Skip to Main Content




"Share-In-Savings" Contracts: Risky Business

June 13, 2001 


A critique of Rep. Tom Davis' Share-in-Savings" federal contracting legislation by Charles Tiefer, Professor of Law, University of Baltimore School of Law and former General Counsel to the U.S. House of Representatives.  He warns that such contracts will, among other things, lead to long-term backdoor appropriating by government agencies.

Contractor groups have put forward a legislative proposal for a new type of contract that would be lucrative for them, but relatively untested in approach, called the "share-in-savings" (SIS) contract. In an SIS contract, the government commits, on a locked-in long-term basis, part of the appropriations for government agency operations to contractors who might invest their own capital up-front in performance of those operations.

SIS contracts could turn out to be off-the-books, sweetheart deals for politically-favored contractors. As a professor of government contract law, and the author of the current legal casebook on the subject (Government Contract Law: Cases and Materials (Carolina Academic Press 1999 & Supp. 2001)), I can readily identify at least ten major categories of legal risks surrounding SIS. On one side are undocumented hopes of SIS enthusiasts; on the other side, the reality of government-funded long-term, locked-in contracts, that may undermine or circumvent existing appropriation policies. SIS contracts could propagate problems similar to those that accompanied deregulation of government-insured savings-and-loans institutions or procurement of defense spare parts in the 1980s by sole-source contracts.

After a basic introduction and some concrete examples of how SIS might work, this paper will describe the following ten major categories of SIS's unexplored risks:

    1. Authorizing major changes to proven approaches before reviewing results from pilot projects

    2. Long-term backdoor appropriating and undermining of Congressional oversight

    3. Needless higher government costs by off-the-books borrowing through contractors

    4. Unbenchmarked, hard-to-audit payouts by the fisc

    5. Non-competitive contracting

    6. Potential to unfairly advantage politically favored contractors

    7. Exposure to open-ended litigation and losses

    8. Potential to undermine policy safeguards for government operations, e.g., protection of citizen data privacy

    9. Hard-to-undo mistaken choices about privatizing

    10. Potential tampering with revenue operations

Basic introduction, concrete examples

In a share-in-savings contract, the government hires a contractor to perform a service with payment measured as a share of the "savings" in an ongoing government activity. That is, the government awards a contract in which it commits to pay the contractor a portion of what would otherwise be appropriated for the procured activity. SIS has not been used frequently, and where it has been used, it has only been in the context of energy efficiency contracts. Section 5311 of the Clinger-Cohen Act authorized pilot SIS information technology (IT) acquisition, but this has not been utilized.

Some of the risks in SIS are inherent in the concept, and would be present even if SIS were implemented cautiously, starting with pilot projects and surrounded by as many of the traditional safeguards for government contracts as possible. The government could award SIS contracts strictly following competition requirements, namely, by putting out formal and public solicitations, competitively evaluating proposals, and affording opportunity for rejected offerors to raise legal protests with courts or the GAO. Further, the government could award SIS contracts that would privatize current government functions using the existing A-76 processes to scrutinize the appropriateness of such privatizing.

However, the proponents of SIS have conspicuously failed to state that they would only implement it without the traditional safeguards. Hence, broad authorization of SIS would expose the government not just to the risks inherent in SIS, but to the exacerbated risks when SIS occurs without safeguards.

Examples of potential SIS contracting may help concretely anchor the discussion that follows. Many of the potential SIS contracts concern IT operations. Suppose the Social Security Administration's large facilities in Baltimore face a situation where they will be performing some important SSA data inputting and processing function on a costlier basis than necessary unless the government keeps pace with the times by investing in new hardware and software and personnel training. And, suppose an IT consulting firm in Northern Virginia offers to manage, to take over, and to provide the up-front modernization capital for this function, in return for a locked-in, long-term share in what would otherwise be appropriated for the performance of the function.

Or, suppose there is some DOD payroll function now performed at facilities around the country in Maryland, New York, and California. And, suppose an IT consulting firm in Northern Virginia is awarded, on a non-competitive basis, an SIS contract to take over this function.

What are the legal risks?

1. Authorizing major changes to proven approaches before reviewing results from pilot projects:

The first risk comes out of the idea of having a sweeping new legislative provision to authorize SIS on a government-wide basis. As noted, there have been limited provisions and initiatives in the past for pilot SIS contracts. Yet, it is proposed incautiously to authorize government-wide SIS contracts in untested areas. As will be seen, SIS contracting involves long-term, hard-to-undo commitments, with major government exposure to risk. That is precisely the kind of contracting that should not occur, but if at all, only after pilot programs are fully evaluated; only to the extent and in the contexts warranted by those pilot programs; and, even then only with safeguards built into those programs. What is proposed is to authorize government-wide exposure to this program, before we learn the outcome of pilot programs that can assist us in avoiding the risks that can accompany SIS contracts.

In the example stated, SIS legislation could authorize a large-scale Social Security Administration SIS contract even though there is no previous experience even with a small-scale Social Security SIS contract. During the Reagan Administration, the House Committee on Government Operations conducted oversight, with which I was personally familiar, regarding an IT procurement at SSA - its procurement of terminals from the Paradyne Company. Instead of proceeding cautiously, the SSA had been persuaded to move faster than experience warranted. The result was a debacle and major chaos in a vital government program.

2. Long-term backdoor appropriating and undermining of Congressional oversight:

The heart of SIS consists of a form of backdoor appropriating of an extreme kind - for long-term, capital-intensive projects. In normal agency operations, the agency must ask Congress annually for appropriations. If SSA plans to initiate a long-term capital-intensive project, such as procuring an upgraded IT system, the agency must come to Congress for appropriations to purchase that system. If approved, Congressional appropriators either provide a full-funding multiyear appropriation for the projected cost over the life of the project or at least sink the initial appropriations into the project mindful of the projections for its lifetime costs. Moreover, any multi-year contract awarded to a consultant on a non-fixed-price basis must include what is called a "Limitation of Funds" or "Limitation of Costs" clause that restricts the fisc's exposure to that which is appropriated.

By contrast, in an SIS contract, the agency will not ask Congress for appropriations to fund the project. It is not at all clear whether the agency will even provide Congressional appropriators with a hard-nosed assessment of its lifetime costs. Because the SIS contractor furnishes capital for the performance of the operation, the agency's position is that it has released itself from having to obtain appropriations for that capital. Yet, the SIS contractor is not making a charitable contribution to the government; the SIS contractor insists on its receiving what would otherwise have been appropriated for the agency operation. The contractor receives that commitment of future appropriations, not through the front door of current decisions by appropriators, but through the "backdoor," by an agency contract.

Moreover, once SSA makes a ten-year contract with an IT firm regarding some SSA function, the funding of that contract is not dependent annual Congressional appropriations. By definition, Congress cannot de-fund the SIS contract during its ten-year duration without paying a substantial cancellation cost. Furthermore, there either is no Limitation of Cost clause, or any such clause would be cause for a lawsuit against the United States for breach, since the payments to the SIS contractor cannot be restricted by limited appropriations, as with a normal cost reimbursement contract.

One particular concern is the undermining of Congressional oversight. Suppose an SIS contract is agreed to and two years later a Congressional oversight committee becomes concerned. Under the normal system in which the agency would receive annual appropriations, Congressional oversight would be meaningful because the agency would worry about the effect of oversight on its funding. However, under a long-term SIS contract, the SIS contractor cannot be removed without substantial cancellation costs essentially removing the threat of oversight. The appropriators lack control, undermining the rest of the oversight apparatus.

A second particular concern is the suggestion that a portion of the "savings" be "kept" by the agency. In other words, if the Defense Department awards an SIS contract leading to less appropriations than were otherwise anticipated, the Department would "keep" the difference. Perhaps the idea is that if the Defense Department has some other project which Congress is reluctant to fund, it would have the right to fund that project out of this "savings". The concept of agencies "keeping" future appropriations and not having to go through the budget process to fund their projects amounts to a sub-delegation, fragmentation, and balkanization of fiscal control which has traditionally been anathema, not merely to Congress, but also to OMB.

3. Needless higher government costs by off-the-books borrowing through contractors:

The point which SIS proponents most often cite is that contractors will provide the up-front capital investments needed to perform government operations. For example, a large IT contract by an SIS contractor might include the capital outlays for substantial amounts of hardware and software dedicated to agency operations. Under a normal contract, the agency would have had to request such appropriations from Congress.

In effect, the agency would be leasing the hardware and software from the SIS contractor. Another way of putting it is that the agency has switched from financing its operation through the Treasury at the government bond interest rate to having the contractor finance its operation through the contractor's interest rate. Then annual appropriations for the agency must be committed to repaying the contractor.

As a general policy matter, the government simply does not finance its operations by borrowing through contractors, thereby creating unofficial debt (i.e., off the books). Government agencies do not sell their working equipment to contractors and lease it back, or have contractors buy major equipment and lease it to the government. The United States has a better credit rating than any other borrower, and it could never benefit by borrowing in such ways through contractors. Such borrowing through contractors always results in greater expense for the United States, and merely moves the cost of borrowing into a different accounting situation, namely, off the books.

Moreover, part of the reason any such leaseback contracting occurs in the private sector is for tax and depreciation purposes, where ironically, it is the United States rendering the transaction profitable by providing the tax break. The United States could never have anything to gain from juggling its own asset use arrangements through leaseback schemes, but much to lose, for any favorable tax treatment means the United States is taking out of its revenue stream what it believed it was putting into its expenditure stream.

In short, SIS could be a sweetheart deal for the contractor, of a kind that the federal government would never otherwise allow itself to get involved in. Careful thought would have to occur about any specific type of SIS contract in order to avoid such schemes. A sweeping, government-wide legislative authorization of SIS contracting is the opposite of such careful thought.

4. Unbenchmarked, hard-to-audit payouts by the fisc:

The SIS proposals so far do not involve legislative specification, in rigorous detail, of how the "savings" by an SIS contractor would be measured. What is distinctive about SIS from other contracts is that it measures the amount to be paid to the contractor by the "savings" from the hypothetical appropriations for the project. If the payout is audited, the audit is measured against that hypothetical appropriation.

Sometimes such hypothetical appropriations may appear concrete. If the government has been performing some very precise operation, and the cost of the operation are easily specified, then these appear concrete. But, there is not necessarily such precision and invariance as it may seem. Frequently, for any particular government operation, the quantity and performance characteristics vary over time, the difficulties of performance and new governmental directives vary over time, the costs include a number of indirect as well as direct costs that are hard to capture, and many of these factors may well be difficult to score against each other in a quantitative way. Moreover, new costs such as contract administration and quality assurance must be added to the SIS contractor's own bill when considering the benefit of having an SIS contract.

Moreover, if this were not complex enough, SIS enthusiasts propose that contractors receive added payment - as a form of "savings" - for improvements they make in non-cost aspects such as rate of performance. That means that quality assurance methods must actually measure the monetary value of asserted, and often disputed, improvements in performance. For those who are familiar with the problems of even rudimentary performance measurement of government operations, the challenge of measuring the payouts and of auditing such payouts, by performance measurements adapted to each particular operation, should be apparent.

In current performance measurement of government operations, if the in-house performers assert they have improved their performance and others would claim that the improvement is illusory or was inevitable, the dispute has its limits. Further, in-house performers do not have a financial incentive to skimp on some aspect in order to maximize performance. In contrast, for an SIS contract, when the contractor demands to be paid the "savings" from disputed improvements in government operations, a government refusal to do so will be treated by the SIS contractor as a breach.

5. Non-competitive contracting:

SIS proponents have suggested authorizing at least some of SIS contracting on a non-competitive bases. For example, an IT consultant could approach the Defense Department with an offer to take over some function and rather than put the potential contract out for competition, the Defense Department could negotiate and award a contract to the initiator.

All the risks of SIS are magnified by non-competitive contracting. At least if several contractors submit SIS proposals in a competitive process, the agency can formally evaluate which offers the best value and quality assurance. Furthermore, protests by the rejected competing contractors might at least lend some oversight to the process, highlighting serious transgressions of procedural rules. Absent these competitive safeguards, SIS becomes a worrisome gap in the protection that the government enjoys through full and open competition.

6. Potential to unfairly advantage politically favored contractors:

SIS contracts are highly incentivized and geographically mobile contracts with the potential for advantaging politically favored contractors. SIS contracts differ in this regard both from market-controlled contracts, which are not so highly incentivized, and from agency service operations, which are not so geographically mobile.

In a market-controlled contract - say, a government contract to purchase office equipment - the price to the government should be the market price, and there is no reason for the politically favored contractor either to especially seek or especially benefit, from receiving the contract. The contractor charging the lowest cost receives the contract, without excess profit. In contrast, since an SIS contract may occur without competition, and may well take insider access to arrange, it is entirely possible that the successful awardee will turn out to be one that has political favor. Further, since it is already difficult to objectively gauge the payout in an SIS contract, the benefit to the contractor may be affected if the awardee turns out to have political favor.

Agency service operations tend not to change locations without some reason. Ordinarily, there is little reason for a contractor to propose to alter the location of agency service operations, or to especially seek or especially benefit from such an attempt. In contrast, an SIS contract may be geographically mobile. That opens the possibility that contractors, located in some particular place, who have political favor may offer to enter into SIS contracts around the country. Thus, the SIS contract becomes a tool to assist a politically favored contractor in wresting appropriations away from some currently fixed site, motivated by the highly incentivized contract.

SIS contracts would not be the first ones that went to contractors who, for geographic or other reasons, are politically favored. However, in normal contracts a proposal to move resources from stable sites in California, New York, or Maryland, to Northern Virginia, would run into scrutiny at senior administrative levels and in Congress. An SIS contract renders it much easier for a contractor to obtain a transfer for agency operations to its own location, and because SIS contracts provide a large incentive to do so, we can see the opportunity materialize for the politically favored to attempt such transfers. Moreover, the "backdoor appropriations process" of an SIS contracting circumvents the normal scrutiny that a transfer would attract.

So, in short, SIS provides a vehicle by which favored contractors could siphon away appropriations from other locales, potentially creating a greater role of political favor than we are used to - or than we may want to see.

7. Exposure to open-ended litigation and losses:

Government contract law for ordinary contracts affords valuable tools to the government for preventing certain kinds of litigation and losses which might not constrain SIS contracts. Ordinarily, government contract law provides the powerful tool of government termination for convenience as a way of precluding litigation and limiting losses. A contractor can rarely challenge a government termination for convenience, and there is a well-established machinery for resolving disputes regarding how much the government owes the contractor. That body of law specifically avoids paying contractors "unearned profits" - the future profits of a contract that have not yet been earned.

In contrast, the essence of a "share-in-savings" contract is that the government commits itself in ways that prevent it from walking away, inherently disallowing termination for convenience. Any "unearned profits," are already committed to the contractor. However, without this powerful tool for termination, the government has exposure to kinds of open-ended litigation and losses with which it has limited experience - and has not traditionally wished to develop such experience. A broad legislative authorization for SIS would be "Exhibit A" for contractors in a suit for termination of an SIS contract.

Also, it is not clear how the allocation of risks operates if an SIS contract should run into unexpected situations. What if the SIS contract's performance should be accompanied by what the government sees as dissavings, like a savings-and-loan financial institution than goes deeply into the red? Does the contractor commit to cover the losses? The discussions of SIS contracting to date have assumed savings rather than losses, but the first rule of sound government contracting law is to anticipate, before the government makes a contractual commitment, what would occur if the project produced losses.

8. Potential to undermine policy safeguards for government operations, e.g., protection of citizen data privacy:

Various government policy safeguards apply when agencies perform their operations themselves. The operations are performed by civil servants, in government facilities, surrounded by a host of explicit and implicit guarantees that performance occurs in the public interest. In procurement generally, a wide variety of explicit policy safeguards apply, from the Buy American Act to the wage and hour protections of contractor employees. In particular kinds of agency operations, specific policy safeguards may apply, such as Privacy Act protection for IT government operations involving the sensitive data of individual citizens.

A contractor who undertakes to perform a task on a fixed-price or a cost-reimbursement basis normally will not demand a high degree of flexibility to make changes in the existing mode of government operations. However, the SIS contractor may demand an unusual degree of flexibility in order to produce the "savings" that their profit will be based on. The SIS contract may impose much less in the way of controls and safeguards, including less protection for privacy, consistent with the contractor's insistence on flexibility of method in order to produce "savings." SSA beneficiaries could lose their privacy protections as a result. Other important government policies could face similar potential undermining.

9. Hard-to-undo mistaken choices about privatizing:

Currently, for non-SIS contracting, an elaborate A-76 process occurs before an in-house function can be contracted out. SIS proponents have left some indications that SIS could be an exception to A-76. Moreover, while it is not inherently necessary that SIS contracting be an exception to A-76, one can anticipate the arguments that SIS contractors will make as they seek to be an exception. The heart of A-76 is a comparison between the cost of the private contractor proposal, and the cost of continuing to perform the operation in-house. An SIS contractor would insist that, by definition, its proposal commits to producing savings hence, the very process of making an SIS offer and receiving an SIS award, demonstrates that the SIS alternative must be superior to performing the function in-house.

However, the debate over A-76 is a large one, beyond the confines of SIS, and to use SIS as an end-run around A-76 would require resolving the A-76 debate first. The reason for the careful scrutiny in an A-76 operation is to identify the potential downsides of privatizing. An SIS arrangement entered into without A-76 scrutiny exposes the government to the downsides of privatizing, without the appropriate prior scrutiny.

Moreover, if an ordinary privatization turns out to be ill-advised, it is hard enough to bring the work back in-house; the in-house employees and facilities may no longer be available. However, at least the usual rules about the government's power to terminate for its convenience apply. In contrast, because SIS can make it very costly for the government to cancel, a mistake in privatizing - say, because the contractor does not perform the operation fully satisfactorily, or because agency policy evolves in a few years in such a way as to require a major change in the operation so that the contract is no longer the best way to proceed - is considerably harder to undo.

10. Potential tampering with revenue operations:

One of the most prominent areas of federal in-house operations, particularly IT work, is revenue operations. SIS proponents have not expressly excluded revenue operations, indicating that a general legislative authorization of SIS would encompass the use of SIS contracts for such operations. After all, there is often a public wish that the Internal Revenue Service would move more quickly as to IT, and the proponents of SIS may very well be suggesting it as the way to jump-start improved revenue operations.

However, in many respects, applying SIS to revenue operations would involve extra types and levels of risks beyond what is found in applying SIS in other areas. Congress has legislated changes to IRS operations in recent years precisely out of concern that taxpayers can be subject to abuse. Imagine if revenue operations are turned over to SIS contractors with the instruction that their own profits will increase when they produce "savings" in revenue operations. Those "savings" might take the form of extracting more from taxpayers by undesirable methodologies, spending less on helping taxpayers, or devoting less effort to protections for taxpayers. Such concerns are why Congress has traditionally sought more oversight, not less, of revenue operations. At the very least, one would expect to develop some history of satisfactory experience with SIS in non-revenue operations, and to learn how best to safeguard against problems, before considering extending SIS to the sensitive context of revenue operations.


Founded in 1981, the Project On Government Oversight (POGO) is a nonpartisan independent watchdog that champions good government reforms. POGO's investigations into corruption, misconduct, and conflicts of interest achieve a more effective, accountable, open, and ethical federal government.

# # #