chevron-down Created with Sketch Beta.

Public Contract Law Journal

Public Contract Law Journal Vol. 49, No. 4

Empowering the End-User as Procurement Agent Through E-Commerce

Abraham Young


  • Discusses the "agency problem" in public procurements
  • Considers how procurement reform, such as creating a government e-commerce platform and reducing the regulatory burden on small value government purchases creates solutions to the "agency problem"
  • Proposes mitigating the "agency problem" by granting procurement authority directly to the end-user of the product
Empowering the End-User as Procurement Agent Through E-Commerce
Westend61 via Getty Images

Jump to:


Public procurement is an example of the principal-agent paradigm, in which a principal employs an agent to make decisions on behalf of the principal. This agency relationship has been scrutinized from countless academic angles because of the intractable problems it creates, referred to generally as the “agency problem.” In this problem, almost invariably, the agent will diverge from the principal’s optimal outcome, out of the agent’s selfish interests. The agency problem inevitably causes residual losses, despite extensive efforts and costs in monitoring the performance of procurement agents and sanctioning aberrant agents. Whereas in the commercial sector, costs and losses due to the agency problem are borne by the principal, and thus efficiently managed, the residual loss in government procurement often falls on the end-user (that is, the individual who actually uses the procured item). The agency problem creates large inefficiencies that are inconspicuous precisely because the burden is on the end-user. Granting procurement authority directly to the end-user could eliminate multiple levels of cumulative agency costs. By allowing users to purchase directly from electronic marketplaces, the government could make more efficient use of public funds and acquisition resources.

I. Introduction

Public procurements pose a practical conundrum because of the “agency problem,” in which agents employed by the principal to help accomplish the agency’s mission almost invariably stray from the agency’s optimal outcome due to selfish interests. Specifically, in the U.S. federal procurement system, the equipment acquired is often not what users would have chosen themselves, despite the efforts of countless members of the acquisition workforce. This is particularly the case for individual equipment used in the military services, because these purchases do not attract the media coverage of large procurements, and no functional feedback system ensures consistent high quality. Often, the quality of individual military equipment is so poor that service members prefer to purchase commercially available substitutes with their own money — often from electronic marketplaces. This opens the door to an obvious question: why not allow users to purchase directly from commercial marketplaces, bypassing an inherently flawed procurement system?

The government currently can work only through agents in purchasing items, such as military uniforms and pieces of equipment, for individual service members. This article describes the costs of agent-centered procurements in the military. It argues that unrecognized burdens caused by the agency problem are borne by individual end-users and thus are organizationally invisible in the procurement system. Ultimately, this article argues for empowering individual end-users in appropriate circumstances to make individual bestvalue determinations — which is precisely what the federal government proposes to do in the current initiative that would open electronic marketplaces to government users using government purchase cards. This approach would eliminate layers of aggregated agency costs and merge the agent’s incentives with the end-user’s particular preferences. This approach would also help ensure a well-considered choice by trading a disinterested purchasing agent for a user with personal experience and an anticipation of actually using the product. This would allow the government to benefit from a thriving competitive market that already exists, while simultaneously providing users with high-quality goods and services.

Eliminating layers of agents has remarkable implications for efficiency in procurement. To say that this would double efficiency in the procurement of individual equipment might seem like a bold claim, but the author believes it is possible to achieve this with a simple, readily available approach. However, it requires jettisoning an antiquated procurement system in exchange for a twenty-first-century approach.

To make sense of this potentially massive shift in procurement strategy, this article begins in Section II with a case study of the procurement system for individual military equipment from the perspective of a U.S. service member. Section III explores the various costs and burdens caused by the principal-agent procurement paradigm, focusing on the detrimental impacts on the final outcome. These detrimental impacts, referred to as residual loss, are the result of a series of agents’ decisions, each of which diverges from the optimal outcome. This article argues that the residual loss is borne by the end-user in the government procurement context, rather than the principal, thus creating a “moral hazard” — an opportunity for the agent to exploit his position. To dispel the common notion that agent’s performance can be corrected simply with better monitoring or sanctions, this article, using the lens of other disciplines, highlights the less understood pitfalls of agent performance. Section IV discusses current trends in procurement reform, such as creating a government e-commerce platform and reducing the regulatory burden on smaller value government purchases, as well as considering how these trends pave the way for a solution to the agency problem. Section V ultimately proposes that the agency problem could be mitigated by granting procurement authority directly to the end-user. Government users could make purchasing choices based on the experience of others through an e-commerce platform. Doing so would reduce waste and unnecessary burdens on the acquisition workforce, and increase competition-based quality and innovation. To use an American aphorism, walking a mile in a service member’s boots is a good place to begin.

II. Disenfranchised End-Users and Disinterested Purchasing Agents

In our case study, a fictional but representative Private First Class (PFC) Jones is deploying to Afghanistan. Private Jones is an infantryman and has been in the Army for about a year since he completed basic training. In preparation for deployment, PFC Jones is issued extra boots, uniforms, gloves, and protective eyewear to augment his initially issued equipment. While receiving his boots, he is told not to pull the laces too tight, because the eyelets tend to rip out. When he first washes the gloves, the hook and loop fasteners disintegrate, and the seams begin to split. His new eye protection fits so poorly that it falls off his face when he looks down, even though the temple pieces are adjustable.

Rather than face the inconvenience of turning in equipment for replacements that he knows will suffer the same problems, PFC Jones takes a cue from more senior soldiers in his squad. He purchases substitutes for his issued equipment online in the commercial marketplace. This is not easy on his monthly salary of less than $2000, but it is the only fast and reliable solution. It is, after all, a common practice.

Instead of the issued equipment, PFC Jones now wears equipment that he bought himself. The equipment he bought, readily available in the commercial marketplace, is designed to meet his uniform requirements. In fact, the Army recognizes that soldiers frequently choose to purchase their own equipment and, as a result, specifically authorizes the wear of “commercial boots of a design similar to that of the Army combat boot.” The only drawback that PFC Jones faces is that he had to pay for the equipment himself. However, rationalizing the expenditure is simple: a few hundred dollars is well worth it to PFC Jones, because his life depends on the fit, function, and comfort of his personal equipment. The boots, gloves, and eye protection that PFC Jones received for deployment sit unused in his locker back in the United States during his deployment.

PFC Jones now joins approximately 2.7 million U.S. service members who have deployed in support of recent overseas conflicts. Like countless service members before him, he fights for the United States wearing boots and other uniform items he purchased for himself. According to some estimates, half of U.S. service members purchase their own boots, despite being issued government-purchased boots. The problem is that some issued equipment is so poor that service members would rather pay with their own money in exchange for better equipment.

The failure to consistently provide useful equipment is particularly puzzling considering that the acquisition workforce — the agents who perform the procurement function — costs the Department of Defense (“DoD”) over $60 million per day. On top of that, the United States has the best-funded military in the world, with a defense budget nearly three times that of the closest competitor. How does the agency problem contribute to this?

Back in the United States in our fictional case study, Mr. Smith works for the Defense Supply Center Philadelphia in the Textile and Clothing Directorate (C&T). He is a purchaser for DoD uniforms and gear — a busy job as Mr. Smith handles several contracts out of the thousands handled by C&T as a whole. This is especially true as new equipment needs and uniform turnover have recently increased his workload.

Mr. Smith’s job is to find the “best value” contract that meets the end-user requirements. Though it has been years since Mr. Smith himself was in uniform, he purchases standard equipment for tens of thousands of service members. Mr. Smith uses taxpayer funds to purchase equipment that he will never use. These items are then distributed to Central Issue Facilities and the Rapid Fielding Initiative across the military before being issued to service members such as PFC Jones. Since the end-users are so highly dispersed, Mr. Smith will likely never ask PFC Jones whether he is satisfied with his boots, gloves, and eye protection, or why he is not using them.

This disconnect between purchasing agents and end-users is well known, though less well understood. Congress has even taken steps to increase interaction between the acquisition workforce and end-users like PFC Jones in an attempt to get purchasing agents to better understand the mindset and requirements of the individuals who will use the equipment.

Institutional mechanisms are in place for determining what features are necessary for new equipment items, and field testing is conducted. What is lacking is a timely feedback mechanism by which PFC Jones can contact Mr. Smith about problems with his equipment to improve the equipment. Improvements to issued gear can take years to implement. Thus, from the service member’s perspective, problems with equipment go unresolved, and the institutional link between user and buyer is non-existent.

The key to understanding the conundrums that plague public procurement is to better understand the relationships between the parties involved in each procurement. As many have observed, the principal-agent model, and its social-science-correlates, identify the many causes of inefficiency suffered by traditional military procurement:

The procurement officer is not the end-user of the procured goods and services, does not manage the resulting contract, and of course does not pay for the contract deliverables. Nor does the procurement officer pay for any problems that might arise during performance, such as change orders, litigation, or procuring a brand-new contractor to replace a failed one. The principal-agent problem is therefore clear: The procurement officer internalizes the costs of the instant procurement and negotiation process, but does not internalize the costs related to contractor performance.

By examining the disparities between the parties’ goals, incentives, burdens, and information, the true costs of agent-centered procurement become clearer. Reducing — or even eliminating — these disparities through structural changes to the procurement system would increase the overall efficiency of government procurement by reducing exposure to the agency problem itself.

In some situations, procurement may be better done by the end-user, eliminating the need for an agent altogether. This is both a feasible and desirable outcome. Advances in information technology can revolutionize markets simply by reducing transaction costs and eliminating intermediaries; in procurement, information technology such as electronic marketplaces can allow users to purchase directly and thus bypass the agency problems that plague public procurement. In certain circumstances, the time has come to empower the end-user to make individual best-value determinations for uniform items. This modification would eliminate layers of superfluous agency costs, enhance competition, harness commercial inventiveness, and reduce waste.

III. Agency Theory

Scholars have applied agency theory to government purchasing as a tool to analyze costs inherent in procurements performed by agents. Agency theory explains the economic problems associated with a scheme in which a principal empowers an agent to make decisions on that principal’s behalf. Agency theory’s usefulness is that it provides a framework for methodically analyzing and then aligning the incentives between a principal and an agent in order to increase overall efficiency. Analysis of government procurement under agency theory has a subtler utility as well, because the same principal-agent paradigm has been examined from the unique viewpoints of other social sciences. The social sciences provide fundamentally different perspectives concerning the agency problem. By considering these perspectives, one can assess the depth and complexity of the agency problem in government procurement.

The traditional military procurement system in the case study above involves layers of complexity. In our simplified example, the Military Department (principal) empowers Mr. Smith (agent) to make purchasing decisions on behalf of the principal, for the benefit of, and use by, PFC Jones (end-user). However, in reality, the principal-agent relationships in this procurement are far more complex.

The author argues that, under our Constitution, the people of the United States are the true principal. Due to the nationwide diffusion of responsibility and control among the population, most of these principals are neither knowledgeable nor interested in the acquisition of military equipment and have delegated their responsibilities as a principal to Congress. Congress then, in this article, is treated as the initial principal because of the constitutional power of the purse, even though individual legislators are, in a sense, the agents of their constituents.

Each year, typically in a two-step legislative process, Congress first authorizes spending and then appropriates funds for the military to use for the national defense. Congress enacts provisos and requirements for the spending of the appropriated funds through the annual National Defense Authorization Act (NDAA). Each military department is then the agent for spending the funds appropriated by Congress for the national defense, within the boundaries set by Congress, the principal, through the NDAA. Each military department maintains contracting activities, as agents to accomplish required procurements. These contracting activities, in turn, each employ multiple agents, called contracting officers. Through this cascade of delegated authority, contracting officers have authority to bind the United States contractually.

In this way, Congress (the initiating principal) delegates spending authority with statutory caveats and restrictions, through a hierarchy of agents and sub-agents. Each sub-agent then delegates a gradually decreasing amount of authority to yet another sub-agent to perform the actual procurement. Finally, a sub-agent (in this case, Mr. Smith) chooses a contractor to supply either goods or services to the military. Though a contractor is not formally a sub-agent in the contractual sense, the contractor relationship includes many features of the agency problem as well. Thus, this relationship, too, can be viewed through agency theory, with the contractor as a final sub-agent whose principal is the contracting officer.

The end-user’s needs drive the principal’s procurement requirements, making the end-user most aligned to the principal’s priorities. Nonetheless, the end-user is neither involved in the delegation of procurement authority nor, typically, in the agent’s procurement decision. Thus, though the end-user’s needs are the cause of the procurement in the first place, the end-user is not a prominent part of the current procurement process.

The magnitude of the military procurement enterprise is stunning. The acquisition workforce maintained by the DoD is approximately 165,000 strong. Of these, about ten percent are uniformed personnel who serve predominantly in managerial positions. The acquisition workforce includes roles ranging from engineers and program managers to contracting officers and auditors. While some of this workforce has prior DoD or even active duty military experience, the majority of new acquisition workforce entrants each year have none.

At an annual fully-burdened personnel cost of approximately $15.8 billion for this defense acquisition workforce, the first and most apparent cost of agent-centered procurement is apparent: the government must pay its agents. However, the actual cost of employing an agent is much greater than simply the value of pay and benefits. The agency problem is not only about the explicit costs of hiring the agent, but rather about the transactional burdens caused by delegating a decision. This sequential use of agents and subagents to perform purchasing decisions has distinct costs.

A. Agency Costs in Government Procurement

The academic literature includes three different cost categories associated with employing agents to make decisions: monitoring costs, bonding costs, and residual loss. Monitoring costs are those incurred to ensure that an agent’s performance is in line with the principal’s priorities. Bonding costs, however, are more directly borne by the agent — they reflect the resources forfeited or the punishment endured if the agent strays from the principal’s goals — to ensure that choices made are in line with the principal’s priorities despite differing incentives. In the context of a government principal, it is not surprising that these bonding costs take the form of punitive sanctions.

Monitoring and bonding are sufficient to ensure optimal agent performance in a world of perfect agents. Unfortunately, a perfect agent, who reliably maximizes the principal’s outcome regardless of personal motivations, is a fiction of economics. Even well-crafted systems for monitoring and bonding agent performance can be opportunistically manipulated by agents with different incentives.

Finally, residual loss accounts for the difference between an ideal outcome from the principal’s point of view, and that which the agent actually chooses on behalf of the principal. This loss occurs in spite of the monitoring and bonding costs inherent in the principal-agent relationship, both of which are specifically intended to reduce the magnitude and frequency of the residual loss.

Despite these inevitable costs of the agency relationship, it is nevertheless efficient to use an agent to make decisions where the benefits of employing an agent outweigh the cumulative costs. The benefits of employing agents in public procurement — aggregating purchasing expertise in a few persons, concentrating a procurement decision in one person rather than among diffuse stakeholders and focusing accountability in one person — help explain why governments rely too much upon procurement officials. The common use of procurement officials as intermediaries also means, though, that procurement systems bear common costs in ensuring that agents follow the government principal’s purposes. To employ an agent efficiently, however, it is necessary to be able to calculate (or at least estimate) the magnitude of the total costs of employing an agent. While a government may usually be circumstantially bound to operate through agents, there are occasions when the agency problem may be bypassed. Before exploring this issue, a closer understanding of the types of costs involved in an agent-centered procurement is required.

B. Monitoring Costs

Monitoring costs are incurred by the principal to ensure that an agent does not deviate from the principal’s priorities. These costs include the burden of communicating the principal’s priorities and ensuring that the agent complies with them. Monitoring costs differ throughout the principal-agent hierarchy, with differing monitoring schemes applying to each.

For example, Congress appropriates funds out of the national revenues to provide for defense, including equipping service members. Separately, Congress also spells out in detail, through the annual NDAA, what the DoD may and may not purchase with those appropriated funds. The extensive detail in the annual military authorization and appropriations legislation is akin to a limited delegation of authority to the DoD to make spending decisions for the national defense. Communicating a principal’s priorities is part of the monitoring process. This can be viewed as an ex ante monitoring cost.

Communicating the principal’s priorities to the agent is only the first part of monitoring costs associated with employing an agent to carry out a decision. Authorizing legislation also typically mandates annual reporting of the expenditure of public funds, which can be considered an ex-post monitoring cost.

Though beyond the scope of this article, the cause of Congress’s extreme fidelity in monitoring is the subject of debate. This apparent lack of trust in the agent may undermine agent’s intrinsic motivation, further justifying and reinforcing the extensive use of monitoring and sanctioning measures to combat risk aversion. Whether the explanation is a lack of trust or even the principal’s own moral hazard, behavioral economists have documented a powerful bias against delegating authority. This bias is true even when delegation would seem to be optimal. Retention of control is of value to the principal, even in situations requiring an agent. The cost of this extensive monitoring scheme puts added burdens on the agent, both before and after the procurement, to comply with the principal’s guidance and to report results.

Monitoring continues throughout the principal-agent hierarchy from the highest to the lowest level. At the final sub-agent level, there are extensive monitoring schemes to diminish the information disparity between the contracting officer and the contractor. For example, cost-reimbursement contracts of over $50 million, with some exceptions, are subject to the Cost Accounting Standards (CAS). The CAS are a set of nineteen accounting standards requiring disclosure to the government of contractor accounting practices and consistency thereafter in these practices. Contractor failure to abide by these accounting regulations may result in financial penalties. This allows the contracting officer to gain far more information than normally available to a principal regarding the cost of performance on a contract to ensure that a contractor is not acting opportunistically. Even with the range of statutory and regulatory tools at the principal’s disposal to monitor performance, in our example, the contractor does not fully meet the requirements to the satisfaction of the end-user.

C. Bonding and Sanctions

Bonding ensures that the agent makes choices in line with the principal’s priorities despite differing incentives. In the government context, legislative bonding takes the form of sanctions at every level of the principal-agent–subagent relationship. These sanctions ensure that departures from the guidance of the legislature are reported and could be administratively or even criminally punished.

Sanctions augment both intrinsic and extrinsic incentives. Especially for government employees, where monetary performance bonuses are rare, intrinsic factors are important motivators. Social norms and personal moral values such as duty, integrity, and altruism may motivate an agent to expend full effort in service of the principal’s aims.

Less obviously, sanctions also function as an inverse performance bonus. Whereas agent incentives are usually tied to the success of an outcome, a similar outcome can be achieved by paying the agent a salary assuming a successful outcome, and then sanctioning the agent for failure. Notably, even if the bonus and sanction are equal but opposite, resulting in mathematically similar outcomes for success or failure, agents respond differently to prospective bonuses or punishment. Due to loss aversion, the fear of a sanction would tend to impact agent behavior more than a bonus of equal magnitude. At the organizational scale, widespread loss aversion among agents brought on by the threat (real or imagined) of punitive sanctions, could metastasize into an organizational norm of risk aversion.

The U.S. military procurement system, as created by Congress, also is replete with examples of sanctions at every level of the principal-agent hierarchy. At the executive agency level, an expenditure of funds in violation of a legislative appropriation is punishable under the Anti-Deficiency Act (ADA). This statute requires the executive agency to report violations to the President, Congress, and the Comptroller General. The ADA even calls for criminal penalties for purchases that violate an appropriation, including confinement for up to two years and a $5,000 fine. The annual report by the Government Accountability Office lists ADA violations by agency, amount, and result. Each agency letter explains the spending violation and the steps taken to ensure that such an event never happens again.

Though no officials have been convicted of an ADA violation, adverse administrative personnel actions frequently arise after an investigation into the unauthorized expenditure of funds. It is possible that an unrecognized and unquantifiable consequence of this agent-sanctioning measure is the often-cited risk aversion in the procurement community. Even if a prison sentence is not likely for an ADA violation, it is still likely to have adverse career impacts.

Multiple statutory sanctions exist at the final sub-agent level and are aimed at ensuring contractor compliance with government requirements. For any fraudulent information in billing the government for services, there are civil or criminal sanctions under the False Claims Act. The contractor may face large monetary penalties for every false claim and treble damages stemming from the fraudulent claim. Should a contractor fail to fulfill the contracted duties for which they are the agent, the contractor may also be suspended or debarred at the executive agency’s discretion. A contractor who is suspended or debarred cannot receive any additional contracts from any component of the federal government, which can force some businesses into bankruptcy.

As discussed, the procurement system includes extensive monitoring and sanctioning regimes at every echelon of the principal-agent hierarchy to ensure agent compliance with the principal’s priorities. Even so, the actual choices of the agent on behalf of the principal may deviate from an optimal outcome. This discrepancy is called the residual loss. Whereas monitoring and sanctioning increase the transaction costs of procurement, residual loss detracts value from the outcome itself.

D. Residual Loss in Agent Procurements

Despite taking monitoring, bonding, and even sanctioning efforts to ensure perfect compliance with a principal’s priorities, an agent’s performance may nevertheless diverge from what the principal considers an optimal choice. Referred to as “residual loss,” this divergence includes all unpreventable sub-optimal results from employing an agent.

Residual losses stem from agent motivations, incentives, and biases left unaddressed by the monitoring and bonding efforts. These counterproductive incentives, biases, and motivations, and the losses they cause, are inherently hard to detect, making them also difficult to prevent.

A cross-disciplinary examination of the agency problem reveals situations when similar pitfalls are likely to occur in government procurements. The following section explores the impact of economic incentives, inherent biases, and sociological tendencies as they impact principal-agent transactions.

1. Economic incentives

It has been said that it “is inconceivable that purposeful action on the part of human beings can be viewed as anything other than responses to incentives.” This broad statement epitomizes the core of the problem explored by agency theory, namely that the principal and agent inevitably have different motives that drive the outcome. Even in situations where one would expect an agent’s interests to align nearly perfectly with the principal’s priorities, there are still measurable differences in agent performance caused by the divergence of incentives.

For example, an agent may have a strong incentive to ensure quick or easy, rather than optimal, performance of the principal’s wishes. That is, despite having the knowledge of priorities, and the expertise to achieve them, an agent may perform less than his ability otherwise would allow, because of divergent priorities. This happens in contexts outside of the government procurement realm as well.

In a study involving real estate agents, researchers Steven Levitt and Chad Syverson discovered quantifiable evidence of residual loss. When real estate agents sell their own homes, as opposed to representing a seller, real estate agents, on average, keep their homes on the market longer and achieve 3.7% higher sale prices. This figure demonstrates the concept of divergent priorities: when not acting as an agent for someone else, and instead focusing solely on their own priorities, real estate agents consistently achieve superior results.

This study’s analysis is based on a sample size of almost 100,000 home sales, meaning that, collectively, the real estate agents in this study may have foregone nearly $1 billion when selling for others, due to agent underperformance. This outcome is not very promising, given the pure simplicity of the principal’s priorities in selling a home: sell the house for the highest price possible. Despite potentially conflicting with industry codes of ethics, this seems to be a common occurrence, for the same reasons that we see in other agency problems.

When an agent sells another person’s house, the agent receives only approximately 1.5% of the sale price. The seller, however, who is the principal in a real estate transaction, receives the majority of the sale price and, therefore, benefits far more from an increased purchase price. Conversely, the agent bears the burden of the additional time and effort required to achieve that optimal sales price. By selling homes more quickly, an agent may achieve greater commissions in the aggregate by selling more homes overall. However, when an agent sells his own home, the agent is more incentivized to invest both time and effort to achieve a greater sales price. This is true even after tying the agent’s fee directly to the sales price. If agents truly sought optimal performance for the sake of their principals, there would be little or no difference between the prices achieved by real estate agents when selling their own homes versus those of sellers they represent.

Though there are clear differences between a contracting officer and a real estate agent paid on commission, there are also important similarities. The contracting officer bears the burden of monitoring contractor performance, just as the real estate agent bears the burden of ensuring the best sales price. Additionally, there is no incentive for the agent to take on the additional work required to obtain a better outcome because procurement officials are paid the same, regardless of whether he successfully chooses the best value product. Further, ensuring the best value product is a complex effort, including inspecting the goods or services and confronting the contractor for poor quality. Because the agent does not use the goods himself, the job of inspecting the products on a given contract depends on outside resources and reporting, further straining these divergent priorities.

In the case of Mr. Smith, the burdens of ensuring quality are clear. He must collect user satisfaction data from soldiers, like PFC Jones about the equipment that he bought, and analyze this data to correct any functionality or quality issues, such as acting against the contractor or changing the requirements of the contract. Even then, the only way to ensure that quality is maintained is to continue soliciting and reacting to feedback from the end-users. Just like real estate, however, any negative impact of poor functionality or quality is borne by others, while Mr. Smith, as the agent, would personally bear the burden of ensuring quality. Thus, no financial incentive exists for Mr. Smith to do the extra work to ensure functionality and quality.

When the principal is not the consumer of the products or services, the agent’s performance is more difficult to assess through traditional methods. This is because the burden of the residual loss is shifted from the principal to the end-user. Residual losses are caused by more than simply misaligned economic incentives.

2. Biases and transaction costs

Another cause of residual loss stems from the biases of each agent. Accumulated biases quickly sap the value of a transaction, contributing to both the magnitude and intractability of the residual loss. The relationship between an individual investor and the investment manager is illustrative of how inherent agent biases increase transaction costs without any overall increase in value. The world of finance offers a stark and persistent case study in unproductive agent biases. An investor (the principal) relinquishes control over investment decisions to a professional manager (the agent) who invests the funds on the investor’s behalf. These mutual fund managers then charge the investors a percentage of the amount invested as an annual fee.

In his now-famous study, Dr. William Sharpe showed that these mutual fund fees are negatively correlated to investor returns. Fifty years after Dr. Sharpe’s study, and presumably after industry efforts to improve manager performance, Dr. Sharpe’s observation still holds true. Investment fees are the best predictor for investment returns — but they cause the opposite effect from what would be expected. That is, higher fees led to lower returns for every quintile of cost across six categories of funds studied. On average, active investment fund managers fall short of the annual market returns, and performance from year to year is inconsistent. Simply put, the more you pay an agent to invest your money, the lower your expected average returns. Studies of the magnitude of this effect estimate that an investor who pays the lowest fees will earn between twenty percent and sixty-five percent more in returns than a similar investor who pays the highest fees hoping for higher returns. It turns out that when agents are allowed to exert a lot of control over a situation, they are likely to erroneously draw a causal link between their efforts and any beneficial outcome. Worse yet, they are likely to irrationally retain as much control over their decisions as possible.

This phenomenon can be explained by what Dr. Ellen Langor calls “the illusion of control.” The illusion of control is when a person believes that his personal probability of success is greater than that warranted by the actual statistical likelihood of success. This illusion occurs in situations where a person chooses his own position in a situation otherwise dictated by random or other uncontrollable factors, such as the stock market. This phenomenon may also explain the perennial popularity of lotteries, in which players choose their own numbers, despite the odds remaining constant even if numbers were chosen randomly instead.

Similarly, as the principals (at each cascading level of the principal-agent-sub-agent hierarchy) in government procurement exert control over each subordinate agent, this influence increases the overall cost and difficulty of the transaction, but not necessarily its value. Worse yet, as each agent employs a sub-agent, the inherent bias is to retain as much control as possible even when more complete delegation would be most appropriate given the circumstances. As a result, the cascading agents assert inefficient control over the contract, thereby collectively increasing the likelihood of ineffective procurement.

This phenomenon also reflects another inherent bias at work in agents: the status quo bias. When faced with a choice, most people are concerned more with the potential downsides of change rather than with the potential benefits. Thus, when there is a choice to act or not, the tendency, without an overwhelming reason to act, is to not.

The Berry Amendment is an example of how retention of control may create anti-competitive impacts, engender inefficiency, and ultimately result in additional residual losses. When Congress appropriates funds for the DoD to buy uniforms for service members, the uniform procurements are separately subject to the Berry Amendment. Originally enacted during World War II, the Berry Amendment requires, with few exceptions, that when the DoD purchases items made from textiles, the products must be made in the United States.

When the Berry Amendment was initially passed, market conditions were different than they are today. Twenty years after the Berry Amendment, in the 1960s, ninety-five percent of clothing used in the United States was still made domestically. Today, a mere three percent of clothes and two percent of footwear used in the United States are made domestically. Due to the Berry Amendment, those few remaining manufacturers have a statutory oligopoly as military suppliers and understandably prefer the status quo. Thus, by authorizing the purchase of equipment, the Berry Amendment limited the field of available vendors to a mere two percent of manufacturers, acting as an extraordinary restraint on the delegated authority to purchase military uniforms and footwear. Similar to other protectionist statutes, this anti-competitive effect eliminates much of the remaining vendors’ economic need for innovation, quality, and efficiency. Any residual loss stemming from the inefficiency engendered by the Berry Amendment is a burden that is grudgingly borne by the agency but is then literally distributed to and worn by the end-user.

As a result, whether driven by lack of innovation or competition or simply stagnant specifications, multiple versions of combat boots have gone decades without meaningful improvements. Meanwhile, the commercial marketplace, which must compete on both cost and quality, continues to innovate. In one particularly apt example, a rubberized over-boot for extreme cold weather, affectionately known as “mickey mouse” boots, are made in such an antiquated fashion that the industry simply does not use the technology any longer. This has finally forced the military to look for substitutes for a boot that has not changed in nearly sixty years. Not surprisingly, extreme cold weather boots are readily available in the commercial sector, with ratings down to negative 148 degrees Fahrenheit. The Berry Amendment’s dramatic example of retention of control by Congress over the purchasing decisions of contracting officers blocks the purchase of better quality equipment from foreign manufacturers.

While economic incentives (or lack thereof) and personal biases may contribute to potential residual loss in a principal-agent situation, interpersonal and even inter-group sociological tendencies may as well.

3. Unwarranted Cooperation with Contractors

Another source of residual loss is unwarranted cooperation between the agent and the contractor. Interestingly, the urge toward social cooperation is a result of biological factors, and the circumstances of repeated interactions with contractors, in addition to other common motivations. Further, circumstances may play a critical role in determining whether an agent subordinates social pressures in pursuit of the principal’s interest.

Cooperation and social conflict are both neurologically processed in ways likely to sway an agent toward cooperation. One study showed that the drive for mutual cooperation and reciprocity is a powerful neural force operating in the instinctual regions of the human brain. Cooperation triggers the same neural networks of human brains that are triggered by other pleasurable experiences, thereby reinforcing social cooperation. In experiments in which one person is given the authority to unilaterally grant a sum of money to another player, the strongest tendency is to do so evenly, even when that means that the one deciding takes a smaller portion. This is true even when the cooperation only benefits the other party, as is the case in purely altruistic reciprocity. In contrast, social conflict triggers the same neural networks associated with physical pain.

On the one hand, when a contracting officer makes a decision that affects the contractor, the urge to cooperate begins instinctually. If carried out, even altruistically, neural reward centers register pleasure, and the behavior is reinforced. On the other hand, if the contracting officer must enforce contractual obligations to the detriment of the contractor, social conflict registers in the same brain regions as physical pain. This neural tendency to cooperate is reinforced by one of the most important findings in game theory: cooperative behavior may emerge in long-term interactions.

Cooperation can occur even in inherently non-cooperative competitive interactions, so long as the number of iterations of the “game” is unknown by both parties. In long-term interactions among competitors, some form of cooperation is likely to develop. One famous example of this occurred World War I. Opposing sides created informal truces despite the contrary orders of superiors. This phenomenon culminated in the unsanctioned Christmas Truce of 1914, which saw an unofficial temporary peace that brought about the intermingling of combatants from opposing armies.

However, if the same primal urge to cooperate with contractors existed, in spite of best interests of the government, there must be some evidence of it. Consider incentive and award-type contracts. In this type of government procurement contract, the fee or profit that a contractor is paid is tied to technical or performance objectives. These performance measures may be tied to timely performance, keeping total costs down, or to other suitable goals. Using incentive-type contracts, a contracting officer is able to craft a scheme of payment to reward superior performance and deter mediocre performance.

The potential benefits of this type of contract have been well understood and studied. Nearly forty years ago, scholars studied the optimal use of these incentives to motivate superior contractor performance. Given the purpose of these incentives, however, a government contracting officer could not use this power effectively to motivate contractor performance without additional oversight over the contractor’s performance.

A 2005 General Accountability Office (GAO) study of incentive contracts showed that contracting officers paid award fees regardless of contractor performance. The GAO study found that the DoD paid a median award of 90% of the available fee to contractors whether or not their performance fell below, met, or exceeded expectations. Over 90% of the total possible award was paid in 320 of the 572 evaluation periods studied. This resulted in an estimated $8 billion in award payments not appropriately tied to performance. The GAO’s conclusion was that the DoD was unwilling to deprive contractors of any meaningful portion of the award fee, even in the shorter context of any individual evaluation period.

This initial failure of incentive contract mechanisms led some to question whether incentive contracts were a viable method to improve contract outcomes. These findings prompted renewed efforts to provide additional guidance for the use of incentive contracts, reinforcing the intentional use of these incentives to motivate high contractor performance. One year after the GAO study, in March of 2006, the DoD issued revised guidance for these incentive contracts. The DoD emphasized that the award component of the fee must be linked to performance metrics. In the most telling example of how specific these policies had to become to prevent inappropriate usage of incentives, DoD made it impermissible to give award fees for unsatisfactory contractor performance. This policy also limited the award for mere satisfactory contractor performance to no more than fifty percent.

These changes in the practice of award fee payments required a cultural change both among procurement officials as well as contractors. Due to additional guidance and scrutiny brought on by incentive type contracts, the frequency of using cost-plus award fee type contracts declined between 2005 and 2008. Four years later, the Acting Deputy Under-Secretary of Defense (Acquisition and Technology) was still testifying before Congress about wasteful contractor bonuses, and, more than a decade later, the DoD continues to publish more restrictive guidance concerning how and when to use incentive fees. These renewed efforts are additional monitoring costs incurred in response to unanticipated residual losses from procurement officials who are too generous with incentive fees.

This scheme demonstrates that when contracting officers are initially given a tool to leverage contractor performance, it instead adds to the residual waste rather than effectively motivating contractors, likely due to the inherent human preference to avoid conflict. This unexpected residual loss requires additional monitoring and repeated guidance revisions to force contracting officers to use the tool effectively, rather than wielding it for the sake of the contractors.

Consider also Mark Granovetter’s observations about the power of weak social ties. An individual’s broadest opportunities, professional and otherwise, come through what he calls weak social ties. Whereas the professional networks of co-workers largely overlap, allowing them to share information about relevant opportunities, the same is not true of a mere acquaintance’s network. Counterintuitively, for a contracting officer, the weak social connection to the contractor represents a vital link to information about opportunities. The value to the contracting officer of this connection with the contractor may be another possible cause of undue cooperation.

The core of the agency problem is its intractability. An agent’s performance, no matter the costs expended by the principal in monitoring and sanctioning, will often deviate from the optimal outcome, causing residual loss. A myriad of factors contribute to the residual loss. Economic incentives may skew an agent’s motivation to perform in unintended ways. Biases, like the illusion of control, or an irrational failure to delegate, may simply compound transaction costs and undermine competition. To complicate matters further, procurement agents unduly cooperate with contractors for a myriad of reasons. These are but a few of the factors that cause the residual loss in a principal-agent relationship. The cumulative effect of these fluctuating factors is the residual loss, which, because it is borne by the end-user, constitutes a moral hazard.

E. The Moral Hazard of Residual Loss

A moral hazard occurs in a situation where the risk or negative impact caused by one person’s decisions are shared with others. Usually, within the principal-agent context, the risk of an agent’s decisions (that is, the residual loss) is borne by the principal. This can create a tight feedback loop that naturally limits the amount of residual loss that a principal will tolerate. This feedback loop thus also restricts the magnitude of the moral hazard problem in the relationship because the agent’s decision will normally be easily observable to the principal.

This is not the case for government procurement of individual equipment for military personnel. The risk of residual loss is predominantly borne by the end-user, rather than the principal, thereby aggravating the moral hazard. Whereas any monitoring and sanctioning costs are borne by the principal and agent, the residual losses ultimately accrue to the detriment of the end-user. Whether the residual loss is due to a procurement agent being reluctant or unable to collect and act on quality information, because of a Congressional mandate to purchase from only two percent of available contractors, or due to some other reason, the poor quality product impacts only the end-user. Due to the structural distance between the principal and end-user, this burden borne by the end-user may never be recognized by the principal.

Furthermore, these residual losses are largely opaque to the principal-agent hierarchy so long as the end-user is not part of the procurement decision or quality feedback system. Without an active part in the procurement system, these end-users are effectively disenfranchised but nevertheless bear the negative results of the residual losses that accumulate through the hierarchy of agents and sub-agents. To turn this problem around, delegating the procurement choice directly to the end-user would drastically reduce the moral hazard as well as the residual loss.

F. Cumulative Agency Costs in Federal Procurement

A simplified example of the agency relationships within the DoD Acquisition System is illustrated in Table 1 below.

Table 1 illustrates the flow of delegated authority from Congress down through the procurement bureaucracy of each military department to spend money for the benefit of the end-user. Each level of the principal-agent hierarchy creates a unique combination of monitoring costs, sanctioning costs, and residual losses. Table 1 gives examples of each. The cumulative agency cost for procurement is the sum of all monitoring, bonding, and sanctioning costs, plus the residual loss at every level of the hierarchy. All of these costs are in addition to the cost of employing agents in the first place.

At the first level, Congress can be seen as both the agent of the people of the United States and the principal, through the constitutional power of the purse. Congress incurs monitoring costs (for example, through authorization, appropriation, and oversight) and by requiring reporting of procurement results. Congress also enacts sanctioning measures, such as the Administrative Procedure Act, to ensure that subordinate agents do not deviate from legislative restrictions. Finally, assuming for a moment that the population of the United States is the principal, this first level of principal-agent hierarchy may incur a residual loss. An example of this could occur when a special interest group, like manufacturers, incidentally benefit from protectionist legislation and lobby for its enforcement.

Table 1: Principal/ Agent/ Sub-agent in Military Acquisitions (with examples of agency costs at each level)

At the second level, agency heads must both ensure that subordinate agents comply with legislative requirements and report acquisition results to Congress. A possible residual loss suffered at this level, and lower levels, is that these agents are concerned more with avoiding negative attention from the principal than with efficient execution of procurement priorities. This could be classified as risk avoidance.

These examples of costs related to the agency problem highlight that agency costs impact the cost of a transaction in two distinct ways: monitoring and bonding. Monitoring and bonding costs add to the overall cost of a transaction without increasing its value. Rather, they are transaction costs and can be analyzed as a matter of efficiency. These transaction costs mount within an organization as the complexity of a procurement increases. In the government context, natural market limits on transaction costs do not exist due to the government’s inherent insensitivity to financial incentives. This combined effect can expand the overall transaction costs for a simple acquisition to outweigh any benefit derived from employing an agent to perform the procurement.

In contrast to monitoring and bonding costs, the residual loss associated with the agency problem impacts the actual value of the transaction’s outcome. Whereas the other agency costs amount to paying more for the same outcome, the residual loss results in a worse outcome. When a principal is also the end-user, the principal would necessarily be aware of any sub-optimal outcome and bear the residual loss inherent in employing an agent. In the government context, however, that is seldom the case. The cumulative residual loss may be invisible to the principal because it is borne by the end-user.

Our examination of these costs in detail demonstrates two fundamental features of the agency problem. First, the costs associated with employing an agent cannot be reduced to zero. Agency costs may be minimized but never completely eliminated. If efficiency is a fundamental goal of government procurement, then it behooves a government to reduce these agency costs as much as possible.

Second, these agency-related costs and losses are cumulative as each agent, in turn, employs a sub-agent. Congress incurs monitoring costs, bonding costs, and residual loss in employing an agent to accomplish spending. Similarly, at the other end of the agency hierarchy, when Mr. Smith contracts with a sub-agent for the provision of goods, this relationship also results in monitoring costs, bonding costs, and residual loss. While some costs do not overlap, others may, and therefore do not duplicate the costs incurred at a higher echelon principal-agent relationship. That is, Congress incurs different monitoring costs than the contracting officer. Thus, the true cost of an agent-facilitated procurement is the aggregated sum of all monitoring costs, bonding costs, and residual losses that occur between each principal, agent, and sub-agent relationship at every level of the hierarchy.

IV. Ongoing Efforts to Reduce Transaction Costs in Military Procurements

Government procurement reform in the United States has a long and tortured history of trying to lower transaction costs while still ensuring that opportunistic contractors do not take advantage of the government. Recently, Congress has again implemented changes and begun exploring further possibilities to increase procurement efficiency. The discussion below focuses on two of those reform initiatives: a blue-ribbon panel, which recommended that purchases of commercially available items be radically simplified, and a statutorily mandated pilot in allowing government users to purchase directly online, through commercial marketplaces. Both would facilitate the solution to the agency problem proposed in this paper — authorizing end-users of personal military equipment to purchase the equipment directly.

A. Blue-Ribbon Section 809 Panel

Congress established a blue-ribbon panel through Section 809 of the NDAA for Fiscal Year 2016 (the Section 809 Panel) to review the existing DoD procurement regulations and processes. The purpose of this panel was to improve the overall functioning and efficiency of procurement practices in order to increase flexibility and effectiveness in acquisitions. In making its recommendations, the Section 809 Panel emphasized that the DoD benefits from commercial market-based approaches. The Section 809 Panel argued that by dropping many of the burdensome procurement requirements, safeguards, and checks — developed over the previous century — the government would be able to take advantage of a commercial marketplace that itself provides significant protections from opportunistic vendors. Deregulation, the panel argued, would not only increase flexibility for government acquisitions but also allow the DoD to become a more attractive business partner to commercial vendors, thus increasing overall competition for DoD business.

Among other things, the Section 809 Panel recommended restructuring the procurement system to increase integration with the commercial marketplace. This integration would be most beneficial where the commercial sector already offers products and services that meet government needs. The Section 809 Panel called these goods and services “readily available” and recommended a radically simplified procurement process for these readily available items, for purchases up to fifteen million dollars. Under the panel’s recommendations, purchases of “readily available” goods and services could be made (among other methods) directly from electronic marketplaces, with no prior notice and very limited structured competition or transparency. In recommending this highly streamlined approach for “readily available” goods and services, the Section 809 Panel noted the unintended consequences of over-regulation in government procurement.

Challenges persist, in part, because decades of legislation and policy initiatives that governed, and often attempted to reform, the acquisition system continue to rely on unique terms, conditions, and processes better suited to the industrial age, not the information age, much less the rapidly approaching artificial intelligence age. These industrial age artifacts are not agile, do not value time, and serve as barriers to small and nontraditional businesses.

The Section 809 Panel recognized that market forces might be harnessed to increase efficiency. In making its recommendations for radically simpler purchases up to fifteen million dollars, the Section 809 Panel in some ways facilitated the more fundamental changes that General Services Administration’s (GSA) “electronic marketplaces” pilot could bring in terms of prioritizing user experience, discussed below in Section V.

B. Simplified Acquisition and Micro-Purchase Thresholds

The Section 809 Panel’s recommendations were made against the backdrop of other recent reforms to streamline procurement. Congress has broadened exceptions to competition requirements for low-value contracts, for example. These broadened exceptions allow purchasers to leverage commercially competitive environments. This flexibility is especially important because the transaction costs that burden traditional procurement are often disproportionate to the overall value of low-value contracts. Because burdening low-value contracts is particularly inefficient, Congress has long-established competition and transparency regulatory exemptions for contracts below certain value thresholds. These thresholds are called the “simplified acquisition threshold” and the “micro-purchase threshold.” By statute, acquisition-related dollar thresholds must be adjusted for inflation on October 1 of each year divisible by five.

The simplified acquisition threshold was first established at $100,000 in the Federal Acquisition Streamlining Act of 1994, which identified a suite of legal requirements that are not to apply to acquisitions below the threshold. Since then, the threshold has been both raised and broadened to cover additional categories of goods and services. Most recently, Section 805 of the NDAA for Fiscal Year 2018 increased the simplified acquisition threshold from $100,000 to $250,000. This raised the simplified acquisition threshold by one hundred and fifty percent. Congress has exempted transactions below the simplified acquisition threshold from traditional legal requirements in order to reduce transaction costs, including administrative costs of contracting, inefficiencies and delays, and unnecessary burdens on both the agency and contractors.

The smallest purchases in the U.S. system — micro-purchases — enjoy even more exceptions from normal procurement requirements. For acquisitions below the micro-purchase threshold, the Federal Acquisition Regulation allows purchase cardholders (which include thousands of government users who are not procurement officials) to buy supplies and services without competition, so long as the purchaser considers the price to be reasonable. This is likely as close as current federal procurement regulations come to allowing for direct end-user purchases. Notably, the exception for micro-purchases has been broadened substantially by Congress in recent years.

Section 806 of the NDAA for Fiscal Year 2018 increased the micro-purchase threshold from $3,000 to $10,000. This new amount is an approximately 300% increase in the value of contracts that are exempt from nearly all normal procurement restrictions. Despite the enormous increase, the U.S. GSA already has recommended increasing the micro-purchase threshold even further to $25,000 for purchases made on GSA-approved electronic marketplaces (discussed further below). The justifications given for this meteoric rise in the micro-purchase threshold are to streamline simple acquisitions and preserve procurement resources for more complicated projects.

These recent expansions of the simplified acquisition and micro-purchase thresholds reflect a growing consensus that, in some situations, overly restrictive procurement regulations cost more than the benefits that they claim to add. The consensus is clear: allow lower value purchases to be subject to fewer regulatory restrictions, and allow normal marketplace competition to provide protection from opportunistic vendors.

The Section 809 Panel recommendations and recent liberalization of low-value purchases have focused on the regulatory sources of inefficiency. Commercial electronic commerce platforms — the subject of GSA’s pilot initiative — provide a proven technology to do exactly this. In fact, however, a major source of inefficiency in government procurement is the agent-based procurement system itself. Indeed, might a transaction be both more efficient and suffer from less residual loss if the end-user made the procurement decision?

V. The End-User E-Commerce Paradigm

Agency theory is a critical tool for discovering and analyzing the inefficiencies hidden in agent-based transactions. Just as important, other social sciences lend insights into the scope of the agency problem with equally valid implications for how to minimize the agency problem. Generally, whether a principal decides to use an agent to accomplish a purchase should depend on whether the benefits of using an agent outweigh the costs. Would it be possible to eliminate the agency problem altogether by employing the end-user in appropriate circumstances as the procurement agent?

Section 846 of the NDAA potentially created an intriguing solution to this agency problem. This provision called for testing and eventual adoption of e-commerce procurement platforms for commercial items. Indeed, GSA has recently issued a contract solicitation to fulfill this very goal. Current models of Internet-based marketplaces have several features that would greatly reduce agency costs in government procurement. The most advantageous feature of this model, however, is that the “best value” determination could be made on an individual basis. That is, end-users could make individual choices regarding the best value on an item-by-item basis. This would entirely eliminate the layers of monitoring and bonding costs. Most promisingly, it could entirely erase the residual loss as well.

Deputizing the end-user as the procurement agent for certain items of their own equipment more accurately aligns the authority of the agent with the knowledge and experience of the end-user, and tempers inherent biases. In other words, the end-user agent is inherently interested in the outcome of the procurement in a way that a disinterested purchasing agent cannot be. The end-user has fundamentally more accurate information on which to make a best-value determination decision regarding the choice among options than does an agent who will never use the gear himself. This is especially true for individual equipment items where fit varies widely according to the user. An effective e-commerce platform could simply incorporate features already available to enable end-users to make individual procurement decisions.

A. Central Billing and Central Data Collection and Central Curation

An e-commerce solution could incorporate central billing for purchases made at the end-user level. This payment system is both similar to, and distinct from, the government travel card system, which currently allows end-users to make certain limited procurement choices. Unlike the travel card, an electronic marketplace would necessarily include transaction data for every item purchased by every individual. Rather than being forced to apply for an institutional credit card based on personal credit with personal credit impacts, users in the e-commerce system could be individually added as authorized users. This clarity of transaction detail would work to prevent fraud.

Additionally, by shifting potentially billions of dollars in sales to an electronic marketplace, GSA’s initiative could allow acquisition workforce personnel to be prioritized for more complex procurements — a key goal of the Section 809 Panel recommendations. Available commercial choices could be screened by current acquisition workforce personnel, ensuring minimum uniform requirements are met by the available products. An e-commerce portal could do more than simply allow end-users to make their own procurement choices; it would restore user experience as the driver of innovation.

B. Experience-Based User Feedback

Despite obvious problems and risks, commercial electronic marketplaces provide a convenient and tested method for collating feedback directly from end-users, and for responding nimbly and flexibly to individual users’ needs. The GSA solicitation calls for this feature. While a government-centralized purchasing agent must work to collect data about the quality of a supply purchased that is widely distributed, a commercial electronic marketplace is suited to better respond to each end-user’s intimate knowledge of the quality, fit, and durability needed for procured items.

Collectively, the experience of countless end-users in a commercial marketplace guides individual decisions of other purchasers with real experience. User ratings from purchasers are frequently used by buyers on e-commerce platforms to guide purchasing decisions, spawning a new field of marketing research. The fact that eighty-five percent of individuals seek this type of information before making a personal purchase is an indicator of how important this type of experience-based information can be. From an economic perspective, these reviews reduce the information asymmetry between the suppliers of goods and prospective users. That is, experience-based product reviews increase the access potential buyers have to information about the quality, durability, comfort, and other factors regarding a certain product.

This user-feedback mechanism also provides a convenient method for monitoring the performance of suppliers. If a product’s ratings consistently fall below a certain level, for example, if durability is frequently complained of in reviews, this product could be removed from the list of available options. Quality-based information provided by users directly to the government public-procurement marketplace would add pressure on suppliers to maintain quality.

C. Efficiency and Equity in Individual Equipment Purchasing

In addition to providing a mechanism for collecting user product reviews, granting choice to users would also reduce waste in two ways. Systematic underutilization of issued equipment due to poor fit and function is a twofold problem. Recall, for example, that half of the soldiers in the example above stated that they prefer to purchase boots in the commercial market rather than wear the boots issued to them. This fifty percent utilization rate equates to paying twice for the equipment that is actually used, and the math is not encouraging. If half of the 2.7 million deployed service members are not using either of the two pairs of approximately $100 boots they are issued, then the military will have spent over $270 million on boots that went unused.

Perhaps just as bad, purchasing poor quality equipment effectively offloads the expense of adequate equipment onto individual service members. Surely, when Congress is appropriating funds to equip service members, it is not the intent that those funds be used to purchase equipment that no one will use, and it is not the intent to foist equipment costs onto individual service members.

D. Continuous Market-Based Competition

An equally promising component of a web-based procurement platform for end-users to use directly is the effect on competition. User feedback ratings not only provide a metric to inform the individual procurement decisions, but they also maintain pressure on vendors to maintain appropriate levels of quality assurance. High ratings over time for quality, fit, function, and durability are the vendors’ way to continue attracting individuals to choose a particular vendor’s products over the competition. This is in contrast to the current incentive to create an initial high quality offering in order to get a mass production contract, which thereafter must only pass a representative sample inspection, for which the contractor is typically responsible.

The pressure to maintain high user ratings as the mechanism for achieving sales amid flexible individual choices also drives commercial innovation. No longer would individual equipment go for decades without improvement. Market forces would force vendors to maintain high quality in current production, as well as drive innovation in order to capture market share of enduser choices.

VI. Conclusion

The modern commercial electronic marketplaces that GSA has embraced (with Congress’s support) will enable individual end-users to make informed choices regarding the goods that they buy to ensure that those goods meet the users’ special needs. The contracts, as GSA has made plain, would focus on the “user experience” and incorporate end-user reviews. Absent overriding needs for modularity, this system of procurement has the potential to increase the reliability of the purchasing agent’s “best value” determination and restore competitive market pressure on suppliers. Further, this would allow the acquisition workforce to concentrate on more complex procurements. This approach would greatly reduce intractable costs due to the principal-agent procurement system that currently imposes large residual losses on individual service members. Granting procurement authority directly to end-users in appropriate situations bypasses the agency problem, thereby decreasing waste, decreasing acquisition workforce costs, and most importantly, bypassing the otherwise unavoidable burdens of residual losses caused by agent-centered procurements, currently borne by end-users.

This article is based on a thesis that was submitted in partial satisfaction of the requirements for the degree of Master of Laws in Government Procurement at The George Washington University Law School. The views expressed in this paper are solely those of the author and do not reflect the official policy or position of the United States Army, Department of Defense, or U.S. Government.

This article is dedicated to the author’s wife, Gabriela, and children, Daniela, Jonathan, Evangeline, Liliana, and William, and to military families like them, who steadfastly bear the burdens of military service.

Many thanks are due to Professor Christopher Yukins for his insightful discussions on the wide-ranging disciplines covered in this article, and for patience while reviewing multiple drafts. His advice and guidance increased both the clarity of the ideas and the precision of their expression. Any remaining oversights, of course, are the author’s.