Connecting the Dots beyond Counterterrorism and Seeing Past Organizational Failure

The Critique of Security and the Prose of Pacification

On September 9, 2001, a Maryland state trooper stopped Ziad Jarrah for a traffic violation. Having no reason to do otherwise, the trooper sent Jarrah on his way. Two days later, he was the hijacker-pilot of Flight 93, which crashed in rural Pennsylvania. Citing incidents like this stop, the 9/11 Commission found that the intelligence community had failed to "connect the dots." In response, the new Department of Homeland Security (DHS) promised to link the entirety of domestic intelligence — from municipal police departments to the federal intelligence community — with a new National Network of Fusion Centers. At the same time, the spread of intelligence-led policing (ILP) complemented the rise of these networked intelligence hubs with a "smarter" approach that uses intelligence to preempt threats.

The resultant process of intelligence fusion starts with information: massive databases decades in the making; open-source data gleaned from the web and social media; and streams of information created by new surveillance systems like automated license plate readers. To "fuse" data into useful information, analysts often use powerful computers and specialized software to "connect the dots" and, in theory, draw out the signal from the noise of data. Different tools offer different insights. With specialized software, analysts can turn unintelligible and interminably long lists of phone calls into a pattern of use, and, from there, a social network analysis. They can map unwieldy agglomerations of information — such as geospatial data drawn from police files, the census, and other public records — to create "predictive" heat maps to anticipate where the next shooting is likely to occur. Intelligence fusion also "connects the dots" in simpler ways. Often, fusion centers operate as data brokers, providing investigative support to law enforcement partners. Data brokerage can also mean doing even less: in so-called "pass-throughs," fusion centers simply disseminate another agency's intelligence reports.

For the most ardent proponents, the intelligence fusion never really ends. Traditionally, there is an iterative tendency built into intelligence: a cycle in which decision-makers demand information, officers collect information, and analysts process the data into intelligence. The feedback executives provide helps orient the next turn of the intelligence cycle. Intelligence-led policing tries to transcend this reactive model with a proactive approach. It collapses intelligence collection and analysis into a conjoined and continuous activity. Intelligence producers strive to maintain the situational awareness necessary to preempt and disrupt behaviors deemed criminal and disorderly. Hence, their ever-creeping reach: first, government records and private data brokers; next, the integration of old surveillance systems like closed-circuit TV cameras and new ones like automated license plate readers; and, most recently, wholesale data-mining of social media and other forms of open-source intelligence.

Intelligence-led policing is also an administrative philosophy. The goal is efficiency. About three decades ago, police executives started using crime mapping to manage police departments. With crime hot spots identified, they knew where to direct patrols and investigators and which middle manager to hold accountable. Today, ILP strives for proactive crime control with increasingly individual targeting, a shift from "hot spots" to "hot people." Under this pressure, officers target "chronic offenders." Detectives try to refine leads out of data. Analysts work to stay ahead of events and otherwise divine the future. They collaborate with police on long-term investigations, providing a variety of services from simple database searches, to routine crime analysis and mapping, to in-depth criminal profiles and social network analyses. In theory, intelligence fusion and ILP will produce more proficient policing.

One decade and upwards of a billion dollars later, the results are unimpressive. In October 2012, the US Senate excoriated fusion centers. After two years of investigation, they could not identify any "reporting which uncovered a terrorist threat ... [or any] contribution such fusion center reporting made to disrupt an active terrorist plot." The report brought uncomfortable national attention to fusion centers. "DHS 'fusion centers' portrayed as pools of ineptitude and civil liberties intrusions" read the Washington Post's headline. The New York Times had a more subdued title but opened with an equally damaging assessment: "One of the nation's biggest domestic counterterrorism programs has failed to provide virtually any useful intelligence." "It's brutal," one federal official involved in funding and management of fusion centers later told me. "It's one-sided. Definitely. But it's not totally wrong. We have some problems to work out." Seven months later, the Senate's findings were confirmed in spectacular fashion by the Boston Marathon bombing. In the preceding two years, the FBI and CIA had neglected to share information about Tamerlan Tsarnaev, the elder brother implicated in the attack, with the Boston Regional Intelligence Center. Even if they had, Boston's fusion center was preoccupied with other matters: spying on Occupy Boston.

The wider conversation on fusion centers reflects the major themes of the Senate report: dysfunction, mission failure, and abuse. From the very start in 2004, when DHS began encouraging state and local governments to create fusion centers, journalists criticized the new program for its ineffectiveness, the potential for mission creep, and civil liberties violations. By 2008, government researchers and auditors identified the factors contributing to these problems: an ill-defined, vague mission, poor coordination, over-classification, and incompatible information systems. Policy advocates repeated many of these concerns and recommended reform. Liberal organizations like the American Civil Liberties Union (ACLU) and the Brennan Center for Justice focused on protecting civil liberties and recommended greater oversight, while conservative groups like the American Enterprise Institute argued that centralization could reduce costs and increase information sharing. Only practitioners in law enforcement and a later report from the House of Representatives found that fusion centers were an effective and worthwhile addition to law enforcement. Some journalists, activists and civil liberties groups inverted this stance, charging that fusion centers are effective, not at counterterrorism, but at suppressing dissent.

Despite all the criticism and bad press, neither...