Features – Reinventing the Empire of Secrecy: An Agenda for the First DNI

(as approved by the CIA Publications Review Board)

Lee S. Strickland currently serves as a professor at the University of Maryland’s College of Information Studies (CLIS) and Director of the Center for Information Policy. The Center – jointly sponsored with the School of Public Affairs – is a multidisciplinary research organization that assists the public and private sector by facilitating solutions to current policy issues surrounding the convergence of information and technology. He also maintains consulting relationships with the federal government and holds a current TS SCI security clearance. Strickland has written extensively and been a frequently invited speaker as well as commentator for the media on cutting-edge issues. His articles have appeared in the Journal of Homeland Security, the Bulletin of the ASIS&T, the Journal of the ASIS&T, Virginia Libraries, Government Information Quarterly, the Information Management Journal, the Notre Dame Journal of College and University Law, and EDUCAUSE Quarterly. Strickland is a member of the District of Columbia and Virginia State Bars and numerous professional organizations. He is a graduate of the University of Central Florida (Bachelor of Science in Mathematics and Computer Science, magna cum laude), the University of Virginia (Master of Computer Science), and the University of Florida (Juris Doctor with honors).

This article was reviewed and approved by publication by the CIA Publications Review Board after a contentious process that initially denied publication in its entirety. The review process does not constitute official acknowledgement, confirmation or endorsement by the CIA.


Four years after the attacks on New York and Washington, and as Ambassador Negroponte builds-out of his office as the new Director of National Intelligence, there is no shortage of judgment concerning the requisite changes in the Intelligence Community (IC) in order to prevent another 9/11. Two high-level reviews (the National Commission on the Terrorist Attacks Upon the United States (the “9/11 Commission“) and the Commission on the Intelligence Capabilities of the United States Regarding Weapons of Mass Destruction (the “WMD Commission”)), one new law (the Intelligence Reform and Terrorism Prevention Act of 2004, Public Law 108-458), and two sets of proposals, one each from the CIA and the FBI, have offered detailed findings and recommendations.

Yet, to a substantial degree, these efforts have offered broad policy objectives or the time-honored solutions of more centralized management, organizational changes and additional resources for the IC. Notably absent were specific actions to resolve precisely the dysfunctions in American intelligence that allowed the first major terrorist attack on the homeland since a remarkably parallel, politically-inspired attack on the civilian population of Columbus, New Mexico 85 years ago. To be fair, the 9/11 Commission accomplished a very thorough and highly detailed assessment as to specific shortcomings and the WMD Commission followed with a very correct judgment on the systemic intelligence problems — human collection that is severely limited in terms of quantity and quality as well as finished intelligence production that is marred by analytic errors and poorly communicated. And although the WMD Commission spoke to remediation efforts such as integration across the IC including the appointment of mission managers, the fact is that the nation lacks an action plan predicated on a root cause analysis of specific functional failures that permitted terrorists to enter and operate with impunity inside the United States.

And this action plan is even more necessary when we consider the improvement proposals by the FBI and CIA that were ordered by President Bush in November and commented upon by the WMD Commission on March 29, 2005. The essential parameters: the FBI plan focuses on the creation of a new intelligence directorate while the CIA plan focuses on increasing the ranks of intelligence analysts and overseas operations officers by 50% through FY 2011. The assessment by the Commission: “a business as usual approach” that falls short, “timid” experimentation with new approaches, and designs “too general to create accountability.” But we know from the professional literature and experience in the business sector that organizations in today’s information age must employ systems thinking (see, e.g., Peter Senge’s Fifth Discipline) where the totality of inputs, processes and outputs – and their interrelationships – effect the desired outcomes. Moreover, it is apparent from these sources that it is the information proficient organization that is best positioned to reach outcomes and this results largely through a comprehensive mission and functional re-engineering in order to optimize the use of information and hence performance directly and over time by incorporating a lessons learned capability.

What this suggests is that successful organizations in the public and private sector recognize that their business model and structures must continually be re-invented if errors of the past are to be avoided and new threats met. It follows that the Intelligence Community, essentially a business conglomerate with independent divisions offering overlapping product lines, must also embrace innovation – new strategies, approaches and technologies – that can achieve better results. However here, we are not concerned with consumer goods and returns for voluntary investors but something vastly more important — national intelligence and the necessary outcome of deterring and preventing terrorism. Notably, the WMD Commission commented similarly by finding the FBI and CIA plans as proposing a “business as usual approach to intelligence” that will not serve to make America safer. The one thing that is certain is that the terrorist threat today is even graver and presents a more difficult intelligence target. Quite simply, our intelligence and military successes against al Qaeda have created a more diffused, polycellephous organization with less central control and more autonomous action. Indeed, its very role may be changing from directional of one organization to inspirational for many.

Thus in this evolving risk environment, is the U.S. Intelligence Community fundamentally flawed or is it a matter where past failings can be remedied by the creation of new, specialized intelligence centers and more resources for existing agencies? In other words, what should the agenda be for the new DNI? However one might answer these questions, there is the critical need to acknowledge that the IC, with the events of 9/11, had experienced a classic failure of knowledge management. In the private sector and scholarly literature, we define this as a failure to bring together an organization’s collective explicit information and tacit knowledge and, through rigorous analysis, create new knowledge so that business threats can be anticipated, identified and mitigated. This is the field of competitive intelligence – deriving from the historical realm of government intelligence – where the issues and needs are functionally identical.

But given the Congressionally-noted failures, that can best be characterized as an array of missed opportunities from lacking the required human intelligence assets to failing at analysis by not developing and rigorously evaluating a full range of hypotheses, where does the nation go? Is the direction for the future demonstrated by the Intelligence Reform Act of December 2004 and the previously discussed plans of the agencies? Many experts, as reported in the media, believe that the legislation, with its ambiguous powers for the DNI and its preservation of prerogatives for the Defense Department, sets the stage for business-as-usual if not additional bureaucracy and costly duplication of effort. Others are more sanguine that central, effective control can be achieved. But perhaps a truly strategic, finely granular approach is required that specifically addresses the problems that have so aptly been highlighted. What follows is a multifaceted plan that considers root causes and proposes fundamental changes that could make a difference within a year and quite likely free substantial funding to address targets of opportunity.

1. Rationalize and restructure, not reorganization and bureaucracy:

This objective is fundamentally different than organizational change that involves the creation of new centers and additional management layers that focus on coordination activities. Rationalizing and restructuring is a zero-based review of the lines of business in order to eliminate overlap, achieve economies of scale and, perhaps most significantly, encourage growth in new areas – all with the objective of achieving the best use of available resources. If the Intelligence Community is going to perform better, the first priority is to rationalize and functionally align the structure to the new era of intelligence threats (targets) and hence requirements (business needs) thus eliminating the competing capabilities that continue to grow including, for example, the human collection programs in the Department of Defense. In other words, today the Intelligence Community organizes, budgets and operates around capabilities – “INTS” such as HUMINT or SIGINT – and then proceeds to duplicate those functions at member agencies. Better would be a management strategy that organizes in terms of intelligence priorities and applies intelligence sources and methods as necessary to the unique tasks. Today, although there exists a National Intelligence Priorities Framework (NIPF) that is intended to focus collection and analysis activities on particular and evolving threats, we nevertheless organize and operate largely in terms of “INTS” and not threats. The result is that the collection machinery, not the national needs, drives the intelligence business.

Moreover, and this is true of both collection and analysis, there is the belief in many quarters that current security paradigms in the Community should reflect private sector business practices – undertaking quantitative assessments of threats, risk, potential and cost of mitigation. With this, the full benefit of new technological and human capital could be achieved, bureaucracy reduced, and agility to respond to new demands and changed environments vastly enhanced. Congress and others have observed that Intelligence is the most change resistant and risk adverse organization in government. Indeed, to prove this allegation, one would need only research the calls, studies and proposals over the years for change and the singular lack of results (see, e.g., Fixing Intelligence for a More Secure America by General William E. Odom). In business terms, any organization today must become information proficient through appropriate tools and policy that foster agility. Business dominance comes from this approach that produces quality intelligence and this is true whether the business mission is profit or homeland security. Indeed, applying this theory, one can argue that the real intelligence failure leading to the war in Iraq was not the flawed analysis (more about this later), but rather the lack of substance and depth in the collection and analysis of this target. While Iraq did not present a transparent government, it was not a closed, totalitarian society in the character of North Korea. Iraq government officials, businessmen, politicians, military officers, and well-placed citizens traveled the world and there is little doubt among experts that the intelligence product over the last ten years should have been prodigious. With a business focus and an optimized organization, that could have been the case.

The bottom line is that the Intelligence Community must be managed as a joint command with integration and cooperation beyond the boundaries of today’s individual agencies. This translates to a division of labor and the development of individual competencies much like the military has accomplished with joint structures where each service has assumed lead responsibility for a particular aspect of a defense priority (e.g., the Air Force lead in decontamination for the Joint Chemical Biological Defense Program). Such a restructuring for intelligence might well divide responsibilities even for activities such as covert human collection and it would almost certainly divide analytical focus along both functional and geographical lines. The point is that operational tradecraft knowledge and analytical subject matter expertise are severely limited and must be developed – a process that is not consistent with the current structure where agencies are so expansive in their focus and individual position tenure so limited. In sum, a joint structure, built on a plan from demonstrated and nurtured capabilities, would ensure efficiency and true integration of both the activities and the resulting products. This is the path toward homeland security.

2. Institute a performance-based management system:

At least in part because of inherent secrecy in the intelligence process, the IC is a remarkably insular organization that has avoided most of the Congressional initiatives for measurement. For example, the 1993 Government Performance and Results Act (GPRA, or “The Results Act”), P.L. 103-62, was intended to enhance the effectiveness, efficiency, and accountability of government programs by requiring agencies to focus on results rather than traditional concerns such as staffing and activity levels. Under the GPRA, agencies must publicly set goals, measure performance, and report on their accomplishments; specifically required are Strategic Plans, Annual Performance Plans and Annual Program Performance Reports. In doing so, agencies need to be able to articulate their mission, identify goals, identify activities that will achieve those goals, identify performance measures, and identify how that measurement information will be used to make improvements.

Yet, inside the IC, very little if any measurement activities take place at the project or even higher levels and it thus impossible for senior management to prudently allocate funds or understand the efficiency of expenditures. Now, few would argue that this objective is simple — as evidenced by the 10 March 2004 report of the Government Accountability Office (GAO-04-38 available at http://www.gao.gov/new.items/d0438.pdf) that states the government is doing better but still has “difficulty setting outcome-oriented goals” and developing measures for these rather than simplistic measures of outputs. Why measures of outputs such as speeding tickets issued, human sources recruited or terrorists eliminated are insufficient is evident from the definitions: outputs look to merely to the process while outcomes look to results, goals and allow systems improvement — in our case the outcome-oriented goal of “eliminating the terrorist threat to the homeland.” In other words, ignoring measures or focusing merely on outputs is a guarantee of strategic policy failure. In sum, the IC needs to elevate measurement to a strategic imperative with respect to operations, analysis and the myriad support functions including information technology.

3. Engineer an effective information sharing policy and system:

There is a basic truth, seemingly anomalous, that secrecy and intelligence efficiency mutually conflict. Secrecy inhibits the efficiency of analysis by limiting access to information but the lack of secrecy can also harm analysis by disclosing sources and methods. And often this latter concern is more bureaucratic than fact-based especially when secrecy impairs sharing within the IC among security-cleared personnel or between federal and local government agencies where there is a critical need for the vertical sharing of information.

This suggests there should be some optimal balancing point but little thought or study has taken place in this regard and the current culture of secrecy clearly disfavors information sharing and hence hinders knowledge development in the IC as well as the universe of homeland security personnel. Although some progress has been made (e.g., better sharing between the FBI and CIA), the lack of information architectures and a focus on technology to permit sharing rather than policy to enable sharing confirm there is much to do. One answer is to create by statute a Chief Knowledge Officer at each IC agency, to be appointed by the President and confirmed by the Senate, whose responsibility would be to organize and proactively share institutional knowledge. Quite simply, effective sharing depends on changing the culture and creating new functions to manage knowledge through a new lifecycle – from creation through dissemination to all prospective users including the 95% of homeland security personnel that work at the local and state level. Yet today sharing all too frequently focuses on technical capability to move data, is constrained by deficiencies in the information architectures at the various agencies, and almost never addresses the implicit knowledge in the organizations. Just as the statute that created the Inspectors General at federal agencies in a prior generation helped solve rampant waste, fraud and abuse, so dedicated responsibility for knowledge sharing can improve intelligence analysis and hence homeland security.

4. Revitalize the scientific discipline of intelligence analysis:

If there is one improvement that can be accomplished with certainty, it is ensuring that the analytical errors leading to 9/11 and persisting today are abolished and that a collaborative, sound process and environment for analysis is created throughout the Intelligence Community. Largely forgotten in the bureaucracy and proliferation of agencies is the fact that intelligence analysis is much more than thinking about a problem, embracing the conventional wisdom, or offering judgments influenced by political considerations; quite simply, analysis is the scientific method in action.

The synonymy of analysis and science developed from the work of two members of the academic community who together occupied the role of chief intelligence analyst from 1941 to 1967 – first Dr. William Langer and second Sherman Kent. Together they began in the Coordinator of Information’s (COI) Research and Analysis Branch, oversaw the transition to the wartime imperative of the Office of Strategic Services (OSS) and returned to serve in the newly created Cold War-era Central Intelligence Agency at the request of then-DCI Lt. Gen. Walter Bedell Smith in November 1950 in order to re-establish a powerful analytical base. Langer, as the first director of the new Office of National Estimates (1950-1952), and Kent as the second (1952-1967), created the modern intelligence function of analysis as a scholarly discipline with a defined methodology. Together, in the words of CIA historian Don Steury, they established intelligence analysis as a process characterized by “intellectual rigor, conscious effort to avoid unconscious analytical bias, willingness to hear other judgments, collective responsibility of judgment, precision of statement, systematic use of some of the country’s best experts as consultants and checks against in-house blunder, and candid admission of shortcomings.”

Of course this is not to say that the rigorous process was infallible – Kent, for example failed to predict the deployment of Soviet, nuclear missiles in Cuba — but it is to say that the process, as with all science, learned from its failures, publicized those learnings and improved. In the proper execution of analysis, judgments are developed through a rigorous, scientifically-based human process of analysis of competing hypotheses (ACH). In the seminal work to discuss this approach in detail, Psychology of Intelligence Analysis (available at http://www.cia.gov/csi/books/19104/), Richards J. Heuer, Jr. explains the basics of analysis, the ACH process and mechanisms by which it avoids many of the classic pitfalls in judgment. He posits that the predicate to effective analysis is decomposing and externalizing complex analytical problems where decomposition is defined as breaking the issue into its component parts and externalization is defined as documenting the decomposed problem on paper or other media in order to interact effectively with the totality of the issue. Without this approach, issues quickly exceed the ability of working memory. And inherent is this process of structuring an analytical problem, is the importance of recognizing “evidence” which is data validated by ensuring that it is accurate, precise, sufficient, representative and authoritative.

Thereafter, effective analysis is accomplished by a process through which a complete set of hypotheses are generated and rigorously evaluated – an approach that produces more accurate judgments by avoiding decision-making strategies that are identifiably faulty (e.g., satisficing), can overcome cognitive bias in the evaluation of evidence as well cause and effect, and is particularly useful for controversial issues. In sum, ACH proceeds from the identification of a full range of hypotheses (prospective answers to the research question), the evaluation of the evidence and assumptions for consistency or inconsistency with each hypothesis, and thus the identification of most likely outcomes.

The importance of generating the most complete set of hypotheses, of course, cannot be overstated and was recognized by the WMD Commission as a key objective as was the use of innovative analytical methodologies and tools generally. Indeed, a related method of note is Bayesian analysis — where the probabilities of competing hypotheses are revised by new events according to a mathematical computation based on their relative likelihood — has also been known within the Intelligence Community for many years (see, e.g., “Bayes’ Theorem for Intelligence Analysis” by Jack Zlotnick, Studies in Intelligence, 1972). Although this approach does not eliminate the potential for cognitive and experiential bias, it has been incorporated into commercial software today that facilitates the creation of complex models (Bayesian belief networks) required by intelligence including multiple hypotheses and interactions between events. Moreover, these models may be continually updated with new evidence thus facilitating the fusion of all available data into the evaluation process (a point made by the WMD Commission) and demonstrating the quantitative impact on probabilities by particular evidence.

But readers may wish to consider the evidence available today from independent reviews demonstrating that the Intelligence Community has slipped away from such requisite rigor and analytical approaches. Indeed this is the general conclusion in a recent work by Dr. Rob Johnston, Analytic Culture in the U.S. Intelligence Community: An Ethnographic Study (available from GPO in print), who was invited by the Director of Central Intelligence in August 2001 “to identify and describe conditions and variables that negatively affect intelligence analysis.” As evident, this work began contemporaneous with the events of 9/11 and this fact contributed to the great interest demonstrated by the subjects that participated in some 489 interviews and focus groups. The key findings and recommendations for improvement (synthesized from the views of Dr. Johnston and the author of this article) follow:

Time demands inhibit analysis: For several years experts have recognized that the glut of information was overwhelming the abilities of analysts but it is now also clear that the demands for current production of intelligence are so daunting as to impair “group interactions and the analytic process.” In the words of one analyst, “people seem to have confused writing with analysis.” And in the words of another: “We’ve got Bayesian tools, simulations, all kinds of advanced methods, but when am I supposed to do any of that? It takes all my time to keep up with the daily reporting as it is.” Thus, an information overload with a work focus on the short term, supported by a reward system that is perceived to support this dichotomy, has inhibited quality analysis.

The loss of scientific methodology: Although the culture and literature of intelligence generally focuses on operations rather than analysis, the fact is that analysis, as recognized by General William Donovan, is the essential heart of modern intelligence. Yet the focus on operations has led to a view that the “tradecraft” concept in operations is equally applicable to both arenas. But as we have discussed, analysis is neither craft nor art but rather a part of the scientific process and while the process is taught, few managers and analysts are practitioners. The result can be seen in the analysis leading to the war with Iraq.

5. Develop and implement a strategy against terrorism:

The continuing and escalating incidence of terrorism, from New York and Washington four years ago to the streets of Baghdad, London and Madrid more recently, symbolize the fact that the threat to homeland security today is not al Qaeda per se but rather a society within a society that detests but exploits Western culture for its operational success. If fact, the intelligence and military success against al Qaeda can be perceived as creating the opportunity for this new threat signature. Consider, for example, Abu Musab al-Zarqawi, a Bedouin common criminal turned jihadist from Jordan, who has adopted the banner of al Qaeda and has formed ad hoc groups into an effective terror machines with multiple kidnappings and beheadings to their credit and a growing media capability that not only enhances the effects of their actions but also facilitates recruitment.

This equates to a need to move intelligence policy to a broader, strategic approach to defeat terrorism – a concept advocated, in fact, by the 9/11 Commission with four recommendations to develop a global counter-terrorism strategy and nine others to prevent the continued growth of Islamic terrorism through the use of information, collaborative relationships with other countries and economics to win the battle of ideals in the Middle East. Yet there remains a drift in national policy in this regard. And the threat is exacerbated by the educational system that remains in the hands of extremists in broad areas of the Middle East and by the continued existence of training grounds in Iraq, Afghanistan and Pakistan that are effective exporters of terrorists to countries throughout the world. Watch how National Security Advisor Townsend either shapes or reflects policy in this regard and whether a new presidential directive issues. In short, the threat to homeland security is broad and rooted firmly in economic deprivation and cultural issues. The intelligence agenda must focus as brightly here as it does on eliminating individual and group threats.

The bottom line: The report of the 9/11 Commission that documented the shortcomings in American law enforcement, intelligence, leadership and Congressional oversight also found that this attack should not have come as a surprise and that the primary governmental failure “was one of imagination.” The accuracy of this conclusion is amply established in the evidence and Staff Statements of the Commission (e.g., No. 11, p. 6) as well as extensive media reporting. In fact, the government at large, as an information space, did know that such a scenario was possible but was unable to marshal its assets in mitigation. What each of these recommendations says in substantial part is that the Intelligence Community must change from its insular past and make innovation and learning part of the intelligence culture. And in this regard there is promise; for example, the CIA has initiated an Innovation Center and is considering ways to create an institutional learning capacity. But much remains to be done including a fundamental restructuring and the development of an information proficient Intelligence Community in terms of collection, sharing and analysis. The question is whether the political establishment can make the difficult and complex business decisions that are clearly necessary.

Posted in: Cybercrime, Features, Government Resources, Information Management