23rd July 2010: Drezner and Kaplan on Top Secret America

Fred Kaplan and Dan Drezner are two very savvy commentators on intelligence and national security issues.

So it’s intriguing to compare their divergent¬† responses to The Washington Post‘s Top Secret America series which argues the United States national intelligence system has grown into a vast institutional bureaucracy. Has “too big to manage” or “too big to lead” become the intelligence community’s problem comparable to “too big to fail” banks and financial institutions after the 2007-09 global financial crisis?

Drezner suggests intelligence analysts in different agencies may benefit from “redundancies” that enable them collectively to piece together a puzzle. Kaplan counters that the rapid proliferation of government agencies, companies and private contractors means that few people have the expertise or ‘span of control’ to integrate data from compartmentalised systems into “actionable intelligence.” This is despite collaborative data-sharing initiatives like Intellipedia; the Open Source Intelligence community — named for publicly accessible data and not for the software movement; and the use of other vehicles like the venture capital firm In-Q-Tel, which makes ‘last stage’ investments in promising technologies.

Several things are going on in the Drezner-Kaplan debate that academic researchers in the Intelligence Studies subfield of Political Science have discussed for over 15 years. Enterprise Resource Planning vendors such as SAP AG and Tibco have mapped the defence and intelligence supply chain in multi-level ways beyond Sherman Kent‘s original intelligence cycle for policymakers. However, this emphasis on vendor systems also reflects an institutional bias towards signals intelligence (SIGINT), imagery intelligence (IMINT), and measurement and signals intelligence (MASINT) over human intelligence (HUMINT). Many of Kaplan’s integration problems might be solvable through procurement, supply chain integration and service-oriented architecture policies between government intelligence agencies and firms. The HUMINT problems however are not: which intelligence problems or key questions are prioritised, who interprets the data, by what methods, and how ‘responsible’ decision-makers use the conclusions. Finally, RAND’s Gregory Treverton notes these intelligence problems may be solvable problems, Bayesian mysteries or complex-plus dilemmas that pose a variety of challenges to ERP vendors and the national intelligence system.

Drezner, Kaplan and others may have overlooked one final problem that analysts and ‘responsible’ decision-makers must already grapple with: vapourware. Science journalist Sharon Weinberger highlighted vapourware in her book Imaginary Weapons: A Journey Through the Scientific Underworld (Nation Books, New York, 2007) about the research and development (R&D) games around fringe projects and experimental weapons systems. Post-doctoral student Doina Collins later created a ‘rejoinder’ site , to defend the scientists and researchers who Weinberger ‘named and shamed’.

The Washington Post‘s Top Secret America series contends the public faces a new intelligence bureaucracy. Yet the ‘responsible’ decision-makers, especially those charged with budget and resource allocation, will continue to face further problems: ‘spin’ for defence R&D dollars and vapourware projects that fail to deliver on their hype cycle promises.