SIX-PERCENT TAX INCREASE WOULD BE NECESSARY TO FINANCE BALANCED BUDGET AMENDMENT, SAYS CRS.

                          Donald W. Kiefer

The President and Congress have often avowed their support for deficit reduction, a policy that requires current sacrifices for future benefits. As progress repeatedly has faltered, increasingly rigid constraints on the budget have been imposed by law since 1985. Because deficits continue nonetheless, many policy makers are turning toward a constitutional amendment requiring a balanced budget in hopes that it can suppress the tendency for short-term demands to overwhelm long-term considerations. This report assesses the economic effects of such a constitutional amendment.

If a balanced budget were to be required in FY96, and if all parts of the budget except interest payments and deposit insurance were subject to change, spending cuts and tax boosts averaging 6 percent of currently projected levels would be needed. With no new taxes or cuts in social security, however, balancing the budget in FY96 would require cuts of about 15 percent from the smaller remaining amount of unprotected spending. In real terms, discretionary spending would be reduced by 25 percent from today's level. In either case, deficit reductions averaging about $90 billion annually for four consecutive years would have disruptive effects on the economy as well as on Federal programs -- an argument for a longer phase-in period for the balanced-budget requirement.

While elimination of deficits would benefit the economy in the long run, the form of balanced-budget amendment currently proposed also would bring reductions in national income and welfare by inhibiting economic policymaking. In most instances it would neutralize the "automatic stabilizers" that expand outlays and reduce tax collections during economic downturns, leaving stabilization to monetary measures by the Federal Reserve. Lack of a prompt automatic stabilizing mechanism would entail some loss of jobs and income. Prohibition of borrowing to finance Federal net investment that yields services over several years also runs counter to the principle that services should be paid for when received and is likely to reduce public investment, hampering productivity and income growth.

The budget cost of Federal deposit insurance during the recent recession affords an illustration of the destabilizing potential of a balanced-budget amendment and the case for defining with care the budget to be balanced. Deposit insurance costs rose by $44 billion from 1989 to 1991. Meanwhile the recession increased the deficit by some $72 billion. Maintaining a balanced budget would have required tax boosts and spending cuts of $116 billion, which would have brought on a deeper recession.

State balanced-budget requirements differ fundamentally from the current Federal proposal by allowing borrowing for net investment in capital facilities, requiring balance only in the "general fund" for current services. Even this requirement often is circumvented. It is common for some current services to be financed in "special funds" that need not be balanced annually or by loans from special funds experiencing surpluses. Debt limitations also are routinely circumvented by establishing public corporations and authorities and issuing debt not guaranteed by the taxpayer. These actions transfer substantial control over public policy from elected officials to appointed officials and the courts.

                              CONTENTS

The economic case for reducing the Federal budget deficit is compelling. Despite this fact, neither normal processes of Government nor extraordinary statutory restrictions imposed on the budget process since 1985 have succeeded in reversing the deficit's long- term upward trend. These failures have lent new support to proposals for a constitutional amendment requiring a balanced Federal budget.

A. ARGUMENTS FOR REDUCING THE BUDGET DEFICIT

Economic arguments for reducing the Federal budget deficit rest on dissatisfaction with the Nation's economic progress and prospects for the future. They are rooted in a belief that sacrifices to reduce consumption and hence to increase saving are warranted to accelerate future gains in living standards.

The Federal budget deficit (dissaving), together with low private saving, causes a shortfall of saving relative to investment in the United States. Viewed another way, total demand for funds by Government and private investors exceeds domestic private saving. The resulting high interest rates attract foreign capital to cover the shortfall.

Capital inflows result in spreading foreign ownership of U.S. assets and mounting payments of investment income to foreigners. They also have resulted in trade deficits of unprecedented size. So long as capital inflows continue, trade deficits are likely to persist. If foreign investors should become unwilling to provide needed capital under prevailing conditions, U.S. interest rates would rise and/or the dollar's exchange value would fall, perhaps steeply, to make U.S. assets cheaper to foreigners and more remunerative. High interest rates resulting from the saving shortfall raise capital costs in the United States and tend to curtail total investment. Lower exchange rates would impose a drag on living standards by raising prices of imports and worsening the Nation's terms of trade.

Over the long run, therefore, large Federal deficits and low private saving imply increasingly costly and precarious dependence on foreign capital and less investment to modernize and expand the economy. They imply smaller gains in productivity and in income for American families. Slow growth since 1973 -- indeed, no growth or declines in income for many workers and families -- already is a source of frustration and anxiety for many Americans and, if continued, may permit other nations to surpass our living standard in the next few years. Large persistent deficits today to finance consumption will make it impossible for future generations to finance expected Government services without elevated taxes and will make their living standards lower than they otherwise would have been.

The goal to enhance productivity means that the budget deficit should be reduced in ways that reduce consumption, not investment, public or private. Closing the deficit by cutting public investment in R&D, education, training or infrastructure, so long as these investments are well conceived, would not serve the economic objectives of deficit reduction.

B. CHANGES IN BUDGET PROCEDURES, 1974 TO 1990

If Congress passes the balanced budget constitutional amendment, it will not be the first time it will have attempted to impose rules to force attainment of a balanced budget. The first such effort was the Congressional Budget and Impoundment Control Act of 1974, which established the foundation of the modern congressional budget process. The Act established the Budget Committees and for the first time required passage of annual congressional budget resolutions. Prior to 1974, Congress passed separate revenue and appropriations bills each year, but passed no single piece of legislation specifying budget totals or the amount of the deficit. There was no vehicle for assuring that the receipts and spending totals and the deficit which resulted from the individual revenue and appropriations bills conformed to any overall budget plan or were appropriate for the economic conditions.

The 1974 act was expected to result in smaller budget deficits because Members of Congress would be reluctant to vote for budget resolutions specifying large deficits. It did not work out that way. The 1974 deficit (which amounted to 0.4 percent of GDP) was, with the single exception of the 1979 deficit (1.7 percent of GDP), the last one to be less than 2.5 percent of GDP. The deficit ballooned in 1975 and 1976 in response to the 1973-75 recession. It diminished in the late 1970s, but ballooned again in the early 1980s due to back-to- back recessions, a large tax cut, and a big increase in defense spending.

In 1985, with the deficit exceeding 5 percent of GDP, Congress enacted the Balanced Budget and Emergency Deficit Control Act, better known as the Gramm-Rudman-Hollings Act (GRH). This act established annual deficit targets which decreased to zero in FY91 and created a procedure of across-the-board spending cuts, termed "sequestration," to enforce the targets.

By 1987, the prospective deficit for the following fiscal year exceeded the GRH target by about $60 billion. Spending cuts and tax increases enacted that year closed about half that gap. Rather than allowing sequestration to achieve the remainder of the required deficit reduction, GRH was revised to raise the deficit targets and stretch the period for eliminating the deficit to FY93.

In 1990, with the projected FY91 deficit $120 billion in excess of the target, GRH was revised again in the Omnibus Budget Reconciliation Act. This Act, combined with other legislation, implemented $42 billion of deficit reduction. It also revised the deficit targets upward substantially. Furthermore, the new deficit targets are to be adjusted in the President's budget submission for revised economic and technical assumptions (the adjustments are required through FY93; they are at the President's discretion for FY94 and FY95). The target specified in the 1990 act for FY93, for example, was $236 billion; the revised target -- adjusted to take account of the impacts of recession, halting recovery, unforseen deposit insurance outlays, and other factors -- is $407.2 billion.

The deficit targets have been made largely irrelevant at least through FY93, however, by new limits on discretionary spending (separate limits apply to defense, international, and domestic spending for FY91-93) and a "pay-as-you-go" limitation on revenues and entitlements (social security is excluded from the pay-as-you-go limitation). The pay-as-you-go limitation requires any deficit- increasing action to be offset by an equal deficit-reducing action. Spending designated by both the President and Congress as "emergency" spending is exempt from the limits. ¹

Under these revised budget procedures a balanced budget is not expected to be achieved any time in the foreseeable future. The Congressional Budget Office projects deficits exceeding 3 percent of GDP into the next century (and rising, rather than falling, toward the end of the 1990s).

Hence, over the past seven years Congress has adopted increasingly complex limitations on the budget process that seemed to promise deficit reduction in the future. But two or three years down the road when the limitations become constraining, the process has been revised to relax the current and near-term limitations and offer a new prospect of deficit reduction further in the future. During this period of frequent changes in the budget process to achieve deficit reduction, the deficit initially diminished somewhat but more recently has grown to one of the highest levels in the post-war period.

Against this background, the apparently increased support for a balanced budget constitutional amendment is, perhaps, understandable. A simple rule requiring a balanced budget by a fixed date in a form that cannot easily be amended has an obvious appeal.

C. A CONSTITUTIONAL AMENDMENT TO REQUIRE A BALANCED BUDGET

This subsection summarizes the balanced budget constitutional amendment proposals and the rational behind them.

1. The Proposals

The two principal proposals for a balanced budget constitutional amendment, H.J. Res. 290 and S.J. Res. 18, are similar but not identical. S.J. Res. 18 would require that "Total outlays of the United States for any fiscal year shall not exceed total receipts to the United States for that year, unless Congress approves a specific excess of outlays over receipts by three-fifths of the whole number of each House on a rollcall vote." ² H.J. Res. 290 has a similar requirement except that it refers to estimated receipts rather than actual receipts and requires Congress and the President to agree on an estimate of receipts by enactment of a law devoted to that subject. H.J. Res. 290 also would require a three-fifths vote by the whole number of each House to raise the limit on debt of the United States held by the public. S.J. Res. 18 has no similar provision.

Both resolutions would require the President to submit to Congress prior to each fiscal year a proposed budget in which outlays do not exceed receipts (the resolutions would not prohibit the President from also submitting a budget with a deficit if he chose to do so).

Both resolutions would require a vote of a majority of the whole number of each House by a rollcall vote to increase revenue. The Senate resolution contains an exception allowing adoption of a revenue-increasing bill by unanimous consent. ³

Both resolutions would allow the balanced budget requirement to be waived for any fiscal year in which a declaration of war is in effect. The Senate resolution would also allow waiver in a fiscal year in which the Nation "is engaged in a military conflict which causes an imminent and serious military threat to national security."

Both resolutions would put the constitutional amendment into effect beginning with the second fiscal year after its ratification. If the amendment passes Congress and the States ratify it quickly, it could become effective as early as FY95. This would allow only two fiscal years in which to achieve the required budget balance, or three years if progress toward balance is begun in the FY93 budget, which is currently being written.

The resolutions raise a number of issues regarding interpretation, procedure, and enforcement, most of which are not addressed directly in this report. ⁴

The effect of the constitutional amendment in the resolutions would be to allow a 41-percent minority in either House to block adoption of a budget with a deficit. The deficit referred to is the gap between "total outlays" and "total receipts." While it is not entirely clear how these terms would be interpreted, the intended applicability does not seem to be limited to those outlays and receipts "on budget" or for particular purposes (e.g., current operations). The amendment would make it somewhat more difficult to raise revenue because of the required majority of all Members of each House rather than the normal requirement of a majority of those present and voting. Finally, the amendment would not provide for waiving the balanced budget requirement if the economy turned downward, as do the current budget limitations ensuing from the GRH act. To set aside the balanced budget requirement in a recession would require a three-fifths vote in each House.

2. The Rationale for the Proposals

The rationale for a constitutional amendment to require a balanced budget is that the current budget decision-making process shortchanges the welfare of future generations in favor of current generations, and the only way to achieve an optimal result is to impose strong constraints on the process. This line of reasoning contains two apparent paradoxes. First, the Government already has the power to make all the decisions necessary to balance the Federal budget; yet, despite apparent commitment to that objective, it has not used the power to that end. Why not? Second, why would the Congress voluntarily impose limitations on its own decision-making power?

The answers may be found in the economic theory of self-control, which has previously been used to explain some aspects of individual savings behavior. ⁵ Why, for example, do some people authorize overwithholding of income taxes as a way of "forced savings" (on which no interest is earned)? Why were Christmas clubs popular for many years? These were special accounts at financial institutions into which members deposited fixed amounts weekly; the accounts earned no interest, and the funds could not be withdrawn until December.

The explanation offered by the theory of self-control is based on the notion that individuals actually have two separate economic decision-making personas. One of these "selves" is essentially a myopic consumer, interested primarily in buying as much enjoyment as possible at any given moment. The other "self" is a planner who considers long-term objectives such as the need to save for a child's future college education or his own retirement. The individual, aware of both of these inner selves, realizes that the only way to constrain the consumer is for the planner to adopt rules for the consumer to follow. The theory posits that individuals are most likely to impose constraints on their own behavior regarding those decisions for which the benefits and costs occur at different times, especially decisions involving early costs and delayed benefits.

If the individual possesses sufficient will power (the consumer within is readily controlled), simple rules of thumb may suffice: "I will save X percent of my salary each pay period;" or "I will not borrow to finance consumption." If the consumer is more difficult to control, the rules may be more rigid, allowing little if any flexibility. An example is a professional athlete who hires a financial manager to handle all finances and investments and to put the athlete on an allowance to support current spending. ⁶

This theory has obvious relevance to the budget decision-making problem. There is widespread agreement that the deficit should be reduced. The President and Congress have gone on record many times supporting deficit reduction as a policy goal. But deficit reduction requires current sacrifices in exchange for the promise of future benefits. Despite the apparent commitment to deficit reduction, budget decisions have not made consistent progress toward this objective.

Attempting to constrain the decision-making options, Congress has adopted increasingly tighter limitations on the budget process in the Gramm-Rudman-Hollings legislation and its revisions. So far, however, the rules have not resulted in a balanced budget or a budget that is even nearly balanced. Each time the rules have imposed a serious limitation, they have been changed to lessen the constraint. The theory of self-control suggests that in this situation, rules that are more rigid and difficult to evade are likely to be adopted. Hence, support for a constitutional amendment requiring a balanced budget.

Rigid rules, however, are a second-best or even third-best approach to decision making. The best situation is an unconstrained decision process which properly takes into account both short-term and long-term considerations. In this situation, the decision-maker is free to optimize fully. If this approach fails, a second-best approach is to adopt rules, limitations, or incentives which give greater weight to long-term considerations, but which also allow flexibility in responding to short-term contingencies. Finally, if other options fail, a third-best approach is to impose strict limitations on the decision-making process to minimize the degree to which short-term concerns are allowed to determine the outcome.

Limitations on the decision-making process, of course, have both benefits and costs. The benefit is that (if they work) the limitations result in decisions which, in general, are more consistent with long-term objectives. The cost is that, in some particular circumstances, the optimal decision (based on a proper balance between short-term and long-term considerations) may not be attainable because it is ruled out by the limitations.

The remainder of this report examines these issues with regard to the proposed constitutional amendment to require a balanced Federal budget, specifically: 1) what costs are associated with this approach?, and, 2) would it really achieve its objectives? Sections II and III discuss economic policy costs associated with the proposal. Section IV examines the experience with balanced budget requirements and debt limitations in the States to assess whether a Federal requirement is likely to achieve its objectives.

II. REACHING A BALANCED BUDGET UNDER THE AMENDMENT

Ratification of a constitutional amendment requiring a balanced Federal budget will not make achieving this goal any easier than it was before. The hard decisions on what to cut, whose interests to protect, and whose to sacrifice still must be made.

The Congressional Budget Office (CBO), in its January 1992 assessment, projected that the Federal deficit under present policies would decline from $352 billion in FY92 to $178 billion in FY96 and then rise again to exceed $200 billion in FY97 and thereafter. This "baseline" projection assumes that several revenue-raising measures expire as scheduled in 1995 and 1996 and that discretionary spending, no longer limited after FY95 by the Budget Enforcement Act of 1990, resumes rising with inflation. ⁷

This is the projected deficit that would have to be eliminated under the proposed constitutional amendment. For purposes of this analysis, we assume that the amendment requires the Federal budget to be balanced by FY96.

A. BASELINE PROJECTIONS BY MAJOR BUDGET ELEMENTS

The general composition of spending from FY92 through FY97, as projected in CBO's January baseline, is shown in Table 1. Certain characteristics of these projections should be noted. First, under the Budget Enforcement Act of 1990, so-called discretionary spending (spending under annual appropriations) is scheduled to be reduced by about 1-1/2 percent per year in FY93 and FY94 and to remain virtually unchanged in FY95. In terms of real purchasing power these appropriations would decline faster. Discretionary spending includes defense spending, however, which is targeted for substantial cutbacks. Some substitution of domestic discretionary spending for military spending is expected to take place under these limits.

 TABLE 1. Baseline Federal Outlays, Receipts and Deficits Projected

Net interest payments are projected to increase by 6-1/2 to 7 percent per year, assuming that interest rates on three-month Treasury bills go no higher than 5.6 percent in the mid-1990s and that rates on ten-year Treasury notes remain at their 1992 level. Mandatory spending and interest are the only two major categories of spending projected to rise as shares of GDP in the baseline projections. Receipts meanwhile are projected to rise by roughly 7 percent per year during the economy's rebound from recession in FY93 and FY94 but then to slow to 5.4-percent annual gains as economic growth subsides to a sustainable rate.

B. BALANCING THE BUDGET IN FY96 WITH UNIFORM PERCENTAGE SPENDING CUTS AND TAX INCREASES

Table 2 shows what would be required to move to a balanced budget by FY96 from the projected baseline budget using uniform percentage tax increases and spending cuts in all major categories. If the entire budget is on the operating table, so to speak, spending cuts and tax increases averaging 6 percent from CBO's baseline would suffice in FY96; 7.3 percent in FY97. The numbers appearing for FY93 through FY95 are intended to illustrate a steady phase-in of the cuts required by FY96.

      TABLE 2. IMPACTS OF BALANCING THE BUDGET BY UNIFORM PERCENT

Spending would be cut by $87.9 billion, bringing it down likewise to 20.3 percent of GDP from an FY96 baseline level of 21.5 percent and from 24.9 percent at present. Despite the uniform percentage changes in spending and taxes to comply with the constitutional amendment, most of the change from today's GDP shares would be made on the spending side. This fact results in part from the restraints on discretionary spending in the Budget Enforcement Act of 1990, which are embodied in the baseline projections.

These spending cuts by FY96 would include $33.3 billion from the baseline level of discretionary spending, $21.4 billion from social security, and $33.2 billion from other mandatory spending such as Medicare, Medicaid, Federal employee retirement programs, food stamps, income support for the poor, and veterans' and farm programs. Offsetting receipts would be raised by $4.6 billion. The only outlays not slashed would be those for interest payments and deposit insurance.

Required spending cuts and tax boosts actually could be somewhat smaller than outlined here because such major deficit reductions should lower Federal interest payments concurrently as the amount of Federal debt to be financed falls below baseline assumptions and as monetary policy is eased to offset the impact on the economy of the dramatic fiscal contraction. On the other hand, budget forecasts in the past often have proven to be too optimistic, and drastic budget cuts themselves call into question the plausibility of steady economic growth. ⁸ If the baseline budget forecasts should prove overly optimistic, then the spending cuts and tax increases required to reach a balanced budget by FY96 would be greater than indicated.

C. IMPLICATIONS OF EXEMPTING TAXES OR CERTAIN SPENDING CATEGORIES FROM BUDGET-BALANCING MEASURES

Complicating deficit reduction considerably are the desires of President Bush and various groups in Congress to cut the deficit while placing major components of the budget off-limits for change. One major group, including the President, vows to avert any changes that would increase taxes, placing the entire burden of deficit- cutting on Federal spending. The proposal for a constitutional amendment introduced by Rep. Barton (Texas) and Senator Kasten (H.J. Res. 248/S.J. Res. 182) would require, for instance, that receipts not increase faster than national income unless three-fifths of the whole membership of each House of Congress approves a faster increase. Baseline receipts are projected to rise in step with income. Another large group, also including the President, wishes no reductions in social-security benefits. Obviously putting parts of the budget off-limits for deficit-reduction purposes would require correspondingly larger changes in the remaining parts.

Table 3 shows the significance of such exclusions in terms of the percentage cuts that would be required in other parts of the budget. The numbers entered for FY93 through FY95 are again merely illustrative of a four-year phase-in. Sparing social security from reduction would require only moderately larger changes in other accounts than in the case with no restrictions. Social security will be about 22 percent of Federal spending by FY96 and, although it is very large for a single program, exempting it would reduce the base for across-the-board deficit reduction by only about 12 percent. Excluding tax increases from the budget-balancing package, however, would nearly double the size of required spending cuts relative to the unrestricted scenario. Cuts of this size would reduce spending by 2.4 percent of GDP from the FY96 baseline level and bring it down to a 19.1-percent share of GDP, a share lower than in any year since 1966.

       TABLE 3. IMPACTS OF BALANCING THE FEDERAL BUDGET BY FY96

Table 4 indicates what these cuts would mean in dollar terms. Cuts in discretionary programs would have to reach more than $80 billion in FY96 and more than $100 billion in FY97, nearly 2-1/2 times the size required in the case without restrictions. Mandatory spending other than social security and deposit insurance obligations would have to be reduced by about the same amounts. Needless to say, such deep cuts would necessitate basic restructuring of the programs affected. This would be especially true for the discretionary programs, for which baseline appropriations already involve reductions in purchasing power under the Budget Enforcement Act of 1990. If inflation proceeds at the 3 to 3.4-percent rate assumed in CBO's economic forecast, the purchasing power of discretionary budgets would be reduced by an average of 25 percent between FY92 and FY96 and nearly 28 percent by FY97. Discretionary appropriations fund multifarious administrative and program functions of the executive branch, including defense, law enforcement, transportation, education and training, and health research, as well as the operations of Congress and the Judiciary.

           TABLE 4. IMPACTS OF BALANCING THE BUDGET WITHOUT

Hence, this budget-balancing program would impose by far the largest and most persistent multiyear contraction of the Federal budget in history and can be expected to seriously disrupt Federal programs as well as private economic activity dependent on them, and to exercise a heavy drag on overall production and employment over this period. While the Federal Reserve would attempt to offset this drag via stimulative monetary policy, it may well not be able to counteract promptly such a rapid swing in the budget's balance. This fear calls into question the plausibility of the healthy underlying economic scenario assumed here if a balanced budget is phased in over only four to five years, and provides arguments in favor of a longer phase-in period. After the phase-in is complete, a balanced Federal budget would have long-run benefits for the economy as outlined on page 1 above.

III. EFFECTS OF A BALANCED BUDGET REQUIREMENT OVER THE LONG TERM

The balanced budget amendment is based on the premise that, except in special and rare circumstances, the appropriate objective of Federal fiscal policy is balance or surplus in the total budget each year. This objective raises at least two issues: 1) Should the budget be balanced every year?, and 2) Is the deficit in the total budget the appropriate variable to control? This section of the report examines these issues.

A. A CHANGE FROM MACROECONOMIC STABLIZATION TO DESTABILIZATION?

Changes in the Government deficit have macroeconomic effects; generally, an increase in the deficit stimulates the economy and deficit reduction restrains the economy. Based on this relationship, fiscal policy was used actively during the 1960s and 1970s in attempts to counter the business cycle. Tax cuts and increased spending were used to try to stimulate the economy out of a recession or a period of slow growth; opposite policy changes were used to try to dampen inflation. For a variety of reasons these policies were not as successful as had been expected or hoped, and active counter- cyclical fiscal policy has largely been abandoned during the last decade. ⁹

The Federal budget still includes important AUTOMATIC stabilizers, however. Automatic stabilizers operate without requiring any policy decision to activate them. Unemployment compensation is an example of an automatic stabilizer. As the economy slows or turns down, the program automatically increases Federal spending as more unemployed workers apply for benefits. Welfare and other assistance programs have similar effects. The tax system is also an automatic stabilizer. As the economy enters a recession, income, wages, and sales transactions decrease, reducing revenue from the corporate and individual income taxes, the social security tax, and excise taxes. Income taxes are particularly responsive to changes in the economy.

The automatic operation of these stabilizers is important. One of the reasons active counter-cyclical fiscal policy did not work as well as expected is the difficulty of determining the state of the economy early enough to counter its cycles. Economic indicators are imprecise and are available only with a lag. It is never clear how lengthy and severe a recession will be. The process of formulating and enacting a budget policy response is also involved and time- consuming. Consequently, by the time an economic downturn is diagnosed and a policy response is developed and implemented, the recession can be over (this happened with the 1975 tax cut, for example).

The automatic stabilizers do not require knowledge of the state of the economy. They respond directly to economic changes. The Congressional Budget Office estimates that a one-percentage-point decrease in the growth rate of real GDP would decrease tax revenues by $22 billion and increase outlays by $4 billion (a deficit increase of $26 billion) in the first year. ¹⁰ A one-percentage-point increase in unemployment, which would entail a larger slowdown in economic growth, would reduce revenues by $42 billion and increase outlays by $8 billion (a deficit increase of $50 billion). The automatic deficit increase resulting from an economic slowdown helps moderate the downturn.

Since the budget deficit changes automatically as the economy moves through different stages of the business cycle, the actual deficit at any given time is not purely an indicator of budget policy. The deficit reflects both budget policy and the state of the economy. It is useful to have a measure which separates these two effects. The so-called standardized-employment deficit was devised for this purpose. It is an estimate of the deficit under current policies if unemployment were maintained at a level consistent with a constant inflation rate. If budget policies remain unchanged and the economy turns downward, the actual budget deficit will increase but the standardized-employment deficit will remain constant.

From the perspective of stabilization policy, the appropriate variable to control is the standardized-employment deficit, not the actual budget deficit. To illustrate this point, consider the following hypothetical example. Assume the unemployment rate is 5.6 percent (the rate currently estimated to be consistent with stable inflation). Further assume that the budget is balanced at this unemployment rate, which implies that both the actual deficit and the standardized-employment deficit equal 0.

Now assume that economic growth slows and unemployment increases 1 percentage point to 6.6 percent. According to the CBO estimates, the budget deficit will increase to $50 billion as a result of the slowdown. The standardized-employment deficit, however, will remain at 0 because budget policy has not changed. To reduce the budget deficit to 0 would require a combination of spending cuts and tax increases of $50 billion; these budget changes will retard economic activity and FURTHER INCREASE the unemployment rate. This further increase in unemployment, in turn, will result in a new budget deficit which will also have to be offset to return the deficit to 0. This sequence will continue until a new equilibrium is reached. At that point, the economy will be more depressed and unemployment will be higher than if budget policy had been left unchanged. That is, A REQUIREMENT TO AVOID A DEFICIT INCREASE AS THE ECONOMY SLOWS IS DESTABILIZING, RATHER THAN MODERATING AN ECONOMIC DOWNTURN, BUDGET POLICY WOULD MAKE IT WORSE. To avoid this outcome, the standardized- employment deficit would have to be kept at a constant level, not the actual budget deficit.

It might be argued that such destabilizing effects could be avoided through use of the escape clause in the proposed constitutional amendment which would allow for a deficit upon a vote of 60 percent of the Members of each House of Congress. In practice, however, this is likely to work imperfectly because, in effect, it converts the automatic stabilizers into instruments of discretionary fiscal policy. As mentioned above, experience in the 1960s and 1970s indicated that discretionary fiscal policy was not very successful, and one of the reasons was the inability to recognize and respond to economic changes in a timely manner. This problem would be compounded by the 60-percent vote requirement. Furthermore, passage of the constitutional amendment would probably add considerable stigma to a budget deficit and make Members reluctant to vote for a deficit without very clear evidence of an economic downturn. ¹¹ The likely result is a less stable economy.

The increased instability would place more of a burden on monetary policy. Monetary policy can also help stabilize the economy by making credit cheaper when the economy slows and more expensive when it overheats. But monetary policy also has limitations. Discerning the state of the economy is no less difficult for the Federal Reserve Board than it is for the Congress, so the recognition lags that affect discretionary fiscal policy also affect monetary policy. Once monetary policy is changed, moreover, its effect on the economy occurs only with a long lag. Hence, while monetary policy may be able to offset some of the increased economic instability resulting from a balanced budget requirement, it should not be expected to offset all of it.

In sum, the balanced budget constitutional amendment would largely nullify the automatic stabilizers in the budget that expand outlays and reduce tax collections in economic slowdowns and recessions. The budget would no longer serve to moderate business cycles, leaving stabilization to monetary measures by the Federal Reserve. Lack of prompt automatic countercyclical mechanisms is likely to entail some loss of jobs and income through more severe recessions.

B. IS IT THE RIGHT DEFICIT TO CONTROL?

A balanced budget requirement raises at least two issues regarding the appropriate variable for control. The first involves government investment. The second involves special adjustments in the deficit measure. These issues are discussed in turn.

1. THE ISSUE OF GOVERNMENT INVESTMENT

Total outlays of the Federal Government include not only expenditures for current services -- such as funds for social security, health care for the elderly, unemployment compensation, and farm subsidies -- but also outlays for investment projects -- such as the construction of roads, dams, and office buildings, or the procurement of military equipment. Most finance theory (both public finance and private finance) suggests that it is appropriate to finance current services from current revenues but it is inappropriate to finance investment projects from current revenues.

Efficient allocation of resources requires that costs and benefits be matched so that those receiving the benefits of a service must pay its cost. ¹² For investments undertaken by private businesses, this matching is achieved by depreciation, which accounts for the cost of gradually using up capital assets in the production process. Depreciation is included as a cost of production when businesses determine prices to charge for goods and services. As a capital asset is depreciated, revenue equal to the amount of depreciation may be recouped to repay its financing (debt and equity financing in the case of private businesses). By the end of the asset's useful life (when it will be fully depreciated), its financing will be fully repaid. By this procedure, the cost of a business asset is spread among all the consumers of the goods or services produced by the asset.

While depreciation accounting is not generally applied to government investments, similar principles apply with regard to efficient resource allocation. If the government invests in a machine, a building, a road, or an airplane, for example, matching the cost of its use with the services it provides requires spreading the cost over the asset's useful life. The Federal Government makes no systematic attempt to achieve this objective. The States in some cases more closely approximate matching costs and benefits for capital assets through two fiscal devices: a capital budget and debt finance tied to asset lives.

A governmental unit that utilizes a capital budget divides total outlays into two components: current services and capital outlays (investments). While the current services component of the budget is to be financed by current revenues, the capital component is financed by debt with maturity approximating the useful lives of the investments (a 20-year bond issue sold by a highway authority to finance construction of a highway expected to last about 20 years, for example). ¹³ This results in charges for capital services in the public sector that are analogous to depreciation in the private sector (State budget practices are discussed in more detail below).

The Federal Government does not follow these procedures. There is no attempt to match the size of the Federal deficit each year with the amount of investment outlays. Furthermore, the length of maturity of Treasury bonds sold to finance the deficit is based on assessment of financial market conditions, not the useful lives of Government investments. ¹⁴

Use of a capital budget by the Federal Government has been proposed several times in the past, but has never been adopted. ¹⁵ The issue may be considered again, however, within the context of a balanced budget constitutional amendment as a means of preserving debt financing for investment outlays. It could be argued that, if taxpayers are required to pay at the outset the full cost of public investments that will provide services over a long period, underinvestment will be the likely result. While this argument is valid, its practical significance is greatest when investment varies substantially from year to year or when a large increase in investment is being considered. ¹⁶ A significant increase in Federal infrastructure investment, for example, is probably less likely if it must be financed from current tax revenues than if it can be deficit financed, as would be done in an analogous case by a private company or even by States that have balanced budget requirements.

Assuming the concept of capital investment were defined narrowly, a Federal capital budget with a requirement to balance the current services budget would compel substantial deficit reduction. The increase in debt to finance investment should match NET investment, which equals gross investment (investment outlays) minus depreciation. This raises a technical problem in capital budgeting: depreciation would have to be calculated for Federal government assets. While estimates are currently made for depreciation of physical capital and research and development, the data on which they are based are incomplete and imprecise, so the estimates are regarded as only approximations. In FY93, net Federal investment in physical capital, including military capital, is expected to be about $6.7 billion, and net Federal investment in R&D is projected to be $17.8 billion, a total of $24.5 billion. ¹⁷ For these two categories of investment, therefore, a capital budget would justify less than 10 percent of the current Federal deficit. Of course, a major increase in Federal investment spending would raise the level of debt financing under a capital budget, at least during the initial years.

Opponents of a capital budget fear that pressure would build to expand the concept of "investment" and dilute the impact of a balanced budget requirement. The concept of investment spending by the Federal Government is vague. Building highways, dams, and airports is clearly investment. But what about spending for education, training, and health programs for children and young adults? This type of spending is not the prototypical investment in buildings, machinery, and equipment as normally thought of in the business world. Nonetheless, spending on education, for example, creates "human capital" which will yield "dividends" in the form of greater productivity, higher earnings, and a more educated citizenry in the future.

Similarly, opponents fear that advocates of new Federal programs might be successful in arguing that the programs belong in the capital part of the budget rather than the current services part and thereby avoid the discipline of having to pay for the programs on a current basis (the discussion in the next section indicates that this has been a problem in the States). If the balanced budget amendment were adopted and applied only to the current services portion of the Federal budget, frequent struggles over what would be included as a capital expenditure would be likely.

It is difficult to surmise the effect of a broader concept of investment because no depreciation estimates are available for education, health care, and other such types of human capital. The area most likely to be included in a broader concept is probably education and training. Federal outlays for these functions in FY93 are projected at $42.5 billion. Allowing for an offset for depreciation, even if education and training were included as investment, a capital budget would still require substantial deficit reduction.

To summarize, it is appropriate to debt-finance investment projects so the cost will be spread over the period during which the investments provide benefits. A requirement to finance all government spending, including capital spending, from current revenues would probably make significant new capital investment programs -- such as a large increase in infrastructure investment -- less likely, even if they are economically justified. While the distinction between current operations and capital spending is difficult to determine for some Government programs, use of a capital budget and allowing debt finance normally only for investment outlays would still require very substantial deficit reduction from current levels.

2. SPECIAL ADJUSTMENTS TO THE DEFICIT MEASURE

The total Federal budget deficit for FY93 is projected by the Congressional Budget Office at $327 billion. It can be argued, however, that this is not the most meaningful figure in terms of gauging the economic effects of the deficit.

A commonly made adjustment is to exclude the costs of deposit insurance from the deficit (the costs of closing or merging failed savings and loan institutions and banks). These costs increase the deficit but their effects on the economy are really more like those of debt refinancing than new debt, because they involve paying off existing obligations (the insurance backing for depositors in the failed institutions) rather than creating new ones. The outlay for deposit insurance in FY93 is expected to be $68 billion; the deficit excluding this amount is $258 billion. The deposit insurance exclusion will not always reduce the deficit, however. Eventually, the proceeds from selling the institutions, loans, and properties taken over by the government will exceed the costs of closing additional institutions, and the revenues generated by the operation will reduce the deficit. In FY96, for example, CBO estimates that excluding deposit insurance would increase the deficit by $44 billion.

Another adjustment suggested by many analysts to obtain a more meaningful figure for policy purposes is to exclude the trust funds from calculation of the deficit. Prior to the mid-1980s, the trust funds had a relatively small effect on the deficit in the total budget. Currently, and for the next two decades, however, the retirement trust funds (social security, military retirement, and civilian retirement) are building up large surpluses in anticipation of retirement of the "baby boom" generation beginning about year 2010. The total surplus in the trust funds in FY93, for example, is expected to amount to about $120 billion. If the total Federal budget were balanced, this surplus in the trust funds would be matched by a deficit in the rest of the budget, and the budget would have no net effect on the credit markets.

Many analysts argue, however, that the surpluses in the trust funds should be used to add to total savings in the economy, increasing investment and productive capacity, so that higher future output will be available to help finance the future retirement costs. Achieving this objective would require excluding the trust funds from the deficit calculation and balancing the rest of the budget. The deficit excluding the trust funds is projected at $446 billion in FY93; if deposit insurance is also excluded, the projected deficit is $378 billion.

Other adjustments to the deficit measure may also be appropriate. Some analysts, for example, maintain that effects of inflation should be accounted for in measuring the deficit. ¹⁸ As another example, CBO excludes the contributions from foreign governments to help defray the costs of the Desert Storm operation to obtain a better measure of long-term deficit trends.

It is not necessary to conclude here which adjustments yield the most useful measure of the deficit. The point is that legitimate arguments can be made that the total Federal deficit is not the most meaningful number for policy purposes. It appears that no adjustments of the deficit measure would be allowed in the proposed constitutional amendment; the requirement to balance the budget refers to the unadjusted deficit in the total budget. A requirement to balance the budget based on a deficit measure that is inconsistent with the economic effects could result in the budget stimulating or restraining the economy (depending on the nature of the inconsistency) when the intention is for policy to be neutral.

3. AN ILLUSTRATION: THE CASE OF DEPOSIT INSURANCE

Payments ensuing from Federal deposit insurance provide a dramatic illustration of both the destabilizing potential of a balanced-budget amendment and the case for defining with care the concept of the budget that is to be balanced.

Rising Federal payments to protect depositors of failing thrift institutions and banks do not stimulate the economy like rising Federal payments for goods or services. Instead they avert a contraction of the money supply and the economy that would occur if deposits were wiped out when banks failed. If it were not for depositors' trust in the Federal commitment to protect them from losses, mass withdrawals probably would have ruined still more institutions as capitalization at some declined to worrisome levels. As stated above, Federal financing to save failing thrifts and banks is more like refinancing old obligations than like new debt. It is for this reason that economists, including Chairman Greenspan of the Federal Reserve Board, have recommended that deposit insurance payments be financed by borrowing and not by raising taxes or cutting other spending.

If a balanced-budget requirement like those proposed had been in effect in recent years, rising outlays stemming from deposit insurance would have posed a case in which permitting unbalanced budgets for several years would have been highly desirable. If a waiver of the balanced-budget requirement had not been passed by the stipulated 60-percent majorities of the whole membership of each House of Congress, then attempting to cut other spending and/or raise taxes to finance these outlays would have imposed large contractionary forces on the economy of the kind that providing deposit insurance was designed to avert. These downward forces would have been reinforced by the income-multiplier effect, magnifying budget deficits, and requiring still larger spending cuts and/or tax increases unless approval for an unbalanced budget had been obtained.

To make the illustration even more daunting, one might note that the costs of protecting depositors increased sharply as economic growth slowed in 1989 and mushroomed as recession gripped the economy in 1990 and 1991. Even though some of the failing institutions had long since been insolvent, recession made additional borrowers unable to repay their loans and caused the value of collateral, especially real estate collateral, to fall. An illustration was offered in subsection III A above of the cumulative consequences of trying to balance the budget in the face of recession itself. An attempt to maintain a balanced budget in the face of soaring deposit insurance expenditures in addition would have made the recession more severe.

The cost of deposit insurance in the Federal budget rose by $44 billion from FY89 to FY91. At the same time, unemployment rose from 5.2 percent to 6.7 percent as the economy sank into recession. Based on the CBO estimates of the sensitivity of Federal revenues and expenditures to changes in the economy, this higher unemployment increased the Federal deficit by $72 billion. Hence, maintaining a balanced budget while confronting these two forces would have required tax increases and spending cuts aggregating $116 billion. This sizeable fiscal contraction, occurring over a 2-year period when the economy was turning downward into recession, would have resulted in a much steeper downturn.

In such a case deficits presumably would have been approved, hopefully on an adequate scale before great damage had been done. If so, a constitutional amendment might not have yielded smaller increases in deficits since 1989 than existing processes have done. The repeated inability or reluctance of Congress and the President to recognize the true size of the deposit insurance obligations, however, spotlights the potential difficulty they also could have in foreseeing the size of deficits that would need to be authorized in such circumstances.

IV. THE STATE EXPERIENCE WITH BALANCED BUDGET REQUIREMENTS AND DEBT LIMITATIONS

Almost all States have imposed constraints to assist in curbing the appetite of taxpayers and elected officials for financing consumption of current public services with future tax revenue. One of the two most important constraints imposed by the States is a balanced budget requirement, the other being a limitation on borrowing. No matter how persuasive may be the skeptics' gloomy predictions concerning the effects of adopting a Federal balanced budget requirement, ultimately they must confront the fact that States manage to operate successfully under such a requirement.

This enthusiasm for having the Federal Government follow in the wake of the State experience may reflect the fact that the actual operation of these State constraints is not well understood. In fact, the States' budget balancing requirements and constraints on debt issuance, and the response of State elected officials and taxpayers to these limitations, bear little resemblance either to the balanced budget requirement proposed for the Federal Government or to the hoped-for response of Federal elected officials and taxpayers.

As is to be expected with any group of 50 independent entities, the specific methods employed to implement these constraints are diverse. The purpose here is not to provide the reader with the details of this diversity, but rather to examine the States' experiences with these constraints for relevance to the proposed Federal policy.

A. HOW DO THE STATE BALANCED BUDGET REQUIREMENTS WORK?

What does a balanced budget constraint mean? In a literal sense, the answer is that this year's government spending must be financed with this year's revenue. If no distinction is made between spending for public services that are consumed this year and spending for capital facilities that will provide services over many years, then all capital facilities must be financed from current tax revenue.

Most States have not chosen to subject themselves to such a constraint, because it would require them to forego the use of borrowing for the financing of capital services. State taxpayers lay claim to the benefits from capital facilities by dint of residency and relinquish their claim to benefits when they move. Given the demands a market-oriented society places on labor mobility, taxpayers are reluctant to pay today for State capital services to be received in the future. The rational response of the State official concerned with satisfying the preferences of constituents is to match the timing of the payments to the flow of services, precisely the function served by long-term debt financing. Any attempt to pay for capital facilities "up front" is likely to result in a less-than- optimal rate of public capital formation.

Accordingly, the condition for State budget balance is not comparable to the proposed Federal equality between total revenue and total outlays. All States have established General Funds for which current expenditures cannot exceed current revenue plus any balance (surplus) carried over from previous years. All States have also established a multitude of other Operating Funds or Special Funds whose balance can include borrowed funds. Thus, in practical terms the purpose of State balanced budget requirements is to prohibit or discourage the financing of current consumption with long-term borrowing. Put another way, current consumption is not to be financed with future tax revenues.

This is precisely the behavior one would expect from the discussion of a capital budget presented earlier in this report (see section III B 1). In theory, the General Fund should account for spending on current services and the taxes and fees raised to finance this spending, usually including interest on the outstanding debt. Included in the Special Funds should be tax and fee revenue equal to the dollar value of the current services being provided by capital facilities, what in accounting terms is called a depreciation allowance. These taxes and fees are being used to retire a portion of the principal on the long-term bonds issued to finance the construction of capital facilities.

An additional deviation of State budget practice from the proposed Federal requirement to equalize total revenue and outlays is the use of short-term borrowing. States are faced with the necessity of planning their General Fund budget for the year (or in some cases for two years). This requires a balancing of revenue forecasts against forecasts of the demand for services and spending. Not infrequently, unforeseen circumstances cause a revenue shortfall which must be financed with short-term borrowing (repaid within one year). Technically, the General Fund is balanced only because funds were borrowed. Because the borrowing is short-term, the next year's budget must appropriate money to retire the short-term debt.

Such short-term borrowing cannot be viewed in any realistic sense as a violation of a balanced budget requirement. Without such borrowing, the only policy options to balance the budget are mid-year reductions in services or increases in taxes and fees. It is likely that short-term borrowing is much less disruptive to the orderly functioning of the public sector. A pattern of persistent and growing short-term borrowing for the General Fund may, however, be an indication of growing difficulties with balancing the budget, and may suggest a need to examine the State fiscal structure.

B. CIRCUMVENTION OF STATE BALANCED BUDGET REQUIREMENTS

This budget structure sounds impressive -- States balance the current services budget and maintain a capital services budget in which assets balance liabilities. Financing current services with future tax revenues is limited to budget gaps produced by forecasting errors, which are corrected by adjustments in the next fiscal year's tax revenue and spending. Unfortunately, the reality of State budget practices does not attain this ideal.

The share of total State budgets assigned to the General Fund varies substantially across States. In 1990, Connecticut included 70.8 percent of its total revenues and 71.8 percent of its total expenditures in its General Fund. In contrast, Wyoming included only 21.9 percent of its total revenues and 21.0 percent of its total expenditures in the General Fund. The national average for the 50 States was 52.6 percent of total revenues and 54.3 percent of total expenditures. The actual division of the budget between current and capital services among States is unlikely to vary between 71 percent and 22 percent. Thus, these numbers imply that some current services probably are being provided in the various Special Funds, which suggests that some current services might be financed with long-term borrowing.

Pension funds for State employees provide a good example of the potential for financing current services outside the General Fund. Part of the compensation package for State employees is a pension, to which the State makes some contributions. A balanced General Fund budget would require full funding of the current services State employees provide to taxpayers. This would require a tax-financed pension Fund payment whose future value (after investment earnings) equals the expected future pension liabilities earned by State employees during the fiscal year. If the tax contribution is too small, the pension Fund has unfunded accrued liabilities. The funding shortfall does not appear as an unbalanced General Fund budget, and current services are financed with future tax revenue. The Federal balanced budget proposal would not prohibit such practices.

The Pension Fund can also be used to illustrate another method States use to circumvent their balanced budget requirement. Suppose the Pension Fund has a substantial cash surplus (which can occur even if unfunded accrued liabilities are present). Suppose a recession causes tax revenue to decline unexpectedly and threatens the provision of essential public services. And suppose the State's precarious financial condition makes borrowing expensive and accompanied by belt-tightening dictates from lenders. States have been known to use the accumulated surplus in the Pension Fund (or some other fund with a surplus) to make loans to the General Fund, thereby avoiding the necessity of borrowing from the private capital markets. This technique also enables current services to be financed with long-term borrowing.

What this brief discussion tells us is that the States are well aware of the dangers of allowing the taste for current consumption to be financed with future tax dollars. Accordingly, these future tax dollars are protected with a balanced budget requirement. However, the absolute protection seemingly promised by the requirement is more apparent than real. The ebb and flow of popular sentiment and powerful interests have produced State budgetary systems that do not classify all current consumption in the General Fund, making it possible to finance these current services with borrowed funds. And should a particularly popular current service in the General Fund be threatened by temporary economic difficulties, the budgetary systems have the flexibility to borrow from the Special Funds to finance current consumption in the General Fund, thereby "hiding" some long- term borrowing for current consumption.

C. CONSTITUTIONAL AND STATUTORY RESTRICTIONS ON STATE BORROWING

As indicated by the discussion in the preceding section, taxpayers have good reasons to permit State governments to borrow funds -- for efficient financing of capital services, for short-term adjustments to forecasting errors, and to circumvent their own balanced budget requirements in hard economic times. This is reflected in data on total outstanding State debt, which has grown steadily from $353 million in 1870, to $21.6 billion in 1962, and to $315.5 billion in 1990. This growth has occurred in spite of the fact that almost all States impose some restriction on debt issuance, whether it be a requirement for a referendum, a legislative vote by a majority or a super-majority, or a dollar limit on the outstanding stock of debt. ¹⁹

This growth has occurred for two reasons. First, although some may think debt restrictions were adopted to prohibit debt financing of public capital services, the historical record does not support this view. That record suggests the restrictions are a legacy of the nineteenth century reaction to defaults and repudiations of bonds issued to finance PRIVATELY OWNED investments in roads, canals and railroads. ²⁰

Second, growth has occurred because taxpayers, elected officials, and, most importantly State courts have devised ways to circumvent the restrictions on debt issuance. Thus, the lessons from the State experience with debt restrictions, should the balanced budget amendment now being considered by Congress be adopted, are (1) that debt restrictions ignoring the reality of the public's taste for consumption of capital services will be circumvented, and (2) that the courts will play a much more important role in making decisions about financing capital services.

D. CIRCUMVENTION OF STATE DEBT RESTRICTIONS

The first step in circumventing State debt restrictions was for State courts to rule that the restrictions apply only to guaranteed debt -- general obligation bonds whose debt service is backed by the taxing power of the issuing government. Nonguaranteed debt is not subject to restriction, and depends upon a revenue source other than taxes. Nonguaranteed debt grew from 14.2 percent of total outstanding State debt in 1942, the first year for which data are available, to more than 50 percent in 1962, and to 76 percent in 1990.

This nonguaranteed debt usually is financed with fees and charges for the use of the capital facility being built with the bond proceeds. Examples are hospitals whose debt service comes from payments for use of the hospital, room and board payments from students living in college dormitories, and tolls from toll roads. This financing technique seems quite reasonable and might even represent an improvement in economic efficiency. It substitutes the benefit principle of taxation for the ability-to-pay principle, which is appropriate when the benefits from the public service are enjoyed primarily by those directly using the public service.

But not all capital projects needing financing in States constrained by debt limits are suitable for user-charge financing. This has led to the issuance of a considerable amount of nonguaranteed debt that is actually a subterfuge for guaranteed debt. One commentator has suggested such debt is nonguaranteed "only upon tortured reasoning" by the State courts. ²¹ For example, a State building authority may issue bonds to build a facility and finance debt service from rental payments made by the State's General Fund. In some States, such tax-supported bonds are considered to be nonguaranteed because rents are not debts as a matter of common law. In other States, the courts have held that the rental payment is a contingent State obligation that must receive an annual appropriation, which means it is not a debt and therefore not subject to the debt limitation.

One study of State borrowing restrictions concludes that State ". . . courts have transformed absolute constraints on legislative power to incur debt into more flexible limitations that leave the judiciary with a substantial role in determining the fate of proposed borrowing schemes." ²²

A second technique has been used by the States to circumvent debt limitations that prohibit financing of capital services desired by taxpayers. States moved a considerable amount of the public's business out of the budget and into special districts, special funds, and what the Census Bureau classifies as "dependent agencies" of governmental units ("constituted authorities" specifically authorized under State law, or corporations formed under general State nonprofit corporation law). These authorities and corporations are often managed by appointed boards rather than elected officials. Questions have been raised about the extent to which the public manages to exercise control over the public business being conducted by these quasi-public entities. ²³

Thus, the State experience suggests that if Federal taxpayers decide necessary investments in physical capital cannot be financed with the borrowing authority permitted by the balanced budget amendment (equal to the amount freed by debt retirement), the Federal Government may begin to use the same techniques employed by States to circumvent State borrowing restrictions. These techniques involve issuing debt that is not guaranteed by the tax base, and setting up off-budget agencies to issue debt. This circumvention will not come cheaply. First, elected officials may be required to sacrifice a considerable degree of direct control over capital service financing decisions. Judges on Federal courts may become arbiters of the legality of the nonguaranteed Federal debt and of the dependent agencies created to issue this debt. And these dependent agencies and authorities may have to be controlled by appointed and relatively independent boards that are conducting the public's business off budget. Second, higher interest costs are inherent in issuing debt that lacks the backing of the public's taxing power.

² "Total receipts" are defined to exclude receipts from borrowing, and "total outlays" exclude outlays for repayment of debt. Hence, the Federal Government could continue to borrow up to the amount of debt retirement each year without violating the restriction.

³ H.J. Res. 248 and S.J. Res. 182 propose similar balanced budget constitutional amendments with one additional requirement: a vote of a three-fifths majority of the whole number of each House would be required to allow total receipts to increase by a rate greater than the rate of growth in national income in the second prior fiscal year. A full analysis of the potential effects of this provision is beyond the scope of this report.

⁴ For discussions of some of these issues see U.S. Library of Congress, Congressional Research Service, Comparison of Two Proposed Balanced Budget Amendments, H.J. Res. 290 and S.J. Res. 18, general distribution memorandum, by James V. Saturno, April 28, 1992, 5 p.; and U.S. Library of Congress, Congressional Research Service, H.J. Res. 268, a Proposal for a Constitutional Amendment to Require a Balanced Budget: Potential Questions of Interpretation, general distribution memorandum, by James V. Saturno, July 10, 1990, 11 p.

⁵ Richard H. Thaler and H. M. Shefrin, An Economic Theory of Self-Control, Journal of Political Economy, April 1981, p. 392-406.

⁶ Areas of behavior other than the consumption/saving choice show similar patterns. Many programs for dieting or quitting smoking, drinking, or drug habits, for example, involve individuals voluntarily placing themselves in situations in which outside constraints are imposed on their choices.

⁷ U.S. Congressional Budget Office, The Economic and Budget Outlook: Fiscal Years 1993-1997, January 1992, p. 27-30. Note: the Budget Enforcement Act of 1990 is Title XIII of the Omnibus Budget Reconciliation Act of 1990.

⁸ For further discussion of this question, see p. 12 below.

⁹ See the discussion in U.S. Library of Congress, Congressional Research Service, Tax Cuts and Rebates for Economic Stimulus: the Historical Record, Report No. 92-20 E, by Donald W. Kiefer, January 2, 1992. 11 p.

¹⁰ The CBO estimate assumes the economic change would begin in January 1992; the revenue and outlay estimates are for fiscal year 1993, which would begin nine months after the change. See The Economic and Budget Outlook (cited in footnote 7).

¹¹ Another potential approach to avoiding the requirement to offset the effects of the automatic stabilizers in a downturn would be to run a large budget surplus, say $100 billion, when the economy is operating at full employment. In a downturn, the automatic stabilizers would reduce the surplus, thereby stimulating the economy, but would not result in an immediate deficit requiring offsetting fiscal action. There are arguments in favor of maintaining a surplus (see the discussion in subsection B 2 below), at least during the next 20 years or so. Once the "baby boom" generation begins retiring, however, a budget surplus or even a balanced budget may be inappropriate policy. Given recent history, a continuous large surplus in the budget seems unlikely.

¹² This statement ignores any redistributional objectives.

¹³ As discussed in more detail below, the balanced budget requirements in the States apply to the current services portion of their budgets, not the capital component and not the total budget.

¹⁴ It is possible that the term structure of Federal debt could approximate the average useful life of Government assets, but this would only be coincidental.

¹⁵ While the Federal budget has never been divided explicitly into current services and capital components, during some periods a distinction was made between "ordinary" and "extraordinary" expenditures. The extraordinary category was never well defined, however, and included at one time or another investment outlays, non- recurring expenditures, relief payments, some defense outlays, and programs for economic stabilization. To move away from the concept of separate budgets for different purposes, the unified budget approach was adopted in the late 1960s. See the discussion in Herbert Stein, The Fiscal Revolution in America, University of Chicago Press, Chicago, 1969, p. 123-130.

¹⁶ If investment is stable, growing at a small nearly constant rate each year, there will be relatively little difference between the annual cost of paying for new investments when they are made versus paying for the depreciation of existing assets.

¹⁷ Source: Physical and Other Capital Presentation, Budget of the United States Government, Fiscal Year 1993, Table 29-2, p 3-38, and Table 29-4, p. 3-40. The data, which are reported in constant 1987 dollars, have been adjusted for the average inflation rate between 1987 and 1991 as reflected by the GDP deflator for fixed nonresidential investment. By way of comparison, Federal outlays for investment in physical capital in fiscal year 1993 are projected at $133.1 billion; research and development expenditures are projected at $75.1 billion, for a total of $208.2 billion.

¹⁸ For a discussion of this and several other potential adjustments, see U.S. Congressional Budget Office, The Federal Deficit: Does It Measure the Government's Effect on National Saving?, March 1990, 60 p.

¹⁹ A detailed listing of restrictions for 1986 is available in Advisory Commission on Intergovernmental Relations, Significant Features of Fiscal Federalism, 1989 Edition, Vol. 1, no. M-163.

²⁰ For a more complete account of the adoption of constitutional restrictions on debt issuance, see Dennis Zimmerman, Private Use of Tax-Exempt Bonds: Controlling Public Subsidy of Private Activity, Washington, D.C., The Urban Institute Press, 1991.

²¹ C.R. Morris, Evading Debt Limitations with Public Building Authorities: Costly Subversion of State Constitutions, Yale Law Journal, 1958, p. 234-68.

²² Stewart E. Sterk and Elizabeth S. Goldman, Controlling Legislative Shortsightedness: The Effectiveness of Constitutional Debt Limitations, Wisconsin Law Review, 1991, p. 1301-1367.

²³ Annmarie Hauck Walsh, The Politics and Practices of Government Corporations, Cambridge: MIT Press, 1978.