Staff Paper Prepared for the President's Commission to Study Capital Budgeting

June 19, 1998

In what ways could budget policies improve macroeconomic outcomes?


Is there evidence of a fundamental accountability issue -- the improper use of long-term borrowing for operating purposes?

Federal debt held by the public has increased five-fold since 1980. At the end of FY 1980, it was $710 billion; by the end of 1997, it was $3,771 billion. The growth of Federal debt held by the public accelerated during the early 1980s due to very large budget deficits. The deficits have declined since 1992, but they continued to exceed $200 billion until 1994 and to exceed $100 billion until 1996; consequently, debt continued to grow substantially. The ratio of Federal debt to GDP rose from 26 percent in 1980 to 50 percent in 1993, the highest ratio since the mid-1950s. The ratio of Federal debt to credit market debt also rose, though to a much lesser extent, from 18 percent to 26 percent. Interest outlays on debt held by the public, calculated as a percentage of both total Federal outlays and GDP, increased by about two-fifths.

During this 1980-1997 period, there was a continuing effort to constrain spending that limited the growth of discretionary programs and reformed many mandatory ones, including the fast-growing Medicare and Medicaid programs. Partly because of this constraint, but largely for other reasons such as the end of the Cold War, Federal investment in tangible and intangible capital for its own operations and for the Nation fell from 20 percent of Federal outlays to 14 percent; major physical capital investment fell from 11 percent to 7 percent. Meanwhile, payments for individuals rose substantially during this period from 47 percent of outlays to 59 percent. Transfers and other current Government consumption were the principal use of resources during this long period of excessive deficit financing -- not public investment. As a result, while the real Federal debt held by the public grew by $2.1 trillion (in 1992 dollars), the federally financed capital stock according to OMB estimates grew by only $1.3 trillion.

Should investment net of depreciation be financed by borrowing?

Some have argued that the Government ought to balance the operating budget and borrow to finance the capital budget -- capital expenditures less depreciation. The rationale is that if the Government borrows for net investment and the rate of return exceeds the interest rate, the additional debt does not add a burden onto future generations. Instead, the burden of paying interest on the debt and repaying its principal is spread over the generations that will benefit from the investment.

This argument would not justify net borrowing in excess of net investment. To the extent that capital is used up during the year, there are no additional assets to justify additional debt. If the government borrows to finance gross investment without repaying debt incurred on the worn out capital, the additional debt would exceed the net addition to capital assets. This rationale would not currently justify much additional Federal borrowing, if any, because the net Federal capital stock has leveled out with the decline in defense investment, and the growth in the federally financed national capital stock is small.

Because the Federal Government is responsible for the general welfare of the Nation, to maintain and accelerate national economic growth and development, the Government needs to consider how its decisions affect private investment as well as its own investment. For more than a decade, net national saving has been low, both by historical standards and in comparison to the amounts needed to meet the demographic challenges expected in the decades ahead.

When the Government finances its own investment in a way that results in lower private investment, the net increase of total investment in the economy is less than the increase from the additional Federal capital alone. The net increase in total investment is significantly less if the Federal investment is financed by borrowing than if it is financed by taxation, because borrowing primarily draws upon the saving available for private (and State and local) investment whereas much of taxation instead comes out of private consumption. Therefore the net effect of Federal investment on economic growth would be reduced if it were financed by borrowing. This could be the result even if the rate of return on Federal investment was higher than the rate of return on private investment. For example, if a Federal investment that yielded a 15 percent rate of return crowded out private investment that yielded 10 percent, the net social return would still be positive but it would only be 5 percent.

How, if at all, should financing principles be related to countercyclical policy?

The "high employment" or "structural" surplus/deficit is defined as the surplus/deficit that would occur if the rate of unemployment were maintained at the lowest sustainable level over time. Unemployment in the United States has been above this level for most of the past two decades while the budget was in deficit, so the high-employment deficit was smaller than the actual deficit during this period. Had the actual unemployment rate been lower, budget receipts would have been higher; for this reason, balancing the high-employment budget would have required less drastic action than balancing the actual budget. Recently, however, this pattern has been reversed. Currently, the actual rate of unemployment is 4.3 percent, which is significantly below the rate that is thought to be sustainable. As a result, the high-employment or structural budget surplus is smaller than the projected surplus.

Balancing the structural budget would produce a surplus when times were good and unemployment was low as they are now; it would permit deficits to occur, however, were the unemployment rate to rise above the sustainable level. If unemployment fluctuated randomly around that level, the budget would show no tendency toward either surplus or deficit.

The advantages of this goal for fiscal policy are that it maintains long-run fiscal discipline while avoiding the pro-cyclical fiscal tightening that would sometimes be required to maintain an annually balanced budget. However, there are still technical difficulties in achieving this goal because the level of unemployment consistent with stable inflation is not stable over time, and economists do not agree on its exact magnitude. Thus, the size -- and at some times even the direction -- of the change in the fiscal policy needed to maintain structural balance is uncertain.

Should the Federal Government attempt to add to national saving by running a surplus in order to encourage private investment?

Another question is whether the appropriate goal is balance in the structural budget, or whether the best policy -- in view of anticipated fiscal stress starting in about a decade -- might not be to run a structural surplus. The coming retirement of the baby boom generation is cited as one reason why such a policy might be advisable. More fundamentally, the expected growth in longevity, the low birth rate, and the prospect that growth in the labor force will slow to a crawl in the next century will mean that retirees and workers will continue to be dependent on deepening capital intensity and rising productivity to support an adequate level of retirement income.

The main economic argument for controlling the budget deficit and increasing the budget surplus hinges on the effect that deficits and surpluses have on private investment. Deficits are a claim on saving. Funds used to finance deficits would otherwise be available, through the financial markets, to support private investment. Surpluses will add to the pool of private saving and augment the funds available for private investment. Thus, the connection between the Federal deficit/surplus and private capital formation is relatively loose. Nevertheless, the connection exists, and it would be possible, in theory, to conduct fiscal policy by setting a goal for private investment, and using the surplus/deficit to reach it.

This raises the question: What is the optimal rate of investment? In the very long run, according to the view of most macroeconomists, the growth rate of the economy is determined by technical progress and demographic trends. Additional saving can raise growth, but eventually the growth of both the capital stock and the economy reverts to that determined by technical progress and demography as the law of diminishing marginal returns takes hold. (Each year's increment of capital adds a little less to output because there is more capital relative to labor and output, which reduces the usefulness of the marginal increment of capital(1).) This might appear to limit the role of fiscal policy, but that conclusion would not be correct.

During the transition from less capital to enough capital to result in little marginal return -- a transition period which can last for many decades -- a fiscal policy that encourages more investment does affect growth. Thus, increased investment, made possible by smaller deficits or higher surpluses, can raise the long-run level of income. But there is a natural limit to this process. The path of maximum sustainable per capita income is reached when the rate of return to capital is equal to the economy's "natural" rate of growth, given by the sum of population growth and technical progress. Currently, the natural rate is usually estimated to be in the range of 2 percent to 2-1/2 percent, with population growth and technical progress contributing in roughly equal proportions to the total; while the average real rate of return to capital is estimated to be around 7 percent for all types of capital.

A policy that drives the stock of capital to a point where the marginal rate of return equals the natural growth rate has been called the "Golden Rule" of growth; at this point further saving no longer adds to an economy's per capita consumption possibilities. Any more saving would be wasted, but whether it is optimal to raise the saving rate to the level dictated by the Golden Rule depends on how impatient consumers and voters are. While maximizing long-run consumption is a laudable goal, it is not possible to do so without sacrificing some current consumption. A dollar tomorrow is worth less than a dollar today for most consumers -- and for most voters and their elected representatives. Respecting this time preference would mean permitting the rate of return on investment to exceed the Golden Rule rate by the amount of time preference. Under such a modified Golden Rule, targeting a structural Federal surplus between 1 and 2 percent of GDP would seem justified, at least for a time until the rate of return to capital sinks closer to the long-run growth rate.

Does the Federal Government over- or under-invest in Federal capital? In non-Federal capital? Is there compelling evidence that the existing budget process adversely affects the macroeconomy?

The other main channel which the Government can use to influence the economy in the long run is though its own investments and through the investment it subsidizes by other levels of government, individuals, and private organization. It is very difficult to assess whether, in the aggregate, the level and growth of Federal investment is high enough.

Most of the net capital stock in the United States is in the private sector. Private fixed capital in 1996 was $14.6 trillion or 3.4 times as much as the $4.3 trillion of fixed capital for all levels of government combined (both in chain-weighted 1992 dollars). Private capital has grown more rapidly than public capital since 1970, but most of the differential occurred before 1985. Since then, the growth rate of public nondefense capital has been as fast as that in private capital.

Most government fixed capital is owned by State and local governments -- $3.1 trillion in 1996 or 2.6 times the $1.2 trillion owned by the Federal Government (also in chain-weighted 1992 dollars). Looking at non-defense public capital only, the State/local capital is 7.6 times the $0.4 trillion of Federal capital.

A forthcoming CBO report will evaluate available studies on the return available to public investments, updating its 1991 report on "How Federal Spending for Infrastructure and Other Public Investments Affects the Economy." The partial work we have done so far indicates that returns may be high in selected areas. Is there a useful principle for determining appropriate aggregate Federal investment in Federal assets? In non-Federal assets?

The categorization of investment that most affects analysis of the appropriate amount of investment is by ownership. The Government invests in federally owned assets that are combined with other inputs to produce Federal services, and it finances investment in assets that it does not own that add directly to the wealth and well-being of the Nation.

Investment in federally owned physical capital is 29 percent of Federal investment. This capital is combined with other inputs, including labor and support services, to produce the goods, services, transfers, grants, and credit that the Federal Government provides to the public in carrying out Federal missions. The goal in selecting such investments is to produce each funded service as efficiently and effectively as possible; there is no reason why capital-intensive operations are always better. Moreover, although the Federal cost of capital is always lower, it is sometimes efficient to lease capital for short-term use or to shift technological risk to the private sector.

It follows that there is no "appropriate" level or growth rate for total investment in federally owned capital. There is, however, a useful set of principles in deciding whether the amount of Federal capital is appropriate. The sequence is, first, to allocate resources among the many Federal programs based on policy judgements as to their relative importance. Second, to bring analysis to bear as to what is the most efficient way of achieving the programmatic goals; for example, should the program hire more staff, buy more computers, or purchase services from the private sector? And finally, to make the decisions on the appropriate combination of inputs based on a level playing field in which the programs pay the fair cost of each resource used and in which they are free to select the desired resources and providers. The optimal amount of investment in Federal capital derives from these decisions.

Of all Federal investment, 71 percent does not result in assets that the Federal Government owns; it yields intangible capital -- human capital or knowledge from research and development -- or assets owned by State and local governments. When the Government finances investment in assets that it does not own, the investment itself is the program. The goal for much of this investment is to support and accelerate sustainable economic growth for the Nation as a whole, and sometimes for specific regions or groups of people.

One way to determine an appropriate level of such investments would be to try to compute the rate of return on them. If the rate of return on a project or program exceeds the rate of return on the best available alternative use of the resources, then the investment is worth undertaking. This is the standard benefit-cost criterion applied to Federal investments: compute the benefits, calculate the implied rate of return, compare that with the return on competing uses of the resources, and proceed if the computed rate of return in benefits is larger than the opportunity cost.

The collective total of all investments whose benefits exceed their opportunity costs can be thought of as setting an implicit goal for total investment. The Issue paper on Investment Targets discusses various options for using such targets in the budget process. Setting any such target ought to be grounded in an assessment of the volume of available investments that would meet the benefit/cost test, and the investments chosen should meet this criterion. Setting an aggregate target, per se, would not improve the process of selecting Federal investments so that those with the highest returns are chosen.

1.  The diminution of the marginal product of capital can be slowed if technological change is capital intensive or if it is "embedded" -- requiring new investments for its implementation.

President's Commission to Study Capital Budgeting