Last week, my Brookings colleague Daniel Byman and I released our fun little Disposition Matrix App–on The Atlantic‘s web site. The rather surprising response—who knew that so many Americans were anticipating catching citizen terrorists and needing to figure out how to process them?—has made me start toying with the idea of expanding the flowchart beyond citizens. Dan and I limited the flowchart to citizens because we’ve been writing a paper about the problem of U.S. citizens’ going abroad to fight against their country. The flowchart started as a napkin drawing of what I called a “rough decision tree” our work on the had suggested. Dan found it compelling and suggested we elaborate and publish it as a piece of op-art. Because the paper was limited to citizens, however, we never considered trying to reverse-engineer the larger disposition matrix, which included much material beyond the scope of our inquiry. Now I’m thinking that maybe we should take a stab at it.
I’m interested in reader input on what this larger decision tree looks like—particularly from readers who have active involvement in different corners of that larger tree. I’m, of course, not looking for anything classified. But I am interested in tracing the aggregate cross-agency process that goes from suspect/target identification through dispositions of various types. In other words suppose someone in the U.S. government plausibly identifies Person X as a terrorist suspect. How precisely or generally—based on the public record—can we identify the chain of decisions that would lead from that identification to any of a bewildering array of possible outcomes: targeting, prosecution, deportation, extradition, rendition, etc.?
Several aspects of this question strike me as very tricky. First, the decision must differ a lot depending on which government component does the initial identification. The unitary executive being rather plural in practice, that is, it matters a lot in the way a case proceeds if, say, the military identifies a potential target or, by contrast, if the FBI identifies a potential suspect. This may be less true for major league suspects—the UBLs and Anwar Al-Aulaqis of the world—who necessarily trigger an inter-agency and presidential level response. But surely for the garden-variety suspect, the decision tree varies greatly depending on which agency does the identification and takes the lead in seeking the disposition. Second, the Disposition Matrix necessarily blends a variety of different considerations—ranging from legal questions (what may we do?) to questions of strategic judgment (how much do we care about this guy?). One can cull some of the higher-altitude legal and policy judgments that make up the tree from the administration’s national security law speeches. And folks like Greg McNeal have done amazing work fleshing out targeting standards. But I suspect there are unclassified materials that shed a lot of light on granular aspects of the tree. What are they? Moreover, the integration of these legal and policy rules with bureaucratic structures across government and with certain ineffable strategic question into a coherent decision tree strikes me as very difficult to trace. I’d welcome reader input on any components of what an expanded Disposition Matrix App might include—and pointers to any unclassified source material that might plausibly inform it.