Incentives In Procurement Contracting

Jim Leitzel

• Publisher: Routledge
• Year: 2019
• eBook ISBN: 9780429722929
• Edition: 1
• Topic: Politica e relazioni internazionali
• Subtopic: Politica

This volume presents a nontechnical treatment of issues that arise in procurement contracting, with an emphasis on major weapons systems procurement. Employing the economic theory of agency as their analytical framework, contributors assess the incentives that arise, for both buyers and sellers, in different contractual settings. Procurement contra

1
Introduction

Jim Leitzel and Jean Tirole

This volume has emerged from two diverse developments of the past decade. The first development was the rise on the public agenda of U.S. defense procurement practices, unfortunately driven by a host of procurement scandals involving the now familiar trio of waste, fraud, and abuse. The second and more wholesome development was the coming of age of economic models of agency relationships, resulting in a better understanding of incentive contracting. The confluence of these two streams of activity, the application of agency models to defense procurement, has the potential to improve both procurement practice and the economic theory of contracts. The realization of some of these potential improvements is the goal of this volume.

The papers that compose this volume are largely non-technical accounts drawn from the authors' technical research. The economic theory of agency, which is notationally and sometimes technically demanding, is the common framework from which the authors draw their ideas. An unintended side effect of this volume is that it can serve as a non-technical, and far from exhaustive, introduction to agency theory. Readers unfamiliar with the agency literature may get a sense of both the power and the limitations of the economic approach to principal-agent relationships.

While the authors draw on U.S. defense procurement for inspiration, the models that they present apply much more generally. Public regulation of industries, government contracting in non-military sectors, and planning mechanisms in centrally planned economies are just three areas where the agency models presented in this volume have relevance. From within this general framework, many of the authors are able to draw policy recommendations for U.S. defense procurement. Possibilities for improved policies exist in other areas of application as well.

David Baron's paper, "Defense Procurement: Politics, Management, and Incentives," serves as a call to arms for economists interested in defense procurement. He notes that the current incentives in defense procurement are extremely distorted. Applications of optimal contract theory to defense procurement are therefore potentially quite useful, but the realities of the defense contracting world must inform the modeling. Baron discusses these realities and how the models must be altered to incorporate them. The incentive problems are particularly acute with respect to the procurement of major weapons systems. The defense realities in this realm include: (1) an evolving product that cannot be well specified in advance; (2) technological uncertainty; (3) incomplete contracts; (4) unanticipated events; (5) negotiated and renegotiated contracts; and (6) politics. These realities imply, for example, that the Department of Defense (DoD) is unable to commit to long-term contracts, and the contractor may have the bargaining power after production begins. Among the lessons for economic researchers are that optimal contract theory must focus on restricted types of contracts, earlier stages of the procurement process, and ex post efficiency evaluations.

William Rogerson's paper, "Inefficiently Low Production Rates in Defense Procurement: An Economic Analysis," provides one example of the type of analysis that Baron calls for. Production rates for most major weapons systems are too low relative to capacity. Rogerson presents a procurement game between Congress and the military that offers an explanation for this oft-cited fact. In his model, the military wants to maximize the number of units (of a weapon) that are produced. They move first, by choosing capacity, and then Congress chooses the production rate. In equilibrium, the military expands capacity until Congress is just willing to produce (instead of cancelling the weapon program), more output is produced than Congress would like, and the installed capacity is inefficiently large relative to the actual output. The model also offers an explanation of why DoD planning incorporates unrealistic projections of future congressional funding. Empirically, production rate estimates are not overstated once capacity is in place. The overestimates may therefore be part of the military's strategy in the capacity expansion game. Such considerations may also explain DoD reluctance to move to more flexible production technologies.

The next paper, "Incentive-Based Procurement Oversight by Protest," by Robert Marshall, Michael Meurer, and Jean-Francois Richard, also responds to Baron's challenge, this time by paying close attention to the institutional features concerning federal purchases of automated data processing and telecommunication equipment. The paper is primarily concerned with the protest mechanism, wherein a losing firm can challenge a procurement award. Prior to the creation of the protest mechanism, the main problems with data processing and telecommunication equipment procurements were overly restrictive minimum specifications, biased evaluation criteria, inappropriate exclusions of potentially bidding firms, and assessments by the procurement officer that were inconsistent with the stated criteria. Protests help to alleviate these problems by allowing losing firms to act as "private attorneys general." The authors identify some potential negative aspects of the protest process as well. These negative aspects include overdeterrence, where procurement officers behave over-conservatively in order to avoid a protest, and cash settlements, where the winning firm or the procurement agency pays off losing firms to settle protests. Marshall, Meurer, and Richard offer policy prescriptions that are designed to enhance the positive effects of protests while reducing the negative effects.

The next five papers are all concerned with the issue of quality in procurement. If the quality of a procured item is observable to the contracting parties and verifiable to a court, and if a firm that supplies the wrong level of quality can be sufficiently punished, then the buyer should be able to procure the appropriate quality. These five papers address the outcomes in procurements when the conditions for procuring quality are not as favorable.

In "Procurement and Quality Monitoring," Tracy Lewis and David Sappington examine procurement situations when quality is observable to the contracting parties but not verifiable to third parties such as courts. The actual quality of the procured item depends on "innate" quality and contractor effort, both of which are private information to the contractor. The timing of their contract model is: (1) the procurer offers a menu of two-part tariffs; (2) the contractor chooses a tariff, and chooses the actual quality of the item; (3) the procurer observes the actual quality and chooses the number of units to purchase. If actual quality is verifiable, ex post production is efficient, but quality is too low because of moral hazard. The contractor will not engage in enough effort to raise quality to the first-best level. If quality is not verifiable, ex post production is no longer efficient, as the contractor must be paid more than her marginal cost for her to be willing to produce a high-quality item. With price exceeding marginal cost, production rates are inefficient from a first-best standpoint. Lewis and Sappington note that both the contractor and the procurer may therefore be better off when quality is verifiable.

Julio Rotemberg's comments on the Lewis and Sappington paper are of sufficient independent interest to warrant a separate chapter. His discussion demonstrates that the key assumption driving the Lewis and Sappington result concerns cost information asymmetry. Specifically, the result requires that there be a greater asymmetry of cost information between the contractor and the procurer for the high-quality item than for the low-quality item. Also, in the case of symmetric information, it is demonstrated that the seller is made worse off with verifiability in a repeated setting, as opposed to the Lewis and Sappington single period result of buyer gains from verifiability. Finally, Rotemberg argues that "gold-plating" of weapons systems may be a result of a positive correlation between a firm's ability to produce complex systems and the usefulness of the firm's wartime production capabilities. The production of complex weapons therefore may serve as a useful signalling device.

"The Provision of Quality in Procurement," by Jean-Jacques Laffont and Jean Tirole, continues the examination of quality issues when quality is unverifiable. Laffont and Tirole analyze the influence of unverifiable quality on procurements both before and after the production source has been chosen by the procurement agency. The analysis in the post-source selection case hinges on whether quality is a search or an experience characteristic. (A search characteristic is one that the buyer can learn before purchase. An experience characteristic is only learned by the buyer after purchase and use.) For experience characteristics, the incentives to supply quality and incentives to reduce costs are inherently in conflict. Why maintain a reputation as a high-quality supplier if you bear all the costs of maintaining that reputation? With search characteristics, however, the number of units purchased can be used to induce quality, while an appropriate cost reimbursement scheme can be used to provide incentives for cost reduction. In the pre-source selection case, a hierarchical model is employed, in which the DoD knows quality but Congress does not, and there are two firms that can bid for the production contract. With soft (unverifiable) information, if the DoD can collude with either bidder, Congress imposes a symmetric auction and ignores quality claims. If quality information is hard (verifiable) and DoD can collude with only one bidder, the optimal auction favors the other bidder when no quality information is provided.

"The Choice of What to Procure," by Jim Leitzel, also concerns situations where potential sellers may have better information than the buyer about which projects are high quality. The case of a single seller is first examined. The seller has different inherent productivities in producing the various projects. These productivities and the seller's cost-reducing effort are private information. The existence of multiple projects may reduce the usual agency costs from this private information, however, as the projects compete with one another. A seller that has high productivity on one of the projects may be unwilling to claim that it is low productivity if the buyer can credibly threaten to procure an alternative project. With multiple sellers, the model involves two stages of competition. The first stage results in the choice of what project to procure, and the second stage involves the choice of the producing firm. Competition in the first stage is valuable to the procurement agency by allowing it to choose better projects. To encourage first-stage competition, an "incumbency advantage" for the second-stage competition has to be offered to the first-stage winner. This advantage, however, reduces production contract competition, so there is a tradeoff to the procurement agency between choosing better projects and procuring them at low cost.

The contribution of Benjamin Hermalin and Michael Katz, "Defense Procurement with Unverifiable Performance," also affords an interpretation in terms of the procurement of quality. The seller's effort helps determine the quality of the procured item, so the buyer cares about the seller's effort. Assume that the seller's effort or performance is observable, but that it is verifiable only with error, i.e., the court will sometimes make the wrong decision. Consider contracts that are written on such observable but only partially verifiable actions. After production, contract renegotiation arises, and if there is disagreement, the court's noisy decision is sought. As long as the court's decision is informative (in a statistical sense) of the seller's action, the full verifiability outcome can be achieved. The parties never go to court because the renegotiation takes place with symmetric information. Katz and Hermalin then ask: Given this result, why is defense procurement seemingly inefficient? They provide two potential explanations. First, the evidence on defense procurement inefficiency is often not inconsistent with the results of optimal contract theory. Second, various assumptions, like the government's ability to precommit, its representation as a unitary actor, and the possibility of asymmetric information at the renegotiation stage all limit the applicability of the analysis to the defense procurement realm.

Lars Stole and Michael Riordan each provide papers concerned with cost reduction in defense procurement. Stole's paper concerns the possibility of second sourcing via licensing, while Riordan examines the case for auditing the costs of defense contractors.

In "Licensing and Technology Transfer," by Lars Stole, the government chooses among three options: (1) the developer produces; (2) the second source produces; and (3) the second source produces using technology licensed from the developer. If licensing is unavailable, the optimal auction is biased towards the second source to induce the developer to bid more aggressively. Licensing acts as an additional competitor for both the developer and the second source, causing them to bid more aggressively. An extension of the analysis concerns the situation where the quality of the product depends on the developer's unobservable research and development expenditures. In this case, the same qualitative results continue to hold, but the bias against the developer in the optimal auction decreases with project quality.

Michael Riordan's paper, "Incentives for Cost Reduction in Defense Procurement," is primarily concerned with the benefits of auditing the costs of defense contractors. With a fixed price contract and a single period, contractors receive the entire benefit from cost reduction, and so make efficient effort choices. With multiple periods, however, the negotiated prices in subsequent periods depend on audited costs in the current period. With monopsony power, the government is able to appropriate many of the benefits of cost-reducing effort. A reduction in audited costs, even though it may increase the markup over audited costs, probably reduces the base price enough so that the contractor is worse off. Anticipating this, the contractor will bid less aggressively ex ante. The benefit of auditing is reduced monopsony distortions. The costs include effort distortions and administrative costs. Riordan concludes that the case for auditing, like the case for second sourcing, is narrow, and for the same reason: Smaller ex post rents result in less competitive ex ante behavior.

The final contribution, "Defense Procurement and Economic Theory," by Richard Stubbing, complements David Baron's paper in that it also offers an agenda for future research. Stubbing draws on his extensive defense experience to identify areas within DoD procurement that deserve the attention of economists. He notes that the defense industry does not operate in a market environment, leading to perverse incentives and results, particularly for major weapons systems. Problems include overstated requirements (gold-plating), concurrency in R&D and production, and the politicization of contract awards. Nine areas are suggested for improvement, including dual sourcing, inventory control, and contracting regulations.

The papers in this volume were presented at a conference held at the National Bureau of Economic Research in Cambridge, Massachusetts, on September 22-23, 1989. The conference was sponsored by the Program for the Integration of Economics and National Security of the Pew Charitable Trusts. The discussants' comments from the conference as well as the general discussion from the floor are included in distilled form after the relevant papers.

2
Defense Procurement: Politics, Management, and Incentives

David P. Baron

Introduction

The political context of defense procurement and the bureaucratic and management processes of program initiation and management have considerably distorted incentives. Fox (1988, p. 317) concludes:

The Defense Department customarily does business with an inverted system of rewards and penalties. Contractors are often rewarded for higher than planned program costs with increased sales, contributions to overhead, and profits. The system encourages government managers to place a far higher priority on gaining congressional approval to begin a new weapons program or obtain additional funding for an ongoing program than on controlling costs for existing programs. The acquisition cost problems of the 1980s are not aberrations; they are the result of many government and industry participants reacting in accord with the distorted rewards and penalties inherent in the acquisition process.

This characterization of defense procurement is so different from the world of the models, and their derived implications, used by economists in studying procurement contracting that two conclusions seem warranted. The first is that there is considerable opportunity for optimal contract theory to contribute to improved efficiency in defense procurement. The second is that the realities of defense procurement have important implications for how economists should study contracting incentives and efficiency in defense procurement. Instead of focusing on the implications of contracting and incentive theory for defense procurement, this paper focuses on the implications of the political, bureaucratic, and managerial dimensions of defense procurement for research on contractual incentives intended to improve procurement efficacy.

Defense procurement includes a wide variety of goods ranging from standard items such as uniforms and ammunition to major weapons systems whose acquisition may take 10 to 15 years to complete. Procurement of standard items is amenable to competitive bidding and fixed-price contracts, but competition and fixed-price contracts are less, and negotiated contracts more, appropriate for major weapons systems that are complex, involve yet-to-be-developed technologies, and have performance objectives that may be unattainable or unattainable at reasonable cost. Particularly in the case of systems that involve research and development, the ability to anticipate future technological developments is limited, and even if the possibilities can be identified, it may be impractical to provide contractual contingencies for all possible events. In addition, defense procurement involves not only DoD and defense contractors but also political actors in the Congress and the executive branch, contributing to DoD unwillingness or inability to commit credibly to the terms of contracts. Consequently, in the case of the acquisition of the major weapons systems which are the focus of this paper, it is necessary to work within a framework of an evolving "product," technological uncertainty, incomplete contracts, unanticipated events, negotiated and renegotiated contracts,...

Table of contents

• Cover
• Half Title
• Series Page
• Title
• Copyright
• Contents
• Acknowledgments
• 1 Introduction
• 2 Defense Procurement: Politics, Management, and Incentives
• 3 Inefficiently Low Production Rates in Defense Procurement: An Economic Analysis
• 4 Incentive-Based Procurement Oversight by Protest
• 5 Procurement and Quality Monitoring
• 6 The Provision of Quality in Procurement
• 7 The Choice of What to Procure
• 8 Defense Procurement with Unverifiable Performance
• 9 Licensing and Technology Transfer
• 10 Incentives for Cost Reduction in Defense Procurement
• 11 Defense Procurement and Economic Theory
• About the Book and Editors
• About the Contributors