Reducing “Waste, Fraud, and Abuse”: One Percent Of What?, 9/2/03

Reducing “Waste, Fraud, and Abuse”: One Percent Of What?
Double-Counting and Other Budget Committee Mistakes Will Require
Some Committees to Cut Entitlement Programs More than One Percent
By Richard Kogan

Today (September 2, 2003), House and Senate Committees are scheduled to submit recommendations on how to reduce “waste, fraud, and abuse” in federal “mandatory,” or entitlement, programs. In May, the House and Senate Budget Committees directed 12 Senate committees and 15 House committees to find ways to reduce entitlement programs; the Budget Committees issued dollar targets for each of these committees, supposedly equal to one percent of the total entitlement spending in each committee’s jurisdiction. The Budget Committees’ calculations were flawed, however, in major and minor ways. As a result, most committees are required to find savings that represent substantially more than one percent of the entitlement spending under their jurisdiction, and a few are required to find savings of less than one percent.

The flawed calculations result in combined committee savings targets that are too high. They exceed one percent of entitlement costs by $27 billion over ten years, or by 26 percent on average. Because the savings targets are overstated, most committees will have to cut their programs by more than one percent — in some cases, far more — to achieve the specified dollar reductions.

The Budget Committees’ largest error was to double-count the cost of certain “trust fund” programs. Double-counting makes these programs look costlier than they really are, and so makes the savings targets for committees with jurisdiction over these programs higher than one percent. The Budget Committees double-counted the cost of all programs that have trust funds into which the U.S. Treasury transfers money, with the trust fund then using the transferred funds to make payments to (or on behalf of) program beneficiaries. Programs with such a funding structure include the military retirement program, the civil service retirement program, and Part B of Medicare (the Supplemental Medical Insurance program, which pays for visits to physicians, rather than hospital bills).

Table 1
Double counting certain trust funds
In billions of dollars
Committee	Ten-year target is too high by…
H & S Armed Services	$4.0	77%
H Government Reform	$2.5	35%
S Governmental Affairs	$2.5	34%
H Ways and Means	$17.5	32%
S Finance	$17.5	21%

The cost of these programs is the amount of payments made each year from the trust fund to, or on behalf of, beneficiaries. Those are the amounts that the Congressional Budget Office and the Office of Management and Budget count as federal expenditures in their analyses of the federal budget. In assigning savings targets to the various Congressional committees, however, the Senate and House Budget Committees incorrectly counted as a cost to the government both the payments from the Treasury to these trust funds and the payments from the trust funds to the beneficiaries. This is double-counting, and is scrupulously avoided by CBO and OMB. In total, double-counting of this nature by the Budget Committees produces ten-year savings targets that are $24 billion too high.

For example, in 2004 the Treasury will pay $98 billion to the trust fund for Part B of Medicare (the SMI trust fund). That trust fund will also collect $30 billion in premiums from elderly enrollees in Medicare Part B. It will use the combined income from the Treasury and the premiums to make payments to physicians and other providers of $126 billion on behalf of Medicare patients. The cost of the Part B program is thus $126 billion, which is the amount CBO and OMB count as the program’s cost when recording federal expenditures. The “one-percent target” for the Ways and Means Committee accordingly should equal one percent of this amount, or $1.26 billion. Instead, the Budget Committee has added the $98 billion paid from the Treasury to the $126 billion paid from the Trust Fund, acting as though Medicare Part B will cost $224 billion in 2004, and included $2.24 billion in its “one-percent target” for the Ways and Means Committee. This approach double-counts the $96 billion paid from the Treasury to the SMI trust fund.

By analogy, if my wife hands my sons $4 at an amusement park and they use the money to ride the roller coaster, the cost to the family is only $4. The Budget Committees say that the cost of this transaction to my family is $8: the $4 handed to my sons plus the $4 they paid to the roller coaster’s barker.

As a result of this type of double counting alone, the House Ways and Means Committee, House Armed Services Committee, and House Government Reform Committee have been given targets that range from 32 percent to 77 percent higher than their “one-percent target” should really be. The Senate Finance Committee, Senate Armed Services Committee, and Senate Governmental Affairs Committee were unfairly treated in the same manner. See Table 1.

The Budget Committees also erred with regard to programs where the federal government acts as an agent in “passing through” to other entities certain funds that it collects for specified purposes. An example involves foreign military sales. In 2004, foreign governments are expected to buy $12 billion of military hardware from American defense contractors. This $12 billion in payments from foreign governments is paid to the federal government and deposited in the Foreign Military Sales trust fund, which in turn pays the defense contractors. The $12 billion in payments from the FMS trust fund to the contractors cannot be reduced below the agreed-upon sales price, so these payments cannot be cut one percent. Yet the Budget Committee added one percent of the cost of such pass-through transactions to the amounts they assigned to the various congressional committees, as though these pass-through transactions somehow include fraud, waste, or abuse that can be reduced as a way to shrink the deficit. In fact, if the U.S. government could in some way arrange to have American defense contractors charge foreign governments one percent less than $12 billion so that “spending” from the FMS trust fund would be one percent lower, the foreign governments would simply deposit one percent less into the FMS trust fund to begin with, and the budget deficit would not be reduced one penny. When a budget account is not really a government “spending program” but essentially is a pass-through for other transactions, there is no possibility of reducing the costs by one percent.

This analysis explains the major problems that underlie the dollar targets set for the House and Senate committees, belying the claim that each committee has been asked to find savings that represent the same proportion (i.e., one percent) of the cost of its mandatory programs. The final table in this analysis (Table 5 on page 14) shows the dollar amount and percentage by which the savings targets for each House or Senate Committee are overstated. The savings target for the Senate Finance Committee was overstated by the largest amount, almost $19 billion.

Both the impetus for the entitlement reductions discussed in this analysis and the errors in calculation discussed here derive from the budget plan that the House of Representatives passed in March. That plan would have required most Congressional committees with jurisdiction over entitlement programs to pass legislation reducing the costs of those programs. The House-passed budget contained a specific dollar target for such reductions for each of the committees in question; the House Budget Committee said these targets equaled one percent of the cost of the entitlement programs under each committee’s jurisdiction and were intended to reduce waste, fraud, and abuse. The dollar targets in the House budget plan contained significant methodological flaws, particularly the double-counting of the cost of certain entitlements. The targets thus generally exceeded one percent of the cost of the entitlement programs under the affected committees.

Cutting Benefits versus Cutting Waste, Fraud, and Abuse

Reducing waste, fraud, and abuse is a compelling goal, one that many Administrations have striven for with varying degrees of success. (For example, during the late 1990s the amount of Medicare over-billing appeared to drop substantially after the Administration devoted more attention to it.)

But the distinction between reducing waste, fraud, and abuse and reducing program benefits must be kept clearly in mind. Consider, for instance, the Veterans Compensation program, which compensates veterans who have suffered service-connected disabilities. The size of the monthly payment is based solely on the degree of disability, as judged by the Department of Veterans Affairs. For many years, CBO’s annual list of possible budget reductions has included an option to pay compensation only for future disability claims rated 30 percent or higher; milder disabilities suffered in the future would not receive benefits. But eliminating those benefits would not be a reduction in “waste, fraud, or abuse,” even if some might use the word “waste” to refer to these payments. Rather, it would be a policy decision to provide a smaller benefit.

In contrast, reducing “waste, fraud, and abuse” generally refers to reducing payments made in excess of what the law intends — for example, payments to veterans who have died, or to the same person at two addresses, or because a disability rating is higher than appropriate. (Such overpayments often stem from program complexity or human error rather than fraud. The eligibility criteria for some programs make the tax code look simple.) Errors can also cause veterans to have their claims denied, their disabilities under-rated, or their checks lost. It is a mistake to think that all errors produce overpayments and cost the government money.

In the conference on the budget plan, the House provision requiring Congressional committees to write and approve legislation now — as part of a reconciliation bill to be passed this year — was substantially altered. Under the conference agreement on the budget, which Congress approved in April, the Chairmen of the Budget Committees were to issue overall dollar targets in May for savings to be achieved by House and Senate committees with jurisdiction over entitlement programs, and those committees would, by September 2, submit recommendations (rather than legislation) on ways to achieve the targeted budget savings. “After receiving those recommendations,” the conference agreement stated, “the Committee on the Budget may use them in the development of future concurrent resolutions on the budget.”[1]

On May 19, Senator Don Nickles, Chairman of the Senate Budget Committee, submitted targets for Senate committees, and on May 21, Representative Jim Nussle, Chairman of the House Budget Committee, submitted corresponding targets for House committees.[2] The targets were said to equal one percent of the cost of the entitlement programs under the jurisdiction of each affected committee. The targets were set using the same methodology as had been used by the House in establishing its initial entitlement-reduction targets.[3]

When the Budget Committees issued their savings targets for each House and Senate committee, they acted mechanically — that is, they attempted to set targets equal to one percent of the cost of entitlements and other mandatory spending under the jurisdiction of each committee, so that each committee would share proportionately in the task. But their mechanical calculations were flawed in a number of ways.

As noted above, these targets reflect substantial double counting in various programs that are financed through trust funds. Certain trust fund programs derive some or all of their resources from the general fund of the Treasury. Payments are made from the general fund to the trust funds. The government then makes payments from the trust funds to, or on behalf of, the beneficiaries of those programs. For example, the general fund transfers resources to the Medicare Part B trust fund and the civil service and military retirement trust funds, which use these resources to make payments for the provision of physician services to Medicare patients or to issue retirement or disability checks to retired or disabled military or civilian employees of the United States.

You Often Have to Spend Money to Save Money

The approach to waste, fraud, and abuse taken by the Budget Committees has weaknesses beside flawed calculations. To begin with, the government frequently needs to invest in better program administration in order to weed out wasteful practices, but the budget plan does not make room for extra up-front expenditures to hire more staff to reduce inappropriate spending, to improve staff training for this purpose, or invest in improved computers and information systems to detect improper payments. Such investments are in the purview of the Appropriations Committees; the entitlement committees generally do not have the authority to fund the additional resources needed for such resources.

Suppose, hypothetically, that either through error or fraud, some beneficiaries are paid twice, at two different addresses. How can the Department of Veterans’ Affairs (for example) reduce the number of duplicate payments? In general, the Department would have to spend more on staff, training, and computer technology to identify the erroneous claims and either prevent them from being made or reduce the amount of overpayments. Reducing waste, fraud, and abuse often is dependent upon increased funding for program administration. Here, however, problems arise with the Budget Committees’ approach.

The Veterans’ Affairs Committee does not have the ability to increase funding for DVA administration; only the Appropriations Committee can do that.

Recent congressional budget plans have made a point of squeezing the amount of money allocated to the Appropriations Committees. Moreover, those Committees are always under political pressure to increase funding for popular programs, such as veterans medical care, rather than the boring but essential job of checking the validity of claims.

To obtain ten years’ worth of savings (as the Budget Committees’ targets require), the Appropriations Committees would have to increase DVA administrative funding for a decade, but the budget and appropriations process is implemented annually.

CBO is reluctant to credit increased appropriations with reducing the cost of entitlement programs, in part for institutional reasons — the Appropriations Committees wish to be credited or blamed only for the amount of funding they directly provide, not for some ancillary effects estimated by an unelected CBO staffer.

Finally, experience with increases in the IRS budget in the early 1990s shows that increased administrative funding may be self-defeating for the first year or two — initially, the most experienced employees are diverted from their normal tasks to train new employees.

In short, for many entitlement programs, a key method to reduce waste, fraud, and abuse is to increase resources available for the administration of these programs and sustain it over time. The Comptroller General implied as much in recent testimony, when he stated:

… many of these problems and risks are chronic and long standing in nature and their ultimate solution will require persistent and dedicated efforts on many fronts by many actors. Some will require changes in laws to simplify or change rules for eligibility, provide improved incentives or to give federal agencies additional tools to track and correct improper payments. Continued progress in improving agencies’ financial systems, information technology, resources, and human capital will be vital in attacking and mitigating risk to federal programs integrity. Some areas may indeed require additional investments in people and technology to provide effective information, oversight and enforcement to protect programs from abuse. … Continued persistence and perseverance in addressing the high risk areas will ultimately yield significant benefits for the taxpayers over time.*

Yet a sustained increase in annually appropriated funding for administering federal entitlement programs is exactly what the authorizing committees, tasked in May to reduce waste, fraud, and abuse in entitlement programs, cannot by themselves accomplish.

________________________________

* David Walker, Opportunities for Oversight and Improved Use of Taxpayer Funds, Testimony before the House Budget Committee, June 18, 2003, pp 5/6. (GAO-03-922T)

In the federal budget, such transactions generally are portrayed in three budget accounts: as a general-fund transfer to the trust fund, as a receipt by the trust fund of the general-fund transfer, and as a trust fund payment to beneficiaries. Since the first two budget transactions cancel each other by netting to zero, total government expenditures reflect only the trust fund payments to beneficiaries, which is as it should be.

Table 2
Double-Counting In Committee Savings Targets
Caused by Error in Treatment of General-Fund Transfers to Federal Trust Funds

(Ten-year savings target is too large by the following amounts, FY 2004-2013, in billions of dollars)

Mandatory General Fund transfer	Amount over-counted[1]	House Committee	Senate Committee
to Medicare Part B	15.213	Ways and Means	Finance
to civil service retirement and disability fund	2.505	Govt. Reform	Govt. Affairs
to Social Security (taxes on benefits)[2]	2.240	Ways and Means	Finance
to military retirement fund	2.184	Armed Services	Armed Services
to uniformed services retiree health care fund	1.790	Armed Services	Armed Services
to supplemental DC pension fund	0.034	Govt. Reform	Govt. Affairs
to foreign service retirement and disability fund	0.021	Internat. Relations	Foreign Relations
to UMWA benefits fund (from Medicare*)	0.011	Ways and Means	Finance
to UMWA benefits fund (from coal industry)	0.009	Energy & Commerce	Resources
to UMWA benefits fund (from Interior*)	0.007	Energy & Commerce	Resources
to Judiciary trust funds	0.004	Judiciary	Judiciary
to Fish and Wildlife Service funds*	0.004	Resources	Environment
TOTAL, amount double-counted	24.021
[1] The dollar amount is one percent of the ten-year total of the funds transferred from the general fund to the trust fund in question, and represents that amount by which a committee’s target for savings was set too high. [2] A portion of Social Security benefits are considered taxable income and so are subject to the federal income tax. The general revenues collected as a consequence are transferred from the Treasury to the Social Security Trust fund in part, and to the Medicare Part A (Hospital Insurance) trust fund in part. Because of an anomaly in the CBO database, the payments to the Social Security trust fund were (erroneously) counted by the Budget Committees in creating their savings targets but the payments to the Medicare Part A trust fund were not. * These payments are from mandatory trust or special funds (rather than the general fund) to other mandatory trust funds, and produce the same type of double-count for the same reasons.

When CBO, OMB, and the Budget Committees display overall budget totals, only the payments to beneficiaries constitute government expenditures. Simply transferring money from one part of the government to another is not itself a government expenditure. When CBO estimates the cost of new legislation, it also reflects this basic reality. If legislation were to reduce the cost of a program financed in this two-step manner, CBO’s cost estimate would reflect only the reduction in payments to the beneficiaries, even if the transfers from the general fund to the trust fund were reduced accordingly.

Yet, in each case like those just described, the Budget Committees counted both the mandatory transfer from the general fund to the trust fund and the mandatory payment from the trust fund to beneficiaries, thereby double-counting the size and cost of the program.[6] They should not have counted the transfer from the general fund to the trust fund. Having incorrectly counted these mandatory general fund transfers to trust funds as though they were government expenditures, the Budget Committees then took one percent of the overstated totals when issuing their savings targets for the congressional committees. Table 2, on the previous page, lists the cases in which mandatory general fund transfers to trust funds were counted but should not have been, producing a double-count in the totals. Not only does this double counting put the listed committees at a disadvantage relative to other committees, but it places the trust fund programs in question at a disadvantage relative to other programs.

The Budget Committees also overstated the amount of funds under a committee’s control, and therefore overstated the size of that committee’s “one-percent target,” in cases where the federal government receives payments as an agent for others and simply passes through the funds received to the intended recipients.

Certain mandatory expenditures constitute payments that “pass through” the federal government from one source to another. An example is the Foreign Military Sales program, in which foreign governments purchase arms (such as fighter airplanes) from U.S. corporations. To increase the chances that these transactions will not violate federal security regulations, the Federal Government acts as the middle-man, collecting the payments from foreign governments and passing them through to the American manufacturers.

In the budget, one account reflects the incoming funds from foreign governments, while another account reflects the outgoing funds. The net cost to the government and the taxpayer of acting as a middle man for these transactions is zero, and the payments from foreign governments are recorded as “offsetting receipts.”[7] However, the payment of outgoing funds is assigned to the House International Relations Committee and the Senate Foreign Relations Committee in CBO’s database, while the incoming amounts from the foreign governments are not technically assigned to any committee (although they are counted in the budget as offsets to the spending and thus result in a net cost of zero for these programs).

The Budget Committees’ calculations err in this area. The Budget Committees assigned the International Relations and Foreign Relations Committees a savings target equal to one percent of the gross cost of the payments made to the U.S. manufacturers. Yet the cost of this activity is really zero, and Congress cannot produce savings by forcing manufacturers to accept one percent less than the price that foreign governments pay to purchase the weapons.

In another example, the government collects fines from foreign companies or countries that “dump” their goods in American markets and passes the proceeds through to American industries that are harmed by the dumping. Here, as well, the Budget Committees mistakenly calculated one percent of the proceeds paid to American industries and added that amount to the waste, fraud, and abuse targets.

A number of other federal programs also are pass-through programs. For example, the government administers many “shared receipt” programs, in which the government collects rents, royalties, or fees for the private use of public land and then passes a fixed percentage of them (sometimes 100 percent) back to the states or counties where the land is located, as a type of payment in lieu of taxes. Here, as well, the pass-through payments to states and counties show up as mandatory payments in federal budget accounts and are assigned in the CBO database to certain congressional committees, while the payments made to the government in the first place are not technically assigned to any committee. The Budget Committees mistakenly took one percent of these pass-through payments in calculating their fraud, waste, and abuse targets, as though these programs were taxpayer-supported benefits. In reality, these programs constitute a net reduction in the federal deficit because the rents, royalties, and fees the federal government collects often exceed the payments made to states or counties.

Needless to say, these pass-through programs do not lose money through waste, fraud, and abuse that increase the federal deficit. The federal government knows exactly how much it has collected in rents, royalties, and fees and makes payments by formula to recipient states and counties. There are no “ineligible” states or counties “erroneously” receiving payments for which they are not eligible.

Similar pass-through requirements exist for certain “special funds” such as the crime victim’s fund, the assets forfeiture fund, and the like. In each such case, existing law requires 100 percent or some other specified percentage of certain federal income to be distributed. Even in those cases where there is some possibility of fraud or error (e.g. a payment to a person who is not really the victim of a crime), improved compliance will not save any money or reduce the deficit; money that is saved by weeding out an ineligible recipient will result automatically in more money being paid to the eligible recipients.

These pass-through programs should not have been included in the Budget Committees’ calculations when the savings targets were constructed and cannot be a source of savings in waste, fraud, and abuse. Table 3, on page 10, lists these programs and the amounts by which the savings targets were correspondingly overstated.

Table 3 also lists other miscellaneous instances in which the Budget Committees have overstated the amount of expenditures within a committee’s jurisdiction and thereby overstated the amount of savings achievable from a one-percent reduction in mandatory spending. For example, the Budget Committees rightly excluded from their calculations most Federal interest payments on Treasury borrowings. But they mistakenly allowed some small interest payments to remain in their calculations.

Another such example occurs with the Supplemental Security Income program. It, like most benefit programs, recovers overpayments made to beneficiaries when the overpayments can be identified. The cost of the program equals the benefit payments it makes minus the overpayment benefit amounts it recovers. In the case of the SSI program, however — unlike all other comparable programs — the Budget Committees mistakenly failed to net out these recoveries. In setting the targets assigned to the House Ways and Means and Senate Finance Committees, the Budget Committees took one percent of the gross cost of SSI payments without subtracting the overpayments that are recovered.

Still another example involves the payment of salaries to Members of the U.S. House of Representatives. These mandatory payments cannot be subject to error: the House Administration Committee cannot discover through more careful compliance with statutes and scrubbing the records that there are really only 431 Members rather than 435. Yet the House Budget Committee assigned one percent of these costs as well.

An additional category consists of programs that are first cousins to the pass-through programs. These closely related programs receive special earmarked receipts from the public (such as proceeds from the sale of public land or gifts to the Library of Congress), which are then allowed to be spent by the recipient agency on its normal operations (often those related to the income), as a supplement to the annual discretionary appropriations the agency receives to cover its costs. In these cases, even if the committees could find ways to spend the funds more efficiently, all of the income is allowed to be spent. Greater efficiency could allow the funds to go further or allow the Appropriations Committee to divert some of its allocation from agency administrative costs to other priorities, but would not reduce total federal expenditures.

Table 3 Over-Counting In Committee Savings Targets Caused by Inclusion of Pass-Through Payments and by Other Anomalous Accounting Ten-year savings target is too large by the following amounts, FY 2004-2013, in billions
Name of Budget Account	Amount over-counted		House Committee	Senate Committee
FMS trust fund		$1.153	International Relations	Foreign Relations
Various railroad retirement accounts[8]		$0.866	Transportation	Finance
Interest paid to credit financing accounts[9]		$0.493	Financial Services	Banking
Interest on IRS refunds		$0.300	Ways and Means	Finance
Supplemental Security Income		$0.215	Ways and Means	Finance
Interest payments to REFCORP		$0.194	Financial Services	Banking
Immigration examination fees		$0.100	Judiciary	Judiciary
Mineral leasing payments		$0.085	Resources	Energy and Resources
Funds for strengthening markets (“§ 32”)		$0.070	Agriculture	Agriculture
Immigration user fee		$0.068	Judiciary	Judiciary
Crime victims fund		$0.060	Judiciary	Judiciary
Unemployment trust fund, repayable advances		$0.059	Ways and Means	Finance
Allied contributions and cooperation		$0.050	Armed Services	Armed Services
Sport fish restoration		$0.041	Resources	Environment
Tribal special fund		$0.040	Resources	(Indian Affairs*)
Assets forfeiture fund		$0.038	Judiciary	Judiciary
Continued dumping and subsidy offset		$0.031	Ways and Means	Finance
Forest service permanent appropriations		$0.031	Agriculture	Agriculture
Judiciary information technology fund		$0.030	Judiciary	Judiciary
Education benefits fund		$0.030	Armed Services	Armed Services
Treasury forfeiture fund		$0.029	Ways and Means	Finance
IRS collections for Puerto Rico		$0.027	Ways and Means	Finance
Federal aid in wildlife restoration		$0.025	Resources	Environment
Judiciary filing fees		$0.021	Judiciary	Judiciary
Payment to copyright owners		$0.020	Judiciary	Judiciary
Forest Service trust funds		$0.015	Agriculture	Agriculture
Compensation of Senators and Members		$0.013	House Administration	(Rules*)
BIA misc. permanent appropriations		$0.011	Resources	(Indian Affairs*)
Health and Human Services: misc. funds		$0.011	Commerce	Health
All other smaller such accounts		$0.085	various	various
Total		$4.210
*Because this Committee was not assigned a savings target by the Senate Budget Committee, its target was not overstated.

A final category consists of “mandatory” programs that receive payments from discretionary appropriation accounts to perform work on behalf of those discretionary programs. For example, the Public Health Service constructs and repairs health facilities on behalf of the Indian Health Service, and the Indian Health Service reimburses the Public Health Service for the costs. If these “mandatory” expenditures by the Public Health Service can be reduced by improving efficiency, the Indian Health Service will be able to pay less to the Public Health Service. The Indian Health Service is a discretionary account and therefore is annually funded with a fixed-dollar appropriation; if it can pay less to the Public Health Service, either it will have more to spend for other purposes or the Appropriations Committees, within the fixed allocations they are given under the annual congressional budget resolutions, will be able to reduce the Indian Health Service appropriation and provide more funding for some other purpose or program. Either way, total expenditures will not be reduced (although the money that is spent will go further). Achieving greater efficiency in such programs is surely desirable if it can be achieved. But there are no scoreable mandatory savings that can be achieved by such efficiencies, and the Budget Committees consequently should not have included such programs in their calculations of committee saving targets.[10]

Understating Costs by Including Credit “Liquidating Accounts”
and Certain Other Revolving Funds

The Budget Committees also understated some committees’ targets by a small amount by erroneously including “credit liquidating accounts” as “negative expenditures” when calculating the one-percent targets for waste, fraud, and abuse. The Budget Committees should have excluded these accounts from their calculations because these accounts do not represent new budgetary activity by the federal government.

Liquidating accounts record the current cash flow of direct loans or loan guarantees made before 1992. For the most part, the current cash flow of these liquidating accounts constitutes loan repayments to the government from a portfolio of old loans, for example for rural housing or rural electrification. These repayments on old loans show up in the budget as “negative expenditures” — offsetting proprietary receipts — and thereby appear to “offset” or reduce the total amount of new program costs over which a committee has jurisdiction. Where these liquidating accounts exist, the Budget Committees took one percent of a committee’s total expenditures after subtracting these offsetting proprietary receipts. As a result, the Budget Committees understated the scope of possible savings from other mandatory programs in a committee’s jurisdiction, programs from which improved program administration could as reasonably generate savings as from any other mandatory program. The committee targets for such committees were understated because the credit liquidating accounts were included (as offsets or “negative expenditures”) in the calculations, thereby making the total costs of ongoing programs in the committees’ jurisdiction appear smaller than they actually were. As a result, the one-percent waste, fraud, and abuse targets for a few committees with liquidating accounts are understated, as reflected in Table 4.

This is not to say that such portfolios of old loans are immune from savings. For example, it may be possible to achieve scorable savings by improving the government’s ability to identify and collect delinquent outstanding loans, including loans made before 1992. But the Budget Committees have mechanically reflected these outstanding loan portfolios as though they reduced the scope of possible savings in waste, fraud, and abuse. At a minimum, these liquidating accounts do not reduce the scope of possible savings, and they may increase it.

The same is true of a small number of mandatory credit accounts that are estimated to produce “negative subsidies.” These credit programs issue loan guarantees and are estimated to receive more in origination fees then they will ultimately pay in default claims.[11] Leaving aside administrative costs, these programs thus appear to make money for the government. But, as with the liquidating accounts, the existence of such a “negative subsidy” does not reduce the scope of possible savings from other programs within a committee’s jurisdiction, and here the Budget Committees should not have included the “negative spending” in their calculations.

Similarly, the Budget Committees included in their calculations the income of certain mandatory revolving funds, such as the Pension Benefit Guarantee Corporation, the Bank Insurance Fund, and the Exchange Stabilization Fund. In these cases, the income (or “negative spending”) is relatively independent of the spending for operations or benefits in these mandatory accounts. In these programs, spending for operations or benefits is presumably a possible source of waste, fraud, or abuse. The Budget Committees in effect understated the possible scope of savings by netting the income from such programs against their expenditures.

In all three cases — liquidating accounts, negative subsidies, and certain revolving funds — the Budget Committee’s inclusion of “negative spending” in their calculations in effect treated the income as a source of “negative waste, negative fraud, or negative abuse.” This is an inappropriate approach to these programs.

Table 4
Understatement of Committee Savings Targets Caused by Inclusion of Certain
“Negative Expenditures” in Budget Committee Calculations
(Ten-year savings target is too small by the following amounts, FY 2004-2013, in billions)

Overall, however, the double-counting and other overstatements discussed above and reflected in Tables 2 and 3 vastly exceed these understatements.

For the reasons discussed in this analysis, a majority of House and Senate committees have been assigned savings targets that are higher than (or in a few cases, lower than) the one percent that was intended to be assigned to them. Table 5 consolidates this information and shows the amount and percentage by which each committee’s ten-year target exceeds or falls short of the expected one percent. To a large extent, this table illustrates the pitfalls of looking at budget figures in a purely mechanistic way without reflecting the nature of budget programs or the fact that they are financed in many different ways.

Because most committees were given targets that represent more than one percent of the costs of mandatory programs under their jurisdictions, attempts by committees to meet these targets will mean that committees are supposed to find ways to reduce programs by more than the stated one percent. Moreover, because the degree of overstatement varies quite widely from committee to committee, some programs are much more likely than others to be affected by the errors in the Budget Committees’ calculations.

	Given savings target	Dollar over- or under-statement	Percent over- or under-statement
Senate Finance	$98.735	$18.926	24%
House Ways and Means	$71.428	$18.058	34%
House and Senate Armed Services	$9.165	$4.055	79%
House Government Reform	$9.859	$2.539	35%
Senate Governmental Affairs	$9.904	$2.539	34%
House Int. Relations, Senate Foreign Relations	$1.181	$1.099	1,333%
House Transportation and Infrastructure	$1.578	$0.869	123%
Senate Judiciary	$0.618	$0.356	136%
House Judiciary	$0.644	$0.354	122%
House Resources	$0.297	$0.257	643%
Senate Energy and Natural Resources	$0.201	$0.144	252%
Senate Environment and Public Works	$0.195	$0.076	63%
House Energy and Commerce	$26.523	$0.040	0%
House Financial Services	$0.095	$0.037	64%
Senate Banking, Housing, and Urban Affairs	$0.092	$0.033	57%
House Administration	$0.024	$0.016	188%
Senate Commerce, Science, and Transportation	$0.807	$0.012	1%
House Science	$0.003	$0.003	n.a.
Senate Veterans’ Affairs	$3.852	$0.000	0%
House Veterans’ Affairs	$3.850	-$0.002	0%
Senate Health, Education, Labor, and Pensions	$0.944	-$0.004	-0%
Senate Rules and Administration*	$0.000	-$0.008	-100%
House Education and the Workforce	$2.431	-$0.020	-1%
House Agriculture	$4.945	-$0.146	-3%
Senate Agriculture, Nutrition, and Forestry	$6.251	-$0.156	-2%
TOTAL*	$131.984	$27.115	26%

* The total has been divided by two, reflecting the fact that each mandatory program is within the jurisdiction of one House committee and one Senate committee. The Senate Rules and Administration Committee was not assigned a target, while the House Administration Committee was. For each of these committees, a ten-year target of $8 million would equal one percent of the cost of mandatory programs under their jurisdiction potentially subject to waste, fraud, and abuse.

[1] See §301(b) of H. Con. Res. 95, the concurrent resolution on the budget for fiscal year 2004.

[2] See Congressional Record, page S6598, May 19, 2003, and page H4512, May 21, 2003. The Senate and House Intelligence, Senate Rules, Senate Indian Affairs, and House Homeland Security committees were not assigned targets for any outlay savings.

[3] The reductions that would have been required under the House budget plan were said to equal one percent of the cost of entitlements under various committees’ jurisdictions in fiscal year 2004. These reductions climbed to higher levels in subsequent years, averaging almost four percent over the ten-year period 2004 through 2013. The budget plan the House Budget Committee approved required reductions in Medicare, but these reductions were removed before the full House passed the plan. The targets in the final budget plan that emerged from conference, which committees are now called upon to respond to, are supposed to equal one percent of entitlement costs both in fiscal year 2004 and over the next ten years.

[4] The dollar amount is one percent of the ten-year total of the funds transferred from the general fund to the trust fund in question, and represents that amount by which a committee’s target for savings was set too high.

[5] A portion of Social Security benefits are considered taxable income and so are subject to the federal income tax. The general revenues collected as a consequence are transferred from the Treasury to the Social Security Trust fund in part, and to the Medicare Part A (Hospital Insurance) trust fund in part. Because of an anomaly in the CBO database, the payments to the Social Security trust fund were (erroneously) counted by the Budget Committees in creating their savings targets but the payments to the Medicare Part A trust fund were not.

[6] The Budget Committees’ double counting occurred in all cases where a) the transfer to the trust fund is made from a mandatory budget account (rather than from a discretionary account over which the Appropriations Committees have control), and b) the payment from the trust fund to, or on behalf of, the beneficiary also is made from a mandatory budget account.

[7] The existence of this “offsetting receipt” is the common feature that relates these “pass-through” transactions to the double-counting of general fund transfers to trust funds, explained in the previous section. The payments from the general fund to the trust funds also generate “offsetting receipts” or “negative expenditures” that the Budget Committees failed to account for in their calculations.

[8] Only “Tier II” railroad retirement benefits and mandatory administrative costs should be included in the calculations. Tier I benefits are Social Security benefits, and so should be excluded because all other Social Security benefits have been excluded. The remaining railroad retirement transactions represent the investment or re-investment of trust fund assets, and so are not “expenditures” in any meaningful sense

[9] These costs should be excluded from the calculations because they constitute interest payments. As noted, interest payments should be excluded (and largely were) because they are not subject to waste, fraud, or abuse. Even if one disagrees with that conclusion, the Budget Committees should have excluded this interest account for two other reasons. First, while some “credit financing accounts” receive interest from the Treasury, others pay considerably more interest to the Treasury; in net, the interest costs to the Treasury from transactions with credit financing accounts are negative (that is, the Treasury makes money on these transactions). The Budget Committees inappropriately included only the Treasury’s interest payments in its calculations, excluding its interest receipts. Second, the various “credit financing accounts” with which the Treasury transacts business are within the jurisdiction of various committees, not just House Financial Services Committee and the Senate Banking Committee; as a result, those committees do not have the jurisdiction to make the required savings included in their savings targets (if savings were in fact achievable in these accounts) .

[10] Technically, in the example given and in all such similar transactions, if committees find ways to decrease waste and improve efficiency in the construction of facilities (for example), the mandatory budget account would spend less but the mandatory “offsetting receipt account” (which represent the receipt by the mandatory account of payments from the discretionary account on whose behalf the facilities are constructed) would be correspondingly smaller, as well. Because net mandatory spending in this area is zero to begin with — the mandatory spending and the offsetting receipt net to zero — net mandatory spending would still be zero after efficiencies are achieved. This is why CBO will not score legislation achieving efficiencies in these accounts as achieving mandatory savings, in net. It is, in fact, debatable whether the accounts that perform work on behalf of discretionary accounts, and are reimbursed by them, should be classified as “mandatory” at all.

[11] An origination fee is an up-front payment made by a borrower to the government as part of the agreement in which the government guarantees, or co-signs, the borrower’s loan from a bank. It is akin to a payment of “points” when taking out a mortgage. A default claim is the payment by the government to the bank if the borrower (for example, a student, farmer, or small business) defaults on a loan that has been guaranteed by the government; the terms of the guarantee require the government to make good on the loan if the borrower defaults.

[12] The dollar amount is one percent of the ten-year total of net payments to liquidating accounts under the jurisdiction of the indicated committees, i.e., it represents that amount by which the committee’s target for savings is too low.

[13] In the case of the Banking Committees only, the Budget Committees appear to have removed a portion of the “negative expenditures” under those committees’ jurisdiction from their calculations; they seem to have removed some or all of the amounts associated with the years 2009-2014. Since the Budget Committees have already removed part of the negative expenditures from these calculations, Table 4 reflects the remaining amounts that should be removed from the calculation of the targets.

House / Senate Committee	Undercounted[12]
Financial Services / Banking, Housing, and Urban Affairs[13]	$0.650
Agriculture / Agriculture, Nutrition, and Forestry	$0.273
International Relations / Foreign Relations	$0.075
Ways and Means / Finance	$0.064
Education and the Workforce / Health, Education, Labor, and Pensions	$0.019
various other committees	$0.003