THE OUTSOURCING OF U.S. INTELLIGENCE Robert Akira Watson (110001586) This dissertation is submitted in part requirement for the Degree of M.A. (Honors with International Relations) at the University of St Andrews, Scotland, and is solely the work of the above named candidate. 24 April 2015 12,600 words 1 TABLE OF CONTENTS Abstract………………………………………………………………………….……………………………………...2 I−Introduction………………………………………………………………………………………………………..3 II−Intelligence Contracting……………………………………………………………………………………...5 III−The U.S. Intelligence Community and its Place in Government……………………………..11 IV−Bureaucracy and Bureaucratic Politics……………………………………………………………...14 V−Privatization and Principal-‐Agent Theory…………………………………………………………..21 VI−Issues with Intelligence Contracting………………………………………………………………….28 VII−The IC’s Demand for Expertise…………………………………………………………………………45 VIII−Conclusion…………………………………………………………………………………………………….52 Appendix: Glossary of Acronyms……………………………………………………………………………..53 References…………………………………………………………………………………………………………....54 2 ABSTRACT Private contracting in U.S. intelligence is a $50 billion per year industry that has received little public and academic attention. This paper aims to shed light on this topic by evaluating what happens when contractors, as private actors, are deeply involved in as sensitive an activity as intelligence, traditionally reserved to government. To support discussion, frameworks for public and private provision are developed, and principal-‐agent theory is extensively used. Intelligence contracting is studied by considering the problematic issues that arise in the private provision of different intelligence functions. These issues taken whole provide a substantive argument against outsourcing intelligence to contractors. The Intelligence Community’s unique demands for expertise provide the thrust of the argument in favor of bounded intelligence contracting. Through its dialectical approach, this paper is able to generate normative prescriptions for the appropriate use of private contractors in intelligence. 3 THE OUTSOURCING OF U.S. INTELLIGENCE I. INTRODUCTION This paper examines the subject of intelligence contracting. The involvement of private contractors in the most sensitive functions of government has been a rising silent trend over the past two decades, most prevalent in the United States; as such, this paper exclusively deals with contracting in within the 17-‐member U.S. Intelligence Community (IC). Intelligence contracting raises serious, troubling questions about conflict of interest, democratic accountability, and the nature of government responsibility, to name a few. This paper seeks to understand the dynamics at work when contractors, as private actors, participate in what most would consider inherently governmental functions, and from this, derive a normative conclusion on whether this involvement is appropriate. The weight of the legal and ethical issues associated with intelligence contracting contrasts with the dearth of public and academic attention the subject has received. Pervasive secrecy surrounds intelligence, extending to the use of contractors. Official figures are difficult to come by; intelligence budgets are classified and contracting companies are usually not at liberty to disclose their work with the IC in much detail (Shorrock, 2008: 17). Secrecy creates real barriers to scholarship (Crampton, Roberts, and Poorthuis, 2014: 197). There has been scant research devoted specifically to intelligence contracting (Shorrock, 2008; Chesterman, 2008; Crampton et al., 2014; Hansen, 2006; Krishnan, 2011). The topic 4 does sometimes receive passing mention in the larger study of private military contracting as well as intelligence (Chesterman and Lednardt, 2009; Isenberg, 2009; Singer, 2008). This paper makes full use of the above material, as well as unclassified government reports on IC contracting. However, achieving this paper’s objectives requires going outside the strict purview of international relations and political science. I enlist scholarship in public administration, bureaucracy, and privatization to develop a theoretical framework for approaching intelligence contracting. Models of public and private provision, including their respective advantages and disadvantages, are conceptualized; intelligence can be construed as a service to be provided by either. The general role of intelligence in U.S. government is used to provide a frame of reference for private contracting. Principal-‐agent theory undergirds much of the paper’s discussion. With these conceptual instruments in place, intelligence contracting is dissected according to the various issues and dilemmas it entails. These issues are weighted against the IC’s particular demands for expertise, from which a final normative assessment of intelligence contracting can be made. 5 II. INTELLIGENCE CONTRACTING Some good scholarship has been written about private military and security companies (PMSCs) amidst their rising use in the 2000s. PMSCs may offer their services to state and non-‐state actors, including national militaries and private businesses, for a fee. Intelligence contractors form a distinct subsection in the PMSC industry. Their use by the U.S. government is the exclusive focus of this paper. Writing in 2008, Tim Shorrock estimated the domestic private intelligence market to be worth $50 billion per year (Shorrock, 2008: 12). According to an unclassified 2007 PowerPoint presentation by the Office of the Director of National Intelligence (ODNI), spending on private contracts accounted for 70 percent of the intelligence budget at the time (Shorrock, 2008: 18-‐19). Across the IC, which includes civilian and military intelligence agencies, contractors are believed to account for 18 percent of the total workforce (RAND, 2013: 16-‐17). In a follow-‐up to a 2007 ODNI inventory of the IC workforce, the ODNI’s Chief Human Capital Officer (CHCO) Ronald Sanders cited contractors as comprising 27 percent of the total IC workforce (ODNI, 2008). The discrepancy between the 2007 and 2013 figures reflects two factors: First, intelligence contracting peaked around 2005-‐07, largely in response to surfacing IC and congressional concern over the practice’s extent (ODNI, 2008). Second, different intelligence agencies’ commonly use different methodologies for defining what constitutes a contractor, and more specifically a core contractor (GAO, 2014: 18). Core and non-‐core contractors. This paper breaks down intelligence contractors by their use in core and non-‐core capacities. Non-‐core contractors supply a diverse 6 range of mainly uncontroversial services, such as staffing cafeterias and grounds maintenance. Non-‐core also denotes commodity contractors, which sell products like computers and satellites to the IC (ODNI, 2008). The aforementioned 70 percent figure includes spending on non-‐core contractors (ODNI, 2008). Due to the relatively mundane nature of most non-‐core contracting, this paper focuses predominately on core contractors. Core contractors, as defined by the ODNI, “provide direct support to core IC mission areas such as collection activities and operations (both technical and human intelligence [HUMINT]), intelligence analysis and production, basic and applied technology research and development, acquisition and program management, and/or management and administrative support to those functions” (Kennedy, 2006: 1). In October 2009, the ODNI issued Intelligence Community Directive 612 (ICD 612) to guide IC agencies in their hiring of core contractors, by stipulating appropriate roles for their use. Directly referenced in ICD 612 is the concern that core contractors “perform functions closely associated with…inherently governmental functions” (GAO, 2014: 2). This basic concern undergirds much of the criticism of military and intelligence outsourcing in general. I return later in the paper to the inherently governmental standard, which is a federal regulation supervising contractible activities. Intelligence contracting companies. A diverse range of firms of different sizes and specialties define the multibillion-‐dollar private intelligence industry. Shorrock subdivides the intelligence industry into four classes of companies, led by top tier systems integrators (Shorrock, 2008: 23). These companies are worth several billion dollars, possess thousands of security-‐cleared staff, and are involved at nearly every 7 level of intelligence, obscuring the public/private sector division in intelligence (Shorrock, 2008: 23-‐24). Second tier companies provide specialized technical and analytic services to the IC, often as subcontractors for the systems integrators (Shorrock, 2008: 24). Third, are small technical firms providing more singular expertise to intelligence agencies; many of these companies have sprung up since 9/11 and are regularly based around the Washington, D.C. area. They are colloquially referred to as beltway bandits (Shorrock, 2008: 25). Shorrock’s fourth class of contractors includes “companies, large and small, that are known primarily for their achievements in information technology, communications, or satellites [which] have made major inroads as the [IC] has opened up to contracting” (Shorrock, 2008: 26). This class of companies includes major telecommunications providers enlisted by the government after 9/11 to assist in its mass electronic surveillance and data mining programs (Shorrock, 2008: 307). A common denominator for these companies that should be apparent is their servicing of the IC’s enormous appetite for technology. In the proceeding paragraphs I discuss some of the factors that have helped generate this distinctive demand, which defines contemporary intelligence. External causal factors for intelligence contracting Peace dividend and 9/11. With the Cold War’s end, the IC suddenly found itself missing its primary raison d'être: checking the global communist threat. The peace dividend of the 1990s saw dramatic cuts in military and civilian intelligence budgets, leading to equally substantial cuts in personnel. According to CHCO Sanders, the IC’s workforce peaked between the late 1980s and early 1990s, declining by 40 percent before bottoming out in 2001 (RAND, 2013: 43). Many of these former government 8 employees found work in the private sector, bringing with them their experience, institutional knowledge, and security clearances (Hansen, 2014: 73). Whether a fully staffed U.S. intelligence apparatus would have potentially thwarted the 9/11 attacks is totally open to speculation. What is clear, however, is that in the immediate aftermath of the attacks there was a massive surge in demand for intelligence services. Owing to the decimation of its ranks during the 1990s, the IC found itself unable to meet this demand with in-‐house supply (RAND, 2013: 45). Much of the expertise the IC needed was now in the private sector, which is where the IC turned, rapidly expanding capacity with contracted personnel. Between 2002 and 2005, the IC’s spending on contracts rose from $32 billion to $43.5 billion, before the government began concerted efforts at hiring its own intelligence personnel internally (Shorrock, 2008: 113-‐114). Partly because of this post-‐9/11 contracting binge and the dependencies it seeded, it is now virtually impossible for the IC to wean itself off completely from contracted labor. New wars and new actors. Different modes of conflict are historically linked to the chosen means of executing military ends. Singer argues there is a relationship between mass military demobilization and increased conflict in weak state zones (Singer, 2008: 38). In this respect, the end of the Cold War (which included a set of organizing principles for international relations) and the peace dividend “produced a vacuum in the market of security” (Singer, 2008: 49-‐50). The post-‐9/11 security paradigm, as part of a broader post-‐Cold War paradigm, views threats as being unbound to traditional state actors and emanating from asymmetric – often non-‐state – sources, largely in the form of international terrorism and rogue states (Halperin and Clapp, 2006: 10-‐11). Michael Hayden, a former director of the CIA and the NSA, 9 framed the difference between Cold War and post-‐9/11 intelligence as having to confront an enemy (the Soviets) that “was easy to find [but] hard to finish” and now facing a jihadi enemy that “is easy to finish, [just] very, very hard to find” (Shorrock, 2008: 193-‐194). Two-‐edged technologies. A major factor driving the ongoing shifts in contemporary threats and conceptions of security is the proliferation of two-‐edged technologies. I borrow this term from the ODNI’s 2006 “Five Year Strategic Human Capital Plan” (ODNI, 2006: 3-‐4). Advances in computers, the internet, imagery, and other information technologies (IT), as key examples, both produce great societal benefits and complicate the state’s ability to provide security within and outside its borders because of these technologies’ empowering of non-‐state actors (ODNI, 2006: 3-‐4; Avant, 2009: 182). Technological advancements in encryption, IT, communications, and imagery have fundamentally altered how intelligence is gathered today. Since the 1990s, these advancements have predominately been made in the commercial sector – outstripping the government and IC’s ability to keep up (Shorrock, 2007). By the 1990s, “commercial developments in computing power, cryptology, and high-‐speed telecommunications [had] surpassed” the NSA’s capacity to collect signals intelligence (SIGINT) and break codes (Shorrock, 2008: 199-‐199). In meeting its obligations as an agency, the NSA under Hayden turned to the private sector, with its technical expertise, to help create a SIGINT system able to perform in a world of cell phones and fiber optics (Shorrock, 2008: 201). The IC has a crucial demand for technological solutions that can decipher valuable intelligence from the reams of information available from open sources. In satellites and imagery software, the center of gravity 10 today for technical expertise firmly rests in the private sector – boosted in large part by Clinton-‐era policies to enhance U.S. competitiveness in high technology (Shorrock, 2008: 245). Contemporary technologies offer powerful tools for intelligence agencies to conduct their work. However, these same technologies are two-‐edged, in that they offer similar benefits to previously weak actors with goals contrary to American interests. There is an overwhelming drive within the IC to harness technological advances to suit intelligence needs. The commercial-‐driven nature of today’s technological innovations has meant the IC has had to enlist the help of the private sector to accomplish this. 11 III. THE U.S. INTELLIGENCE COMMUNITY AND ITS PLACE IN GOVERNMENT This section describes the structure of the U.S. Intelligence Community and the general role of intelligence in American government. The Intelligence Community. The U.S. Intelligence Community consists of 17 organizations structured in a loose federation headed by the Director of National Intelligence (DNI), whose office, the ODNI, is one of the 17 IC members (GAO, 2014: 1). The IC is broadly divided into civilian and military intelligence, represented by the National Intelligence Program (NIP) and Military Intelligence Program (MIP). Most government intelligence personnel are military, and 80 percent of the IC’s budget is under the Pentagon’s jurisdiction, extending to the NSA, National Reconnaissance Office (NRO), and National Geospatial-‐Intelligence Agency (NGA) – the three most important collection agencies (Shorrock, 2008: 154). There are significant differences between the organizational structure of the NIP and the MIP. The MIP’s primary concern is intelligence for the “planning and conduct of tactical military operations” (RAND, 2013: 52). The MIP follows a formal chain of command with the Secretary of Defense at the top (RAND, 2013: 53). In contrast, the civilian-‐based NIP lacks clear hierarchy and is largely decentralized. Prior to the intelligence reforms of 2004, the CIA’s director served simultaneously as head of the agency and Director of Central Intelligence (DCI), with responsibility for coordinating the activities of the entire IC but lacking power to fully exercise this mandate (RAND, 2013: 10). The 2004 reforms, which replaced the DCI with the DNI, were meant to redress this 12 problem; however, the DNI still lacks “chain of command authority,” as CHCO Sanders pointed out in 2007 (RAND, 2013: 10). Office of the Director of National Intelligence. The Director of National Intelligence’s position and office were created by the Intelligence Reform and Terrorism Prevention Act of 2004 (IRTPA). Passage of the IRTPA and the ODNI’s establishment were motivated by the 9/11 Commission’s recommendation that the IC develop stronger central leadership to more effectively coordinate intelligence activities (RAND, 2013: 5). The DNI and his office are tasked with broad responsibilities, which include coordinating IC functions, facilitating greater interagency cooperation, and community-‐wide workforce planning (RAND, 2013: 7). However, as previously pointed out, the IRTPA did not give the ODNI sufficient formal authority to truly exert centralized control of the IC. As part of its workforce planning mandate, the ODNI has basic authority to regulate the use of contracted personnel by IC agencies, as illustrated in its 2009 ICD 612, stipulating appropriate uses for core contractors. While intelligence agencies generally do follow the ODNI guidelines, actual execution of workforce policies is at agencies’ discretion (GAO, 2014: 28-‐29). Role of intelligence in U.S. government. Annual U.S. government spending on intelligence (civilian and military) totals almost $80 billion (Crampton, Roberts, and Poorthuis, 2014: 197). As with any bureaucratic organization, intelligence agencies are intended as an executive tool for implementing decisions made by the state’s political leaders (Svara, 1999: 687-‐688). However, intelligence agencies are unique from other bureaucratic actors in that their most basic mission is to provide the information from which their political superiors can make decisions. Hence, while these agencies are situated in the government’s bureaucratic hierarchy under 13 decision makers, their main sphere of activity can be interpreted as existing one step ahead of decision-‐making. In addition to providing intelligence and analysis to decision makers, intelligence agencies are also tasked with conducting covert, or clandestine, operations: this area is a particular specialty of the CIA. In practical terms, the IC’s primary responsibility of providing the sensitive information used as the basis of policy formation and decision-‐making gives it enormous power in the government. Given the structural power the IC has, U.S. intelligence agencies are barred from recommending specific policies; in their estimates and reports to decision makers, including the president, agencies are instead “limited to predicting the consequences of alternative policies” (Halperin and Clapp, 2006: 142). The first step in setting the IC’s priorities comes from the White House’s National Security Strategy (NSS). The broad points conveyed in the NSS are adapted into the classified National Intelligence Priorities Framework (NIPF), first introduced in 2003 (RAND, 2013: 34). The NIPF communicates IC-‐wide guidance for member agencies in determining their resource requirements and matching NSS goals with individual agencies’ own internal organizational vision (RAND, 2013: 35). Just as IC members use the NIPF to plan their organizational functions, the ODNI uses it to develop community-‐wide planning (RAND, 2013: 35). *** Having conceptualized the basic role of the IC in U.S. government, I explore in Part IV principles of model bureaucratic administration, and how actual governance frequently deviates from this model. 14 IV. BUREAUCRACY AND BUREAUCRATIC POLITICS This section considers various theories on bureaucracy, starting with Max Weber’s foundational work on the topic. I then consider defects in bureaucratic administration, which center on the premise of bureaucracy as a politically active agent. Finally, these concepts will be related back to the IC. Weber’s rational legal bureaucracy. Max Weber’s writings in his Theory of Social and Economic Organization underlie the classical conception of modern bureaucratic administration. Weber identifies bureaucratic administration as the purest type of exercising authority on rational legal grounds (Weber, 2012: 333-‐334). It is important to note the pure distinction: Weber implies the systems he describes do not exist in practice as perfect models. Nonetheless, Weber’s work is still normatively and descriptively valuable in studying bureaucracy. The central advantage of bureaucracy is its ability to handle complex tasks in a maximally efficient and consistent manner (O’Connor, 2013). Bureaucracy is able to do this because of its division of labor into different offices, each with a marked sphere of professional competence and responsibility (Weber, 2012: 330). As Weber demonstrates, the exercise of power by bureaucracy is founded on the basis of knowledge; this is what makes it a supremely rational form of authority (Weber, 2012: 339). Candidates for office are selected and promoted on a meritocratic basis by their technical qualifications (Weber, 2012: 333-‐334). Offices are organized by hierarchy, generating a clear chain of command and appeals/feedback process; this hierarchy reflects the bureaucracy’s specialized division of labor (Weber, 2012: 331). 15 Bureaucracy’s capacity to handle complex tasks is finally cemented by its potentially vast fiscal and material resources, including human capital (Hill, 1991: 271). These resources and concentration of competencies tend to put public administration at a distinct advantage over competitors in providing certain goods and services, often resulting in the state being a monopoly provider of said goods and services. Interestingly for this paper, one of services associated by Weber with monopoly provision by the state is the legitimate use of force: military and intelligence contracting plainly challenge this idea. Weber’s formal model of legal bureaucracy is the essential starting point for understanding how bureaucracies and public provision work, or at least are supposed to work. In the following paragraphs, I consider two general criticisms of bureaucratic administration, which are followed by the bureaucratic politics perspective. Two criticisms of bureaucracy Subject to grooved thinking. While Weber idealized bureaucracy for its competence and ability to handle complex tasks, the term bureaucratic has negative connotations for most of the public (O’Connor, 2013). It conjures an image of administration that tends to be remote, inflexible, and slow. The complex tasks bureaucracies are responsible for managing are often rife with uncertainties; this is no truer than in foreign and security policy. In dealing with these uncertainties, large organizations, like bureaucracies, and their staff adopt particular assumptions and images to help instruct their work (Halperin and Clapp, 2006: 21-‐22). How these images are formed, whether other individuals share them, and how they guide thinking and action 16 depends on a wide range of variables; however, a good place to start looking is the actor’s position in context. Most bureaucrats, especially in low to mid-‐level positions, exhibit what is called grooved thinking: in looking at a given situation or problem, they are inclined “to focus on a few key variables and [have] a programmed response to those particular variables” (Halperin and Clapp, 2006: 22). Bureaucratic administrations tend to condition grooved thinking because they regularly confront complex problems and are principled on producing stable, consistent results in dealing with these problems. Bureaucratic administration as antidemocratic? Since the inception of modern bureaucratic administration, bureaucracy’s concentration of resources and technical expertise on a large scale in unelected officials has elicited concerns over its potential threat to democratic accountability (O’Connor, 2013; Svara, 1999). These concerns are related to the general perception of bureaucracy as remote and its treatment as a black box (even among academics), owing to the complexities and scale of its operations (Hill, 1991: 266). The apparent separateness between bureaucratic administration and democratic politics, and the concerns this generates, stems from bureaucracy’s operating principle of exercising authority based on specific neutral expertise – not popular support (Svara, 1999: 677). Sometimes the public regards this separateness positively; in highly partisan American politics, bureaucrats can be viewed as honest, unbiased brokers (Hill, 1991: 273). Adhering to the classical conception of bureaucracy as an administrator but not a decision maker, a certain degree of political disengagement is necessary for the successful implementation of policy. 17 Having considered these two criticisms of bureaucracy, I now examine the argument that bureaucracies are actively political agents, with independent interests and decision-‐making processes not necessarily constrained by Weber’s views on neutral, rationally-‐based authority. Bureaucratic politics. As Hill writes, “in the long run, political issues become bureaucratic issues” (Hill, 1991: 272). In most cases where public decision makers, such as Congress and the president, delegate authority to bureaucratic or executive bodies, they allow these agencies discretion in determining the terms and substance of their mandate (Laffont, 1990: 1344). While bureaucratic agents have a supposed professional obligation to serve the public interest in a neutral fashion, this entrusted discretion means that ultimately it is their choice to decide how to conduct their activities. Moreover, the notion of acting in the public interest is almost impossible to verify because it is too ambiguous a term to feasibly define (Levine and Forrence, 1990: 181). In foreign and security policy, including intelligence, actors are working in the similarly vague national interest (Halperin and Clapp, 2006: 9). The result is that agents in either setting will pursue their own private conceptions of the public and national interest (Halperin and Clapp, 2006: 9). The bureaucratic politics perspective argues that bureaucratic actors formulate their version of the national interest along their interests as an organization. Bureaucracy is not a monolithic actor (Hill, 1991: 274). It may seem obvious, but it is necessary to remember that U.S. government is populated by a multitude of different agencies, whose responsibilities, or turfs, frequently overlap in jurisdictional gray areas. In the American system of government, there is moreover 18 no inherent guarantee of an agency’s permanent survival; in order to secure their present and future position in government, bureaucratic agencies “must be politically proactive” (Hill, 1991: 272-‐273). Bureaucratic actors are acutely aware of the implications of decisions made by their political leaders, especially concerning budgets (Halperin and Clapp, 2006: 26). They will pursue an audience with decision makers when topics at hand intersect with their perceived organizational interests, and will advance their interests in these settings, in concert with other bureaucratic actors sharing similar organizational motivations (Hill, 1991: 287). This process produces outcomes that therefore do not reflect a purely rational pursuit of the public and national interest but rather the result of bargaining among involved parties (Hill, 1991: 287). Application to the IC. In the final portion of Part III, I discussed the IC’s model role in U.S. government. The basic conceptualization presented there more closely resembles Weber’s work on bureaucracy – as a neutral executive instrument – than it does bureaucratic politics. The remainder of Part IV applies this latter model to further develop understanding of the IC’s role in government. Within the IC, interagency conflict is clearly seen in competition between the CIA and armed services over control of covert operations (Halperin and Clapp, 2006: 40). That there exist numerous redundant functions between civilian and military intelligence agencies suggests a prevailing concern for achieving and preserving organizational autonomy. The arrival of new actors and threats in foreign and security policy has a causal relationship to the means chosen to address them: this casual relationship is furthermore impacted on both sides by new technological developments. What means are chosen can create new operational areas and 19 reorganize old ones for competing bureaucratic agencies, producing rippling effects in budgets and instigating turf wars. A contemporary example of these processes in play is the ongoing debate between the CIA and Pentagon about operational control over drone warfare (Harman, 2015: 101). Just as American bureaucracy is not monolithic, neither are the individual bureaucratic agencies themselves. Internal debate over what should be a given organization’s essence is another area of bureaucratic politics. While the NSS and NIPF provide some guidance towards setting IC members’ goals and needed capabilities, the agencies still preserve considerable freedom in ultimately determining these factors (Halperin and Clapp, 2006: 26). Returning to the CIA’s drone program, the use of drones for targeted killings has produced notable apprehension among many CIA officers about what it entails for the agency’s organizational essence (Harman, 2015: 100-‐101). Within the CIA, there are three rough groups of opinion about what should be the agency’s primary focus: intelligence gathering, covert operations, and intelligence analysis (Halperin and Clapp, 2006: 34). Internal and interagency competition impacts how information is transmitted to senior officials and decision makers (Halperin and Clapp, 2006: 49). Although the IC is formally limited to evaluating alternative policy courses and cannot endorse a particular option, intelligence agencies can still effectively signal their given preferences in weighting the consequences of alternative policies (Halperin and Clapp, 2006: 142). Additionally, the predilection for autonomy can hinder sharing of intelligence between agencies, with potentially deleterious consequences. The 9/11 20 Commission’s criticism of the CIA and FBI’s failure to share potentially valuable intelligence prior to 9/11 can be read this way (Halperin and Clapp, 2006: 147-‐148). Concerns about the democratic accountability of bureaucratic organizations are no more relevant than they are with intelligence. Public accountability through congressional oversight of intelligence is severely constrained by the pervasive secrecy that surrounds intelligence activities. Only select members of Congress are permitted to review the doings of the IC, with strict limitations on disclosure of information obtained by these members (Halchin and Kaiser, 2012: 34). This secrecy and deficit of oversight are typically justified on the grounds of national security (Nathan, 2009: 2), which essentially means intelligence agencies get to determine the amount of oversight they receive. One of the few potent oversight mechanisms that exists in intelligence is informal leaks, which I briefly consider in Part VI. *** This section has attempted to demonstrate that government bodies, including intelligence agencies, can and do deviate from norms of ideal public administration. To determine in a balanced fashion whether it is appropriate for the government to outsource intelligence capabilities, it is necessary to have good understanding about how public provision of these services works. 21 V. PRIVATIZATION AND PRINCIPAL-‐AGENT THEORY This section provides basic mechanisms of private provision of goods and services, which go some way in illuminating the motivations for privatizing publically-‐ administered services, like intelligence. I also formally introduce principal-‐agent theory, and its real world application in contracts. Reinventing U.S. government. In the United States, faith in markets and skepticism with government provision cross Republican and Democratic party lines, and have done so more or less since the Reagan administration (Freeman, 2003: 1292-‐1293). Buttressing these beliefs is neoliberalism, an economic and political doctrine skeptical of the concentration of power and resources in the central government and which aims to delegate as many functions as possible to the private sector (Kruck, 2014: 119). Privatization serves as a key vehicle for achieving neoliberalism’s minimal state (Kruck, 2014: 119). During Clinton’s administration, Vice President Al Gore was tasked with leading the National Performance Review (also known as the “Reinventing Government” initiative); this policy’s basic purpose was to assess which government services should remain public and which should be opened to privatization and contractor bidding (Freeman, 2003: 1293). The inherently governmental standard was formally developed at this time. (Chesterman, 2008: 1069-‐1070). According to the Federal Activities Inventory Reform (FAIR) Act of 1998, inherently governmental describes functions “so intimately related to the public interest as to require 22 performance by Federal Government employees” (FAIR Act, 2000). The Federal Acquisition Regulation (FAR), which governs the federal government’s purchasing of goods and services (including contractors), lists in Subpart 7.5 specific functions it considers inherently governmental, for which contractors cannot be used (FAR, 2006). However, the FAR list is “not all inclusive,” which leaves agencies discretion in determining whether a given function meets the FAR standard for inherently governmental (FAR, 2006). The result is that the inherently governmental standard acts less to protect public provision of certain services than it does to justify having them provided by the government in the first place (Chesterman, 2008: 1069-‐1070). This paper returns to the inherently governmental standard in both Parts VI and VII. Private provision mechanisms. Weber describes the profit motive as “the ultimate driving force of all economic activity” in a market economy (Weber, 2012: 320-‐321). Private businesses are purposely organized to maximize profits, which sustain the life of an enterprise (Trebilcock and Iacobucci, 2003: 1424). Businesses, as producers of goods and services, in a model free market compete against one another to satisfy consumer demands (Avant, 2009: 181). A firm’s performance and subsequent net capital accumulation (profit) is determined by how well the firm supplies its consumers’ demands in relation to other firms. Failure to sufficiently meet consumer demands and generally poor performance and waste can jeopardize a firm’s profits and thus its survival (Trebilcock and Iacobucci, 2003: 1436). The profit motive, market competition, and the risk of failure combine to strongly incentivize a firm’s performance: the basic principles of market discipline (Trebilcock and Iacobucci, 2003: 1436). Free market and privatization advocates use this concept of market discipline to argue that markets naturally self-‐regulate, as well as to argue against public provision and government regulations. 23 There is good reason to be skeptical of the proposition that markets self-‐ regulate. Versus the hard controls government regulation provides, market-‐based regulatory mechanisms are essentially voluntary (Cockanye, 2009: 207-‐208). While the profit motive underlies the notion of market discipline, there are certainly situations where it can lead to a firm’s determination that “going for the ‘quick score’ is worth the risks of long-‐term market costs” (Singer, 2008: 224). Private versus public provision. Private businesses are motivated to maximize their profits. As discussed, these profits are contingent in an ideal marketplace upon how well a firm satisfies its consumers’ demands in comparison to the firm’s competitors. Thus, market discipline means that producers (private firms) are broadly accountable to the demands of their consumers, which can be determined quantitatively in the form of the firm’s earnings. It is worth noting that privatization advocates elevate utility as conceived in economic terms, and mostly neglect other potentially important intrinsic values relevant in provision (Freeman, 1296-‐1297). The primary motivation for privatization is to improve efficiency: delivering high-‐ quality goods and services at the minimum possible cost (Freeman, 2003: 1296-‐ 1297). Privatization advocates argue the effectiveness of strictly legal accountability mechanisms should not be overstated, including the notion of hierarchal control embedded in bureaucratic theory (Trebilcock and Iacobucci, 2003: 1447-‐1448). Public agencies typically lack the incentives towards maximizing performance and minimizing costs competitive market discipline conditions (Freeman, 2003: 1297-‐ 1298). Moreover, these agencies have soft budget constraints, as their operations are 24 generally guaranteed by taxpayer dollars regardless of performance (Laffont and Tirole, 1991: 88). In comparison, private providers are strongly incentivized by market discipline to maintain accountability to their consumers’ demands and to deliver goods and services efficiently (Freeman, 2003: 1297-‐1298). In arguing for privatization, Trebilcock and Iacobucci offer the following summary: “Market-‐based accountability is not merely an incidental benefit of privatization; [rather] the benefits of privatization arise largely because of the accountability generated by private markets” (Trebilcock and Iacobucci, 2003: 1447-‐1448). Principal-‐agent theory. Undergirding discussion of private and public provision is principal-‐agent theory. Originally developed in economics, principal-‐agent theory describes when a client or supervisor (principal) tasks a provider or subordinate (agent) to do a job (O’Connor, 2013). The principal is interested in seeing the job completed properly, and grants the agent some independent discretion, called slack, to fulfill the principal’s directive (Cockanye, 2009: 197-‐198). Slack describes the fact that principal-‐agent transactions happen “under conditions of uncertainty and asymmetric information” (Cockanye, 2009: 197-‐198). Slack presents an agent opportunity to pursue its own interests in place of its principal’s. An agent has a broad self-‐regarding interest in shirking, which is any deviation from a task as assigned by the principal. An agent may also be working simultaneously for multiple principals, giving rise to problems of coordination, conflict of interest, and hidden gaming (Cockanye, 2009: 209; Laffont, 1990: 302). Principals must therefore monitor their agents in order to ensure their instructions are properly completed. However, monitoring costs can be high. Agents are the actors involved in actually carrying out their principals’ orders and can leverage this position by controlling the upward flow 25 of information, or feedback (O’Connor, 2013). Monitoring problems are further compounded when tasks or objectives are imprecise and/or difficult to verify. Contracts. A contract is a principal-‐agent transaction for a service. In everyday life, most decisions to contract a service, such as a lawyer or electrician, are made on the basis of needing a contractor’s expertise because the client himself lacks the given skills; this fact points to the asymmetric information that frequently accompanies contracts, as a form of principal-‐agent transaction. An enduring concern in contracting for clients is being overbilled (Singer, 2008: 155). In comparison to the purchase of an overpriced good, which is a one-‐off event, overbilling for a service recurs over the life of the contract (Singer, 2008: 155). The client – who may lack the knowledge and tools to accurately monitor performance – is at the contractor’s mercy to inform him on the contract’s progress and what should be done next (Singer, 2008: 153). Contractual slack exhibits itself acutely in the form of contract incompleteness: when the terms and objectives of a contract are difficult to specify and subsequently difficult to monitor (Singer, 208: 153). Incomplete contracts give contractors discretion “to make judgment calls, trade-‐offs, and [decisions] to fill in the contractual gaps,” which may not be in their client’s best interests (Freeman, 2003: 1343). Overbilling and also cost reductions through sacrificing quality may not technically violate the terms of incomplete contracts (Freeman, 2003: 1345). Most of the contracts in intelligence are necessarily incomplete because the areas they deal with (foreign and security policy) are inherently complex and unpredictable, making contractual terms difficult to specify in full (Hart, Shleifer, and Vishny, 1997: 1155). One could argue that a contractor’s desire for future business and maintaining good reputation should regulate his performance and prevent undue shirking. 26 However, this desire for future business also provides perverse incentives for contractors to create path dependencies, whereby contractors structure future need to retain their services through their current work (Cockanye, 2009: 208). Some jobs, especially in incomplete contracts, mandate long-‐terms of service, which paves the way for principals to become dependent on their contractor/s. Public and private provision similarities. Principal-‐agent theory has found widespread application outside of economics, including in political science and international relations. Government can be conceptualized as a chain configuration of principals and agents in which the ultimate principal is the electorate, whose first agents are their elected officials, which act as principals to given legislative committees, followed by bureaucratic agencies, and so forth (Cockanye, 2009: 210). At each principal-‐agent transaction in the chain, the ultimate principal’s (the electorate’s) interests undergo increased agency slack (Cockanye, 2009: 210). Public servants’ basic mandate to promote the public or national interest is similar to contractual incompleteness or slack. Different public actors can all pursue different conceptions of what they claim is the public interest, and just like with incomplete contracts, none of them are necessarily wrong because of the term’s inherent ambiguity (Levine and Forrence, 1990: 180-‐181). Finally, the bargaining games that occur in bureaucratic politics can be construed as a sort of marketplace exchange, where the goods and services being offered are policy proposals being forwarded to maximize an organization’s given interests. The consumers in this case are decision makers and superior officials. 27 *** This section considered arguments in favor of privatization as well as some relevant basic mechanisms of private provision, importantly the profit motive and market discipline. This section also introduced principal-‐agent theory and its application in contracts, which provide especially valuable tools for this paper’s purposes of examining intelligence contracting. 28 VI. ISSUES WITH CONTRACTING INTELLIGENCE In this section, the theories, principles, and background presented thus far are used to determine what happens when contractors, as private actors, are involved in an activity as lacking in oversight and intertwined with the public and national interest as intelligence. Discussion of intelligence contracting here takes into account the respective advantages and disadvantages of both public and private provision considered in previous sections. Two case examples, the NSA’s Trailblazer project and Abu Ghraib, are used to illustrate the contentious issues associated with intelligence contracting. I discuss these issues in rough order of the increased scope and gravity of the problems they raise. I start by revisiting the inherently governmental standard. What is inherently governmental? As discussed in Part V, FAR Subpart 7.5 lists functions considered inherently governmental, for which contractors may not be used. Listed functions particularly relevant to intelligence contracting are: “the command of military forces,” “the conduct [and determination] of foreign policy,” “the determination of agency policy, such as determining the content and application of regulations,” and “the direction and control of intelligence and counter-‐intelligence operations” (FAR, 2006). Subpart 7.5 also notes functions which “may approach being” inherently governmental. These are predominately activities where contractors perform in administrative and advisory capacities dealing with regulatory compliance and contract management, which may lead to conflicts of interest. Both of these lists, as previously noted, are “not all inclusive” (FAR, 2006). 29 This last feature of FAR Subpart 7.5 presents federal agencies slack, by letting them make the call on functions not explicitly listed. It is understandable that federal regulators may intend the not all inclusive disclaimer to guide agencies to proceed more judiciously in their hiring of contractors rather than act as a loophole. However, this hypothetical intention does not negate what the FAR actually says on paper; agencies are still largely disposed with how to interpret the regulation. Moreover, as discussion here and in Part VII should illustrate, there is evidence to suggest contractors have been employed in technically restricted functions. Two cases of contract miscarriages Presented here are two cases of private contractor involvement in intelligence that resulted in controversy: the NSA’s Trailblazer program and the Abu Ghraib prisoner abuse scandal. The NSA’s Trailblazer program. Starting in the late 1990s, as revolutionary advancements in computing, encryption, fiber optics, and communications were sweeping through the private sector, the NSA began exploring methods to monitor cell phone and electronic communications and to process this information and open source data. Michael Hayden, as director of the NSA, championed the expanded use of contractors to develop the agency’s SIGINT capabilities (Shorrock, 2008: 201-‐202). Well before 9/11, Hayden and agency officials were aware of the NSA’s challenges in adapting to ongoing technological advances (Mayer, 2011); however, the attacks greatly magnified the agency’s impetus to develop an effective SIGINT system for the twenty-‐first century. 30 Hayden chose to pursue the Trailblazer program, over a rival in-‐house program called ThinThread, which was purportedly simpler to understand, more cost-‐effective, and included privacy protections for U.S. citizens (Mayer, 2011). On October 21, 2002, Science Application International Corporation (SAIC), as the lead member of a consortium of contractors, announced it had won a $280 million 26-‐ month contract from the NSA to build Trailblazer (Business Wire, 2002). SAIC is the NSA’s largest contractor, has a long history with the agency (Shorrock, 2008: 215), and ranks among the top tier of intelligence contractors, the systems integrators. The NSA choose SAIC to lead the project in part because of data mining programs in had developed prior to 9/11, such as TeraText, a platform for downloading, storing, and rapidly retrieving massive amounts of textual data (Shorrock, 2008: 216). The NSA believed SAIC would be able to deliver a silver-‐bullet solution to the agency’s SIGINT problems. Trailblazer was intended as a replacement for the NSA’s Cold War-‐era eavesdropping system, which could “capture communications travelling on cell phones, fiber optics, and across the internet” (Shorrock, 2008: 214). Tasked with overseeing Trailblazer was Deputy Director William Black, who had retired from the NSA in 1997 to work for SAIC, and had since returned to the agency (Gorman, 2006). Long story short, Trailblazer was a sweeping failure. By the time the program was cancelled in 2005 by Hayden’s successor, Keith Alexander, its price tag had swelled from $280 million to $1.2 billion (Mayer, 2011). NSA and SAIC mismanagement reportedly mired Trailblazer; the NSA in particular was criticized for taking an essentially hands-‐off approach (Shorrock, 2008: 219). John Pike, director of Globalsecurity.org, believes the program went afoul because of “differing interpretations between SAIC and the NSA over data mining” as well as the 31 requirements of the project (Shorrock, 2008: 219-‐220). Nonetheless, SAIC was awarded the contract for Trailblazer’s successor program, and continues to maintain close ties with the NSA (Shorrock, 2008: 218). Abu Ghraib. CACI International is among a collection of companies known as pure plays, which earn the bulk of their revenue from one industry; in this case, intelligence (Shorrock, 2008: 264). CACI’s involvement in the Abu Ghraib scandal starts in 2003 when it acquired Premier Technology Group (PTG), a smaller company with a preexisting blanket purchase agreement (BPA) with the Interior Department “to supply ‘screening, interrogation, and support functions’ and [HUMINT] at an unspecified site in Iraq” (Shorrock, 2008: 279-‐280). CACI and Titan (another contracting firm, now called L-‐3) contractors arrived at Abu Ghraib prison as it was being overhauled to mimic Guantanamo (Shorrock, 2008: 279-‐280). Private contractors were directly involved in the use of enhanced interrogation techniques (EITs) on prisoners at Abu Ghraib. These contractors were in Iraq to serve as translators and had no training in conducting interrogations (Krishnan, 2011: 199). Claims of abusive practices at Abu Ghraib started to arise in 2003 with reports by Amnesty International, but it was not until spring 2004 after several press investigations that the scandal became widely known to the public (Hersh, 2004). Even after CACI’s improper involvement in prisoner abuse was made known, the U.S. government not only failed “to terminate [their] contract, but actually expanded its terms” (Dickinson, 2009: 224). No formal criminal charges were entered against contractors involved at Abu Ghraib (Chesterman, 2008: 1063). In June 2004, 256 Iraqis lodged a class action lawsuit against CACI and Titan under the Alien Tort Claims Act, which bounced around the U.S. federal court system for several 32 years, ending with case’s dismissal being upheld in appeals court in 2009 (Business and Human Rights Resource Center). Secrecy, market problems, and accountability SAIC’s involvement in the NSA’s Trailblazer program and CACI and Titan’s involvement in prisoner abuse at Abu Ghraib are two extreme cases exemplifying the serious concerns stemming from intelligence contracting. The proceeding paragraphs discuss these issues in intelligence contracting and relate them to the Trailblazer and Abu Ghraib cases. Intelligence market and market discipline. As stated in Part V, the benefits of private provision arise from the accountability fostered by market discipline (Trebilcock and Iacobucci, 2003: 1447-‐1448). This market discipline is predicated on the existence of competition between firms in a relatively free market. However, this sort of market simply does not exist in the private intelligence industry. Intelligence agencies, as clients/principals, are both unable and unwilling to promote an open market for contracting services with more readily procurable supply (Singer, 2008: 152). As demonstrated by the renewal of service contracts in the Trailblazer and Abu Ghraib cases following the contracted companies’ failure to exercise due diligence and their involvement in public controversy, there is insufficient competition in the intelligence contracting market to reward good players and punish bad ones. Favoritism pervades the procurement process for intelligence contracts, 33 with privileged firms often being awarded enormous contracts in a predetermined, non-‐open bidding process (Singer, 2008: 152). It is understood in economics that procurement favoritism can lead to adverse consequences, including “cost overruns, production delays, and performance downgrading” (Che, 1995: 1995). All three of these adverse consequences were clearly displayed in SAIC’s performance in Trailblazer. NSA Deputy Director Black, who was tasked with the project’s management, is a manifestation of the revolving door in practice (this term is discussed later in this section). Black was hired after 40 years of NSA service by SAIC for the singular purpose of soliciting NSA contracts (Shorrock, 2008: 216). He returned to the agency in 2000 to lead the Trailblazer program, still in its infancy at the time, and within two years SAIC had “won the master contract for the program” (Shorrock, 2008: 216). As mentioned earlier, there was a genuine legitimate NSA interest in the technical services SAIC could offer; however, it would be almost impossible to deny SAIC was in a special bidding position because of its close personal relationships to the agency. Whether by intentional design, the inherent nature of the provided service, or some combination of both, much intelligence contracting displays path dependency. The expansion of CACI’s services to the U.S. mission in Iraq following the Abu Ghraib scandal reflects the heavy structural dependence of the U.S. military on contractors in the conflict (Kruck, 2014: 120-‐121). IT services – from offices’ routine computer networks to the NSA’s highly advanced data mining operations – require standby troubleshooting and maintenance, particularly as systems increase in size and complexity. This creates a need to retain contractors for long-‐term technical support, even if they were hired originally only to build a given IT system. The highly specialized nature of advanced data mining programs and other IT systems, like 34 Trailblazer, furthermore means intelligence agencies (as principals) often face prohibitive costs in switching agents, creating an undeniable incentive for contractors to price gouge (Cockanye, 2009: 208). Trailblazer and Abu Ghraib are both cases of incomplete contracts. With Trailblazer, the NSA was tasking SAIC to construct a SIGINT platform on a scale much greater than anything in either parties’ previous experience (Shorrock, 2008: 220). Although it cannot be wholly confirmed here, in all likelihood, the NSA drafted the original contract with a set of broad requirements that were both too vague to effectively monitor and evolved over the course of the contract (Shorrock, 2008: 219-‐ 220). At Abu Ghraib, CACI personnel were under a BPA, which is an especially incomplete form of contract, for use as translators – not interrogation. Under a BPA, the principal does not specify all required services at the original signing, but instead enters task orders as needs arise (Dickinson, 2009: 221). The flexible terms of CACI’s Iraqi BPA had a direct relation to the reassignment – through task orders – of contracted translators to prisoner interrogations (Dickinson, 2009: 221). In many intelligence activities, especially HUMINT, substantial contractual incompleteness should be expected. Functions are often almost impossible to fully contract because they are value-‐laden, complex, and change over time (Freeman, 2003: 1343). Most privatization advocates agree that in situations such as these, government provision is typically preferable to contracting, because it allows public officials to make important judgment calls on their own behalf rather than having private agents do it for them (Freeman, 2003: 1343; Hart, Shleifer, and Vishny, 1997: 1130). There are valid reasons justifying the above problems with the private intelligence market, driven by intelligence agencies’ prevailing concern for secrecy. 35 The IC’s responsibilities in national security fuel the confidentiality surrounding its operations as well as its obtuse procurement process, which heavily favors familiar repeat players versus firms with whom the IC has had little previous experience (Cockanye, 203: 203). IC contract procurement tends to be a long, arduous process due to security protocols; smaller, less familiar firms may simply lack the resources to outlast this process (Shorrock, 2008: 27). Moreover, intelligence agencies tend to make contracts not with a given firm, per se, but with known individuals working for these firms, often former officials (Shorrock, 2008: 32-‐33). Former senior officials, like William Black in the Trailblazer case, bring tremendous value to a contracting firm. Not only do they have intimate institutional knowledge, but these individuals also carry their connections still in government to the company (Shorrock, 2008: 31-‐ 32). Contractor cost effectiveness. In business as well as government, outsourcing is commonly motivated by a desire to reduce costs, by lowering the overhead entailed by in-‐house provision. In hiring a full-‐time intelligence officer, the government bears significant costs in screening and training, as well as sustained costs in providing salary, pension, and benefits. Additionally, intelligence agencies must make their hiring decisions based on their present and future strategic needs in human capital, which are contingent (RAND, 2013: 13). In this context, hiring a contractor, versus a full-‐time public employee, can be fairly attractive. Contractors offer agencies more flexibility in adjusting their supply of human capital and expertise to shifting needs (ODNI, 2008). Entering the contracting business is also an alluring option for public employees, as individuals. A contractor can expect to earn roughly double his public counterpart’s salary: the average cost yearly cost of a government intelligence officer 36 (including pension and benefits) is $125,000, while an equivalent private contractor averages between $200,000 and $250,000 (ODNI, 2008). It is not clear, however, whether contracting is actually more cost-‐effective than in-‐house provision in intelligence and defense, as well as business generally (Singer, 2008: 157). One of the primary reasons intelligence contracting firms are able to garnish their employees’ wages so well is their procurement costs have already been borne by the government (Singer, 2008: 74). The majority of intelligence contractors are former government employees, whether in intelligence or the military, meaning the taxpayer has already footed most of their training costs (Singer, 2008: 77). Owing to the structure of the intelligence market – importantly, its lack of effective market discipline mechanisms, overbilling by contractors is a significant, recurring concern (Isenberg, 2009: 23). As discussed in Part V, overbilling in service contracting can be especially severe when compared to purchasing an overpriced good, because overcharging can occur repeatedly over the contract’s lifetime (Singer, 2008: 155). The existence of path dependencies and contractual slack further multiply incentives for contractors to overbill their clients, adding to the government’s total costs. Contracting professionalism? An oft-‐cited concern with private contractors is they are not upheld to the same standards of professionalism as government intelligence officers. In The Soldier and the State, Samuel Huntington argues, “the modern military officer corps is a professional body,” and asserts that professionalism is the military officer’s most important quality (Huntington, 1985: 7). Huntington defines professionalism by its three distinguishing characteristics of expertise, responsibility, and corporateness (Huntington, 1985: 8-‐9). The professional military officer is duty-‐ 37 bound to practice his expertise in the management of violence in service of the military security of the state (Huntington, 1985: 14-‐15). Huntington plainly states pecuniary rewards cannot be a professional’s primary motivation (Huntington, 1985: 9-‐10); this standard is used to differentiate professional officers from mercenaries (Huntington, 1985: 15). The notional problem with pecuniary versus professional incentives is that the loyalty and performance of individuals motivated by the first only extends as far as the money. Robert Baer, a former CIA officer with extensive experience in the Middle East, argues contractors, particularly in HUMINT, are inherently risk averse. Baer explains that in conducting HUMINT, officers must be willing to engage in potentially life-‐threatening work, “to infiltrate organizations like Hezbollah in Lebanon and recruit spies” (Shorrock, 2008: 152). A public servant, who has taken a professional oath and is interested in promotion within an agency, has greater reason to take risks than a contractor, who is “just coming in and collecting a salary,” according to Baer (Shorrock, 2008: 152). However, it is not clear that pecuniary rewards negate professionalism or are even a contractor’s primary motivation. Firstly, military and intelligence officers also receive salaries; they are essentially distinguished from contractors only by being under public oath. How this oath weighs on job performance is open to debate, but it is an effectively weak monitoring mechanism. As noted earlier, most intelligence contractors have a background in either military or intelligence service. Returning to Huntington, he argues that military officers, as professionals, possess a distinct military mind, which arises because “people who act the same way over a long period of time tend to develop distinctive and persistent habits of thought” (Huntington, 1985: 61). Subsequently, there is no compelling reason why or explanation how an intelligence officer would suddenly change their fundamental values and beliefs 38 simply upon exiting public service for the private sector. Most full-‐time employees and executives of intelligence contracting firms furthermore feel they are genuinely working in the national interest as an extension of the IC, albeit in the private sector (Hansen, 2014; Singer, 2008: 154). Finally, the IC’s prevailing concern for security and the stringent process of obtaining clearances protect against most rogue intelligence contractors and incidentally give rise to certain professional norms within the industry (Hansen, 2014: 75; Avant, 2009: 188-‐189). Accountability. Undergirding the preceding discussion on intelligence contracting is the question of accountability: specifically a concern that contractor involvement undermines oversight and control of activities deeply intertwined with the national interest. As earlier discussed, insufficient competition in the private intelligence industry generally prevents market discipline from emerging to regulate firms’ behavior. Poor performance and shirking are difficult to punish, as they often do not technically violate the incomplete contracts typical in intelligence. In the case examples of Trailblazer and Abu Ghraib it is worth noting that contractor misdeeds were only uncovered because of whistleblowers (in Trailblazer [Mayer, 2011]) and investigations by non-‐governmental actors, like human rights groups and the media (in Abu Ghraib) – not public regulators. These informal oversight mechanisms are a poor basis for delivering systemic accountability. Whistleblowing, as a primary example, is too irregular to sufficiently deter wrongdoing; the government, on the other hand, has a valid concern that whistleblowing, or leaks in its preferred parlance, risk real harm to national security interests and reputation. Moreover, undue whistleblowing, as perceived by the government, motivates it to suppress future leaks. 39 A deficit of effective formal oversight is not unique to the subject of intelligence contracting, however. Rather, it characterizes the entirety of intelligence activity. As highlighted in Part IV, it is largely at the intelligence agencies’ discretion how much oversight they receive, and more basically, whether their activities conform to the law. Not only is congressional oversight lacking, but furthermore, the non-‐disclosure of select intelligence committees’ proceedings provides few political incentives for senators and congressmen to engage in intensive oversight to begin with (Halchin and Kaiser, 2012: 35). Accepting that insufficient oversight extends across the IC, and not just its use of contractors, intelligence contracting still feels problematic. While meaningful oversight of the IC is elusive, intelligence agencies are at least de jure justified in democracies by their grounding in the rule of law and governmental hierarchy (Chesterman, 2008: 1069). Contractors, on the other hand, exist outside the formal hierarchy of public accountability. While they are subject to federal and local regulations as any other business, contractors –as private businesses – are not strictly bound to constitutional obligations, unlike public agencies (Freeman, 2003: 1304-‐1305). Interestingly, it is for this precise reason that governments might sometimes use contractors, to provide a cover of plausible deniability (Singer, 2008: 209-‐210). Public versus private interests Perhaps the most vexing concerns with intelligence contracting stem from the question of how the private interests of contractors interact with the public, 40 specifically national, interest the IC is responsible for promoting. The crux of this question comes down to how the national interest is determined. Conflicts of national interest. The pursuit of profit, as covered in Part V, drives the life of private business. Intelligence contractors may insist they are acting in the national interest in a manner free from balance-‐sheet considerations; however, this claim is problematic. Many would argue it is fundamentally impossible for private firms “to put the public [or national] interest before their private business” (Krishnan, 2011: 203-‐204); this is probably most valid in conflicts of interest where clearly contradictory incentives exist for public and private actors. However, the real problem of the claim contractors can altruistically pursue the national interest resides in the fact, as discussed previously, that the national interest is strikingly vague and contested even among government actors themselves. The notion of “promoting the national interest” is non-‐contractible; it cannot be meaningfully reduced to quantitative measures and is therefore difficult to stipulate in contractual terms. Especially in intelligence, contractors operate in gray areas where defining the national interest becomes increasingly complex (Singer, 2008: 154). Public and private actors in this setting must make judgment calls on what is in the national interest. In highly complex functions, a contractor may distort the national interest to private advantage, whether intentionally or even subconsciously (Singer, 2008: 154). Yet it is somewhat reassuring that the national interest is vague and frequently contested, as it suggests this interest is reached though a harmonization of different actors’ varying conceptions. This is the basic premise of bureaucratic politics. The problem is intelligence is considered privileged information, and is therefore weighted differently by decision makers versus other 41 sources of information; intelligence contracting inserts distinctively private interests into an area of government wielding substantial structural power. Analytical work in intelligence provides opportunities for contractor subversion of the national interest – more so than in highly visible cases of wrongdoing like Abu Ghraib (Chesterman, 2008: 1064). The President’s Daily Brief, an intelligence summary presented by the DNI to the White House each morning, is one of the most sensitive documents in government (Shorrock, 2008: 189-‐190). Over 70 percent of the document’s content is furnished by the NSA (Shorrock, 2008: 189-‐ 90). At the time of his writing, Shorrock estimated 50 to 75 percent of the people working at NSA headquarters and ground stations internationally to be private contractors (Shorrock, 2008: 188). This places private contractors very close to highest levels of intelligence analysis in the IC as well as the United States’ top decision makers. The revolving door. The revolving door describes a common practice of individuals moving between government and private sector employment (Shorrock, 2008: 32). This practice is legal in the United States; senior public officials face only a one-‐year prohibition from lobbying or contracting in their former field (Shorrock, 2008: 32). Interestingly, the term first originated from growing criticism in the 1950s over the growth of the military-‐industrial complex (Mundheim, 1980/81: 709). Since then, the revolving door has retained its mostly pejorative connotation to describe collusion between government regulators and the industries they are supposed to regulate (Meghani and Kuzma, 2011: 578) 42 A central benefit of the revolving door is that it allows specialized knowledge and expertise to be transferred between public and private sectors. In its use in regulatory settings, industry veterans provide their institutional knowledge to government, enhancing effective performance of their regulatory tasks (Meghani and Kuzma, 2011: 576). Industry-‐specific expertise is a precious commodity for government regulators, and it is necessary for good regulatory performance (Che, 1995: 379). On the industry side, regulated firms value former regulators’ expertise to help “minimize the cost of complying with regulations” (Che, 1995: 379). In the intelligence industry, former senior officials are crucial to a contracting firm in establishing potential clients’ trust, which is prerequisite to obtaining big intelligence contracts (Shorrock, 2008: 32-‐33). While demand for expertise on both the public and private side is understandable, the revolving door can produce damaging consequences for oversight and industry accountability. The revolving door can bias public policy in favor of industries’ special interests, degrade public confidence in the government’s neutrality and authority, and give special interests an unfair advantage by providing them an audience other policy stakeholders may not have (Meghani and Kuzma, 2011: 576-‐577, 582). Collusion between public agencies and private firms also enhances procurement favoritism and can deepen path dependencies. The NSA’s Trailblazer program and specifically Deputy Director Black’s former career as a vice president at SAIC illustrate the risks towards cost overruns, poor performance, and mismanagement created by overly close ties between industry and the IC. On balance, SAIC emerged from the Trailblazer episode relatively unscathed. The company received virtually no sanction for its failure to deliver on Trailblazer; its interests prevailed at the expense of the taxpayer. 43 Capture of the national interest. The revolving door is closely linked to the notion of regulatory capture, which appears in economics and public administration literature. Regulatory capture argues public agencies, particularly regulators, have life cycles, starting with a sincere interest in performing their oversight mandate (Hill, 1991: 280). Regulators’ expertise requirements, however, eventually lead them to recruit personnel from regulated industries, and their worldview increasingly conforms to industry’s special interests over time. Ultimately, these public oversight agencies are captured by industry, and the public interest is subverted to special private interest (Hill, 1991: 280). The more slack there is between a given public agency and its government principal/s – Congress for example, the greater risk said agency is captured (Levine and Forrence, 1990: 190). This is because the distance between industry and the agency is shorter than that between the agency and its public principal, providing opportunities for hidden gaming, in the form of collusion, between the former two. This problem is potentially acute in intelligence contracting; public oversight of the IC is lax, and relationships between contractors and intelligence agencies are commonly closer than those between the IC and its government principals. A few voices might argue that private special interest dominance of the policy and decision-‐making process does not necessarily thwart the public and national interest; they could in fact be construed as one in the same. However, this proposition is limited to settings in which these two interests coincide. Special interest is firstly self-‐regarding; the public and national interest serves the polity as a whole, and should equitably balance different and diverging private interests (Levine and Forrence, 1990: 176). Excessive special interest influence in governance 44 privileges select self-‐regarding interests over the polity’s general interest, importantly in cases where they deviate. Understanding this, undue contractor influence in the IC threatens intelligence agencies’ ability to fulfill their mandate to promote the national interest. Furthermore, public perception that state institutions are dominated by special interests can degrade the public’s overall trust in these institutions and weaken their authority by calling into question their legitimacy (Zahra and Kuzma, 2011: 581). *** In this section, I presented various problems that can arise from the use of contractors in intelligence. These problems supply good arguments against intelligence contracting. In the next, penultimate section, I weigh these arguments against the most cogent counterargument in favor of intelligence contracting: the IC’s demand for specific expertise from the private sector. 45 VII. THE IC’S DEMAND FOR EXPERTISE Intelligence contracting peaked sometime between 2005 and 2007 after surging dramatically following 9/11. Many of the contractors enlisted during the early 2000s were to replenish the IC’s personnel ranks decimated by budget cuts in the 1990s (ODNI, 2008). The effort to return more intelligence work in-‐house was spurred by congressional and more importantly IC concern about the consequences of private contracting (ODNI, 2008). The ODNI’s 2006 human capital plan noted in particular “the IC [increasingly] finds itself in competition with its contractors for [its] own employees” (ODNI, 2006: 6). In comparison to the previous two administrations, the Obama administration has instituted policies pressuring the Pentagon and IC to be more selective in using contractors (Isenberg, 2009: 15-‐16). Thus, it is apparent there is some level of recognition by the U.S. government and the IC of the issues surrounding intelligence contracting. For its part, the IC would do well by reforming its agencies’ hiring practices to enhance their in-‐house human capital capacity and subsequently reduce dependence on outside contractor expertise. At present, agencies’ security protocols make them highly risk averse in their hirings (McConnell, 2007). Even first and second-‐ generation U.S. citizens face considerable difficulty in landing a job in the IC, denying agencies highly-‐sought foreign language skills (McConnell, 2007). Greater IC hiring flexibility, however, still would not totally resolve the structural need for private expertise in some areas of intelligence. The very nature of contracted work ensures it will always be more flexible than what the best IC hiring reforms could accomplish 46 (ODNI, 2008). Flexibility is essential for intelligence agencies because of changing threats and technologies they must adapt to, and the capabilities contractors provide in unique skills and surge capacity to the IC are noteworthy (ODNI, 2008). In the proceeding paragraphs, I break down intelligence activities into areas which are: (1) inappropriate for contractor involvement, (2) appropriate for contractor involvement, and (3) areas where the appropriateness of contractor involvement is unclear or contingent. To do so, I weigh the risks, as discussed in this paper, raised by contractor involvement versus the particular demand for a given expertise, which is also evaluated by whether agencies can sufficiently procure it internally. Exclusively governmental functions. Subpart 7.5 of the FAR on inherently governmental functions naturally provides a good basis for judging which activities private contractors should be excluded from. Broadly summarized, the policy expresses concern over contractor use in functions where they have equal or greater discretion or command than their government principals (FAR, 2006). This applies especially to procurement activities. I would maintain that intelligence agencies should comply more diligently with the existing law on using contractors, and would also designate a few more inherently governmental functions not included in FAR Subpart 7.5, applying to intelligence. Any direct use of lethal force in an intelligence setting should be exclusively reserved to public actors. Before continuing further, I would underscore that this reservation does not apply the same to principally military contractors. In drawing a distinction between the military and IC as it relates to contracting, the military 47 possesses a clearer chain of command and greater public oversight than intelligence; wrongdoing involving military contractors is therefore more likely to be exposed. The lack of oversight the IC receives creates a pressing need for it – more than any other area of government – to exercise due diligence and assume responsibility for its actions. In the case of contractor involvement at Abu Ghraib, the fact that CACI International and Titan have suffered no criminal or civil penalties is a failure of accountability falling largely on their government principals. Clandestine operations and most HUMINT functions should be provided exclusively by public agencies. Private contractors in HUMINT have staffed up to three-‐quarters of the CIA’s Islamabad station, held positions “as sensitive as [overseas] deputy station chief,” and according to Robert Baer, have even determined where “CIA officers could go and who they could meet” in Iraq (Shorrock, 2008: 118). The last two, if true, are a direct violation of FAR Subpart 7.5, which designates “the determination and control of intelligence and counterintelligence operations” as inherently governmental (FAR, 2006). In most HUMINT cases, there is no structural justification why intelligence agencies, particularly the CIA, cannot furnish their own capability, as they did for the Cold War’s duration. Recourse to the private sector here reflects more an effort to avoid oversight and a certain indolence on the agencies’ part than it does a demand for expertise that cannot be developed within government. Where contractor expertise is appropriate. The IC must handle daily an immense volume of information across 17 separate agencies. The individual agencies are often massive organizations themselves; the NSA for example has a yearly budget estimated at $10 billion (Shorrock, 2008: 187). Simply coordinating the office activities both within and across agencies is a task in itself. Assuming they can be delivered more cost effectively, the IC should outsource non-‐core functions, such as 48 maintaining workplace computer systems, custodial work, and other office logistics. Agency resources can then be reapportioned to focus more exclusively on core missions in intelligence (Hansen, 2014: 60-‐61); this is the logic of outsourcing in general and the original intent of Clinton’s privatization initiatives. In SIGINT, which is the core mission area of many IC agencies, demand for technical expertise in computers, encryption, and databases – as examples, is a valid reason to use private contractors. Here the first concern is not necessarily cost effectiveness, but rather that government agencies cannot independently reproduce the technology services available in the private sector. There is some crossover in the core and non-‐core technical functions for which private sector expertise is sought by the IC. To illustrate: the ODNI considers commercial satellite vendors to be commodity contractors, rather than core (ODNI, 2008). For an agency like the NRO, however, which devotes 95 percent of its resources to contracts (Shorrock, 2008: 16), purchasing satellites from commodity vendors relates directly to its core mission area of geospatial intelligence (GEOINT). This is a relatively minor nuance that does not detract from the general point to be made about the IC looking to contractors for technical expertise. Even with concerted reforms to its hiring policies, it is doubtful the IC will become the first destination for individuals gifted in IT (Harman, 2015: 104-‐105). As Harman writes, “over the long run…Washington won’t win a digital competition with Silicon Valley” (Harman 2015: 105). As stated in Part II, today’s technologies are two-‐ edged. Intelligence benefits greatly from the modern technology available to it. Advanced imagery and instant communications place within reach Donald Rumsfeld’s vision of network-‐centric warfare, in which all tactical military intelligence is 49 seamlessly integrated (Shorrock, 2008: 162, 234-‐235). The ever-‐increasing speed and storage capacity of computers, as well as data-‐processing algorithms and other IT, have today actualized what critics call the national surveillance state (Shorrock, 2008: 340-‐341). However, these capabilities only exist because of their broad-‐based development in the private sector. Acquiring these capabilities and tailoring them specifically for intelligence, mandates agencies work with private business. Moreover, the IC has impetus for fully harnessing today’s technologies because they are available on the open market for use by friend and foe alike (Shorrock, 2008: 340-‐ 341). This convincing intelligence demand for technical expertise outweighs the risks of contracting as illustrated in the Trailblazer episode, but it does not mean they should be dismissed. The IC’s contracted efforts to “modernize its computer and software systems” have periodically resulted in costly failures (Krishnan, 2011: 198). Although some shirking can always be expected, efforts can and should be made to minimize the problems associated with intelligence contracting while appropriately balancing the imperative of security. Dickinson (2009) recommends using the contracts themselves as the best vehicle for regulating PMSCs; I would agree with her. Contractual terms should reduce slack by including more specific terms of service and monitoring mechanisms, and hiring agencies should enforce these terms more vigorously – including by termination if necessary. Contracting gray areas. The final area to consider is where the appropriateness of using contractors is largely contingent upon case-‐by-‐case evaluation. In the immediate aftermath of 9/11, the Bush administration tasked the CIA with conducting interrogations of suspected terrorists and also running secret prisons 50 (Mayer, 2007). Owing to its lack of existing experience in these activities, the CIA enlisted private contractors, many of whom were retired military psychologists, to develop the now infamous EITs (Mayer, 2007). Earlier in this section, I argued contractors should not be involved in most HUMINT functions; this case forms a potential exception. Without knowing the full details, it is difficult to assess the validity of the CIA’s claim of lacking interrogation expertise. If any organization – public or private – has at its disposal the tools for developing interrogation techniques, one would think it should be the CIA. Even so, this sentiment does not exclude the possibility that a private entity might genuinely be able to furnish a particular HUMINT skill better than the IC. Other related gray areas are mission planning and tactical intelligence – for example, Lockheed Martin contractors assisting with drone strikes (Krishnan, 2011: 202). These functions are indirectly but still closely linked to the use of lethal force. Intelligence analysis is a final gray area for two primary reasons. First, as previously discussed, a single piece of information can be construed by way of analysis to be read multiple ways. As profit-‐driven actors, incentive exists for contracted intelligence analysts to present information that fashions need for their continued services. Second, many contracted intelligence analysts are former government analysts doing almost identical work as before, meaning the government is essentially paying double the cost for a service it can supply in-‐house (Shorrock, 2008: 125). However, there are some justifications for using outside analysts, the best of which is the possibility of gaining fresh perspective in an otherwise closed bureaucracy subject to grooved thinking. This justification naturally assumes the analyst really is an outsider, and not simply a contract rehire. More generally, and in 51 comparison to gray areas in HUMINT, most analysis work is relatively benign and provides few opportunities for serious harm, except cost overruns potentially. *** This concludes the paper’s exploration of intelligence contracting. While intelligence contracting does present some significant troubling issues, a great deal of it has emerged because of legitimate practical necessity. This section has tried to demonstrate it may be possible to outsource certain intelligence functions in a qualified, responsible manner. Short of this standard, however, there are simply some demands for expertise in the IC that warrant the risks of private provision. 52 VIII. CONCLUSION In this paper, I have attempted to capture the motivations, mechanisms, and messy problems at work in private contractor involvement in intelligence, as well as provide some normative judgment on the topic. The recommendations I offer in Part VII on what is and is not appropriate for contracting are informed by the particular demands of the IC for certain expertise as well as the preceding content of the paper. Understanding how government should work and how it does work in actuality, the advantages and disadvantages of private provision, how the IC is structured and its internal politics function, and finally, what role intelligence plays in government: all of these are vital to having a balanced understanding of intelligence contracting. Given its extent and the structural need for certain private expertise, it would appear intelligence contracting – in some form or another – is here to stay. I conclude on one final point. A recurring theme in this paper has been characteristics inherent to intelligence, such as minimal oversight, secrecy, and the need to keep pace with current technology. Discussion has focused solely on how these factors interact with private contracting. Questions have been posed like whether contractors should assist intelligence agencies in conducting mass electronic surveillance. However, what has not been asked is whether intelligence agencies should themselves be engaged in mass electronic surveillance. Many of the most vexing issues associated with intelligence contracting have more to do with the practice of intelligence generally, than with contracting itself. These issues are legally and ethically challenging, and deserve critical examination in their own right. 53 APPENDIX: GLOSSARY OF ACRONYMS BPA – blanket purchase agreement CHCO – Chief Human Capital Officer/Office (ODNI) CIA – Central Intelligence Agency DCI – Director of Central Intelligence DNI – Director of National Intelligence EIT – enhanced interrogation technique FAIR Act – Federal Activities Inventory Reform Act (1998) FAR – Federal Acquisition Regulation FBI – Federal Bureau of Investigation GEOINT – geospatial intelligence HUMINT – human intelligence IC – Intelligence Community ICD 612 – Intelligence Community Directive 612 IRTPA – Intelligence Reform and Terrorism Prevention Act (2004) IT – information technology MIP – Military Intelligence Program NGA – National Geospatial-‐Intelligence Agency NIP – National Intelligence Program NIPF – National Intelligence Priorities Framework NRO – National Reconnaissance Office NSA – National Security Agency NSS – National Security Strategy ODNI – Office of the Director of National Intelligence PMSC – private military and security company PTG – Premier Technology Group SAIC – Science Application International Corporation SIGINT – signals intelligence 54 REFERENCES “Abu Ghraib lawsuits against CACI, Titan (now L-‐3).” Business and Human Rights Resource Center. Web. 24 April 2015. http://business-‐humanrights.org/en/abu-‐ghraib-‐lawsuits-‐against-‐caci-‐ titan-‐now-‐l-‐3-‐0 Che, Yeon-‐Che. “Revolving doors and the optimal tolerance for agency collusion.” RAND Journal of Economics 26.3 (1995). Chesterman, Simon. “‘We Can’t Spy…If We Can’t Buy!’: The Privatization of Intelligence and the Limits of Outsourcing Inherently Governmental Functions.” European Journal of International Law 19.5 (2008). Chesterman, Simon and Chia Lehnardt, editors. From Mercenaries to Market: The Rise and Regulation of Private Military Companies. Oxford Scholarship Online, January 2009. E-‐book. http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199228485.001.0001/ac prof-‐9780199228485 Avant, Deborah. “The emerging market for private military services and the problem of regulation (chapter 10).” Chesterman and Lednardt (2009) Cockanye, James. “Make or buy? Principal-‐agent theory and the regulation of private military companies (chapter 11).” Chesterman and Lednardt (2009). Dickinson, Laura A. “Contract as a tool for regulating private military companies (chapter 12).” Chesterman and Lednardt (2009). “Civilian Intelligence Community: Additional Actions Needed to Improve Reporting on and Planning for the Use of Contract Personnel.” US Government Accountability Office (GAO) (Report to Congressional Requesters). January 2014. Web. 24 April 2015. http://gao.gov/assets/670/660486.pdf Crampton, Jeremy, Susan Roberts, and Ate Poorthuis. “The New Political Economy of Geographical Intelligence.” Annals of the Association of American Geographers 104.1 (2014). “Federal Activities Inventory Reform (‘FAIR’) Act of 1998.” 31 USC § 501 (2000). Freeman, Jody. “Extending Public Law Norms Through Privatization.” Harvard Law Review 116.5 (2003). Gorman, Siobhan. “Second-‐ranking NSA official forced out of job by director.” The Baltimore Sun. 31 May 2006. Web. 24 April 2015. http://articles.baltimoresun.com/2006-‐05-‐ 31/news/0605310010_1_alexander-‐black-‐spy-‐agency Halchin, L. Elaine and Frederick M. Kaiser. “Congressional Oversight of Intelligence: Current Structures and Alternatives.’” Congressional Research Service (report). 14 May 2012. 55 Halperin, Morton, H. and Priscilla A. Clapp. Bureaucratic Politics and Foreign Policy. 2nd ed. Washington, D.C.: Brookings Institution Press, 2006. Hansen, Morten. “Intelligence Contracting: On the Motivations, Interest, and Capabilities of Core Personnel Contractors in the US Intelligence Community.” Intelligence and National Security 29.1 (2014). Harman, Jane. “Disrupting the Intelligence Community: America’s Spy Agencies Need an Upgrade.” Foreign Affairs 94.2 (2015). Hart, Oliver, Andrei Shleifer, and Robert W. Vishny. “The Proper Scope of Government: Theory and an Application to Prisons.” The Quarterly Journal of Economics 112.4 (1997). Hersh, Seymour M. “Torture at Abu Ghraib.” The New Yorker. 10 May 2014. Web. 24 April 2015. http://www.newyorker.com/magazine/2004/05/10/torture-‐at-‐abu-‐ghraib Hill, Larry. “Who Governs the American Administrative State? A Bureaucratic-‐Centered Image of Governance.” Journal of Public Administration Research and Theory 1.3 (1991). Huntington, Samuel. The Soldier and the State. London: Belknap Press, 1985. “Inherently Governmental Functions.” 48 C.F.R. §§ 7.500–7.503 (2006). “Intelligence Community Directive 612: Intelligence Community Core Contractor Personnel.” Office of the Director of National Intelligence (ODNI). 30 October 2009. Web. 24 April 2015. http://fas.org/irp/dni/icd/icd-‐612.pdf Isenberg, David. “Private Military Contractors and US Grand Strategy.” International Peace Research Institute, Oslo (PRIO) (report). January 2009. Kennedy, Patrick, F. “Memorandum M-‐06-‐6019: IC Contractor Inventory.” Office of the Director of National Intelligence (ODNI). 12 April 2006. Krishnan, Armin. “The Future of US Intelligence Outsourcing.” Brown Journal of World Affairs 18.1 (2011). Kruck, Andreas. “Theorizing the use of private military and security companies: a synthetic perspective.” Journal of International Relations and Development 17 (2014). Laffont, Jean-‐Jacques. “Analysis of Hidden Gaming in a Three-‐Level Hierarchy.” Journal of Law, Economics, and Organization 6.2 (1990). Laffont, Jean-‐Jacques and Jean Tirole. “Privatization and Incentives.” Journal of Law, Economics, and Organization 7 (special issue) (1991). Levine, Michael E. and Jennifer L. Forrence. “Regulatory Capture, Public Interest, and the Public Agenda: Toward a Synthesis.” Journal of Law, Economics, and Organization 6 (special issue) (1990). Mayer, Jane. “The Black Sites” The New Yorker. 13 August 2007. Web. 24 April 2015. http://www.newyorker.com/magazine/2007/08/13/the-‐black-‐sites Mayer, Jane. “The Secret Sharer.” The New Yorker. 23 May 2011. Web. 24 April 2015. http://www.newyorker.com/magazine/2011/05/23/the-‐secret-‐sharer?currentPage=all 56 McConnell, Mike. “Overhauling Intelligence.” Foreign Affairs 86.4 (2007). Web. 24 April 2015. http://www.foreignaffairs.com/articles/62643/mike-‐mcconnell/overhauling-‐intelligence Meghani, Zahra and Jennifer Kuzma. “The ‘Revolving Door’ between Regulatory Agencies and Industry: A Problem That Requires Reconceptualizing Objectivity.” Journal of Agricultural and Environmental Ethics 24 (2011). Mundheim, Robert. “Conflict of Interest and the Former Government Employee: Rethinking the Revolving Door.” Creighton Law Review 14 (1980/81). Nathan, Laurie. “Lighting up the Intelligence Community: A Democratic Approach to Intelligence Secrecy and Openness.” Global Facilitation Network for Security Sector Reform (report). April 2009. Web. 24 April 2015. http://epapers.bham.ac.uk/1528/1/Nathan_%2D2009%2D_LightingupIntelligence.pdf O’Connor, Thomas. “Theories of Bureaucracy.” Megalinks in Criminal Justice. 14 October 2013. Web. 24 April 2015. http://www.drtomoconnor.com/4090/4090lect02.htm “Results of the Fiscal Year 2007 US Intelligence Community Inventory of Core Contractor Personnel.” Office of the Director of National Intelligence (ODNI) (conference call). 27 August 2008. Web 24 April 2015. http://fas.org/irp/news/2008/08/odni082708.html “SAIC Team Wins National Security Agency TRAILBLAZER Contract.” Business Wire. 21 October 2002. Web. 24 April 2015. http://www.thefreelibrary.com/SAIC+Team+Wins+National+Security+Agency+TRAILBLAZE R+Contract.-‐a093081931 Shorrock, Tim. “The Corporate Takeover of US Intelligence.” Salon. 1 June 2007. Web. 24 April 2015. http://www.salon.com/2007/06/01/intel_contractors/ Shorrock, Tim. Spies for Hire: The Secret World of Intelligence Outsourcing. New York: Simon & Schuster, Inc., 2008. Singer, Peter W. Corporate Warriors: The Rise of the Privatized Military Industry. London: Cornell University Press, 2008. Svara, James H. "Complementarity of Politics and Administration as a Legitimate Alternative to the Dichotomy Model." Administration & Society 30.6 (1999). Trebilcock, Michael J. and Edward M. Iacobucci. “Privatization and Accountability.” Harvard Law Review 116.5 (2003). “The US Intelligence Community’s Five Year Human Strategic Human Capital Plan: An Annex to the US National Intelligence Strategy.” Office of the Director of National Intelligence (ODNI) (report). 22 June 2006. Web. 24 April 2015. http://purl.access.gpo.gov/GPO/LPS75660 Weber, Max. The Theory of Social and Economic Organization. Translated, A.M. Henderson and Talcott Parsons. Mansfield Center, Connecticut: Martino Publishing, 2012. “Workforce Planning in the Intelligence Community.” RAND National Defense Research Institute (report). 2013. Web. 24 April 2015. http://www.rand.org/pubs/research_reports/RR114.html
Comments
Report "IR4099 - Outsourcing of US Intelligence dissertation v2"