David R. Henderson  

Admiral Mullen vs. Professor Hayek

PRINT
Greens, The Road, and <... The Phillips Curve as a Straig...

I mentioned briefly in a comment on my own blog post a few weeks ago that there is a Hayekian approach to protecting ourselves from terrorists. I elaborated in an article yesterday that I'll quote from here:

Friedrich Hayek's insights on knowledge and information can be used to fight terrorists. The evidence that Hayek is right often stares us in the face, and many politicians - I include the U.S. government's top-ranking admiral as a politician - ignore this evidence. Worse, on Fareed Zakaria's latest GPS (Global Public Square) program, Adm. Mike Mullen, chairman of the Joint Chiefs of Staff, seriously misstated the facts to make it sound as if centralized intelligence works well when what really worked well was, à la Hayek, decentralized "local knowledge." Worse yet, Zakaria let Mullen get away with his spinning, specifically regarding the Times Square bomber and the Christmas "underpants bomber."

Consider terrorism. Notice that in the most celebrated cases in which terrorists were thwarted, people used their "local knowledge." (I'll leave aside cases like the recent Portland, Oregon bomber. Here, the FBI used its local knowledge also, but it's not a clear example because the FBI encouraged the bomber and even provided him with the "bomb.") United Flight 93 on that horrible September 11; Richard Reid, the shoe bomber; the Christmas "underpants bomber"; and the Times Square bomber are all instances of successful anti-terrorist acts by what Hayek would call "the man on the spot." Of course, as the relatives of the victims on Flight 93 can attest, the passengers did not save their own lives. But they likely saved the lives of the terrorists' intended targets in Washington, D.C. Indeed, it's because local knowledge has worked so well that I've long believed that the TSA was, and is, not required.

The article is titled "Adm. Mullen's Spinning vs. Prof. Hayek's Insight." It led to a civil e-mail discussion with an economist on Admiral Mullen's staff.


Comments and Sharing





COMMENTS (6 to date)
Daniel Kuehn writes:

As I've said in the past, I agree strongly with the value of decentralized knowledge for these types of things (you can add the foiled Oregon terrorist to this list - apparently his dad tipped off the policy), but I think you come down too hard on Mullen by assuming he's talking only about centralized knowledge. My wife was looking into an intelligence analyst position at one point, and her understanding was that a lot of the job isn't to take a centralized approach - but precisely to cultivate, sort through, and leverage the sort of decentralized information you describe. Presumably, in the case of terrorism, we want to rely on decentralized information but have at least to a certain extent a centralized response (I think it's more appropriate for the FBI to play a major role in thwarting an attack, rather than just local police departments, for example). This seems to be essentially the mission of the intelligence community. I wouldn't be so quick to dismiss their dependence on decentralized information - I think intelligence analysts would be the first to tell you how important decentralized information is, and I don't see anything in Mullen's quote that denies that.

On TSA - and I've also said this before on here - I really think you're drawing your inferences from a highly selected sample of the attacks that have gotten through TSA in the first place. Of course you don't observe attacks that don't get through TSA, because they never materialize into attempted attacks! You don't observe two important groups of terrorist attacks: (1.) those who are caught by the TSA through security checks, and (2.) those who never even attempt an attack because they don't think they can get past the TSA. Yes, of course decentralized information is extremely important in national security, but I think you're being fooled by the fact that the attacks foiled by decentralized information are just the most visible foiled attacks. To use Bastiat, you're missing the "unseen" and focusing on the "seen". Although the successes from decentralized information are the most visible, it does not follow that we would be fine without the TSA.

Of course, none of that is to necessarily suggest that we need all the more draconian policies the TSA has implemented recently.

david writes:

"If you ask amateurs to act as front-line security personnel, you shouldn't be surprised when you get amateur security." - Bruce Schneier, on asking the public to identify threats.

People have been booted off planes for all sorts of silly reasons already, by people wielding Hayekian local knowledge... and by people who don't like people who wear turbans. And, yes, this includes random passengers grabbing other random passengers in a headlock and demanding they be kicked off the plane. It's a very Hayekian police state, I'm sure.

You won't hear about the people blocked from flying for looking vaguely Muslim - by the fears of their fellow passengers and nothing else - on Fox News. It is one thing to resist someone who has already hijacked your plane. It is quite another to start punching someone because you think he might be a terrorist. You cannot cherrypick the incidents when the punched individual really does turn out to have a bomb in his underpants and declare that scheme a success.

You are surely demanding an excessively high level of security when you not only ask for the public to watch each other, but to attack each other on personal suspicions! The TSA is demanding an excessively high level of security, too, wrt the cost tradeoff, but the problem is not the TSA - the problem is the excessive demand!

fundamentalist writes:

Excellent points, Dr. Henderson! And to go farther, police can't prevent crimes; they can only arrest the criminals. People prevent crimes, if they are prevented at all. Police could actually be making crime worse in that they lull people to sleep in a false sense of confidence that the police can protect them. As a result they don't prepare themselves to deal with criminals and it makes criminal work more fruitful and easier. And in the case of the panties bomber and the 9/11 attackers, the state authorities clearly failed to do their jobs.

Bob Murphy writes:

"Civil" like Civil War?

David R. Henderson writes:

"Civil" as in "polite and respectful"--on both sides.

Jacob Oost writes:

I'd like to add another Hayekian point: that the centralized type of knowledge and "planning" (or decision-making) done by the military brass--the kind discussed in this article and the interview--actually is very useful for those things which central planners can actually plan: a task with a clearly definable and attainable goal, like winning a war. Central planners were able to fight WW2 because they had a goal: defeat the Axis Powers, and a means: militarily. They knew what to do, and they knew when they'd be done. They knew they had to invade *these* countries and defeat *these* armies and depose *these* governments and WW2 would be over.

This is an extremely important point made by Hayek that most libertarians really ought to commit to memory, so they have a real retort when somebody asks "how could war planning work so well, if central planning doesn't work?"

This differs from the War on Terror because it has no defined end, no defined enemy (ideological movements and guerilla fighters versus countries on a map with their own borders and armies), etc. You don't know when you're done, or how close you are to being done. Central planners can't plan that.

Which makes it like economic central planning. Economic growth is all about maximizing prosperity, but only individuals can really say whether they are prosperous or not.

Iraq was successfully invaded, its armies defeated, and its government deposed very rapidly. And centralized intelligence played a big role there. So this type of intelligence isn't worthless, it's just not helpful for tasks which can't be centrally planned.

Comments for this entry have been closed
Return to top