Fat Steve's Blatherings

Saturday, November 26, 2005

Some Thoughts on the CIA and its Persistent Failures of Analysis, Part I


        This is the first of a series of posts on why the CIA in particular, and the USAmerican Intelligence Community in general, so often get things wrong.

  • The CIA's first big mistake is an overemphasis on prediction of what foreign nations will do.

At Length:

        With all the hoorah about the question of Iraq/al-Qaeda ties, pre-War Intelligence, the Curveball case, and whether Bush distorted intelligence, and my ongoing attempt to do a history of Plamegate, I've been thinking about the Central Intelligence Agency a bit.  The history of the CIA is one of screwing up royally by failing to anticipate enemy actions.  Here are some thoughts on why that happens.

        In the first place, anticipating what other countries will do is arguably a low priority for an intelligence service.  States are not people, but in the end, there are people in charge of a state's actions.  And people have minds to think with, and free will to decide with.  Further, there's the 'dialectical' aspect of competitive games: what player A decides to do is likely to be influenced by what A thinks his opponent, B, will do.  A case in point is the Iraqi invasion of Kuwait.  The CIA and the rest of the intelligence community decided that Saddam would not invade.  So, no one told Bush 41 "You'd damn well better think about what we do if he annexes Kuwait, sir," and nobody told Saddam what that action would be.  Which may have led to a grave miscalculation on Saddam's part, for, if James Dunnigan and Austin Bay are correct, Saddam decided to take over Kuwait on the spur of the moment, thinking the U.S. would let him get away with it.

        The first thing an intelligence service should do is lay out what it thinks another nation is capable of doing.  The second thing it should do is say "We aren't mind readers here, and even if we were, people can change their minds."  Intelligence should then lay out the what it thinks foreign states might do, with the rough probability figures of the various plans.

        Another reason it's hard to predict what other people will do is the difficulty of realizing that other people are really OTHER, they're different from you, they have different thought processes and different goals.

        For instance, Miles Copeland, in his book
  The Game of Nations, says that he didn't think Nasser intended to attack Israel in 1967 (Copeland was one of the State Department's Egypt experts.)  Instead, he thought Nasser was "conducting an election."  He'd stir Egyptians up about Israel, then magnaminiously back down at the last second.  If Copeland is wrong, it shows his, Copeland's, problem in getting into other people's minds.  If Copeland is correct, then Nasser erred in not asking himself 'What will the Jews think of my actions?  Will they attack?"

        Another Mid-East example is the Yom Kippur War of 1973.  Israel had a "concept," of what Egypt and Syria would do, and the "concept" stated that Egypt would not attack Israel until certain conditions were met.  The reason they decided that is, apparently, that the "concept" fit with what the Israeli military would have done if they were running Egypt.  The "concept" also stated that the Syrians wouldn't attack unless Egypt attacked at the same time.  (For more on the "concept," see Military Misfortunes by Cohen and Gooch, or The Sword and the Olive by Martin van Creveld.)  The result was that when signs of an impending Syrian attack accumulated, they were dismissed because obviously Syrian would only attack when Egypt did.  When signs of an impending Egyptian attack accumulated, they were dismissed because the "concept" stated that Egypt wasn't in a condition to attack Israel yet.  But Anwar Sadat's conception of when it would be worthwhile to attack Israel was different from Israel's "concept," and the Jewish state was caught flat-footed, in danger of being destroyed.

        I had intended to list all the CIA's errors as I saw them, but this is getting too long, so I'll do one post per major error.

        Technorati tags: , , , , ,, , , , , , , , , , , , , , , , , .



Post a Comment

<< Home