Congressional Budget Office
Director's Blog

Over the past several years, the nation has experienced its most severe financial crisis since the Great Depression of the 1930s. To stabilize financial markets and institutions, the Federal Reserve System used its traditional policy tools to reduce short-term interest rates and increase the availability of funds to banks, and created a variety of nontraditional credit programs to help restore liquidity and confidence to the financial sector. In doing so, it more than doubled the size of its asset portfolio to over $2 trillion and assumed more risk of losses than it normally takes on.

In a study prepared at the request of the Ranking Member of the Senate Budget Committee, CBO describes the various actions by the Federal Reserve and how those actions are likely to affect the federal budget in coming years. The report also presents estimates of the risk-adjusted (or fair-value) subsidies that the Federal Reserve provided to financial institutions through its emergency programs. Unlike the cash treatment of the Federal Reserve in the budget, fair-value subsidies include the cost of the risk that the central bank has assumed. Thus, those subsidies are a more comprehensive measure of the cost of the central bank’s actions.

The Federal Reserve’s activities during the crisis have had a striking impact on the amount and types of assets that it holds. In July 2007, before the financial crisis began, the Federal Reserve held about $900 billion in assets; U.S. Treasury securities accounted for about $790 billion of that amount. The central bank had acquired those securities during its normal operations in conducting monetary policy—the process of influencing the level of short-term interest rates and consequently the pace of U.S. economic activity. By the end of 2008, the value of the Federal Reserve’s assets had grown to about $2.3 trillion; of that amount, loans and other support extended to financial institutions made up $1.7 trillion. At the end of 2009, the amount of direct loans and other support to financial institutions, though still quite high by historical standards, had fallen markedly to about $280 billion, but holdings of mortgage-related securities had risen to just over $1 trillion. There was also a marked shift in the composition of the central bank’s liabilities. Before the crisis, the major liability on the Federal Reserve’s balance sheet was the amount of currency in circulation—about $814 billion as of July 2007. At the end of 2009, the amount of reserves that banks held with the Federal Reserve was the central bank’s largest liability, at more than $1 trillion.

The amount and composition of the Fed’s assets and liabilities are major determinants of its impact on the federal budget. That impact is measured by the central bank’s cash remittances to the Treasury, which are recorded as revenues in the budget. (The amount that is remitted is based on the Federal Reserve System’s income from all of its various activities minus the costs of generating that income, dividend payments to banks that are members of the Federal Reserve System, and changes in the amount of the surplus that it holds on its books.) For fiscal years 2000 through 2008, annual remittances by the Federal Reserve ranged between $19 billion and $34 billion.

CBO projects that the Federal Reserve’s actions to stabilize the financial system will boost its remittances to the Treasury during the next several years. That increase reflects the Federal Reserve’s larger portfolio of riskier assets, most of which are likely to earn a great deal more than the amount the system must pay in interest on reserves and its other liabilities. CBO projects that remittances will grow from about $34 billion in fiscal year 2009 to more than $70 billion in fiscal years 2010 and 2011.

However, that estimated effect on the budget fails to account for the cost of the risks to taxpayers from those actions. When the Federal Reserve invests in a risky security, it increases its expected net earnings because the return it anticipates on that security exceeds the interest rate it pays on the debt used to fund the purchase. Nevertheless, when the riskiness of such securities is fully accounted for, the investment may be projected to produce no net gain or even a loss. If the Federal Reserve purchases the security at a fair-market price, equivalent to what private investors would have paid, then the purchase creates no economic gain or loss for taxpayers; the price compensates the central bank for the risk it has assumed. By contrast, if the Federal Reserve purchases a risky security for more than the amount that private investors would have paid, it gives a subsidy to the seller of the security, creating an economic loss, or cost, for taxpayers.

The economic cost of the Federal Reserve System’s actions to stabilize the financial markets—which incorporates the risks to taxpayers—can be estimated using “fair-value” subsidies. In CBO’s estimation, the “fair-value” subsidies conferred by the Federal Reserve System’s actions to stabilize the financial markets totaled about $21 billion at the time those actions were taken. The gains or losses that will ultimately be realized from the Federal Reserve’s activities will almost certainly deviate from CBO’s estimates of the fair-value subsidies those actions provided. Fair-value subsidies are forward-looking estimates that are based on averages over many possible future outcomes, whereas realized gains or losses reflect a particular outcome.

It bears emphasizing that CBO’s fair-value estimates address the costs but not the benefits of the Federal Reserve’s actions. In CBO’s judgment, if the Federal Reserve had not strategically provided credit and enhanced liquidity, the financial crisis probably would have been deeper and more protracted and the damage to the rest of the economy more severe. Measuring the benefits of the Federal Reserve’s interventions in avoiding those worse outcomes is much more difficult than estimating the subsidy costs of the interventions, and CBO has not attempted to do so. It is likely, though, that the benefits of the Federal Reserve’s actions to stabilize the financial system exceeded the relatively small costs of the fair-value subsidies.

The report was prepared by Kim Kowalewski and Wendy Kiska of CBO’s Macroeconomic Analysis Division, and Deborah Lucas, Associate Director for Financial Analysis.

The federal government helps students finance higher education through two major loan programs—one that guarantees loans made by private lenders, and one that makes loans directly to borrowers. Between 2000 and 2009, the volume of outstanding federal student loans more than quadrupled, from about $149 billion to about $630 billion.

Although both programs offer similar types of loans on similar terms to borrowers, they differ significantly in how they are funded and administered. In the guaranteed loan program, loans are made and administered by financial institutions—such as Sallie Mae, commercial banks, and nonprofit agencies. The government bears almost all of the risk of borrowers defaulting and makes certain payments to the lenders. Those lenders usually raise the money to make loans in the private capital markets. By contrast, in the direct loan program, the Department of Education and its contractors bear all of the risks and manage most administrative functions; those loans are funded through the Treasury. Those differences cause the guaranteed loan program to be more costly to the federal government per dollar of lending than the direct loan program.

Today CBO released a study—prepared at the request of the Ranking Member of the Senate Budget Committee—examining the costs of the federal student loan programs, calculated two different ways. One estimate follows the methodology delineated by the Federal Credit Reform Act of 1990, which governs the treatment of credit programs (including student loans) in the federal budget. The other estimate was done on a so-called “fair-value basis,” which provides a more comprehensive measure of cost by including administrative costs and the cost of market risk (the risk that losses from defaults will be higher during periods of market stress, when resources are scarce and hence most valuable). A fair-value estimate represents what a private entity would need to be paid to assume the costs and risks to the government of providing the loans or guarantees.

CBO’s calculations indicate that:

• For both programs, subsidies calculated on a fair-value basis show substantially higher costs than those based on the standard budgetary treatment.
• Regardless of which methodology is used, for a given set of borrowers and loan types, the guaranteed loans cost the government more than those it makes directly.

For example, on a fair-value basis, a mix of representative guaranteed loans would typically cost the government about 20 percent of the principal amount of the loans. Those same loans, made directly by the government to the same people, would typically cost about 12 percent of the principal amount.

The study also looks at several proposals for modifying the student loan programs. For example, the President’s 2011 budget calls for ending the guarantee program’s authority to make new loans on July 1, 2010, and switching entirely to the direct lending program to realize the savings from that program’s lower costs. (CBO recently provided a cost estimate for that particular proposal in a letter to Senator Gregg on March 15 and in our Analysis of the President’s Budget released yesterday.) The House of Representatives and the Senate have both passed legislation (H.R. 4872, the Health Care and Education Affordability Reconciliation Act of 2010) to carry out a similar change.

Other policy options, some of which would reduce costs to the government and others that would increase those costs, include:

• Reducing the cost of the guarantee program by cutting government payments to lenders, reducing the guarantee percentage on loans, or auctioning off the right to lend under the program;
• Indexing interest rates to market rates, allowing federal costs to be more predictable;
• Improving the affordability of student loans by lowering interest rates; and
• Lessening the hardships that borrowers face in repaying loans by making repayment of loans contingent on income, for example.

This study was prepared by Deborah Lucas, CBO’s Associate Director for Financial Analysis, and Damien Moore of CBO’s Macroeconomic Analysis Division.

This afternoon CBO released a report presenting its analysis of the budgetary proposals contained in the President’s budget request for fiscal year 2011.  This report provides more detail than the preliminary analysis that CBO released on March 5, which was discussed in a blog entry that same day. Our latest report differs slightly from the earlier one because it incorporates the impact of some legislation that has recently been enacted. It reflects the revenue and spending estimates that the President included in his budget for major health care legislation, but it does not incorporate the specific effects of the health care bill that was signed into law yesterday. It also includes an analysis of the potential effects on the economy of the President’s budgetary proposals and the impact of those economic effects on the federal budget.

As a reminder, this analysis presents CBO’s assessment of the budgetary outlook for the 2010-2020 period assuming enactment of the President’s policy proposals and reflecting CBO’s economic forecast and technical estimating procedures. The analysis compares that outlook with CBO’s baseline projections, which—unlike the President’s budget—assume that current laws and policies that affect federal spending and revenues remain unchanged

CBO’s analysis indicates that if the President’s proposals were enacted:

• The federal government would record deficits of $1.5 trillion in 2010 and $1.3 trillion in 2011. Those deficits would amount to 10.3 percent and 8.9 percent of gross domestic product (GDP), respectively. (The deficit in 2009 totaled 9.9 percent of GDP.) Compared with CBO’s current-law baseline projections, deficits under the President’s proposals would be about 2 percentage points of GDP higher in fiscal years 2011 and 2012, 1.3 percentage points greater in 2013, and above baseline levels by growing amounts thereafter. By 2020, the deficit would reach 5.6 percent of GDP, compared with 3.0 percent under CBO’s baseline projections.
• Under the President’s budget, debt held by the public would grow from $7.5 trillion (53 percent of GDP) at the end of 2009 to $20.3 trillion (90 percent of GDP) at the end of 2020, about $5 trillion more than under the assumptions underlying the baseline. Net interest would more than quadruple from 1.4 percent of GDP in 2010 to 4.1 percent in 2020 in nominal dollars (without an adjustment for inflation).
• Revenues under the President’s proposals would be $1.4 trillion (or 4 percent) below CBO’s baseline projections from 2011 to 2020, largely because of the President’s proposals to index the thresholds for the alternative minimum tax for inflation starting at their 2009 levels and to extend many of the tax reductions enacted in 2001 and 2003 that are scheduled to expire at the end of 2010. Other proposals in the President’s budget—including those associated with significant changes in the nation’s health insurance system—would, on net, increase revenues.
• Mandatory outlays under the President’s proposals would be above CBO’s baseline projections by $1.9 trillion (or 8 percent) over the 2011–2020 period, about one-third of which would stem from net additional spending related to proposed changes to the health insurance system and health care programs. Discretionary spending under the President’s budget would be about $0.3 trillion (or 2 percent) lower than the cumulative amount between 2011 and 2020 in CBO’s baseline, reflecting reduced funding for the wars in Iraq and Afghanistan.

The President’s budgetary proposals would have effects on the economy, which would in turn influence the budget through changes in such factors as taxable income (which affects the amount of revenues collected), employment (which determines outlays for programs like unemployment compensation), and interest rates (which affect the government’s borrowing costs). Estimates of economic effects depend on many specific assumptions, and there are several approaches to estimating those effects, so CBO used a number of different models to project the impact on the economy of enacting the President’s proposals. There is, however, a high degree of uncertainty about the economic effects of government policies, so the ranges of possible budget effects are quite wide.

In sum, CBO’s analysis of the interactions between the budget and the economy indicates the following:

• For 2011 to 2015, CBO estimates that the President’s proposals would raise real (inflation-adjusted) output relative to that under the assumptions in CBO’s baseline by between 0.9 percent and 1.2 percent, on average. Those estimates incorporate both supply-side effects (influences on the economy’s potential output) and demand-side effects (temporary movements of actual output relative to potential output).
• For 2016 to 2020, CBO estimates that the President’s proposals would lower real output relative to CBO’s baseline assumptions by between 0.4 percent and 1.4 percent, on average. Those estimates incorporate only supply-side effects because the magnitude of demand-side effects depends on the state of the economy, which is difficult to predict over longer horizons. In addition, the Federal Reserve might offset the effect of policies that are foreseen well in advance in order to maintain economic stability.
• CBO estimates that the economic feedback from the President’s proposals would reduce their cumulative cost over the period from 2011 through 2015—estimated to be about $1.4 trillion excluding any aggregate economic effects—by between 2 percent and 14 percent. From 2016 to 2020, the effects of the proposals on the economy would increase their cumulative cost (estimated to be about $2.3 trillion, excluding any aggregate economic effects) by as much as 6 percent or reduce it by as much as 2 percent.

CBO has not modified its economic forecast since January, but the agency updated its baseline budget projections early in March to take into account some legislation that has been enacted since the completion of the previous baseline in January 2010 as well as new information obtained about various aspects of the budget since then. The resulting changes, relative to CBO’s January projections, are modest, adding $20 billion to the projected deficit in 2010 and reducing projected deficits over the 2011–2020 period by a total of $57 billion.

Today CBO released the third of its statutory reports on transactions undertaken as part of the Troubled Asset Relief Program (TARP). This report discusses CBO’s estimate of the costs of those transactions initiated as of February 17, 2010, as well as possible future transactions that could be undertaken with the available authority.

CBO currently estimates that the cost to the government of the TARP’s transactions—including investments, grants, and loans—completed, outstanding, and anticipated will amount to $109 billion. Much of that estimated cost is associated with the assistance provided to American International Group (AIG)—at a cost of about $36 billion—and the automotive industry—at a cost of about $34 billion. CBO estimates a very small net gain to the government from the capital purchase program, in which the Treasury purchased more than $200 billion in shares of preferred stock from hundreds of financial institutions.

The Office of Management and Budget (OMB) estimates that the total cost of the TARP’s transactions will amount to $127 billion. OMB’s estimate is $18 billion higher than CBO’s estimate principally because of differences in the estimated cost of assistance to AIG and in the amount expected to be disbursed by the Home Affordable Modification Program (an initiative that provides direct payments to mortgage servicers to help homeowners avoid foreclosure).

Both CBO and OMB value the TARP’s investments by discounting to the present the projected cash flows stemming from each investment, using a discount rate that captures both the time value of money and the premium that a private investor would require as compensation for the risk of the investment or commitment. The resulting “net present value” is the cost or gain projected for the investment and represents an estimate of its market value.

Currently, the Secretary of the Treasury has the authority to purchase and hold up to $699 billion in assets at one time. CBO estimates that $344 billion of that authorized amount is outstanding or will be disbursed before the program expires on October 3, 2010. (That figure includes an estimated $45 billion that is projected to be used for purposes not yet specified.) 

This report was prepared by Avi Lerner of CBO’s Budget Analysis Division.

I spoke yesterday at the annual economic policy conference of the NABE, the National Association for Business Economics. The theme of the conference was “The New Normal? Policy Choices After the Great Recession,” and naturally I discussed fiscal policy choices. My slides and remarks were based on CBO’s January report on the budget and economic outlook and preliminary analysis of the President’s budget released last Friday.

The first part of my presentation focused on the next few years. CBO forecasts that the economic recovery will be fairly slow, with the unemployment rate returning to near its long-run sustainable level of 5 percent only in 2014. One factor underlying that forecast is declining support for economic activity from fiscal policy. CBO’s baseline budget projection, which follows current law, shows the budget deficit dropping from about 9 percent of GDP in this fiscal year to about 4 percent of GDP two years from now. That decline of roughly 5 percentage points would be the sharpest two-year reduction in the budget deficit that we have seen since the end of World War II.

Most of that decline can be attributed not to improving economic conditions (although those play some role) but to the diminishing impact of last year’s stimulus legislation and the scheduled expiration of earlier tax reductions. The effects of last year’s stimulus package on government outlays and receipts peaks in fiscal year 2010, as can be seen in the following picture, and CBO estimates that the effects of the package on output and employment will begin to wane later this year.

Estimated Budgetary Effects of ARRA (Billions of Dollars)

In addition, economic growth will be dampened, under current law, by the scheduled expiration of the 2001 and 2003 tax cuts and by the increasing reach of the AMT.

Therefore, the key choice for near-term policy is whether to enact additional tax cuts or spending increases to provide more stimulus to the economy or to allow stimulus to be withdrawn quickly as under current law. That choice involves a trade-off between potential short-term benefits and long-term costs. CBO’s January analysis of various policy options found that further fiscal actions, if properly designed, would promote economic growth and increase employment in 2010 and 2011. However, despite the potential economic benefits in the short run, such actions would add to the already large projected deficits and, all else equal, make future incomes lower than they otherwise would be.

The second part of my presentation looked beyond the next few years. CBO projects that the budget deficit and debt are on a trajectory that poses significant economic risks and ultimately becomes unsustainable. U.S. government debt in relation to GDP is quickly entering territory that is unfamiliar to us and unfamiliar to most developed countries in recent years. Therefore, the key choices for medium-term and long-term policy are how quickly and in what way to restrain federal borrowing. In thinking about those choices, it is helpful to understand the evolution of different components of the budget over time.

I showed the following table comparing budget components in 1970, 2007 (before the recession), and 2020 (based on CBO’s current-law projection). The figures are expressed as shares of GDP because overall output and income are the ultimate source of resources that can go to government activities.

Components of the Federal Budget

In 2007, both revenues and outlays represented about the same shares of GDP that they had in 1970. (Revenues were down a little, and outlays were up a bit, making the deficit somewhat larger relative to GDP.) The most striking increase in outlays was for Social Security, Medicare, and Medicaid, which used 4.4 percent of GDP more resources in 2007 than they did in 1970. Defense spending moved in the opposite direction, declining by 4.2 percent of GDP, on balance, between those years. All of the other programs of the government represented about the same share of GDP in 2007 that they did in 1970.

One should not minimize the variation in some of those budget components (relative to GDP) during the past several decades. However, looking across the whole 40-year period, the basic story of U.S. fiscal policy is fairly simple: The country financed an increase in Social Security, Medicare, and Medicaid spending by reducing defense spending relative to the size of the economy. Essentially, the increased costs of those three big entitlement programs were absorbed by a decline in spending (as a share of the economy) that was not very noticeable to the public because it occurred in an area that most people do not directly observe.

Because defense spending is much smaller relative to those programs today, that approach to funding growth in those programs is not feasible in the future. Furthermore, the projected increases in spending for Social Security, Medicare, and Medicaid are greater than those we have experienced in previous years. So, we will have to finance most of the future growth in Social Security, Medicare, and Medicaid through a noticeable increase in the tax burden or a noticeable reduction in other domestic programs relative to the size of the economy—or we will have to take noticeable policy actions to reduce the growth of those programs. The alternative of continuing large increases in federal borrowing would pose a serious threat to the future of the U.S. economy.

Last week I made presentations on the budget and economic outlook to both the Prosperity Caucus and the National Economists Club.  The Prosperity Caucus was described in a recent article in The Hill newspaper as a group founded in 1986 with the purpose of “gather[ing] libertarian-minded economists, Hill staffers and academics … to hear someone talk about the burning policy issues of the day;” The National Economists Club was founded in 1968 and describes itself as “aim[ing] to encourage and sponsor discussion and an exchange of ideas on economic trends and issues that are relevant for public policy.” I was very pleased to be invited to talk with both groups, and to answer a variety of questions that they raised.

In my comments (see the slides), I highlighted these points drawn from CBO’s January Budget and Economic Outlook report:

• CBO expects only a gradual recovery in the labor market. Output growth will probably be slow in light of the continuing fragility of some financial markets and institutions, the restraint on household spending stemming from slow income growth and lost wealth, and declining support from monetary and fiscal policy. Moreover, even when demand for labor picks up, the movement of unemployed workers into new jobs will be difficult in many cases, and improvements in labor-market conditions will draw people who have stopped looking for work back into the labor force. All told, CBO projects that the unemployment rate will fall below 8 percent only in 2012 and will return to near its long-run sustainable level of 5 percent only in 2014.
• Under current law, the budget deficit will drop from about 9 percent of GDP this year to about 4 percent in 2012—the largest two-year decline in the deficit since the end of World War II. However, policymakers are considering changes from current law that would keep deficits from falling so quickly. For example, if policymakers extended all or part of the 2001 and 2003 tax cuts beyond their scheduled expiration at the end of this year, indexed the alternative minimum tax (AMT) for inflation, or boosted spending for transfer programs or government purchases of goods and services relative to the levels projected under current law, the budget deficit would not decline as rapidly.
• Beyond the next few years, CBO expects the nation’s output to return to its so-called potential level (the output that could be produced if all labor and capital were fully employed) and then to rise at a solid but unspectacular pace. In particular, we expect that output growth will be slower than its average pace of the past 60 years, primarily because of slower population growth and a downtrend in the labor force participation rate that has been apparent during the past decade.
• Between 2013 and 2020, CBO projects that the budget deficit will run around 3 percent of GDP and debt held by the public will exceed 65 percent of GDP—under current law. However, deficits would be much larger if policymakers extended the tax cuts mentioned above, or increased discretionary spending in line with GDP (which is about what actually happened in the past 20 years, leaving aside the effects of the stimulus package) rather than only with inflation as assumed in CBO’s baseline. For example, if the tax cuts were extended, the AMT indexed for inflation, and no other changes made in spending or revenue, the deficit would be about 6 percent of GDP in the second half of the coming decade rather than 3 percent, and debt held by the public would be nearly 90 percent of GDP by 2020. Both deficits and the federal debt would be on an upward path.

Today CBO responded to Senator Grassley’s questions about the President’s proposal for a “Financial Crisis Responsibility Fee.” The President proposes to assess an annual fee on liabilities of banks, thrifts, and security dealers, as well as U.S. holding companies controlling such entities. The fee, which would apply to firms with consolidated assets of more than $50 billion, would be approximately 0.15 percent of a firm’s total liabilities—excluding federally insured deposits and certain liabilities related to insurance policies.

Preliminary estimates by the staff of the Joint Committee on Taxation identified approximately 60 bank holding and insurance companies with assets in excess of the $50 billion threshold, which include most of the institutions that are likely to pay the fee. For the most part, the firms paying the fee would not be those that are directly responsible for losses realized by the Troubled Asset Relief Program. However, all of the institutions that might be covered by the fee benefited to varying degrees from the program’s contribution toward stabilizing the nation’s financial system and overall economy. The cost of the proposed fee would ultimately be borne to varying degrees by an institution’s customers, employees, and investors, but the precise incidence among those groups is uncertain. In general, the effects of a 0.15 percent fee on the economy – specifically on the growth rate of gross domestic product, the availability of credit, the stability of financial institutions, and future government outlays to cover losses on those institutions—would probably be small because the fee represents a small fraction of the cost to financial institutions of providing credit to their customers. 

Federal laws try to protect taxpayers and recipients of government benefits from the effects of rising prices by specifying that dollar amounts in many parts of the tax code and in some programs be automatically adjusted—or indexed—for inflation. Without such indexing, a rise in the general level of prices would alter the effects of federal policies even in the absence of action by lawmakers. For example, if a Social Security beneficiary’s payment remained the same over time (in other words, not indexed for inflation), the value of goods and services that the beneficiary could purchase would go down.

Many federal programs and parts of the tax code are currently indexed to increases in the consumer price index (CPI), a measure of inflation calculated by the Bureau of Labor Statistics (BLS). According to many analysts, however, the CPI overstates increases in the cost of living because it does not fully account for the fact that consumers generally adjust their spending patterns as some prices change relative to other prices. One option for lawmakers, as discussed in a brief released today, would be to link federal benefit programs and tax provisions to another measure of inflation—the chained CPI—that is designed to account fully for changes in spending patterns. (CBO previously discussed the possibility of using the chained CPI in its August 2009 Budget Options volume.) The chained CPI grows more slowly than the traditional CPI does: by an average of 0.3 percentage points per year over the past decade. As a result, using that measure to index benefit programs and tax provisions would reduce federal spending (especially on Social Security and federal pensions) and increase revenues. A separate appendix to the brief explains the methods and calculations that could be used to index the federal tax system, Social Security benefits, and federal pension benefits for the growth in the chained CPI.

Although many analysts consider the chained CPI a more accurate measure of the cost of living, using it for indexing could have disadvantages. Because the values of the chained CPI for a given month are revised over a period of one to two years, the tax code and affected programs would have to be indexed to a preliminary estimate of the chained CPI that is subject to estimation error. Also, the chained CPI may understate growth in the cost of living for some groups, such as older people.

This brief was prepared by Noah Meyerson of CBO’s Health and Human Resources Division.

I testified this morning before the Joint Economic Committee about policies to increase economic growth and employment in 2010 and 2011. This hearing was originally scheduled for several weeks ago but then canceled because of the snow. My prepared remarks today were essentially the same as those released a few weeks ago and were based on CBO’s January report on this topic and a follow-up letter to Senator Casey.

My comments emphasized three points:

First, although the economy is starting to recover from the most severe recession since the 1930s, CBO and most private forecasters expect a slow rebound in output and employment. In particular, CBO projects that the unemployment rate will average slightly above 10 percent in the first half of this year, fall below 8 percent only in 2012, and return to near its long-run sustainable level of 5 percent only in 2014. As a result, more of the pain of unemployment from this downturn lies ahead of us than behind us.

Second, fiscal policy actions, if properly designed, would promote economic growth and increase employment in 2010 and 2011. However, despite the potential economic benefits in the short run, such actions would add to the already large projected budget deficits. Unless offsetting actions were taken to reverse the accumulation of additional government debt, future incomes would tend to be lower than they otherwise would have been.

Third, different policies that have received public attention would have quite different effects on output and employment per dollar of lost tax revenue or additional government spending. To be sure, significant uncertainty attends any quantitative estimates of the effects of particular policies, and CBO has emphasized that uncertainty by reporting ranges of estimates that we believe encompass most economists’ views about the effects of each type of policy. Still, we think there would be significant differences in the cost-effectiveness of different policies (as measured in our analysis by years of full-time-equivalent employment per million dollars of total budgetary cost).

Policies that would have the largest effects on output and employment in 2010 and 2011 per dollar of budgetary cost would be those that could be implemented relatively quickly or targeted toward people whose consumption tends to be restricted by their income—for example, reducing payroll taxes for firms that increase payroll or boosting aid to the unemployed. The following figure summarizes the estimated effects of various policies on employment in 2010 and 2011.

Cumulative Effects of Policy Options on Employment in 2010 and 2011,
Range of Low to High Estimates

Under the American Recovery and Reinvestment Act of 2009 (ARRA), also known as the economic stimulus package, certain recipients of funds appropriated in ARRA (most grant and loan recipients, contractors, and subcontractors) are required to report the number of jobs they created or retained with ARRA funding after the end of each calendar quarter. The law also requires CBO to comment on those reported numbers. Today CBO released a report to satisfy that requirement.

CBO’s Estimates of ARRA’s Impact on Employment and Economic Output

Looking at recorded spending to date as well as estimates of the other effects of ARRA on spending and revenues, CBO has estimated the law’s impact on employment and economic output using evidence about how previous similar policies have affected the economy and various mathematical models that represent the workings of the economy. On that basis, CBO estimates that in the fourth quarter of calendar year 2009, ARRA added between 1.0 million and 2.1 million to the number of workers employed in the United States, and it increased the number of full-time-equivalent (FTE) jobs by between 1.4 million and 3.0 million. Increases in FTE jobs include shifts from part-time to full-time work or overtime and are thus generally larger than increases in the number of employed workers. CBO also estimates that real (inflation-adjusted) gross domestic product (GDP) was 1.5 percent to 3.5 percent higher in the fourth quarter than would have been the case in the absence of ARRA.

Data on actual output and employment during the period since ARRA’s enactment are not as helpful in determining ARRA’s economic effects as might be supposed, because isolating the effects would require knowing what path the economy would have taken in the absence of the law. Because that path cannot be observed, the new data add only limited information about ARRA’s impact. Economic output and employment in 2009 were lower than CBO had projected at the time of enactment. But in CBO’s judgment, that outcome reflects greater-than-projected weakness in the underlying economy rather than lower-than-expected effects of ARRA.

Limitations of Recipients’ Estimates

CBO’s estimates differ substantially from the reports filed by recipients of ARRA funding. Those recipients reported that ARRA funded nearly 600,000 fulltime-equivalent (FTE) jobs during the fourth quarter of 2009. Such reports, however, do not provide a comprehensive estimate of the law’s impact on employment in the United States. That impact may be higher or lower than the reported number for several reasons (in addition to any issues about the quality of the data in the reports): 

• Some of the reported jobs might have existed in the absence of the stimulus package, with employees working on the same activities or other activities. 
• The reports filed by recipients measure only the jobs created by employers who received ARRA funding directly or by their immediate subcontractors (so-called primary and secondary recipients), not by lower-level subcontractors. 
• The reports do not attempt to measure the number of jobs that may have been created or retained indirectly as greater income for recipients and their employees boosted demand for products and services. 
• The recipients’ reports cover only certain appropriations made in ARRA, which encompass about one-fifth of the total amount spent by the government or conveyed through tax reductions in ARRA during the fourth quarter; the reports do not measure the effects of other provisions of the stimulus package, such as tax cuts and transfer payments (including unemployment insurance payments) to individuals.

Consequently, estimating the law’s overall effects on employment requires a more comprehensive analysis than the recipients’ reports provide.