The U.S. Intelligence Community and Foreign Policy: Getting Analysis Right

September 15, 2009

Executive Summary

Intelligence analysis seeks to provide necessary information
in a timely manner to help policymakers from the president
on down make better decisions. The information and judgments
must be pertinent to what policymakers need to know
but not skewed to support a particular policy outcome. In reality,
this is more of an art than a science, especially because
the manner and means of most effectively informing the president
and other senior policymakers changes with the preferences
and working style of each new administration.

The Intelligence Community (IC) of the United States has been undergoing major reforms since 2005 when President George W. Bush signed the Intelligence Reform and Terrorism Prevention Act. Under the new Director of National Intelligence, the shortcomings in intelligence analysis that came to light in the wake of the 9/11 and Iraq WMD intelligence failures are being addressed through revamped analytic standards, increased resources for the IC, and numerous organizational and procedural changes. These analytic transformation initiatives seek to reduce barriers among organizations and individuals across the IC and to more effectively prioritize missions.

As of now, many of these innovative initiatives are in the development stage. Once completed, given their conceptual and technological complexity, it will be important to continually assess whether these initiatives result in a significantly improved analytic product. Mindsets and cultures of various IC components may prove serious obstacles to the kind of open and collaborative environment envisioned in these efforts; these new capabilities may prove most effective among digitally- savvy analysts in their twenties rather than among more senior analysts and managers.

Against this background of ongoing reform, this study assesses the current state of play, identifies systemic concerns, and offers practical ideas to improve analytic transformation and make the interactions between the analytic community and policymakers more effective. Extensive interviews with current and former policymakers and intelligence community analysts and managers reveal that there are flaws in the current system that require dedicated attention. The most consequential include:

  • Overemphasis on the President’s Daily Brief (PDB) – President George W. Bush elevated the PDB to an unprecedented level of importance, which had the unintended effect of skewing intelligence production away from deeper research and arms-length analysis to being driven by the latest, attention-grabbing clandestine reports from the field.
  • Disappointing National Intelligence Estimates (NIEs) – NIEs are meant to be one of the major products of the IC, yet they are frequently too late, too long, and too detailed to serve high-level policymakers well. Moreover, NIE analytic quality is often compromised by the effort to present a unified analytic position, producing reports that can become the lowest common denominator statement that is able to achieve agreement across the IC silos.
  • Analytic Risk Aversion – In the wake of the Iraq WMD fiasco, the pendulum has swung decidedly toward a tendency for analytical products to focus on amalgamating all potentially relevant data and to present only that to policymakers—leaving it up to them to draw the analytic conclusions. DNI Dennis Blair has recently made a welcome commitment to having opportunity analysis— the identification by analysts of unanticipated windows of opportunity to advance U.S. policies—become a key component of intelligence products.
  • Insufficiently Deep Country Knowledge – Many of the young IC analysts are trained to follow a particular stream of information from ”their” country but lack the deep immersion in the country’s political system, economy, and modern history necessary to produce nuanced, insightful analytic products. Moreover, very burdensome security constraints make it extremely difficult for them to build that kind of analytic depth.
  • Overemphasis on Classified Sources – IC analysts tend to gravitate to information obtained by clandestine means. Yet much of that information lacks context and is substantively rather marginal. As a consequence, analyses overly driven by classified sources may suffer from ignorance of important information in unclassified sources. This is especially notable with the explosion of unclassified material now available on key targets such as China.

This report’s recommendations to address these shortcomings fall into three broad categories.

On improving the capabilities of analysts:

  • Recruit a greater percentage of the incoming class of analysts from those in their late twenties and early thirties who have had extensive experience related to the country of concern – This change can present a security challenge but the added benefits in terms of maturity, life experience, and deeper country knowledge are worth the additional effort and attention needed to clear these individuals.
  • Establish a National Intelligence University with its own campus and faculty – If the vision of a truly integrated analytic corps is to be achieved, there needs to be an academy that allows the IC to not only establish crossagency relationships and cultivate common standards and procedures, but also to better draw lessons from its own historical successes and failures and to incorporate those into training programs.
  • Devote greater time and attention to formal training – To address the question of analytic depth, special short-term courses that draw in specialists from outside of the IC and that test participants’ learning in the course should be conducted on a regular basis. Moreover, analysts should be encouraged to attend programs held by various Washington-area think tanks, not (as is now the case for many) discouraged due to security concerns.
  • Nurture and reward area specialists – There is no substitute for the key analyst with deep substantive knowledge and experience on a single country or issue. The IC may wish to consider assigning some analysts to conduct in-depth studies of major long-term issues in key countries such as China (e.g. study of the long-term evolution if civilian-military relations in the PRC) in order to help a cohort of analysts develop such depth.
  • Break stovepipes in analytic assignments – On National Intelligence Estimates and other key products, consideration should be given more often to assigning individuals from two different disciplines joint leadership in developing the analysis. This would foster, for example, greater integration of political and technical analysis of missile development.

On improving the utility of IC analytical products for policymakers:

  • Provide formal introductory briefings for incoming policymakers on IC capabilities and limitations – Often, new policymakers come into office with very impressionistic and misinformed views on what the IC is able to produce. Senior IC managers should develop introductory briefings that help policymakers think critically about their intelligence needs and how they can best utilize the IC.
  • Assign IC analysts systematically to provide on-site support to policymakers at and above the assistant secretary level – This not only can help the policy maker but also can provide invaluable feedback to the IC about the policy maker’s actual intelligence needs.
  • Develop regular feedback mechanisms from the policymaker to analysts – Periodic meetings can greatly help the IC understand the look-ahead intelligence requirements of policymakers and garner critical feedback on materials sent over since the last such meeting.
  • Allow for NIEs with formal dissenting opinions, similar to Supreme Court decisions – In such NIEs, dissenters can write specific dissenting opinions and even those who agree can pen concurring opinions that articulate a distinctive analytical approach.
  • Train analysts in the power dynamic between analysts and policymakers – The desire of analysts to please the most senior intelligence consumers who are driving to a decision based, in part, on intelligence judgment can lead analysts unintentionally to overstate their confidence in the intelligence. Analysts need to be better trained and equipped to understand the subtle effects of power dynamics between analysts and policymakers, and policymakers need to keep in mind that their power and positions are intimidating to many analysts who brief them.

On improving the ability of policymakers to elicit and utilize high quality IC analysis:

  • Encourage policymakers to better articulate their intelligence questions and priorities – Taking the time to think through the analytic question they want answered will pay dividends for policymakers. Requests that do not assume the form of analytical questions too often fail to motivate IC analysts to think through the implications of their data, debate the relative significant of different factors, and make explicit their levels of confidence in their responses.
  • Elicit what analysts know, what they don’t know, and what they think is likely to happen – Former Secretary of State Colin Powell told his IC briefers that they would be responsible if he took action based on what they said they know and do not know but that he would be responsible if he took action based on what analysts said when asked what they think is likely to happen. As a result, he incentivized analysts to be both rigorous and thoughtful.
  • Provide the IC with the insights the policymakers themselves gain from their meetings with foreign officials – Presidents and many other senior policymakers are experts at “reading” other political leaders—a skill most IC analysts understandably do not share. If such insights are routinely shared they may improve the quality of intelligence analysis, especially as regards elite politics.
  • Avoid as much as possible the temptation to declassify NIEs – When NIEs are likely to be declassified, analysts are prone—either consciously or subconsciously—to pull their punches and hedge their analysis. Moreover, the impulse to declassify NIEs or to leak selectively from NIEs is often based on the faulty assumption that the IC’s analysis can and should authoritatively settle a policy debate.

In the wake of failures early in this decade, the Intelligence Community today has both the opportunity and obligation to transform itself. With fifty percent of the IC workforce hired since 9/11, there is now a large pool of young, technologysavvy talent that is eager to be shaped into a superior new IC. Indeed, cultural shifts based on the information age almost guarantee that many important changes will happen simply because of the nature and talents of this younger generation.

Ongoing IC cultures of insularity and secrecy, though, present major obstacles to realizing the IC’s full potential. For example, some IC managers continue to deny information to other parts of the community because they do not utilize identical security screenings, such as the polygraph. To cite another example, the need for a National Intelligence University has been understood for some time, but the IC’s sixteen disparate agencies still resist merging their educational and training programs. This resistance highlights that the IC still has some distance to go in terms of individual agency cultures and mindsets if it is to be truly unified under the leadership of the DNI.

The division of labor and of tasking among the major components of the IC should remain a concern. Post 9/11 changes created the ODNI and repositioned the CIA and the NIC, among other shifts. In short, key pieces have been moved on the IC chessboard, and such major changes inevitably require a substantial period of time to gel fully. This report does not, therefore, provide specific recommendations on additional changes in the distribution of responsibilities and authorities among the major IC players. But the research suggests that a thoughtful review of current relationships—especially those among ODNI, the NIC, CIA, DIA, and INR—might prove of considerable value again in about two to three years.

Finally, the task of analytic transformation cannot fall on the IC alone. Policymakers can affect the quality of analysis if they do take the time to provide clear and candid feedback to the IC. Policymakers also should understand the process of intelligence analysis to the point that they can read products as well-informed customers. It would be helpful to good analysis if policymakers realized their own value as IC sources. They should in particular inform analysts of relevant discussions with foreign leaders that may shed light on intentions and motivations. Too often policymakers simply assume that analysts know what the policymakers themselves know, and that comes at some cost to insightful IC analysis.