American Spies Not Clairvoyant. Duh.

An interesting article in today’s LA Times focuses on the limits of US intelligence in the Middle East. Top American spies, hauled in front of Congress to report on their performance amid the spreading turmoil, said tha, despite access to huge amounts of information, they were still unable to predict the sudden waves of protest:

Over the last year, the CIA wrote more than 450 reports that discussed “dangerous” factors in the region, such as repressive regimes, economic stagnation and lack of freedoms. Since mid-December, the U.S. intelligence community has produced 15,000 reports from the Middle East and North Africa that followed what was being talked about in local media and on the Internet.

But the high volume of reports could not predict what would trigger the mass protests in Tunisia — a fruit vendor lighting himself on fire — or that President Zine el Abidine ben Aliwould flee the country so suddenly on Jan. 14, [Director of National Intelligence James] Clapper said.

Duh.

It is not really possible to predict events like this, even if you understand perfectly the various factors that lead to them. It’s not all that different from earthquakes. We know a lot about the geophysics of earthquakes, and we can pinpoint places where the risk of earthquakes is particularly high. But we can’t say exactly when they will occur.

In either case–political turmoil and earthquakes–this does not mean we need to give up trying to understand how things work. But it should inform the way that we approach that task. If we believe that we can build a deterministic model that will tell the exact conditions and timing of a successful uprising in, say, Iran, then we’re in danger of wasting a lot of time and resources trying to squeeze relevance out of every little bit of information we can find. The LA Times article suggests that CIA Director Leon Panetta may be headed in that direction:

To better predict what events might trigger such uprisings, the CIA has assembled a 35-member task force to analyze trends on social media websites and events as they develop, CIA Director Leon E. Panetta told the committee. But there’s a “massive amount of data,” Panetta said, noting that there are 600 million Facebook accounts and 190 million Twitter accounts and that 35,000 hours of video are uploaded toYouTube every day.

Trying to use all that data to predict events that are essentially un-predictable is not just a waste of time, it’s potentially dangerous, as another witness testified:

Having to commit more energy and analysts to understanding the political instability in the region has the “potential to take our eye off the ball with regards to the jihadi terrorists themselves,” said Frank Cilluffo, director of the Homeland Security Policy Institute at George Washington University.

We can do a lot better by focusing efforts on understanding the causes of these events, and what we can do about them when they happen, as they inevitably, but unpredictably will.