June 19, 1998
Is there a general bias against investment or in favor of investment as a result of the existing budget process? Are there different types of bias? Is there a bias one way or the other for some kinds of investment but not others?
A great many potential biases affect Federal decisions on the amount and kind of investment. Some of these would tend to reduce or increase all types of investment and others would affect only particular types of investment. These many types of bias are discussed in the sections below.
Is there a general bias against investment spending because its benefits extend into the future?
Nearsightedness on the part of political decision makers could systematically bias decisions against spending that yields benefits far in the future, and thus could create a bias against investment and in favor of current consumption. Members of Congress face frequent reelection; appointed officials in the Executive Branch typically stay for only a couple of years. In effect, politicians may use too high a discount rate in judging the future services of capital assets purchased now. They may give too little weight to the faster travel, cleaner environment, or other services that will continue to be provided beyond the next election or two.
Political benefits(1), on the other hand, weigh heavily in favor of many kinds of investment. Unlike businesses, governments include externalities in their benefit/cost calculations, along with benefits stemming directly from capital services. Congressmen and Senators like to point to buildings, highways, navy shipyard contracts, etc., as concrete examples of their ability to bring home benefits to constituents. This is particularly true when the projects generate jobs and economic development, increase land values, yield favorable publicity, and/or provide an opportunity to name a prominent and lasting landmark. These collateral benefits are front-loaded, and may more than offset the effects of nearsightedness; investments with large collateral benefits account for the great popularity of "pork barrel" spending.
Collateral benefits are not always large; they differ in size from one investment to another. They tend to be large for highways, airports, and other transportation infrastructure and sizable also for buildings. Political benefits and externalities are very much smaller for other investments, such as air traffic control, Y2K compliance, and other information technology. One kind of investment spending with particularly small collateral benefits is maintenance and repair. The tendency to under invest in such work has led to the relatively high economic rate of return that CBO reported in 1991 (and is likely to report again soon) for maintenance and repair compared with new investment.
One way to offset a bias against less-favored investment would be to undertake systematic long-term capital planning (see staff paper, "A Longer Term Perspective for Capital Needs"). Such plans, set in the context of agency missions and program outcomes, can be cited by politicians to demonstrate the benefits of less politically exciting investments to constituents. Benefit/cost analysis, as an integral part of such planning, can be helpful in evaluating major investment proposals, and could be required for capital projects above a certain size in the budget justification process(see staff paper, "Using Benefit/Cost Analysis"). [Cross reference: Circular A-94 includes guidance on an array of issues that arise in benefit/cost analyses. The Capital Programming Guide, issued in July 1997 as a supplement to Circular A-11 -- the Circular prescribing the nature of budget submissions -- provides detailed guidance for capital planning in the context of agency strategic and annual plans. The Guide also requires analysis of the agency's capital in use, its remaining life, condition, usefulness in current operations, and plans for repairs, renovations, and replacement.]
Long-term plans could be particularly useful in encouraging appropriate levels of maintenance and repair -- scheduling significant repairs and replacement of major components over the asset's life cycle (see staff paper, "Capital is Often Poorly Maintained"). Advance funding might even be appropriated for scheduled major items such as roof or equipment replacement. State and local grantees, too, could be asked to submit plans for and report on maintenance and repair -- perhaps even as a condition for the receipt of grants for new construction.
"Spikes" due to the size of investments relative to operating expenditures might create a bias against investment.
Federal investment is relatively small compared with transfers and other expenditures and, in the aggregate, relatively smooth. Substantial amounts of investment are financed by earmarked receipts (e.g., the gasoline and airline ticket taxes). Substantial amounts, too, are provided as grants to State and local governments for physical capital, or as support for education or research, none of which is "lumpy" in character.
"Bulges" in spending over several years may occur when a major national goal involves a large increase in investment. This was the case for space exploration in the 1960s and the defense buildup in the 1980s. In these cases, the increase in investment was so large that it was visible in the budget as a whole. These bulges were accommodated despite their visible effect on the deficit because the investments involved had strong political support.
However, in small agencies and sometimes even in agencies that are not so small, investment in federally owned capital assets for use in providing Federal services can be irregular and large relative to other spending. Moreover, in larger agencies, the Congress and even agency heads often allocate all of their resources by setting budget "marks" incrementally for each program, region, and office; in each tributary, capital acquisitions may be lumpy. Although some "spikes" in BA have been accommodated by the agency, OMB, and Congress, some programs, in the belief that spikes would not be accommodated, have taken pains to make their requests relatively stable at the expense of efficient investment planning.
Much Federal investment avoids this problem by being aggregated in major capital accounts, where the number of projects financed smooths the total. This is true for defense, environmental, and some other investment. The Federal Buildings Fund was created in part to smooth agency spikes for building construction by aggregation. This technique could be expanded by systematically aggregating agency spending on Federal capital assets in one or more "capital acquisition funds" (CAF) (see staff paper, "Capital Acquisition Funds"). These accounts could help to smooth spikes within the agency, improve capital planning, and show the appropriate capital use charges in the budget accounts of each program by "renting" capital to the programs. Since the rent would be an offsetting collection to the capital acquisition account, the budgetary totals for the agency (and the Government as a whole) would continue to show the cost of each capital acquisition or useful segment up front.
Budgeting for "useful segments" is another way in which potential "spikes" may be accommodated -- and risk may be diminished (see staff paper, "Full Funding for Capital Acquisitions"). A useful segment is defined as a component of a capital project that provides either: (1) information that allows the agency to plan the capital project, develop the design, and assess the benefits, costs, and risks before proceeding to full acquisition of the useful asset (or canceling the acquisition); or (2) a useful component asset for which the benefits exceed the costs even if no further funding is appropriated. Risk is diminished by developing detailed plans and workable prototypes first, and proceeding with the project one segment at a time.
Good capital planning, which can foresee the need for capital projects over a longer time horizon, is also useful to avoid spikes. Moreover, agencies have been encouraged to accommodate "spikes" within the agency budget, and cross-cutting fixed asset reviews are held at the beginning of the fall budget review process in part to identify justified spikes that should be accommodated. Finally, advance appropriations can spread the cost of a project that will take more than one year to construct.
Is there a bias against investment in infrastructure as a result of the existing budget process?
Investment in highways, mass transit, airports and airways, and waterways are partly financed by earmarked receipts (see staff paper, "Dedicated Revenues for Selected Capital"). This is intended to ensure steady financing for infrastructure and to shield these expenditures from a comparison with other priorities. Such earmarking, although it would not necessarily be so interpreted from all perspectives, could be seen as a bias in favor of infrastructure.
If less spending on the investments for which receipts have been earmarked can allow more spending on other priorities, then the shield provided by earmarking may be weakened. This is what happened under the Budget Enforcement Act, when spending for infrastructure was included under the discretionary caps and isolated from the amounts of or changes in earmarked receipts, which were categorized as mandatory.
Spending below the earmarked amount can be viewed as a bias against infrastructure because the political agreement to enact these taxes included an agreement to spend the earmarked receipts on the designated purpose: the highways, airports, and other infrastructure for which these funds were designated. This perceived bias led to the recent decision in the Transportation Equity Act for the 21st Century to add highway and mass transit categories to the BEA starting in FY 1999. These categories are significantly different than the others in that they are based on the estimated level of earmarked receipts and are required to be adjusted for actual receipts. This change has greatly strengthened the shield for highway and transit spending -- making it almost impervious to trade-offs with other spending (see staff paper, "Investment Targets").
Would special budgeting for certain capital assets, such as buildings or "brick and mortar projects," create a "measurement bias" against those long-term items, such as human capital, whose "stock" is more difficult to measure and which would not obtain special budgetary treatment? Would special budgetary treatment create a "measurable bias" against non-investment spending?
The Federal budget allocates resources among responsibilities identified as appropriate to the Federal Government on the basis of political priorities. The Congress allocates resources to Federal agencies to accomplish a wide variety of programmatic goals. Because these goals are diverse and most are not measured in dollars, they are difficult to compare with each other. Policy judgments must be made as to their relative importance.
For the same reasons that it is difficult to measure government output and outcomes, it is difficult to measure the benefits of government investments. These problems complicate our ability to measure rates of return on government investments with sufficient accuracy to rank them and to make only investments with returns higher than the cost of capital as a business would do. Thus, it is hard to say what the total amount of government investment should be, and it is hard to compare the advantages from additional spending on capital for one purpose with those of additional spending on capital for a different purpose.
The appropriate amount and kind of capital to achieve a given goal efficiently is much easier to assess. For example, health experts can judge whether it is more cost-efficient to build another VA hospital or to pay for private hospital services. This is why government capital investment is more efficiently selected by allocating resources among the different functions or goals first, and then within each area determining the amount to be spent on capital compared to other objects.
Whether allocating resources among types of capital with the goal of enhancing national economic growth or allocating resources among different inputs to deliver better VA health care, it would be inefficient if some types of capital or other inputs were treated as "special" in any way in comparison with other types of capital or other inputs.
Thus, the President's Commission on Budget Concepts said: "The Commission believes that a further very persuasive argument against a capital budget(2) is that it is likely to distort decisions about the allocation of resources. It would tend to promote the priority of expenditures for "brick and mortar" projects relative to other Federal programs for which future benefits could not be capitalized (including health, education, manpower training, and other investment in human resources) -- even when there is no clear evidence that such a shift in relative priorities would in fact be appropriate."
This is also the reason why the Budgeting for Results Initiative would eliminate the idiosyncratic differences in the kinds and amounts of resources that are centrally funded at the Bureau or Departmental levels or by central agencies, and charge Federal programs uniformly and appropriately for all of the resources that they use (see Issues on CAFs and Labor/Support Services). This would permit program managers to select the most efficient combination of inputs and providers to help them deliver services to the public.
Does the existing budget process employed by the Federal Government (Legislative and Executive) reasonably optimize choices between operating spending and spending with potential long term benefits?
Inherent bias is not evident in the budget requirement that the cash outlay for investment, like all other spending, be recorded as it occurs. General economic theory, cost accounting, and benefit/cost analysis all conclude that an investment proposal should be evaluated in terms of discounted cash outflows and cash inflows (or the equivalent in intangible benefits). Efficient decision-making requires that the full cost of an investment be compared with the present value of benefits to determine whether or not to make the investment.
The present budgetary treatment facilitates such a comparison. It shows the cost of an asset in the same year as it should be recorded for benefit/cost analysis. Thus, it contributes useful information to the decision. Moreover, it provides the right cost at the right time to compare with future benefits, and therefore provides an incentive to make an efficient decision. Even if benefit/cost analysis is not consciously applied because future benefits cannot be precisely quantified, and even if estimation has to be very crude, this is still the correct conceptual framework for analysis.
The purchase price of an asset is a cash outflow; depreciation is not. Therefore, the purchase price is the theoretically correct cost to use in comparison with benefits; it is the cost normally used by businesses in making investment decisions regardless of whether they employ present value, internal rate of return, or cruder payback periods to evaluate the returns on investment. If depreciation or debt service were to be substituted for the cash outflow, there would be an incentive to buy the asset if the expected future benefits were larger than the first year's depreciation or debt service.
It is only in the year the commitment is made to acquire an asset that the budget exerts any control. Most Federal investments are sunk costs and as a practical matter cannot be recovered by selling or renting the asset. At the same time, there is a significant risk that the need for the capital asset may change over a period of years, because either the need was not permanent, it was initially misjudged, or other needs become more important. Since the cost is sunk, however, control cannot be exercised later by comparing the annual benefit of the asset services with depreciation and interest and then selling the asset if its annual services are not worth this expense.
Since the depreciation or debt service in the first year would be much
smaller than the cost of the asset, there would be a large bias in favor
of investment in any asset for which depreciation or debt service were
substituted for cash outlays. If such a substitution were made for physical
assets, but not intangible investments such as R&D or education, there
would be a bias in favor of physical relative to intangible investments.
Such a bias would be economically inefficient.
2. "One category of Federal
expenditures which has sometimes been singled out as sufficiently distinctive
in character to call for separate treatment is investment in physical assets,
linking that investment directly to Government borrowing. A divided budget,
with investment in physical assets excluded from calculations of the budget
surplus or deficit, is often referred to as a capital budget." (Report
of the President's Commission on Budget Concepts, page 33.)
President and First Lady | Vice President and Mrs. Gore