Federal Investment, 1962 to 2018

The federal government pays for a wide range of goods and services that are expected to contribute to the economy for some years in the future. Those purchases, called investment, fall into three categories: physical capital, research and development (R&D), and education and training. There are several economic rationales for federal investment. The federal government can provide public goods that the private sector and state and local governments would not provide efficiently, such as national defense and basic scientific research. Federal investment can promote long-term economic growth—as education spending does by developing a skilled workforce, as R&D spending does by prompting innovation, or as infrastructure spending does by facilitating commerce. And it can support the work of the federal government by, for instance, providing the structures and equipment necessary to perform federal activities.

In 2018, the federal government spent $492 billion on investment, representing 12 percent of federal spending and 2 percent of gross domestic product (GDP). Those shares have remained roughly stable over the past 25 years, though they reached higher levels in the early 2010s when the American Recovery and Reinvestment Act of 2009 (ARRA, Public Law 111-5) temporarily expanded funding for a number of investment programs. When ARRA spending was at its highest levels, in 2010 and 2011, federal investment approached 4 percent of GDP. Even as ARRA spending wound down in 2012, federal investment was $585 billion (in 2018 dollars), almost 20 percent higher than in 2018, and represented 15 percent of federal spending.

Federal investment as a share of the budget and the economy is lower than in the 1960s. In the 1960s, federal investment represented more than 30 percent of federal spending and averaged nearly 6 percent of GDP. Nearly all federal investment is discretionary spending, which is controlled by annual appropriation acts. Federal investment has gradually declined as a proportion of discretionary spending, from roughly 50 percent in the 1960s to about 40 percent today. In addition, discretionary spending as a whole has fallen as a share of total federal spending since the 1960s. Caps on appropriations that were established by the Budget Control Act of 2011 set limits for most discretionary spending from 2012 through 2021 to amounts that were lower than they would have been if annual appropriations had grown at the rate of inflation. Subsequent legislation increased those caps, most recently for 2018 and 2019. Under current law, the caps would return to the levels specified in the Budget Control Act for 2020 and 2021.

Sixty percent of total federal investment in 2018—or $297 billion, which represented about 1.5 percent of GDP—was for purposes other than national defense. Of that nondefense investment, 41 percent provided funding for education and training, 37 percent was for physical capital, and 22 percent was for R&D. Defense activities accounted for the remaining 40 percent of federal investment and totaled $195 billion, which represented 1 percent of GDP. Three-quarters of federal investment for defense purposes was devoted to physical capital and the rest to R&D.

How Does the Federal Government Support Investment?

The federal government supports public and private investment through several different mechanisms. In many cases, it makes the investment directly, such as when the Army Corps of Engineers constructs a dam or when a federal agency purchases computer equipment from the private sector. In other cases, the federal government makes grants to individuals or to universities and other nongovernmental organizations, which then use the funds to make investments. Examples of such grants include the Federal Pell Grant Program for postsecondary education and the National Science Foundation’s research grants. Those direct investments and grants account for 56 percent of nondefense investment, or $165 billion.

The federal government also invests through grants to state and local governments, which in 2018 represented 44 percent of its nondefense investment, or $132 billion. Grants accounted for about two-thirds of federal investment in nondefense physical capital and for nearly half of federal investment in education and training. State and local governments often have some latitude in determining how to spend the grant funds. Many federal grants require state and local governments to spend their own funds as well.

This report focuses on investment that the federal government makes either directly or through grants. However, the federal government also supports investment in other ways. One is through tax expenditures—credits or deductions that reduce the federal income tax liabilities of individuals and firms as a result of certain investments that they make or finance. Those credits or deductions can reduce the cost of investment for state and local governments as well. Defined narrowly, tax expenditures that support investment amounted to $199 billion in 2018. Of that sum, $143 billion supported investment in physical capital, mostly by excluding from taxable income the interest on state and local government bonds and by allowing tax filers to accelerate the depreciation of equipment and therefore to take larger tax deductions earlier in the equipment’s life. An additional $43 billion supported investment in education and training, mostly through tax credits and deductions focused on higher education. The remaining $13 billion supported investment in R&D. About three-quarters of that amount was the cost of a tax credit for increasing research activities, which primarily benefits corporations; and about one-seventh was the cost of allowing firms to deduct expenses for research and experimentation immediately. The remainder was the cost of a research tax credit to develop orphan drugs, medications that would treat rare medical conditions.

Other federal policies can also affect private investment. Tax policies, including individual and corporate income tax rates, can restrain or encourage economic activities by changing their relative after-tax prices. Regulatory policies influence investment by prohibiting or constraining certain activities, such as polluting the air, or by necessitating others, such as complying with federal safety standards. And federal deficits (and surpluses) influence the amount of funds available for private investment and the cost of those funds. For example, when the federal government issues bonds to finance its deficits, the funds that investors use to buy those bonds are no longer available to finance private investment. In response to the increased federal borrowing, bond buyers may also demand higher interest rates from the government, which would generally raise interest rates throughout the economy and make it more expensive for people and firms to borrow for investment purposes.

What Does the Federal Government Invest In?

Observers define investment in different ways. In the view of the Congressional Budget Office and consistent with the categories of the Office of Management and Budget (OMB), the federal government invests in three broad areas:

• Physical capital includes structures, such as government buildings, transportation infrastructure, and water and power projects; major equipment, such as computers, machinery, and vehicles; and software. For spending on physical capital to qualify as investment, the physical capital must have an estimated useful life of at least two years. Most federal investment in physical capital for defense purposes is for purchases of major equipment, such as ships and aircraft. Investment in physical capital for nondefense purposes, by contrast, is dominated by transportation spending, which provides infrastructure that contributes to the functioning of the economy.
• Research and development has three components: basic research, which seeks to discover scientific principles; applied research, which attempts to translate those discoveries into practical applications; and the development of new products and technology. Federal R&D spending supports a wide variety of work in government laboratories, universities, and the private sector, including health research studies, basic research in physics and chemistry, and the development of weapon systems. R&D investment builds the stock of knowledge that helps expand the economy over time, and the academic research that it funds is essential to the training of future generations of scientists. Most of the R&D spending by the federal government that supports defense is focused on development, rather than on basic or applied research.
• Education and training includes early childhood, elementary, secondary, and postsecondary education, which help produce a skilled, capable workforce that contributes to the country’s productivity. It also includes job training and vocational training for veterans and others, which likewise promote a productive workforce. Federal spending on education and training is thus an investment in the nation’s human capital.

In some cases, it is difficult to determine what qualifies as federal investment and what does not. For example, although this report regards spending on instruction and on the construction of school buildings as investment, it does not regard spending on health care and school lunch programs for children as investment, because those goods and services are promptly consumed. Yet keeping children healthy and nourished improves their ability to learn and produces a healthier and more capable workforce in the future.

The Bureau of Economic Analysis (BEA) includes in its calculation of federal investment most of what CBO identifies here, but omits spending on education. The investments in physical assets and R&D presented in this report are roughly comparable to two line items in BEA’s tables of the national income and product accounts (NIPAs). The first line item is gross federal government investment, which includes investments made directly by the federal government in structures, equipment, software, and R&D. The second line item is capital transfer payments, which are mostly grants to state and local governments for the purpose of investing in physical capital or R&D. R&D spending was first included in BEA’s definition of investment in July 2013, when the NIPAs were revised to count expenditures on intellectual property, including R&D, as investment.

How Does the Federal Government Account for Investment?

For accounting purposes, the federal budget treats most investment the same way it treats other spending: on a cash basis. That is, expenditures on investment are recorded as they are made, just as other expenditures are recorded as they are made and revenues are recorded as they are received. Two important advantages of that approach are that transactions are readily verifiable and that the sum of all transactions provides a close approximation of the government’s annual cash deficit or surplus. However, accounting on a cash basis makes investment appear expensive because many of the benefits associated with investment do not arrive until well after the initial investment has been made. For example, building a highway takes a large initial investment, but its benefits are delayed and then last for decades. By contrast, the benefits of other federal spending occur closer to the actual expenditure—for example, when air traffic controllers safely direct flights. Therefore, the current budget system may provide incomplete information to policymakers as they decide how to divide federal resources between investment and competing priorities.

Some policymakers have proposed creating a capital budget for investment that would allocate current capital costs to the future, spreading them over the period when an investment’s benefits would occur. That approach, which relies more on accrual-based accounting than on cash-based accounting, would be similar to the one used in the private sector.

Adopting a capital budget for investments would not be likely to have a noticeable effect on the federal budget balance, because even though the cost of current investments would be spread over future years, the federal budget would also have to show the depreciation of investments made in previous years. Nevertheless, the proponents of a capital budget argue that it would better align the timing of costs with the potential benefits of investment.

Aligning the timing of capital costs with an investment’s benefits, however, could make such a budget more complex and less transparent in several ways:

• The budget process would become sensitive to small changes in assumptions about the depreciation rates of assets within the capital budget and about how those rates should be adjusted over time to account for inflation and for changes in the assets’ replacement cost. A system in which those valuations were not made transparently could encourage manipulation. And it is likely that no depreciation schedule would perfectly track changes in the economic value of an asset.
• Because so much government spending could be viewed as providing benefits over an extended period, it would be difficult to determine what to include in the capital budget. An overly narrow focus would, by leaving some investments out of the capital budget, make them appear relatively expensive and therefore less desirable. An overly broad focus could turn the capital budget into a device for understating the cost of federal spending. The capital budgeting process could lead proponents of particular programs to try to have them classified as capital spending to lower their current cost and to advocate for longer depreciation periods.
• Policymakers would have to decide whether to include within the capital budget assets that the federal government helps fund but does not own. Roads, airports, and mass transit systems, for example, are often paid for in part by the federal government and in part by the state and local governments or independent authorities that own them. Federal investments in those assets could be excluded from the capital budget because the federal government does not own them. However, excluding those investments would make them appear expensive relative to other federal investments that were included in the capital budget.

Although the federal budget does not use capital budgeting, it treats a small amount of investment—investment that occurs through credit programs—on an accrual basis, recording the estimated present value of credit programs’ expenses and related receipts when the legal obligation is first made rather than when the subsequent cash transactions occur. Specifically, the Federal Credit Reform Act of 1990 (FCRA) prescribes procedures to estimate the present value of direct loans and loan guarantees to record in the federal budget.

An alternative way to estimate the budgetary cost of credit programs, which is also an example of accrual accounting, is called the fair-value approach. Fair-value estimates account for the market value of the government’s obligations and reflect the risks of those obligations for taxpayers. The main difference between FCRA and fair-value measures involves their treatment of market risk (sometimes called systemic risk or nondiversifiable risk), which is the risk that an overall market will decline. Most of the risk associated with financial investments can be avoided by having a diverse portfolio of investments; however, market risk remains even after a portfolio has been diversified as much as possible. It arises from shifts in macroeconomic conditions, such as productivity and employment, and from changes in expectations about future macroeconomic conditions. The government is exposed to market risk through credit programs because, when the economy is weak, borrowers default on their loans more frequently, and recoveries from borrowers are lower. With federal credit programs, the associated market risk of those obligations is effectively passed along to taxpayers, who, as investors, would view that risk as having a cost. In CBO’s view, fair-value estimates are a more comprehensive measure than FCRA estimates of the costs of federal credit assistance and can help lawmakers better understand the advantages and drawbacks of specific policies.

What Are the Benefits of Federal Investment?

Most federal investment for nondefense purposes contributes to the economy on an ongoing basis by improving the private sector’s ability to invent, produce, and distribute goods and services. A similar effect is expected from the small portion of defense investment that goes to basic and applied research. Most defense investment contributes to the development and production of weapon systems and other defense goods and is often thought to be less applicable to innovation in commercial or civilian products (notwithstanding some prominent technological advances, such as the integrated circuit and the Global Positioning System, that resulted from defense R&D and defense agencies’ subsequent procurement of those technologies).

Federal nondefense investment can contribute to private-sector productivity in various ways. Without public highways, the cost to the trucking industry of delivering goods would be much higher; if the Internet had not initially been developed through government R&D, whole segments of the economy would not exist; if not for receiving a public education (funded in part by federal spending), many workers would have lower wages than they do. Not all federal investments enhance productivity. For example, some investments may duplicate other efforts or divert federal resources from more productive projects.

In CBO’s view, the government has made higher productivity possible by making investments that the private sector would not have made on its own or would have made in smaller amounts than their broad public benefits would justify. The result of that higher productivity is higher private-sector returns. However, the size and nature of those returns are subject to considerable uncertainty, and some of the factors that contribute to that uncertainty are important considerations for policymakers facing decisions about how—and how much—the federal government should invest:

• It can be difficult to know which outcomes to attribute to which investments. Scientific and technological discoveries often build on prior work, making it hard to determine how great a share of a new product to attribute to a particular earlier investment. Similarly, workers’ skills are the product of education funded not only by the federal government but also by state and local governments, the private sector, and the workers and their families.
• Realizing the benefits of federal investment may take many years, and the timing varies for different types of investment. A new highway can improve transportation as soon as it is built, but it may take longer to realize the benefits of basic research or elementary education—which may also complicate the already difficult task of identifying those benefits.
• The benefits of federal investment are unlikely to be distributed evenly. Firms located near highways will probably enjoy greater returns from those highways than will firms located farther away. Recipients of federal grants for R&D may acquire patents based on their work; though products and innovations based on those patents may benefit consumers, they may also earn returns for the patent owners that are not shared with the country as a whole.
• Federal investment may discourage investment by private entities or by state and local governments by raising the price of investment goods such as machinery and other materials. If that happens, and if the discouraged investment would have had positive economic returns, then the net increase in output resulting from federal investment will be lower. Furthermore, state and local governments may use federal spending to fund investments that they would otherwise have made with their own funds. (In some cases, however, federal spending on investment may increase state and local investment, particularly when grant programs require state and local governments to invest as well.)
• Changes in federal investment may have different effects on private-sector productivity depending on the nature of the changes. Increases and decreases in federal investment may have asymmetric effects; changes in productivity may also be sensitive to the size of federal investments.

Acknowledging those sources of uncertainty, CBO uses a range of effects on output when estimating the effect of federal nondefense investment on the private sector. At the high end, CBO estimates that federal investment yields the same net increase in output in the long run as the average effect resulting from completed private-sector projects. At the low end, CBO estimates that federal investment results in no increase in overall economic activity—that is, has no net effect on future private-sector output. The actual effect on output from a particular investment could lie outside that range; the project might decrease private output or, alternatively, yield a greater overall increase than investment completed by the private sector.

Sometimes, policymakers may support investments not to achieve the largest expected economic returns but to accomplish other federal goals, such as defending the country or reducing inequities. At other times, the federal government may rely on policies other than investment to reach particular ends. For example, instead of investing to expand capacity on busy highways, the federal government might encourage state and local authorities to manage the high demand with congestion pricing—that is, charging drivers higher tolls at busy times and places. Even if an investment’s benefits would have exceeded its cost, the alternative policy may produce comparable benefits at a lower cost, thus allowing policymakers to find other uses for the funds that would have paid for the investment.