonenergy commodities such as metals and raw industrial products as well as the core intermediate materials producer price index) have generally trended upward since late 2001 and early 2002.

Core consumer inflation slowed to an annual rate of less than 1 percent in the second quarter of 2003 after increasing by 2.1 percent (measured on a fourth-quarter-over-fourth-quarter basis) in 2002 (see Figure 2-12). The category of homeowners' equivalent rent, which accounts for 28 percent of the core CPI-U, explains much of that deceleration.⁽⁹⁾ Inflation in homeowners' rents has declined steadily from over 4 percent in the first quarter of 2002 to 1 percent in the second quarter of 2003. Medical care inflation has also moderated, falling by almost 2 percentage points since the fourth quarter of 2002 after accelerating in every year since 1997.
 

Figure 2-12.

Inflation in the Core Consumer Price Index

(Percentage change)

Sources: Congressional Budget Office; Department of Labor, Bureau of Labor Statistics.

Notes: The core consumer price index is the consumer price index for all urban consumers excluding food and energy.

The last data point is the second quarter of 2003.

The reduction in rental price inflation may be largely temporary. The Bureau of Labor Statistics (BLS) estimates the rental index for owner-occupied housing from a sample of rents for similar housing in the rental market. It subtracts utility costs from the rents to isolate the shelter component and then uses that component to impute the cost of the flow of services provided by owner-occupied housing. Utility costs rose in the early part of this year; but contractual rents tend to lag behind such costs, and the imputed growth of homeowners' equivalent rent was thus small. That growth is likely to be higher if energy prices stabilize or fall.

Notwithstanding the steady decline in the core measure in the first half of the year, quarter-to-quarter fluctuations in overall CPI-U inflation have broadly mirrored events in the energy markets. During the first quarter of 2003, uncertainty about the duration of the war in Iraq as well as political crises in Venezuela pushed crude oil prices well above $30 per barrel. Natural gas prices also jumped sharply. The result was an increase not only in energy costs in the CPI-U but also in overall CPI-U inflation, which rose at an annual rate of 3.9 percent in the first quarter after inching up by only 1.6 percent in 2002. Falling energy prices in the second quarter brought CPI-U growth to an annual rate of less than 1 percent for that period. Since early May, however, crude oil prices have picked up, although generally they remain below their prewar highs.

Uncertainty of the Forecast

CBO's two-year economic forecast represents its best estimate, under its baseline assumptions about fiscal policy, of the economy's most likely path in the near term. However, both CBO's experience and that of other forecasters suggest that the range of possible errors in forecasts is large.⁽¹⁰⁾ Economic developments may play out quite differently than CBO's forecast indicates. For example, the imbalances that developed in the late 1990s--low personal savings, past overinvestment by some firms, and increased dependence on foreign financing--may remain for some time. Growth may continue for a few quarters, but then households' desire to rebuild savings and foreigners' unwillingness to hold dollar-denominated assets may severely dampen economic activity. As a result, businesses' investment spending and real GDP growth might be weaker in 2004--and perhaps in later years--than CBO anticipates.

But a brighter scenario is also possible. Strong economic fundamentals, such as the recent robust growth in productivity, may set the stage for rapid growth with low inflation. Given a pronounced rise in investment and labor force participation, potential GDP between 2005 and 2013 may be higher as well. Such growth could eliminate any remaining imbalances without undercutting economic activity. The outlook for the federal deficit would also be better under such a scenario because the growth of tax revenues would then exceed CBO's baseline projections.

A Comparison of Two-Year Forecasts

CBO's assessment of the economy's near-term outlook does not differ markedly from the consensus view of other forecasters. CBO's current two-year forecast is similar to the August Blue Chip consensus, an average of roughly 50 private-sector forecasts (see Table 2-3). CBO's estimates of nominal and real GDP growth are slightly lower in 2003 and slightly higher in 2004 than the Blue Chip's. In addition, CBO expects marginally higher unemployment and slightly higher CPI-U inflation in 2003 and 2004 than the consensus does. The two forecasts for short- and long-term interest rates are virtually identical.

Table 2-3.
Comparison of CBO, Blue Chip, and Administration Forecasts for Calendar Years 2003 and 2004
		Actual 2002		Forecast
				2003	2004
Nominal GDP (Percentage change)
	Blue Chip consensus	3.6		3.9	5.1
	CBO	3.6		3.7	5.3
	Administration	3.6		4.0	5.0
Real GDP (Percentage change)
	Blue Chip consensus	2.4		2.3	3.7
	CBO	2.4		2.2	3.8
	Administration	2.4		2.3	3.7
GDP Price Index (Percentage change)
	Blue Chip consensus	1.1		1.6	1.5
	CBO	1.1		1.5	1.4
	Administration	1.1		1.6	1.2
Consumer Price Indexa (Percentage change)
	Blue Chip consensus	1.6		2.2	1.8
	CBO	1.6		2.3	1.9
	Administration	1.6		2.3	1.7
Unemployment Rate (Percent)
	Blue Chip consensus	5.8		6.1	5.9
	CBO	5.8		6.2	6.2
	Administration	5.8		5.9	5.6
Three-Month Treasury Bill Rate (Percent)
	Blue Chip consensus	1.6		1.0	1.5
	CBO	1.6		1.0	1.7
	Administration	1.6		1.3	2.0
Ten-Year Treasury Note Rate (Percent)
	Blue Chip consensus	4.6		4.0	4.6
	CBO	4.6		4.0	4.6
	Administration	4.6		3.7	4.1
Sources: Congressional Budget Office; Department of Labor, Bureau of Labor Statistics; Federal Reserve Board; Aspen Publishers, Inc., Blue Chip Economic Indicators (August 10, 2003); Office of Management and Budget, Mid-Session Review: Fiscal Year 2004 (July 15, 2003).
a. The consumer price index for all urban consumers.

In general, CBO's current two-year outlook also differs little from that of the Administration. In CBO's forecast for 2003, nominal and real GDP growth are slightly lower than in the Administration's; for 2004, CBO's estimates are slightly higher. Relative to the Administration's estimates, CBO's anticipate higher unemployment in both 2003 and 2004 and slightly higher CPI-U inflation (in 2004) and long-term interest rates (in 2003 and 2004).

In its midyear report to the Congress, the Federal Reserve presented its economic outlook as ranges--known as the central tendency--which include the majority of forecasts of the members of its Board of Governors and the presidents of the Federal Reserve Banks.⁽¹¹⁾ CBO's forecast falls within the Federal Reserve's central tendency for growth of real and nominal GDP for both 2003 and 2004. It is above the central tendency for unemployment for 2003 but within it for 2004. For inflation, CBO's forecast is above the Federal Reserve's central tendency for both years. The divergence can be explained in part by the use of different price measures.⁽¹²⁾
 

The Outlook Beyond 2004

To develop its medium-term (2005 through 2013) projections, CBO extended historical patterns in the factors that underlie the growth of potential GDP, such as the growth of the labor force, productivity, and the rate of national saving. In doing so, CBO does not attempt to forecast business-cycle fluctuations beyond the next two years. However, it does take the possibility of such fluctuations into account in developing the medium-term trends by basing those trends on historical averages and growth rates, including periods of boom and recession. CBO's medium-term projections also incorporate the effects on potential output of recent fiscal policy (see Box 2-3).
 

Box 2-3. How Recent Changes in Fiscal Policy Affect Potential Output The growth of potential output will be influenced by recent legislation, such as the Economic Growth and Tax Relief Reconciliation Act of 2001 (EGTRRA), the Job Creation and Worker Assistance Act of 2002 (JCWAA), the Jobs and Growth Tax Relief Reconciliation Act of 2003 (JGTRRA), and various spending bills. (Potential output is the highest level of production that can persist for a substantial period without raising inflation.) The changes in spending primarily affect potential output through their impact on national saving and investment. Higher spending reduces national saving and tends to crowd out capital investment, which lowers productivity and slows economic growth. The Congressional Budget Office (CBO) estimates the degree of crowding out by using simple rules of thumb based on past relationships between budget variables, saving, and investment. But because the effects of the three recent tax laws are more complex, CBO estimated them separately, as this box describes. Overall, fiscal policy since 2001 has worked to increase aggregate demand (an effect reflected in CBO's short-term forecast). But it may also have a small negative impact on future potential output. Various provisions in the tax laws affect the economy in complex ways: some increase gross domestic product (GDP), and others decrease it. For example, EGTRRA and JGTRRA reduce marginal tax rates on the income from labor and capital. (Marginal tax rates are those that apply to the last dollar earned.) Other things being equal, those lower rates encourage people to work and save more. Other provisions stimulate businesses to invest in capital goods in the near term. However, the laws also boost households' after-tax income, which encourages people to work less and consume more. The legislation's overall impact is determined by the relative magnitudes of those offsetting forces.1 How CBO Estimated the Effects of the Tax Laws Consensus among economists is lacking about the appropriate model and underlying assumptions to use in estimating the effects of taxes on the economy. To address that uncertainty, CBO examined those effects using two different models. The first was a microsimulation model, based on a large sample of taxpayers' returns, that reflects the major provisions of the individual income tax code and incorporates the assumption that workers will respond to changes in their after-tax wages and income as they have in the past. That model does not pick up effects that might arise from what people do when they expect changes in future tax rates or other fiscal policies. By contrast, the second approach, sometimes termed forward-looking, focuses on the possibility that people plan ahead and choose how much to work and spend on the basis of current and future after-tax wage rates, interest rates, and income, among other things.2 CBO also used two different approaches to estimate the effects of tax policy on capital investment. One method included the assumption, based on historical evidence, that a budget deficit of $1 crowds out 36 cents of private investment, on average.3 The other method estimated tax policy effects by simulating how forward-looking firms and households respond to changes in marginal tax rates and to other policies. Effects of the Tax Legislation The revenue measures enacted since 2001 will boost labor supply by between 0.4 percent and 0.6 percent from 2004 to 2008 and by up to 0.2 percent from 2009 to 2013, according to estimates from the two models. Those increases stem from the legislation's reduction in marginal tax rates on labor income, which drop by an average of 1.9 percentage points from 2004 to 2008 and 0.7 percentage points from 2009 to 2013. The rise in labor supply and the fall in marginal rates are smaller, on average, during the 2009-2013 period because the marginal rate cuts expire at the end of 2010, eliminating their positive effect on labor supply. But the tax legislation will probably have a net negative effect on saving, investment, and capital accumulation over the next 10 years. That outcome derives from the laws' provisions that boost private consumption because, in the long run, they reduce the pool of funds available for capital investment in business equipment, structures, and housing. According to the models, the legislation will reduce national saving by between 3 percent and 6 percent from 2004 to 2008 and between 3 percent and 5 percent from 2009 to 2013.4 The tax laws' net effect on potential output is uncertain during the first five years of the 2004-2013 projection period but will probably be negative in the second five years. However, that impact is small, especially compared with the overall uncertainty of the forecast. According to the models, the legislation could boost the level of potential GDP by as much as 0.3 percent or reduce it by as much as 0.1 percent over the years 2004 to 2008. From 2009 to 2013, it could reduce the level of potential GDP by about 0.4 percent. Potential GDP is reduced in the later years because the cuts in marginal tax rates are scheduled to expire at the end of 2010 and the negative effect of higher consumption on investment is compounded over time. Effects of Alternative Financing Assumptions For its simulations, the forward-looking model requires that the ratio of debt to GDP be stable in the long run. That means that policies that reduce revenues must eventually be financed by increasing revenues or cutting spending. In general, the estimated impact of tax policies depends on what offsetting policies households are assumed to expect. Following the same strategy that it used in its recent analysis of the President's 2004 budgetary proposals, CBO estimated fiscal policy effects under different financing assumptions. Because most of the tax legislation enacted since 2001 is scheduled to expire by 2011, the particular assumption about expected financing that is used in the model makes relatively little difference to the estimates. However, alternative assumptions about how policies enacted in 2001 to 2003 might affect future decisions within the 10-year period could substantially influence the results. Effects of Different Expectations About Tax Law Expirations All of the major tax legislation enacted since 2001 expires by 2011 (according to the so-called sunset provisions), and some provisions expire earlier. Thus, people's expectations about those expirations during CBO's 10-year projection period (2004 through 2013) affect its economic projections. For estimates using its forward-looking approach, CBO assumed not only that those sunsets would take place according to current law but also that people would expect them to occur in that way. Of course, people might view the path of future fiscal policy differently. For example, people might expect the tax legislation to be permanently extended rather than allowed to expire. Alternatively, people might expect sunsets for some, but not all, of the provisions. Finally, as noted earlier, the overall impact depends as well on people's expectations about any modifications to fiscal policy that might be needed outside the 10-year projection window to ultimately stabilize debt relative to GDP. If people expected policymakers to make the tax legislation permanent and to finance it by cutting government spending on goods and services in 2014, output would be 0.2 percent less from 2004 to 2008, by CBO's estimates, than it would have been had people expected the laws to expire as scheduled. GDP would be lower for two reasons. First, because people expected low marginal tax rates on labor to persist, they would have had less reason to work more in anticipation of the sunset. Second, the reduction in government consumption in 2014 would increase the resources available for private consumption. That would tend to lead people to work and save less in earlier years. By contrast, if people viewed the tax policies as permanent and assumed that they would be financed by increased taxes that fell equally on everyone (both workers and nonworkers), CBO estimates that output could be 0.2 percent higher than it would have been had people expected the policies to expire as scheduled. Output would be higher under that assumption largely because older workers and retirees would reduce their consumption and save more in anticipation of the taxes to be imposed in 2014. An assumption that financing would entail adjustments to either income taxes or transfer payments would produce similar results. 1.  CBO used a similar approach in An Analysis of the President's Budgetary Proposals for Fiscal Year 2004, published in March 2003. For details on the methodology used in the analysis, see Congressional Budget Office, How CBO Analyzed the Macroeconomic Effects of the President's Budget (July 2003). 2.  For the labor-supply assumptions used in the first model, see Congressional Budget Office, Labor Supply and Taxes (January 1996); and Chinhui Juhn, Kevin M. Murphy, and Robert Topel, "Current Unemployment, Historically Contemplated," Brookings Papers on Economic Activity, no. 1 (2002), pp. 117-125. For a description of the second model, see Shinichi Nishiyama and Kent Smetters, Consumption Taxes and Economic Efficiency in a Stochastic OLG Economy, CBO Technical Paper 2002-6 (December 2002). 3.  That assumption is based on two pieces of evidence about the historical relationships between budget deficits, national saving, and investment. First, increasing the deficit by a dollar reduces national saving, on average, by 60 cents (because private saving increases); and second, a decline of 60 cents in national saving reduces domestic investment, on average, by only 36 cents (because capital flows in from abroad). 4.  JGTRRA includes two provisions--reductions in the rates of taxation on corporate dividends and capital gains--that could in principle increase the fraction of investment allocated to the corporate sector relative to tax-advantaged sectors such as housing, thereby increasing economic efficiency and output. However, those rate cuts are scheduled to expire at the end of 2008. Therefore, CBO estimates that those provisions will have little effect on the allocation of investment, which can produce returns over many years.

Between 2004 and 2013, real GDP will grow at an average annual rate of 3 percent, CBO projects, which is slightly faster than the average annual growth rate of potential real GDP--2.8 percent. During the 2005-2013 period, real GDP will overtake potential GDP, closing the slight gap that remains between them at the end of 2004. Inflation, as measured by the CPI-U, will average 2.5 percent over the medium term, and the rate of unemployment will average 5.3 percent. CBO's projection for the rate on three-month Treasury bills averages 4.6 percent during the period, and the rate on 10-year Treasury notes averages 5.8 percent. Those projections are virtually identical to the estimates CBO published last January.

CBO's Projection of Potential Output

CBO projects that over the 2003-2013 period, growth of potential real output will average 2.9 percent per year, roughly the same rate that CBO projected last winter (see Table 2-4). That estimate is derived from several offsetting changes in the projections of variables that CBO uses to calculate potential output, including the labor force, the capital stock, and total factor productivity (TFP).

In CBO's current projection, potential labor force growth is faster than it was in January's estimate, largely because of an upward revision to the historical data for the civilian labor force. In January 2003, BLS released significant revisions to data derived from the Current Population Survey (for example, data on the civilian population, labor force, and employment). Those revisions resulted from the Census Bureau's higher estimates of the size of the population, based largely on information from the 2000 census. The revisions did not appreciably affect the unemployment rate, but they raised the level of the labor force by about 1.7 million people in 2000 and by about 3 million people cumulatively in early 2003. Consequently, CBO has increased its estimate of the average growth of the labor force during the 1990s by about one-tenth of a percentage point since last winter's forecast.

CBO carried the faster trend growth of the 1990s forward into its projection, estimating that the potential labor force would grow at a rate of 1.0 percent, up a tenth of a percentage point from last winter's estimate. The projection of the potential labor force also includes the effects of JGTRRA, which, among other changes, accelerated the marginal rate cuts previously scheduled for 2006, as enacted in EGTRRA. However, since JGTRRA preserved the sunset provisions of the earlier law, its effects on labor supply are expected to disappear before the end of the projection period and thus make no contribution to the final level of the potential labor force or its 10-year growth rate.

Potential hours worked will grow at a slightly faster pace during the 2003-2013 period than the rate noted in last winter's projection, or by about 1.1 percent on average, CBO projects. Typically, the upward revision in the projection for the potential labor force would invoke a corresponding revision to the projection for hours worked, which is the key measure of labor input used in production. But BLS also revised the historical data for hours worked, shifting the series sharply downward during 2001 and 2002.⁽¹³⁾ Those revisions to the data on hours lowered the growth rate of potential hours worked during the 1990-2002 period and partially offset the effect of the upward revision to the potential labor force.

CBO projects that capital accumulation will proceed at a 3.8 percent pace, on average, during the 2003-2013 period, or about 0.4 percentage points slower than in the winter forecast. Slower capital accumulation is one consequence of the worsening outlook for the federal budget, which will reduce the national saving rate and decrease the pool of funds available for business investment. The rate of national saving will average 15 percent of GDP during the 10-year projection period, CBO estimates, down from 17 percent last winter.

In CBO's projections, potential total factor productivity grows at an average annual rate of 1.3 percent during the projection period, about a tenth of a percentage point faster than in last winter's projection. Since TFP is calculated as a residual--it is defined as the growth of output that remains after subtracting the contributions made by the growth of hours worked and of capital accumulation--the downward revision to the recent historical data for hours worked caused an upward revision to the historical estimate of TFP. That change, in turn, raised the growth rate of potential TFP during the 1990-2002 period and in the projection.

Unemployment, Inflation, and Interest Rates

The rise in inflation expected in the two-year forecast period will taper off during the medium term, CBO projects; inflation will average 2.5 percent in the CPI-U and 2.1 percent in the GDP price index. Those rates are identical to the assumptions used in CBO's January forecast. In general, CBO assumes that inflation is determined by monetary policy in the medium term and that the Federal Reserve will seek to maintain the underlying rate of CPI-U inflation near 2.5 percent, on average.

The difference between the projected rates of growth of the CPI-U and the GDP price index affects CBO's projections of the federal budget. Many spending programs and all income tax brackets are indexed to the CPI-U, whereas taxable income is more closely related to growth in the GDP price index. Hence, for a given rate of inflation in the CPI-U, a higher rate of GDP inflation results in a lower projection of the federal deficit. CBO expects that the wedge between the two measures of inflation will average 0.4 percentage points during the 2005-2013 period, which equals the average wedge between the two rates over the 1990-2002 period.

CBO estimates that over the latter part of the medium term, the unemployment rate will average about 5.2 percent, down from its average projected levels of 6.2 percent in 2004 and 5.7 percent in 2005. In projecting the unemployment rate, CBO uses the nonaccelerating inflation rate of unemployment (NAIRU) as a benchmark. When the gap between GDP and potential GDP has closed completely, the difference between the unemployment rate and the NAIRU is assumed to close as well.

CBO's medium-term projections for interest rates are almost unchanged since January. For most of the 2005-2013 period, the rate on three-month Treasury bills is expected to average 4.9 percent, while the rate on 10-year Treasury notes will average 5.8 percent. Those rates combine the projection for CPI-U inflation and a projection for real interest rates. CBO estimates that the real rate on three-month Treasury bills will average 2.4 percent during the latter years of the projection period and the real rate on 10-year Treasury notes will average 3.3 percent. CBO's projection for real interest rates is based on analyses of historical averages of those rates and historical trends in the real return to capital.

Taxable Income

CBO's baseline revenue projections are closely connected to its projections of national income. Because different categories of income are taxed at different rates, and some are not taxed at all, the projected distribution of income among its various components is a central factor in CBO's budget projections. The categories of corporate profits and wage and salary disbursements are particularly significant because they are taxed at the highest rates.

Two of the various NIPA measures of corporate profits are important for the forecast. Book profits, or before-tax profits, is the measure most closely related to the profits that companies report to the Internal Revenue Service. By contrast, the economic profits measure is designed to reflect the valuation of inventories and the rates of depreciation that economists believe more truly represent the current value of inventories and the economic usefulness of the capital stock. The difference between the two measures is affected by changes in the tax code. Corporations are allowed by law to value inventories and depreciate assets at certain rates. The book measure of profits is designed to reflect those statutory requirements, whereas the economic measure is not.

The outlook for book profits--the closest approximation in the NIPAs to the profits on which corporations pay tax--is likely to be dominated by statutory provisions that affect how companies can depreciate their assets for tax purposes. The partial-expensing provisions of JCWAA and JGTRRA that expire at the end of 2004 allow firms to depreciate some of their capital stock much more rapidly than the rate at which the economic usefulness of that capital deteriorates. Those provisions are expected to lower taxable profits by nearly $150 billion in 2003 and $200 billion in 2004, because companies can take the extra depreciation in those years. Conversely, in 2005 and after, taxable profits will be increased--by about $125 billion in 2005 and declining amounts in subsequent years--because the extra depreciation taken in 2003 and 2004 will no longer be available to firms.

The underlying trend of profits is hard to discern because of the large changes in depreciation, but CBO's projection implies a relatively sanguine outlook for economic profits, a measure that looks past those tax-induced variations. Economic profits were 7.5 percent of GDP in 2002, a level that reflects the effects of the recession. CBO's projection anticipates that in the latter years of the projection period, the average GDP share of economic profits will exceed 8.3 percent, the average share during the 1990s.

Wages and salaries--the other NIPA income category that is particularly important for revenue forecasting--are currently very close to their 20-year average share of GDP. CBO's estimates keep wages and salaries close to that share (47.4 percent) throughout the short-term forecast and the medium-term projection. The projection incorporates an acceleration in the growth of fringe benefits--specifically, employers' contributions to health insurance and pension plans--which implies that the GDP share of total labor compensation, currently somewhat below its 20-year average, will rise toward that average share over the 10-year period.