Give the Director of National Intelligence control over the purse strings! That's the battle cry of critics of the way the U.S. intelligence system is structured. In Washington, authority without budget power is like a computer without RAM. But the DNI's problems, certainly exacerbated by his inability to threaten to cut funding from agencies if they don't comply with his requests, are much more systemic. The truth is -- and this is what scares some counterterrorism and intelligence experts -- is that the magnitude of the intelligence collection tasks that the U.S. faces will inevitably outpace the bureaucratic procedures needed to properly transform that information into useful policy advice.
Testifying before Congress today, Lee Hamilton, vice chairman of the 9/11 commission, put it plainly: another round of intelligence reform wasn't politically or practically feasible. "The DNI's ability to lead depends on the president's leadership," he said, in what I took to be a comment directed at the White House. That is -- when the DNI decides something about policy, the White House should back him up -- with almost no exceptions. This White House hasn't exactly done that; they've instead played mediator between the CIA and the ODNI, essentially setting them up as equals.
The Chief Technology Officer for the DNI has had near-complete budget authority over the entire intelligence community IT infrastructure, but successive CTOs have been hesitant about using it without sufficient cover and buy-in from the major agencies, all of whom can come with persuasive reasons why they need to keep their computer systems and data integration procedures just like they are. I asked a consultant to the government why, when it was so easy for companies like Catalist to collect enormous amounts of data from disparate sources quickly, create architecture to merge the different data types, and extract meaningful and relevant patterns allowing them to adroitly figure out how to best to appeal to individual voters. The answer is that the amount of data that comes in -- about one billion bits per day -- is enormous. The other is that the data isn't standardized, and the ODNI has no long-term plan in place to standardize intelligence information. Indeed, it might be difficult to force signals intelligence databases to talk to databases that contain collected tips from human sources. Human intervention is necessary to figure out which category of tips and sources ought to be prioritized. Indeed, the real reason -- or, rather, the major reason why the national security machine didn't kick into high gear before Christmas was because human beings misjudged the intelligence -- not that the intelligence didn't present itself or wasn't collected.
That said, the US can do better with the foreign intelligence they gather. It can take years in the IC to accomplish what a private company could do in a month. (Just ask anyone at the Department of Defense about standardized e-mail log-ins.) The time it takes to adopt new technology and spread it around -- tech that would make analysts' lives easier -- is too long. That's where a strong DNI who can bust heads when necessary comes in.