Intelligence in Public Media

Intelligence Analysis: A Target-Centric Approach
Robert M. Clark (CQ Press, 2016, fifth edition), 448 pp., notes, figures, tables, appendices, index.

Reviewed by John Sislin and Christopher Marshall

The 9/11 attacks and the search for weapons of mass destruction in Iraq have both entered the American consciousness as archetypal intelligence failures. Major intelligence failures, whether characterized as such or not, produce self-reflection within the Intelligence Community (IC) and external scrutiny designed to drive intelligence reform, so “this will never happen again.” Noting that “[w]e learn more from our failures than from our successes,” Robert Clark opens the fifth edition of his book Intelligence Analysis: A Target-Centric Approach in that spirit. Clark identifies two goals for the book: “to redefine the intelligence process to help make all parts of what is commonly referred to as the ‘intelligence cycle’ run smoothly and effectively, with special emphasis on both the analyst-collector and the analyst-customer relationships” and “to describe some methodologies that make for better predictive analysis.” (xviii)[1] Further, Clark establishes a rationale for an intelligence process: “First, it should make it easy for customers to ask questions. Second, it should use the existing base of intelligence information to provide immediate responses to the customer. Third, it should manage the expeditious creation of new information to answer remaining questions.” (xix) To do this, according to Clark, the process must be collaborative and predictive.

What is the motivation for this book? Clark raises two justifications (in Chapters 1 and 2). The first, as noted above, is based on the assertion that previous, notable intelligence failures were preventable, and that if only intelligence analysts and collectors and their customers did a better job, then at least some future failures could be avoided. Clark identifies three types of intelligence failures: “failure to share information, failure to analyze collected material objectively, and failure of the customer to act on intelligence.” (3) Although it is true that there have been some spectacular failures, this assertion raises a host of questions that Clark (or most anyone else) cannot answer, including: how often do failures occur and what are their proximate and underlying causes? More importantly, scholars and pundits typically do not look at cases of success (mostly because success in the IC tends not to be remarkable) and causes of success. As such, it is difficult to suggest causality and perhaps even correlation.

A second motivation for the book is the change in the nature of current conflict. Clark asserts that 21st century “conflicts call for a different pattern of intelligence thinking, if we in the intelligence business are to provide the support that our customers need.” (19) Clark writes that the conflicts of today stand in contrast to the interstate wars fought in the 20th century. Clark argues that, at least in part because of globalization and the Internet, the two dominant characteristics of contemporary conflict involve the increased roles of networks and of nonstate actors. Networks simply refers to the multiple relationships between various actors on both (or more) sides of a conflict; nonstate actors include insurgents, transnational criminal enterprises, and individuals. Interestingly, the list seems to be missing terrorists, which some would argue is an area that the IC is too focused on. Clark mentions that subsequent chapters will explore “how to provide such support,” that was called for above—but it is questionable whether Clark explicitly answers why different intelligence thinking is called for. (19)

The book is well organized and presented in three parts. As noted above, two introductory chapters set the book’s goals in motion: the first offers examples of intelligence failures as a justification for better thinking about intelligence analysis; the second examines the role of intelligence in contemporary conflict. The latter chapter is a new component of this edition. Part I, in Clark’s words, is “how to do analysis.” (29) The nine chapters in this section focus on defining Clark’s “target-centric” process. Part II, which focuses on estimative modeling, or “creating target models of the future,” (215) contains an initial key chapter that discusses three types of prediction: extrapolation, projection, and forecasting. One methodology Clark explores as a tool for looking at the future is scenario-building, also known as alternative futures analysis. A second is simulation models, which are “mathematical descriptions of the interrelationships that are believed to determine a system’s behavior.” (275) Part III examines analysis as a function (how the analytic unit functions, what types of intelligence it should produce, and some reasons for intelligence failures), as a process, and as a structure (with such topics as what makes a good analyst and collaboration). This section of the book concludes with chapters on the nature of customers and of collection. Appendix 1 examines different analytic approaches in two National Intelligence Estimates—reports which represent an IC-wide view—on Yugoslavia and on Iraqi WMD. Appendix 2 is an example of a project plan, and Appendix 3 offers advice on how to present analysis.

It is important to note here that, within the field of intelligence studies, there is no accepted common definition of intelligence, and hence, no way to develop a theory to understand how it works. Clark, like many other scholars, has his own definition. He writes, “Intelligence is about reducing uncertainty in conflict [emphasis in original].” (19) The conflict does not have to be violent; it can refer to friendly competition. Conflict stems from “the divergence of two or more parties’ ideas or interests.” (19) Further, “[r]educing uncertainty requires that intelligence obtain information that the opponent in a conflict prefers to conceal.” (19) Clark then states, “A typical goal of intelligence is to establish facts and then to develop precise, reliable, and valid inferences (hypotheses, estimations, conclusions, or predictions) for use in strategic decision making or operational planning.” (19) Clark argues that national security intelligence is similar to other types of intelligence, such as market research, with the difference being that there are specialized techniques and methods that are unique to the intelligence field. An overarching theme for Clark is the notion of “conflict,” and although he notes it is a very broad concept, he tends to fall back to a narrower definition, which is connected with violence or the threat of it. Based on this definition, Clark develops his “Target-Centric Approach,” which ultimately should result in better analysis and, most importantly, better serve the needs of the customer. However, Clark makes several assumptions that may not adequately reflect how the IC actually works.

Students (and practitioners) of intelligence often find some dissonance between an ideal IC and the real world one. The current US IC comprises 17 agencies. Some agencies, such as the Central Intelligence Agency (CIA) or the Defense Intelligence Agency, focus more on collecting all sources of information to create assessments, while other agencies, such as the National Security Agency or the National Geospatial-Intelligence Agency, conduct analysis prompted by focusing on a single source of information, such as communications between adversaries or satellite imagery, respectively.

Moreover, the IC is quite large, and to a great extent it evolved organically after World War II. Given the secret world that the IC operates within, it is difficult for anyone outside it to be aware of all the current programs, organizations, etc. Although Clark has the bona fides to tackle this topic and had insider status (having worked at CIA), it is important to note that he is somewhat removed from considerations and challenges related to his recommendations, for example in encouraging greater collaboration among analysts at multiple intelligence agencies. Three examples illustrate this concern. First, most analysts and customers probably do not have the time to do what Clark is proposing. Clark does recognize this, but does not really address it (126). Second, he implies a closer relationship between the customer and the others in the IC than probably actually exists. Third, he assumes the IC is more collaborative than it actually is. Given that time, a close analyst-customer relationship, and collaboration are key to his approach, its viability can be called into question.

In addition to these possibly faulty assumptions, there are other gaps in Clark’s work. First, Clark says good intelligence requires teamwork. This is a testable hypothesis, yet Clark provides little evidence for the assertion. This is not unique to Clark: most scholarship in the field of intelligence is based on reasoning and not empirical evidence. There are numerous works that have identified the pros and cons of working in groups, yet very few of them are referenced or discussed to support Clark’s position. This in turn is a product of having no accepted definition, (and subsequent theory) of intelligence in the field that builds knowledge incrementally. Second, although Clark is well aware of individual cognitive biases and biases in group settings (e.g., “groupthink”), they receive little attention in the book. More important is the unanswered critical question of how the Target-Centric Approach would specifically mitigate such biases.

Finally, Clark does not really explain how to do analysis but rather generally describes various analytic methods and approaches to analysis. He says that “all intelligence analysis methods derive from a fundamental process,” and that this is what he is explaining in the book. (6) The basic process consists of “creating a model of the intelligence target and extracting useful information from that model” or synthesis and analysis. (6) Additionally, Clark is also providing a “general conceptual framework for all types of intelligence problems.” (7) There are several issues here. First, there is no empirical evidence presented that a conceptual framework, structured analytic technique, or method actually improves the quality of analysis either in the strength or accuracy of the argument. It seems reasonable that it would, and, in fact, many theorists believe this to be the case. Yet the field of study lacks hard data to support this position. Second, Clark never really provides a great deal of specificity on how to model a target or actually use these suggested frameworks. Finally, structured approaches are inherently proscriptive and can be complex. Analysts with little experience in using them can go awry, and the resulting analysis can be far less coherent and accurate than if an analyst did not attempt to lock-step through a structured approach.

Nevertheless, the book is valuable for students of intelligence and intelligence professionals for a number of reasons. It highlights many of the complex issues involved in producing quality intelligence analysis. The process is not linear and involves a lot of art in addition to science. As to structured analytic techniques, although they are not panaceas for analytic success, they can be useful and familiarity with them is not a bad thing, particularly with some of the more mathematical techniques like time series.[2] Quantitative frameworks are useful if analysts know how to use them properly. His focus on “targets” and, more specifically, targets as systems is refreshing. Systems are complex and linkages within those systems critical. Anything that improves our understanding of potential targets is welcome. Clark is also to be commended for taking the notion of a new edition seriously—by rearranging the book and updating the discussion and examples throughout. It might have been easier to simply tack on a new preface or new concluding chapter, but Clark seems to have invested some time in trying to improve the book as it evolved through subsequent editions. Finally, the book is well organized and flows logically. Those interested in intelligence and those who are actually intelligence practitioners should read this book for all of these reasons and, if for nothing else, because it highlights the enormous complexity and obstacles to quality intelligence analysis.

[Top of page]


[1] Numbers in parentheses refer to the page numbers on which authors assertions appear.

[2] Times series analyses seek to identify patterns in data over time. A common example would be stock market trend analysis. The analysis can also be used for forecasting.

[Top of page]

All statements of fact, opinion, or analysis expressed in this journal are those of the authors. Nothing in any of the articles should be construed as asserting or implying US government endorsement of their factual statements and interpretations. Articles by non-US government employees are copyrighted.

Posted: Dec 22, 2017 12:16 PM
Last Updated: Dec 22, 2017 12:16 PM