The classified memorandum written by an F.B.I. agent in Phoenix last summer urging bureau headquarters to investigate Middle Eastern men enrolled in American flight schools also cited Osama bin Laden by name and suggested that his followers could use the schools to train for terror operations, government officials said for the first time today.There's no question there were security holes: the INS demonstrated its incompetency beyond a reasonable doubt. But before one uses this memo to condemn the FBI for failing to heed its own agent's warnings, let's ask:
The memorandum said terrorist groups like Mr. bin Laden's might be sending students to the schools as the first step in what could be a concerted effort to place Islamic militants in the civil aviation industry around the world as pilots, security guards or aircraft-maintenance workers.
1. On what level of actual intelligence and what level of actual speculation did the memo come from, and from how high-ranking an agent?It's very easy to have 20/20 hindsight and complain that a particular memo wasn't followed. When I litigated automobile products liability lawsuits, there was almost always one feature that, in hindsight, could have prevented grievous injury, and plaintiffs wanted to assess blame that the car didn't have that particular safety feature (even though the feature might have caused harm in other ways). But one has to evaluate these matters ex ante: we can't turn back the clock, so the question becomes, going forward, what's feasible? Or, if you're allocating blame, in the summer of 2001, what was the feasible prevention efforts that could have been anticipated? What probability does one assign to the actual September 11 scenario (close to unthinkable outside of Tom Clancy novels), and how many other scenarios had equal or greater probability x magnitude of potential damage figures? And what level of resources would it have taken to address all these scenarios, and would Congress have allocated those billions of dollars to do so?
2. How many memos are produced with similar or lower levels of speculation by similarly-ranked or higher-ranked agents?
3. What level of resources would it take to respond to each of those memos with the effectiveness to shut down a speculated operation in a couple of months?
One of my favorite trial exhibits was a response to a plaintiff's complaint that, in addition to the warning in the owners' manual, the manufacturer should have put a warning light or buzzer or sticker on the dashboard, and that might have prevented the horrible accident that was at issue in the lawsuit. We mocked up a dashboard that had a cacophony of warning lights and buzzers and stickers for every warning in the owner's manual. If you try to protect against every little risk, you end up protecting against none of them at all. Not exactly analogous, but the point remains: you can't selectively pick the memo that in hindsight was most prescient and complain that everyone else wasn't equally prescient.
The cheapest way to stop terrorist attacks is to (1) pre-emptively take out the terrorist structure that produces the attacks; and (2) make sure the INS (or replacement agency) is doing its job.