The shadow Pentagon

Introduction

As war fighting came to dominate the news in the wake of the Sept. 11, 2001, terrorist attacks, names like Halliburton and Bechtel became as familiar to the average American as the names of any general, division or soldier in the field. Fallujah first attracted wide public attention when insurgents killed and crowds mutilated the remains of four employees of Blackwater Security Consulting. Employees of CACI International and Titan were accused of taking part in the abuse of prisoners at Abu Ghraib. That the use of contractors on the battlefield and in nation building in Iraq and Afghanistan is front page news comes as a surprise to many, but it is a consequence of a decades-long policy to keep government smaller by relying on the private sector.

What the War on Terror has shown is the extent to which private contractors have become part and parcel of Pentagon operations. Where once contracts went to build ships, planes, tanks and missiles, today the majority of contract dollars buy services—the time of people—and information technology. Increasingly the private workforce works alongside officials, in Pentagon meeting rooms as well as on Iraqi battlefields, performing what citizens consider the stuff of government: planning, policy writing, budgeting, intelligence gathering, nation building.

In March 2002, a year before the start of the Iraq war, then-Secretary of the Army Thomas White told top Defense Department officials that reductions in Army civilian and military personnel, carried out over the previous 11 years, had been accompanied by an increased reliance on private contractors about whose very dimensions the Pentagon knew too little. “Currently,” he wrote, “Army planners and programmers lack visibility at the Departmental level into the labor and costs associated with the contract work force and of the organizations and missions supported by them.”

In April, the Army told Congress that its best guess was that the Army had between 124,000 and 605,000 service contract workers. In October, the Army announced that it would permit contractors to compete for “non-core” positions held by 154,910 civilian workers (more than half of the Army’s civilian workforce) and 58,727 military personnel. It should have been no surprise, then, when contractors were needed to meet the surge of wartime reconstruction, that the Pentagon itself was hard-pressed to estimate the numbers of its contract employees in Iraq.

The inability to keep a count of contractors is only the most obvious of the problems the Pentagon faces. It has been longstanding bipartisan White House policy that “inherently governmental” work—the basic and most sensitive work of government—must be performed by government officials. Yet in Iraq, at Abu Ghraib, where contractors were involved in interrogations and collecting intelligence, this rule was not only violated, but apparently violated without notice to the Commander in Chief in time of war.

Government by contract

The onslaught of service contracting challenges oversight on multiple counts. The Defense Department, along with other government agencies, keeps precise count of the numbers of those employed as officials—civil servants, political appointees, servicemen and women. But Defense itself lacks the most basic information on its contract workforce. When, in 2002, the Secretary of the Army declared that the Army lacked “visibility” over its service contract workforce, he called for its collection. By mid-2004 the data gathering had yet to begin. As the General Accounting Office, the watch dog arm of Congress, observed, “there is only limited visibility or control at the DOD or military department level, and information systems that provide reliable data and are capable of being used as a management tool are lacking.”

Poor data goes hand in hand with inadequate management. In a 2003 report, the GAO noted that “our work, and the work of DOD’s Inspector General, has found that spending on services is not being managed effectively. Too often, requirements are not clearly defined, alternatives are not fully considered, vigorous price analyses are not performed, and contractors are not adequately overseen.”

Many service contracts buy commercial services—maintenance and laundry services, for example. But a significant amount of the Pentagon’s procurement budget goes to purchase a core government workforce—contract employees, consultants, and other service workers. The third largest category of Pentagon spending, the Center found, following research and development and aircraft, is “professional, administrative and management support services.” In the decade from FY 1994 to FY 2003, expenditures on these workers increased from $7.3 billion to $17.1 billion.

Defense’s Inspector General has repeatedly found Pentagon control of this contracting deficient. In 2003, the IG reported that out of 113 service contract actions reviewed (with an estimated value of $17.8 billion), at least 98 had one or more problems, including inadequate competition, lack of surveillance, or inadequate price reasonableness determinations.

Defense does not know the numbers of contractors performing basic government work—that is, drafting rules, policies, budgets, and other official documents. But a measure of its increasingly commonplace nature can be found on the Web sites of the department’s contractors. Private companies announce that they’re hiring analysts to prepare the Defense Department’s budget, or boast of having written the Army Field Manuals on Contractors on the Battlefield.

For insiders in the corridors of the Pentagon, the pervasive role of contractors in the replacement of civil servants is a given. Government Executive reports that the Undersecretary of Defense for Personnel and Readiness—the senior official responsible for the official workforce “acknowledges that he often attends meetings in which he is the only civil servant in a room full of contractors.”

But the role of contractors in the Department’s basic work remains largely and effectively out of public view. For example, in late 2002, the Department’s Total Information Awareness data mining project provoked great (and bipartisan) outcry; government as “Big Brother” was threatening the privacy of untold citizens. Largely unnoted was that the program was being run by contractors, with a thin layer of official oversight. If the program really had been operated by Big Brother, Bill of Rights privacy protections would have limited their work; but it is much less clear how Constitutional provisions govern contractors.

The oversight challenge is not simply one of fewer officials managing more contractors. It also lies in the possibility that in two critical respects—knowledge of information technology (IT) and capability to bridge organizational chart barriers—contractors are not simply the shadow government, but may become the primary government.

Information technologies are the lifeblood of 21st century military command and control. The Pentagon (and taxpayers) paid for the creation of the Internet, and much else that undergirds today’s cyber world. But it paid contractors to learn about the new technology, and did not pay to train and retain a sufficient, official IT workforce. As Government Executive noted,

“Even before the current push to outsource, contractors were doing at least three-quarters of the federal government’s IT work, according to the market research firm INPUT in Chantilly, Va.”

Contractors managing contractors

At the end of the Cold War, the Pentagon downsized civilian and military workforces. By the mid-1990’s it became apparent that the workforce needed to oversee contractors had been disproportionately downsized. As a 2000 Defense Department Inspector General Audit report recorded: “DoD reduced its acquisition workforce from 460,516 to 230,556 personnel, about 50 percent, from the end of FY 1990 to the end of FY 1999; however, the workload has not been reduced proportionately.” Though there was a slight (3 percent) decrease in the dollar value of Defense procurement over the period, “the number of procurement actions increased from about 13.2 million to about 14.8 million, about 12 percent.”

The end result is that, increasingly, contractors must manage one another.

The Pentagon now relies on contractors themselves to collect and analyze basic data on contracting—including data on the dimensions of its service contracts (“spend analyses” performed by Pentagon mainstay Booz Allen Hamilton) and data on the dimensions of the official procurement workforce (Jefferson Solutions).

Defense’s stable of contract procurement advisors includes the Rand Corporation—a private nonprofit that the Pentagon created and funded to act as a think tank. Rand (and its alumni in the department) was central to Cold War defense contracting, famously associated with Secretary of Defense McNamara’s efforts to bring service buying under central control. From 1998 to 2004, Rand received $474 million in noncompetitive contracts from the Pentagon.

But much of this work is done today by private, profit-making companies. And recent efforts to streamline government have accelerated the process. Where once the Pentagon had to go through time consuming, competitive procedures to hire contractors to manage contracting, 1990’s procurement “streamlining” permits it to purchase “off the shelf” “outsourcing” and “privatization” assistance. Booz Allen Hamilton’s Web site illustrates the “soup to nuts” contract management easily procured from contractors:

Acquisition Program Management. In the context of acquisition program management, Booz Allen Hamilton has experience in all dimensions of support for major government acquisition programs…. Booz Allen will assist the Government in developing and establishing acquisition program objectives, strategies, plans, and schedules that will help the acquisition program through the various stages of the life cycle and its associated review process.

“Inherently governmental functions”

As contractors expand their role both on the battlefield and within the walls of government, one might well ask what sorts of things they’re not doing. This was a concern from the very beginnings of government by contract. In the mid-20th century, when White House officials grew concerned that contracting out might go too far, the principle was established that “inherently governmental functions” should never be performed by anyone outside government. The principle has been embraced by every Administration since then. It was most recently reiterated by the Bush Administration in a May 2003 revision of Circular A-76, which declares that, “agencies shall…Perform inherently governmental functions with government personnel.”

Today, that principle may be more fig leaf than bulwark. Decades of personnel ceilings necessarily meant that work was contracted out without regard for whether it was inherently governmental.

In 1998, Congress passed the FAIR Act, which required agencies to inventory civil service work, and identify specific jobs as “commercial” (which may be contracted out) or inherently governmental (which may not be). Iraq shows the workings of the framework established by White House policy and the FAIR Act. The Army not only identified jobs as inherently governmental (or not), but provided reasoning for its determinations.

In December 2000 the Army determined that intelligence work—such as that assigned to contract employee interrogators at Abu Ghraib—is inherently governmental. A memo by the Assistant Secretary of the Army for Manpower and Reserve Affairs explained why sensitive intelligence work must be performed only by government officials:

“Private contractors may be acquired by foreign interests, acquire and maintain interests in foreign countries, and provide support to foreign customers. The contract administration oversight exerted over contractors is very different from the command and control exerted over military and civilian employees. Therefore, reliance on private contractors poses risks to maintaining adequate civilian oversight over intelligence operations.”

The Army memo directed that the rule barring contractors from intelligence work be added to the next edition of the Army Contractors on the Battlefield Field Manual. The volume, which was itself written by an Army contractor, did not include the determination.

In Iraq, the Army engaged contractor employees for interrogation work at Abu Ghraib, disregarding its own determination that outsourcing such work poses risks to national security. As the story of the involvement of contractors in the Iraqi prison abuse scandal unfolded, it became evident that the contract for interrogators (with CACI International Inc.) illustrated the way in which reforms designed to simplify contracting had given way to broad and ready abuse.

The CACI contract was an offshoot of an agreement between CACI (actually a company acquired by CACI) and the General Service Administration (GSA) under what the government calls a supply schedule arrangement. GSA supply schedule contracts are a product of the 1990’s reforms that reinvented government. At their best, they are a (cyber) catalog of services that can be bought by agencies government-wide with limited time wasted. In theory, the initial agreement with GSA is obtained through competition, so the contractor is permitted to sell services under the schedule to other agencies with little or no further competition required. When the Defense Department found itself in urgent need of person power in Iraq, it turned to GSA schedules with a vengeance.

The CACI interrogators were purchased under a supply service contract which, according to the GSA schedule, was to provide for Information Technology services. Not surprisingly, the underlying agreement with GSA did not provide for “interrogator services.” Moreover, GSA had turned over the agreement to the Department of Interior to administer (for a fee), so the Army contract to purchase interrogators for Iraq ran through the agency in charge of the national park service. On July 16, 2004 the Department of Interior’s Inspector General reported that six orders placed with CACI “were issued predominantly for interrogation, intelligence, and security services in Iraq. Neither the GSA nor our review could find any existing schedule that provided for these services.”

Prior to calling on CACI to provide interrogation under the GSA agreement, evidently no one in GSA or Interior or the Army checked to determine whether the new work was permitted under the GSA schedule. (As a technical matter, an attempt to buy services outside the scope of a GSA schedule violates requirements that opportunity for competition be provided.) Nor, it appears, did GSA or Interior— legally responsible for the contract and receiving fees for its administration—feel obliged to monitor the use to which the contract was put in Iraq. Nor, it appears, did GSA, Interior, or the Army check to see if the work had been determined to be inherently governmental and, therefore, something that the Army could not outsource to the employees of a private company.

The misuse of supply service contracts shown by the CACI contract was not unique, but of a piece with many other Iraq contracts, and many further contracts entered into well before the war.

In fact, Defense’s Inspector General noted—in a March 2004 report on 24 contracts awarded for the Coalition Provisional Authority between February and August 2003, by the Department of Defense Contracting Command-Washington—that 18 of the 24 contracts were awarded through use of GSA supply schedules.

The IG reported that Defense contracting officers “misused” the GSA schedules in awarding 10 of the 18 schedule related contracts. The IG portrayed a pattern of abuse, even lawlessness. The IG found that in calling on contractors to provide services under the schedules, Defense contracting officers themselves did not review the GSA schedules, but took the word of contractors as to whether the service called for by the Pentagon was within the scope of the GSA schedule. The IG further found that in 10 instances, Defense was using contractors to buy “personal services” –in essence, buying workers to serve as part of the official workforce, a practice contrary to procurement rules. The IG concluded that of the 24 contracts, 13 did not have adequate surveillance.

Nor was Iraq the first instance of reported abuse of the supply schedules. In 2003, GAO explained that “since early 2000, both the Inspector General and we have found continuing problems with DOD’s use of the General Services Administration’s Federal Supply Schedule program… in September, 2001 the DOD Inspector General concluded that 304 of the 423—or 72 percent—task orders it had reviewed were awarded on a sole-source or directed-source basis and 264 were improperly supported.”

A question of oversight

Throughout the growth of contracting, separate bodies of law applied to officials (and soldiers) and contractors. American governmental bodies reflect a long tradition of laws enacted to prevent abuse of power by government officials. These laws begin with the Constitution, which defines and limits the conduct of officials, but not necessarily of third parties, even where they may act on the government’s behalf.

The official workforce is also governed by a dense web of statutory provisions that do not apply to third parties. The Freedom of Information Act applies to “agency” records. Contractors, in this context, are not “agencies,” even where they perform decisional roles. Similarly, government officials are subject to a body of conflict of interest provisions, pay caps, limits on political activity, and labor rules that do not similarly constrain contractors who perform similar, even the same, work.

The use of “contractors on the battlefield” in Iraq provides dramatic illustration of different rules applied to similarly situated government and contractor employees (although the differences primarily involve rules governing soldiers, not civilian officials). Contract employees may avoid risky combat zone work – and, ultimately, quit – with no fear that their absence will subject them to penalties under the Uniform Code of Military Justice . The differing rules may also disadvantage contractor employees. For example, daunting questions loom regarding how the protections accorded by international law to state combatants apply to battlefield contractors.

More than two centuries of law have been enacted to protect Americans against Big Government. These laws begin with the Constitution and the Bill of Rights, and include ethics and transparency laws, restrictions on the political conduct of officials, limits on official pay, and the uniform military code of justice. These laws apply to officials, not contractors, on the presumption that officials are in control. The rules do not apply—or protect the public—when, as is increasingly the case, contractors are doing the basic work of government, and government lacks the expertise and experience to control the contractor workforce.

These rules do not contemplate a reality where contractors advertise that they will be writing the President’s Defense Budget, where they routinely write official documents including Army Contractors on the Battlefield Field Manuals, or where they are deployed, in violation of the Army’s own rule, to perform sensitive intelligence work in wartime.

Dan Guttman, a Fellow at Johns Hopkins University, is a government contracting expert who serves as a consultant to the Center for Public Integrity. He is a co-author of The Shadow Government.