- By Daniel W. Drezner
Daniel W. Drezner is professor of international politics at the Fletcher School of Law and Diplomacy at Tufts University and a senior editor at The National Interest. Prior to Fletcher, he taught at the University of Chicago and the University of Colorado at Boulder. Drezner has received fellowships from the German Marshall Fund of the United States, the Council on Foreign Relations, and Harvard University. He has previously held positions with Civic Education Project, the RAND Corporation, and the Treasury Department.
I suspect everyone inside the Beltway will be discussing the first part of the Washington Post’s "Top Secret America" series on the intelligence and homeland security apparatus that has mushroomed since the 9/11 attacks. Indeed, the 5,400 word opening salvo by Dana Priest and William Arkin doesn’t pull any punches in its lead:
The top-secret world the government created in response to the terrorist attacks of Sept. 11, 2001, has become so large, so unwieldy and so secretive that no one knows how much money it costs, how many people it employs, how many programs exist within it or exactly how many agencies do the same work.
These are some of the findings of a two-year investigation by The Washington Post that discovered what amounts to an alternative geography of the United States, a Top Secret America hidden from public view and lacking in thorough oversight. After nine years of unprecedented spending and growth, the result is that the system put in place to keep the United States safe is so massive that its effectiveness is impossible to determine.
The investigation’s other findings include:
* Some 1,271 government organizations and 1,931 private companies work on programs related to counterterrorism, homeland security and intelligence in about 10,000 locations across the United States.
* An estimated 854,000 people, nearly 1.5 times as many people as live in Washington, D.C., hold top-secret security clearances.
* In Washington and the surrounding area, 33 building complexes for top-secret intelligence work are under construction or have been built since September 2001. Together they occupy the equivalent of almost three Pentagons or 22 U.S. Capitol buildings – about 17 million square feet of space.
* Many security and intelligence agencies do the same work, creating redundancy and waste. For example, 51 federal organizations and military commands, operating in 15 U.S. cities, track the flow of money to and from terrorist networks.
* Analysts who make sense of documents and conversations obtained by foreign and domestic spying share their judgment by publishing 50,000 intelligence reports each year – a volume so large that many are routinely ignored.
Priest and Arkin are top-notch reporters and analysts, and a lot of the material in this report is pretty damning. It’s well worth the read.
I have one small quibble, however, which is with the "redundancy and waste" argument about multiple agencies doing the same work. This is a standard argument in favor of rationalization, and it’s not always wrong. It should be noted, however, that some redundancy is actually a good thing, particularly on an issue like counter-terrorism.
Say a single bureaucracy is tasked with intelligence gathering about threat X. Let’s say this bureaucracy represents the best of the best of the best — the A-Team. The A-Team does it’s job and catches 95% of the emergent threats from X. That’s still 5% that is missed.
Now say you have another independent bureaucracy with a similar remit. This agency is staffed by different people with their own set of blind spots. Let’s even stipulate that we’re talking about the B-team here, and they’ll only catch 80% of the emergent threats from X.
If thesr two bureaucracies are working independently — and this is an important if — then the odds that a threat would go unobserved by both bureaucracies is .05*.2 = .01 = 1%. So, by adding another bureaucracy, even a less competent one, the chances of an undetected threat getting through are cut from 5% to 1%. That ain’t nothing.
Now, there are a lot of assumptions that need to hold for this effect to hold. Priest and Arkin suggest that some of these assumptions don’t hold (many inteligence analysts relying on the same information). They also note the rise of segemented information, however, which leads me to think that some redundancy might be a good thing.
Admittedly, a world of 1,271 agencies tackling this question is probably one of redundancy run amok. I’m just saying that a little redundancy is a very good thing.