3

Needs of the Intelligence Community

David Honey (Office of the Director of National Intelligence [ODNI]) introduced the first panel of presenters, which included representatives from the intelligence community (IC), who provided insight on how the SBS decadal survey could successfully inform the work of intelligence analysts. They discussed how intelligence agencies use, or would like to use, social and behavioral sciences and what areas might have the most impact on their work.

CENTRAL INTELLIGENCE AGENCY

George Gerliczy (Central Intelligence Agency) noted he has worked with academic experts outside the IC on a variety of projects but is looking for a more comprehensive and consistent interaction to make intelligence analysis stronger. He said the SBS Decadal Survey may make headway against a perennial challenge for IC analysts to stay abreast of the latest findings and advances in the social and behavioral sciences and to draw relevant knowledge into their work.

Gerliczy focused on three messages. First, he highlighted what policy makers, as the customers, expect and need from the IC analysts. Second, he introduced the notion of an analytic framework, which analysts use to make sense of received information and help policy makers understand the issues they face. Third, he suggested ideas for the collaboration between the IC and academics and considered different ways that research can have an impact on intellectual capital building.

Turning to his first point, Gerliczy noted that analysts do three types of tasks. First, they provide policy makers with answers to specific questions. The questions may have been explicitly asked, but more often analysts anticipate and then proactively answer questions. Second, analysts present policy makers with an analytic framework to facilitate the understanding of an issue of interest as well as the processing of new information. As such, the framework has a longer shelf life beyond specific answers to narrower questions. Third, analysts warn policy makers, as appropriate, of impending developments or crises, he said.

Gerliczy noted that the questions policy makers ask analysts vary over time and in terms of specific country, issue, or leader; however, the questions can be categorized in several themes (not an exhaustive list):

• Strength of leaders and governments: assessments of the stability or discretion of a particular government or leader and the consequences of a change in leadership;
• System dynamics: the risks of failure and instability within a system, whether states, societies, alliances, networks, or organizations, or the kind of system likely to emerge out of a current set of dynamics;
• Role of political philosophy or ideology: role of an ideology in motivating political activity or violence, or whether there might be an opportunity for transnational mobilization in support of or against a particular ideology or movement;
• Calculations between and among states or individual actors: consideration of a state’s or a group’s capabilities, plans, or intentions vis-à-vis another state or group, or how a state or group perceives its strategic position in the region or in the world. Questions about deterrence and about action-reaction dynamics also fall into this category; and
• Identification and measurement of threats: considerations of threats across a variety of domains, such as political, cyber, terrorist, and military, as well as how, when, and to what end a threat would hurt U.S. interests and the opportunities to undermine or defeat those threats.

Gerliczy next expanded upon his description of an analytic framework. Intelligence analysts, he said, often spend a lot of effort identifying existing frameworks or developing new ones. Frameworks provide insight. An analytic framework is essentially a description of the primary drivers of an issue, that is, the key variables that determine how a situation is likely to evolve. Analysts often have to take the questions asked by policy makers and narrow them to more focused questions where a framework may apply.

For example, if the question is about the stability of a government or leader, analysts might look for a framework that would help identify the factors that determine or reflect newly installed leaders’ ability to consolidate power. This analytic framework needs to identify several things: (1) the key drivers or variables that facilitate the consolidation of power; (2) what is known about the interrelationship among those variables and their overall predictive power; and (3) what is known not to facilitate the consolidation of power. According to Gerliczy, it is very important to know what research has shown not to matter in order to focus limited resources appropriately. It is also useful to know the leading paradigms for a certain issue, as well as areas of contention or debate (e.g., the gray zones).

Gerliczy pointed out that analysts must be able to apply a framework effectively to a specific context and convey that framework to policy makers along with whatever insights the analysts gleaned from using it. This application takes expertise in two different areas: (1) general, broader understanding of the phenomena (i.e., the issues at play) and (2) specific knowledge of the context or country under consideration.

On his third point, Gerliczy offered suggestions for collaboration between the academic and intelligence communities. Currently, he said, exchanges take place on specific cases (i.e., more narrow applications), but it would be useful to shift to developing frameworks to generate insights with broader, lasting applicability. The challenge is for people outside the IC, including the academic community, to understand what analysts do, what their needs are, and the kinds of assessments they produce. Generally, according to Gerliczy, this is not because of a lack of interest but because of classification restrictions and secrecy and a culture that emphasizes discretion. He suggested the IC become more forthcoming in trying to explain and share its priorities, and he expressed commitment to becoming part of this effort.

Gerliczy acknowledged differences between the way the two communities operate. Often academics, he said, look backward to determine what mattered to a known outcome. The IC looks forward to diagnose a situation as it unfolds and determine implications for policy. There are differences in data, he observed: academics often make use of established datasets or assemble new ones, and the IC typically has either a flood of details or complete lack of information. He also pointed to differences in the two communities’ areas of inquiry (areas of narrow focus compared to those with a broader focus or with considerable ambiguity), consumers (fellow experts in a substantive area compared to policy makers, who are generalists), and time frames and space to convey assessments (i.e., often IC analysts are given only a few days and two pages to present their findings).

Gerliczy suggested ideas for overcoming these differences to facilitate collaboration. As part of the SBS Decadal Survey, he offered the CIA’s com-

mitment to help academics better understand analysts’ priorities and work practices. He said a list of perennial questions will be developed further and prioritized to help those contributing to the project understand what analysts grapple with on a regular basis and where they can use the most assistance.

He also suggested working with the academic community to develop analytic frameworks on pressing problems that could be useful to analysts working in real time. He envisioned a layered analytic framework, such as a 50-page document that comprises a 1-page visual aid (e.g., a graphic or table that frames key factors), 3-page executive summary, and longer literature review. Developing such frameworks would be better if done collaboratively to ensure that the distillation of research knowledge about a particular issue fits the academic community’s and the IC’s perspectives.

In closing, he noted that success in building broader intellectual capital could have a significant and lasting impact on the quality of analysis within the IC, and thus in enhanced national security.

DEFENSE INTELLIGENCE AGENCY

In his opening remarks, Geoffrey Strayer (Defense Intelligence Agency [DIA]) noted that while DIA may have a narrower analytic focus than the CIA, it performs this analysis for three very large Department of Defense (DOD) customer sets: warfighters, policy makers, and the DOD acquisition community.

Strayer explained that the 2015 National Military Strategy¹ defines national military objectives as follows: deter, deny, and defeat state adversaries; disrupt, degrade, and defeat violent extremist organizations; and strengthen the global network of allies and partners. He emphasized that combat is the last line of defense in everything the military does. However, when it does engage in combat, the military aims to succeed. Strayer pointed out that both staying out of combat and being successful in combat require information.

In managing and analyzing information, the goal is often to be in the position of decision advantage. The DIA uses social and behavioral information to understand adversaries and allies and to gauge their reactions to stimuli. It also uses information and research to understand the implications of behaviors toward military objectives.

The military through the DOD intends to defend U.S. citizens wherever they happen to be. As such, Strayer said, the purview of the department

___________________

¹ The National Military Strategy of the United States of America, 2015, can be found at http://www.jcs.mil/Portals/36/Documents/Publications/2015_National_Military_Strategy.pdf [January 2017].

spans the globe. Strayer described DOD operations as a large, complex machine with every part requiring intelligence to operate effectively. Intelligence analysis helps determine whether the right goals were selected, and as situations develop and change, it helps determine what will impact the ability to achieve the goals. Globalization, technology diffusion, and demographic shifts can alter the strategic environment. The IC must constantly look for the best information and techniques to build decision advantage.

Strayer suggested areas where additional knowledge would be valuable to military operations:

• Understanding cultures: to better know adversaries or allies, especially elements of society engaged in combat. What motivates combat or tensions between factions or between forces in local populations? Who are the role models and icons that drive warfighters’ behavior? What rules do they follow in warfare? This is everything from whether they accept collateral damage, their no-strike targets, and their treatment of prisoners.
• Gauging response and tactics: to better understand how adversaries or allies will react to military response and changes in weapon systems. Will they be willing to use a dangerous weapon as a first-strike weapon? Or will they consider that weapon as an option only if the United States does a certain activity? What are the tipping points toward peace compared to those toward war?
• Modifying plans accordingly: to better ensure success of objectives. Strayer noted that the military prepares for contingencies through an adaptive planning process. The planning process has a number of phases, and each phase is predicated on the success or failure of the phase before. It is critical to know an adversary’s culture and tactics in advance of decisions to change plans. What might work as a deterrent for one adversary but not work as a deterrent for another? What might stimulate conflict with different adversaries?
• Making correct decisions, faster: to aid decision-making in real time. In modern warfare, the IC is increasingly separated from the decisions made in the field that are predicated on information immediately around the warfighters. Such reality requires planning and preparing for every contingency. Strayer suggested that useful software models would encompass the best understanding of reactions; permit the addition of stimulus to reflect options; take advantage of historical evidence; and allow the testing of possible scenarios before troops are put in danger.

In closing, Strayer emphasized that warfighters need information in a different form than policy makers. Warfighters, who are often put in harm’s

way and asked to make decisions, could potentially benefit from knowledge from the social and behavioral sciences.

NATIONAL INTELLIGENCE COUNCIL

Charles Gaukel (National Intelligence Council) explained that his organization is small, resides within ODNI, and draws on senior analytic expertise from across the 17 organizations in the IC. It focuses largely on strategic long-term issues of enduring interest or areas of emerging interest to U.S. policy makers. Gaukel said the SBS Decadal Survey was commencing at a critical time and would help address how the academic experts in the social and behavioral sciences could advance understanding of major challenges in national security and how intelligence analysts can make the most productive use of those advances.

Gaukel concurred with Gerliczy that intelligence analysts are typically charged with anticipating and responding to questions about a particular problem or issue. They are rarely asked to address general questions, whereas the academic community looks at broad explanatory theories with the potential to explain significant portions of variance. Gaukel referred to a book written by the late Alexander George.² The author lamented that much of what scholars focus on is of little interest to national security analysts and policy makers. However, he also pointed out that far too many national security analysts and policy makers have little knowledge of scholarly work that actually might assist them.

George’s book identified three areas where more knowledge would be useful: (1) conceptualization of strategies and instruments of policy; (2) limitations and necessary conditions of the successful employment of each of these strategies; and 3) actor-specific behavioral models. These areas remain of interest today, Gaukel said, and are more challenging to understand given the greater dynamism and consequent uncertainty.

Gaukel noted that the potential for strategic surprise is increasing in both pace and diversity. Some areas previously and continuing to be of concern (such as the risk of proliferation of weapons of mass destruction; nonstate actors able to avail themselves of weapons; ideologies that justify extremism and mass terror; and the reemergence of state-on-state war) are being compounded with new concerns (such as social fragmentation and growing inequality; climate change; and cyberattacks that could create global, financial, or social shocks). The IC is only beginning to contemplate the implications of advances in computational science, artificial intelligence,

___________________

² George, A.L. (1993). Bridging the Gap: Theory and Practice in Foreign Policy. Washington, DC: United States Institute of Peace.

and genetic engineering that may raise profound questions about the very meaning of being human.

According to Gaukel, making headway on intelligence analysis of these concerns is complicated by at least four interrelated trends: (1) the increasing rate of change with which these new challenges, threats, and opportunities emerge, making terms like “over-the-horizon threats” almost obsolete; (2) “noise,” that is, early warnings of emerging threats that are based solely on weak, conflicting, and heretofore unrecognized signals, against a background cacophony of noise; (3) a multiplicity of actors who can directly and quickly impact national security, including states, substates, terrorist groups, and other groups such as nongovernmental organizations, international businesses, and financial firms; and (4) events that unfold in complex closely coupled systems that are often poorly or incompletely understood. Gaukel remarked, “It’s not surprising, therefore, that the intelligence community would reach out to the academic community for help in bounding this uncertainty that policy makers face.”

He said the IC needs help developing and deploying methodologies, approaches, tools, and techniques to understand these concerns and anticipate emerging ones. Gaukel emphasized that advances in methodologies will have to be accessible, understandable, and easily utilized by line analysts. In addition to better tools, the IC needs help determining the quality and utility of data. Without advances in the theoretical understanding of social behavior, increasing access to more and more data may, according to Gaukel, distract analysts, diverting analytic efforts on where data happen to be available instead of on critical factors. Needed subject domains cover a huge spectrum and include, but are not limited to, areas such as understanding deterrence and escalation; implications of massive destruction power; risk of pandemics, either naturally occurring or intentionally caused; and conflict or migration exacerbated or perhaps triggered by climate change.

In closing, Gaukel pointed out that intelligence analysts and policy makers accept uncertainty, but in applying research to practice, analysts need to figure out the tipping point or how to sort out the true positives from the false positives and negatives. He offered an illustration: “It won’t be enough for analysts to be able to tell a policy maker that research shows 70 percent of the time this strategy works. [Analysts need to] accompany that important insight with discussions of the conditions, what intelligence analysts would call indicators, that will enable [considering whether the current context is] located in the 70 percent domain or in the 30 percent domain.”

DISCUSSION

Kent Myers (ODNI) raised the question of whether decision makers would withhold action if they are given information or a prediction that would cause blowback.³ Gerliczy acknowledged that intelligence analysts strive to provide policy makers with a sense of the opportunities and potential pathways along with potential reactions. Policy makers may change their plans if there is an expectation of some type of blowback or they may choose to brace for it. Anytime analysts can help illuminate what is likely to follow from a set of actions or decisions helps policy makers. Strayer added that the DOD wants to support plans with historical evidence and understand any ramifications before it takes particular actions.

In a related question, Margaret Polski (George Mason University) asked how tradeoffs are analyzed and reported. Gaukel pointed out that tradeoffs are more a problem for the policy maker than for the analyst. The role of intelligence analysis is to alert policy makers to potential tradeoffs, at times to areas they might not have thought about. The analysts generally report what they think is happening on a given issue and why, and what they think will happen next. Gerliczy expressed that analysts lay out multiple scenarios and discuss the implications of each one. He agreed with Gaukel that the real tradeoffs are normative with long-term and short-term considerations of the risks and rewards; the IC generally does not engage in weighing tradeoffs because that is a policy maker’s purview. Strayer added that the analysts do not always understand what other matters are under consideration and what the policy maker has as tradeoffs.

Valerie Reyna (Cornell University) asked the IC representatives whether their agencies have a formal process for keeping track of when the predictions fail and when they succeed. Gerliczy said the CIA has tried to track outcomes but has been limited in its ability to do so. The vast majority of what analysts provide policy makers cannot be easily framed as a prediction that can be evaluated in terms of success. Ambiguity accompanies how things evolve, and as such, assigning scores to the outcomes has taken tremendous effort. Strayer suggested that developing models that frame steps and potential reactions could be helpful in rethinking how to evaluate analysis work. Gaukel noted that intelligence analysis is not focused on making point predictions. Rather, the important part is giving a policy maker or warfighter a sense of current and potential future events and how the other side might react. He also called attention to the notions of probability and confidence, two related but distinct concepts. The IC tries to review products to make sure analysts are keeping these two notions separate.

___________________

³ The term “blowback” is used to imply the unintended adverse consequences of a political action or situation.