Cover Image


View/Hide Left Panel

understanding will help attenuate communication problems (Fischhoff, this volume, Chapter 10), such as agents talking past each other or dismissing each other’s analysis.

Unfortunately, cross-agency collaboration can be thwarted by the familiar pathologies found in all intergroup endeavors. These include misaligned organizational incentives, ossified bureaucratic policies, and conflicting political pressures. For example, agencies within the IC are “mission specific” (Sims, 2005), meaning they serve different customers who may be asking different analytic questions about the same target or event. Each agency in the IC naturally has “its own terminology, routines, and expectations about what people can ask for” (Simon, 2005, p. 150). Moreover, interagency “turf wars,” which appear to arise from historical inertia and budgetary conflicts (Sims and Gerber, 2005), provide inherent challenges to interagency collaboration.

Discussing any particular bureaucratic, political, and budgetary conflicts is beyond the scope of this chapter, which instead focuses on how the pluralistic structure of the IC can engender cognitive and motivational biases that hinder effective cross-agency collaboration. Special attention is paid to the agency grouping and how social categorization processes can influence individual behavior. Social categorization is the process by which we make judgments about individuals (and who they are in relation to ourselves) based on their group membership. When we encounter other people, we tend to encode not only individual information about that person, but also social information—in particular, whether he or she is a member of one of our “in-groups”—the groups to which we belong—or is an “out-group” member, or outsider. In the context of the IC, where a salient group boundary is agency (or subagency) affiliation, social categorization suggests that analysts register whether other analysts are members of the same or different IC agency.1 This encoding can produce mental roadblocks to coordination by eliciting well-documented intergroup dynamics, including differentiation, ethnocentrism, and integration neglect (explained below). These roadblocks are not unique to the IC, but rather are fundamental


This agency grouping does not mean that other social groups are irrelevant. Indeed, within some agencies there are large distinctions between analysts and collectors (e.g., the CIA). Groups are people’s subjective representations of others vis-à-vis the self. These groups may exist because people interact regularly and share a common fate (Campbell, 1958), such as with workgroups, or groups may be large collectives where individuals are connected by virtue of common symbolic attachments (Geertz, 1973). The motivational and cognitive biases discussed in this chapter would pertain to any social group distinction that is relevant for the actors involved. For simplicity of presentation, however, I choose the agency level as the focal social group because this follows both the formal organizational structure and other work describing agency affiliation as a dominant social category in the IC (see Sims and Gerber, 2005, or Fingar, this volume, Chapter 1).

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement