Current template: single.php

                                                                                                                                                November 7, 2018

A Budget Process without a Resource Constraint and Relevant Costs of Alternatives is Not a Budget Process

Reform of the federal budget process is a perennial topic of interest to analysts and policymakers. This interest is driven by a consensus view that the current budget process performs poorly in terms of the objectives of stability, efficiency, and equity.

Traditionally, analysts in market economies have looked to government for assistance in pursuit of those objectives, primarily by correcting market “failures,” or gaps between market outcomes and constrained, socially-optimal results. However, the practice of budgeting derives from universal scarcity: the beneficial uses to which available resources may be allocated vastly exceed the available limited supply of resources This limit applies to governments as strongly as it does to private institutions because every dollar that government spends must be paid for by someone, sometime. Alternatively, every unit of resources that government uses must be given up by someone. For all, scarcity implies the necessity of choice and a potential gain from systematically choosing an allocative path for resources that maximizes benefits from those resources, i.e. budgeting.

Recognizing the function of budgeting as constrained optimizations reveals its essential information requirements: the resource constraint, the opportunity cost of each considered alternative use of resources, and the expected social benefit from each use.  Although this “revelation” may seem utterly unexceptional and commonplace, the current federal budget process fails to provide this requisite information in salient form to elected policy makers.  Just as oddly, proposals to provide this information in recent years have aimed almost exclusively at improving the measure of benefit performance, as though the requirements for a well-defined constraint and relevant cost measures had already been addressed fully.

This paper demonstrates the absence from the current federal budget process of a relevant resource constraint and measures of cost.

No Constraint

Advancing the claim that the current federal budget process lacks an effective budget constraint risks running afoul of the admonition to “avoid statements with which no one would disagree.”  Nonetheless, the failure of most budget reform proposals to include this feature appears to imply either that such a constraint already exists or–more baffling–is unneeded.

One bit of evidence on this issue is the Congressional Budget Office’s annual long-term (30-year) projection of budget revenues, outlays, deficits and debt, based on the assumption of the indefinite continuation of current law and budget policy. The projections indicate that under those assumptions, debt grows as a share of GDP without identifiable limit. The figure below is the 2018 projection.

It was not always so. As indicated in the Figure below, for many decades the balanced budget norm was a dominate feature of federal budget policy. Deficits were permitted, but only during wars and other extraordinary emergencies. Once the crisis had passed, deficits were replaced by surpluses and the debt was paid down. Since the early 1970’s, however, federal debt has been on a long-term upward trend. Debt has become the favored means of financing increases in government spending and reductions in taxes in good times and bad.

The balanced budget norm imposed a constraining fiscal discipline on the federal budget process by limiting spending to amounts that constituents were willing to pay in taxes.  Both intellectual and political factors played a role in the abandonment of that constraint.

The Keynesian Revolution in fiscal policy–embraced by the US government during the Presidencies of Kennedy, Johnson and Nixon—added a new, major function to budgeting by sovereign governments: namely: to “balance the economy” at high rates of employment and economic growth. Thus, spending and the size of the deficit were to be determined primarily by policymakers’ assessment of the needs of the economy for fiscal stimulus to maintain high employment and growth. The elevated importance of this function necessarily relegated “balancing the budget” and restricting spending to a “willingness to pay taxes” to secondary status at best.

An important contribution to the assent of Keynesian macro-economic policy was economist Abba Lerner’s 1943 essay “Functional Finance and the Federal Debt,” Lerner proposed that the goal of macro-economic stabilization be pursued by a highly active budget policies; specifically that aggregate demand should be managed by frequent changes in federal spending and taxes to maintain high employment and price stability. One implication of the proposed policy was that deficits and debt would be determined by prevailing economic conditions rather than considerations of long-term fiscal balance.  To the extent that the ratio of debt to national income required stabilization, this could be accomplished by central bank control of the interest rate relative to the rate of growth of the economy: interest rate lower than the rate of economic growth would reduce the ratio (Mason and Jayadev. 2016). Further, to allay the concerns of traditional, conventional analysts, Lerner emphasized an essential difference between sovereign governments and other institutions: because government debt is denominated in the currency of the issuer, government can never be forced into an involuntary default on its debt, which it can always redeem with newly created money. Thus, government debt does not pose a threat to government’s solvency in the same sense that private debt can threaten private issuers. Government is not subject to a conventional budget constraint.

Lerner’s vision of functional finance and its implications for sustainable fiscal policies is now regarded as an artifact of a highly simplified model, unsuited to the politics of democratic decision making (Colander. 1997).  It ignores the long lags that hamper effective fiscal management of the economy and the harm to economic and political stability from a policy of repaying debt with an inflationary issue of newly created money.  Lerner’s postulated escape from a long-term debt constraint likely leads to a catastrophic loss of living standards and shock to social stability (Shaviro. 2009).

Nonetheless, the ghost of “functional finance” continues to haunt the fiscal and budget policy process. Policy choice continues to honor the thesis that the principle function of federal budgeting is to maintain the economy at full employment and promote economic growth.  With rare exception, the economy is seen as falling short or in danger of doing so. And the political imperative is to attend to the current need rather than to a distant future problem with federal debt, conveniently denominated in the currency of the realm.

A strategic political decision also played an important role in the abandonment of the balanced budget norm. In 1976, Jude Wanniski, published “Taxes and a Two Santa Theory” in the National Review, (Wanniski. 1976).  Wanniski described the national Democrats as the Spending Santa party and argued, that to survive as a competitive political entity, the Republicans would have to return to their 1920’s Harding-Coolidge Era role as the Tax Reduction Santa. Wanniski’s argument that the Republicans had become the politically-untenable dispensers of pain in raising taxes to pay for free spending enacted by the Democrats resonated among Republicans and sharply reduced the influence of GOP fiscal conservatives.  This appears to have swept away the last effective voice for a balanced budget norm. Deficits and debt now command bipartisan support as the preferred means of financing an increasing flow of gifts to constituents from the cornucopia of the federal fisc.  The unrecognized downside, however, is as Steuerle (2018) has noted, a loan to be repaid is not a gift.

The current federal budget process lacks an effective budget constraint. A budget process without such a constraint is not a budget process because scarcity is no longer an issue. Without scarcity, cost becomes meaningless because nothing has to be sacrificed to obtain desired benefits. Choice is necessary only, because of the scarcity of legislative time, about the sequencing of the provision of benefits.  But, in time, constituents can have it all.

The Universe of Nominal Budget Processes

Virtually every elected government—there must be a hidden exception—has a set of procedures which it describes as a budget process. In the US that process consists largely of a system inherited from historical practice, over-laid with periodic realignments of responsibilities and calendar changes. (Mikesell 2014, Ch, 3).

The core of that process is a cash-basis accounting, initially adopted in the earliest days of the republic when the role of government was much smaller and simpler than today. This accounting method measures inflows and outflows of cash to and from the government when those flows occur. It recognizes changes in one asset: cash and one liability: Treasury debt held by non-federal entities. All other assets and liabilities are excluded from the accounting. The accounting period is one year.

The summary annual budget aggregate flows under this accounting consist of tax revenues collected by the government, outlays disbursed[1], and the deficit or surplus, which is revenues less outlays. That deficit also articulates with the change in net position for a balance sheet consisting of holdings of cash assets and Treasury debt liabilities.

Cost estimates for proposed changes in federal spending or taxes consist of the net effect of those changes on federal revenues, outlays and the deficit for specific fiscal years.

Even though the budget aggregates are not subject to procedural constraint—other than majority agreement—policymakers exhibit a pronounced preference for lower costs and deficits, all else equal. As is well-known, cash accounting systems are well suited to managing reported costs, and especially to deferring cost recognition to future fiscal years. For example, assets sales conducted concurrently with increases in spending can offset the net effect of increases in outlays and the deficit.[2]  Shifting payments from the last day of a fiscal year to the first day of the next effectively lowers the deficit the current year. Collecting fees or taxes for benefits to be delivered later can mask the reported cost of spending increases in the year the collections are received.

Widespread use of such budget maneuvers to manage the salient costs of legislation means that the current federal budget process often lacks meaningful and relevant estimates of the opportunity cost of enacted policies.


Illustrative Cases of “Managed” Budget Cost

Here I offer three cases of strategically mis-measured budget costs whose accounting remedies are well-defined: a more comprehensive measurement focus (what is measured); basis of accounting (when measured); and consistent application of the current, functionally derived “matching principle” for budgetary accounting.


Federal Assumption of District of Columbia Defined Benefit Pension Plans

Although this example is about 20 years old, it provides a likely irresistible precedent for the not-too-distant-need for federal assistance to failed defined benefit plans of states and political sub-divisions.  It is also a case where the “true,” or at least straightforward, cost of a policy action is apparent on inspection of the legislation.

The context for this case is that in the late 1990s, the government of the District of Columbia was under great stress such that the prospect of default of its public debt was a significant risk.  The financial condition of DC was inextricably related to past federal policies affecting the local government, especially before home rule. This linkage was a major factor creating a sense of implied responsibility by the federal government for the District’s dire condition. An obvious point of fiscal stress was the significantly under-funded pension plans for Police, Firefighters, and Teachers, which held assets of $3.2 billion against about $9 billion in earned retirement benefits leaving a shortfall of $5,8 billion.

The most transparent option for federal assistance would have been for the federal government to provide a grant of $5.8 billion to the pension plan.  However, cash payment of that amount would have had a cash-basis budget cost estimate of $5.8 billion. Sponsors were interested in framing legislation to provide the same assistance but at a much lower reported budget cost.

In the Federal Assumption of DC Pension Plans (Title XI, Balanced Budget Act of 1997, P.L. 105-33), Congress found such an approach.  Under that Act, the Federal government assumed responsibility for paying all pension benefits earned to-date. No further benefits could be earned in the transferred plans.

The 10-year CBO cost estimate for this legislation reported zero net effect on mandatory spending for the first 8 years of federal payments, FY 1998-2005, because new pension outlays were expected to be offset by the proceeds from annual sales of the $3.2 billion marketable securities acquired by the federal government. The pension portfolio of assets was projected to be fully liquidated after 8 years. The absence of a further offset to pension benefit payments resulted a projected total of $1 billion in federal outlays for years 2006 & 2007 combined.

The legislation was also credited with savings of $50 million per year for seven years in payments the federal government was to pay to the pension plans under previous law. Thus, the total 10-year outlay and deficit cost of the assumption of the underfunded plans was $1 billion for years 9 and 10 less $350 million in discretionary savings, rather than the $5.8 billion federal loss from accepting $9 billion in pension liabilities in exchange for $3.2 billion in assets.

Note that the actual subsequent timing of security sales by the Treasury would be of no consequence. A cost estimate is not a law, but rather a projection of intent conveyed to budget analysts by policy officials.  Further, for subsequent budgets, outlays for the assumed DC pension plans were included in the current law baseline and no longer salient to future budget allocative decisions.


Title VIII (Community Living Assistance Services and Supports) “Class Act,” Affordable Care Act of 2010

This legislation, establishing a home care health insurance program, included a crucial requirement that the Secretary of Health and Human Services must develop and adopt a premium and benefit structure that would produce revenues sufficient to cover claims.[3] Further, if such a benefit-premium structure could not be identified, the Secretary would be prohibited from initiating the program.  Subsequently, in October 2011, owing to the anticipated effects of adverse selection, the Secretary announced that she was unable to offer assurances that the program could be structured to be entirely self-supporting.  Accordingly, the effort to establish the program was abandoned.

In preparing a cost estimate for the legislation, however, the Congressional Budget Office assumed that actuarial premiums and benefits could be developed and successfully administered. It therefore prepared projections of equal value cash inflows from premium collections beginning with the assumed initiation of the program and deferred outflows for claims payments spread over three decades.

Eligibility for benefits under the insurance required 5-year vesting, employment for three of those years, and payment of premiums through payroll deduction. Accordingly, the 10-year cost estimate, projected net collections from the Class Act of $72 billion. (CBO 2010). It thus provided a salient, but fundamentally false indication that the legislation would produce a new and usable inflow of resources to government.

This cost estimate also flies in the face of the clear legislative logic that the budgetary resource cost of this program was expected to be zero: either premium collections would fully offset projected claims costs or the program would not be adopted. It reflects a commitment by budget staff to a historical method of budgetary accounting that ignores the purpose of public budgeting: to allocate scarce resources across alternative beneficial uses to obtain maximum social benefits.

Of course, within a year of enactment, policymakers would learn that the projected $72 billion in gains was illusory.  But cost estimates are salient and used in decisions at the point of enactment, not when the actual costs and collections are tallied.


Mandatory Spending

About 70 percent of federal spending is mandatory, meaning that it is not subject to annual control though the appropriations process. Instead, this spending is controlled by provisions in authorization laws that specify eligibility and benefits. Social Security and Medicare account for the largest share of this spending, most of which provides benefits to people who for reasons of age, disability, disease, unemployment, or lack of marketable skills are unable to provide for themselves or their families. Funding for the largest of these programs comes from taxes levied on those who are working. Payment of taxes creates an expectation—encouraged by political authorities—that benefits will be available to those paying taxes when or if they themselves become eligible for benefits.  This expectation promotes a willingness to pay and planned reliance on future benefits by the current generation of workers.

Cash basis-basis accounting for these transfer programs defers recognition of cost until benefits are paid to beneficiaries. That is, to a time when it would be social and politically unthinkable to withhold payments to those in a condition of need. Deferral of cost recognition for benefits being earned in expectation today also suggests that the government on behalf of taxpayers may be facing a future, but not a present, financing short fall.  The absence of a salient, current period cost of accumulating benefits increases the difficulty or managing those cost—until they are paid and it is too late to withhold payment.  Accordingly, mandatory spending under current law is projected to rise relative to national income without identifiable limit.

Cash-basis accounting thus renders the costs of mandatory spending virtually unmanageable. Shifting the recognition of mandatory costs from when paid to when earned, or at least when working, is feasible, but it requires replacing the current cash-basis system with an accrual system. (Phaup 2018). Accruals are presently used in the budget to recognize the current period cost of such activities as the payment of interest on the public debt, federal direct loans and loan guarantees, and lease purchases of assets. Its extension to mandatory spending is likely a necessary condition for including this use of scarce budgetary resources in budget decisions.

In light of the absence of both a resource constraint and relevant measures of the opportunity cost of considered alternatives, the unresolved issue is not why the budget process performs badly, but rather how policymakers manage to credibly describe this activity as the socially important function of budgeting.


Possible Remedies: Adding a Constraint

Proposals to add the requisite information are already at hand.  Many forms of a constraint could be integrated into the existing process, including annual balance rules, intermediate and long-term debt targets and fiscal gap rules. Each has advantages and disadvantages, but if enforced, could improve budget outcomes compared with current policy. The existential question for democratic governments, however, is whether elected authorities can enact and enforce self-denying limitations on themselves.

An annual balanced budget rule and its close relative the Tax and Expenditure Limitation is frequently proposed as simple, widely understood means of forcing federal budget makers to limit resource use.  However, the proposal has been dismissed as unworkable at the national level because of the implied restriction on the ability of fiscal policy to stabilize the economy.  In fact, an annual rule is more restrictive than necessary as demonstrated by the success of the balanced budget norm, under which deficits were permitted during defense or economic emergencies, with debt pay-down in the post-crisis period.

One means of reconciling annual restrictive rules with active macro-stabilization is to require funding of budget contingency or “rainy day” funds.  A key to successful use of this practice is for the government to estimate the annual long-term expected amount of fiscal stimulus needed and to reserve those sums, without recording a politically unsustainable budget surplus. One applicable precedent for achieving this objective is the current accounting for interest on the public debt which is recognized in budget outlays as earned rather than when paid.  Payments to the interest payable account, which is treated as non-budgetary, are reported in budget outlays and the deficit on transfer to the fund, rather than when the payable is liquidated with payment. Similarly, stimulus spending could be scored in good times and effectuated when needed during downturns.

To increase the probability for success in reserving resources in good times, the process could be governed by a rule as automatic as spending for mandatory programs: no Congressional action would be required for the annual transfer of funds to the reserve account or for the disbursement of stimulus. Rather, spending from the fund would be triggered by an index such as the unemployment rate or a specified period of weak growth in real GDP.  Outlays could be earmarked in advance, e.g. for grants to state and local governments which are usually forced to reduce spending during national economic downturns.

A related constraining rule would aim at stabilizing the federal debt as a share of GDP.  Several studies have recommended this approach to adding a resource constraint to the budget process, e.g. CRFB 2018. The basic idea is that by limiting debt as a share of GDP, increases in outstanding debt would be permitted but only consistent with higher income and growth in demand for low-risk Treasury debt.  A limit on the growth of debt would constrain the growth in spending by requiring an increase in taxes if spending rose by more than the permitted growth in debt.

Debt limits are usually expressed in terms of a long-term target with an intermediate target path to the long-term level. The dual challenge to this approach is the desire to permit some short-term budget flexibility to deal with unexpected fiscal events while maintaining a strong commitment to the long-term target. The CRFB proposal would attempt to deal with this tension by making the Congressional budget resolution a law, which would require a Presidential signature and presumably political pre-commitment by both branches of government. Either automaticity of execution or unwavering commitment to fiscal discipline and restraint by policymakers is likely to be required if fiscal policies are to be constrained.

A third approach that could have the advantage of promoting simultaneous enactment of offsets to short-term departures from long-term debt targets would establish a fixed target value for, the long-term, e.g. 25-year, fiscal gap which measures the present value of the difference between projected annual revenues and outlays as a share of GDP. If, for example, the present value of the shortfall over 25 years is currently 2 percent of GDP, Congress and the President might be able to reach agreement to adopt policies, to be phased in over 25 years, that would reduce the gap to a lower level, perhaps zero. In subsequent years, the gap would be re-estimated and the time path of policy adjusted to maintain the target value.  The gain from the use of the fiscal gap measure is that policies—to raise taxes and reduce spending–would become effective in the future when they would have been anticipated and prepared for by taxpayers and beneficiaries.  The risk is that when the time arrives for those measures to take effect, elected officials would yield to political pressure to repeal those measures.  One means that might make reversing course more difficult would be to require that policies to reduce the gap in the future would have to begin their phase-in at some level immediately on enactment.


Remedies: Measuring Relevant Cost at Point of Decision

The key, but missing, information in budgeting for insurance with cash-basis accounting is identical to that for pre-FCRA credit programs: a relevant and salient measure of annual, long-term expected cost.  Without that information, no salient benchmark exists for policymakers to judge the long-term adequacy or deficiency of premiums or of the need to adjust the terms or extent of coverage. Systematic long-term cost management is impossible.


Comply with budget accounting matching principle

  • The budget needs to measure costs accurately so that decision makers can compare the cost of a program with its benefits, the cost of one program with another, and the cost of one method of reaching a specified goal with another. These costs need to be fully included in the budget up front, when the spending decision is made, so that executive and congressional decision makers have the information and the incentive to take the total costs* into account when setting priorities. Budget of the US Government, Analytical Perspectives, Ch. 8, “Budget Concepts,” p.77

* Possible friendly amendment: non-recoverable, marginal cost, or those that cannot be recovered by the sale of acquired assets, e.g. loans, structures, crude oil.  Recoverable resources have not yet finally committed and are still available for possible re-allocation.



Bartlett, Bruce (2010). “Jude Wanniski: Taxes and a Two-Santa Theory,” May 7

Colander, David (1997). “Functional Finance,” An Encyclopedia of Keynesian Economics, Cate, Harcourt, and Colander (eds.), Cheltenham: Elgar

Committee for a Responsible Federal Budget (2018). Memo to the Joint Select Committee: Including Debt Targets in Budget Reform, Washington, D.C., November 2.

Congressional Budget Office (1997). “Budgetary Implications of the Balanced Budget Act of 1997” December.

Congressional Budget Office (2010). “Memorandum to Senator Reid,” March 11. reid_letter_managers_correction_noted.pdf

Lerner, Abba (1943). “Functional Finance and the Federal Debt,” Social Research, 10, February 38-52

Mikesell, John L. (2014) Fiscal Administration, 9th Edition, Wadsworth Cengage Learning: Boston

Mason, J.W and Arjun Jayadev (2016). “Lost in Fiscal Space: Some Simple Analytics of Macroeconomic Policy in the Spirit of Tinbergen, Wicksell and Lerner,” Washington Center for Equitable Growth, Working Paper, November

Phaup, Marvin (2018). Budgeting for Mandatory Spending: Prologue to Reform. Forthcoming in Public Budgeting & Finance. Available at

Phaup, Marvin and Imtiaz Bhatti (2013) “A Case for Adding a Long-Term Budget Constraint to the Congressional Budget Process,” Albany Government Law Review 6,1 110-147

Shaviro, Daniel (2009). “The Long-Term Fiscal Gap: Is the Main Problem Generational Inequity?” The George Washington Law Review, 77, 6/6, 1298-1357

Steuerle, Eugene (2018). “Did the Congress and the President ‘Give’ Us a Tax Cut?” Tax Vox, February 5.

U.S. Congressional Budget Office (2018). The 2018 Long-Term Budget Outlook, June

U.S. Government Accountability Office (2018). Strategic Petroleum Reserve, May.

Wanniski, Jude (1976). “Taxes and a Two Santa theory,” National Observer, March 6 (See Bruce Bartlett reference above).



Social Insurance: OASDI and Medicare.

Disadvantages of Cash-Basis Budgeting for Long-Lived Claims

Cash basis accounting for long-lived obligations (deferring the recognition of cost until cash is paid out) has budgetary disadvantages in addition to understating the cost of providing services in the current budget year: time inconsistency, weak budgetary incentives for mitigation, and an unrecognized increase in debt.

Time inconsistency refers to the change in policymakers’ preference for a policy to be enacted in the future, but which changes into opposition when the time comes to adopt it. Or simply, the appeal of paying for a loss on occurrence diminishes sharply once the loss occurs. Mitigation is any use of resources that reduces losses on outstanding contingent claims. Deferring recognition of the cost of future claims prevents mitigation from being credited with savings.  And, failure to recognize a current cost effectively shifts that claim on resources to the future, which is the essence of borrowing and an increase in debt.

Those undesirable features of cash-basis budgetary accounting apply with equal force to long-term insurance and long-lived loans and loan guarantees. Thus, it is relevant to note that the House Budget Committee’s 2016 proposal for budget process reform included a provision to extend FCRA-type, accrual accounting to federal insurance programs (and federal defined benefit pensions, which are also subject to deferred recognition with cash-basis accounting).  The Committee advanced its proposal on grounds that doing so would improve the ability of the Congress to anticipate, plan for, and finance the long-term costs of in programs.[4]

The key, but missing, information in budgeting for insurance with cash-basis accounting is identical to that for pre-FCRA credit programs: a relevant and salient measure of annual, long-term expected cost.  Without that information, no salient benchmark exists for policymakers to judge the long-term adequacy or deficiency of premiums or of the need to adjust the terms or extent of coverage. Systematic long-term cost management is impossible.


[1] Gross outlays are netted against collections from commercial-type transactions such as sales of assets or services.

[2] For specific instances of the strategic budget use of asset sales, specifically discrete sales from the Strategic Petroleum Reserve, conditional on high prices, see the Bipartisan Budget Act of 2015 (PL 114-74) Sections 403 and 404 and the Bipartisan Budget Act of 2018.

[3] For an accessible, detailed description, see




No Constraint and Cash Basis Accounting

Leave a Reply