The Effectiveness of Electronic Surveillance

It's widely assumed that electronic intelligence gathering techniques, the likes of which are being put to use in the NSA domestic surveillance program, are effective at detecting and thwarting terrorist plots. But Eric Martin at American Footprints links to an article by Ron Suskind which makes you wonder exactly how useful such surveillance is in stopping terrorists:

Here's another lesson from London. Human intelligence routinely trumps fancy and often legally problematic surveillance techniques. The key to discovering the plot was apparently a citizen from Britain's diverse Islamic community who, in the days after last summer's bombings in London, overheard something troubling. He contacted authorities. An investigation took root.... The U.S. intelligence community is in a poor position to replicate that. Concerned citizens in the Muslim world who are close enough to radicals to see or hear something pertinent seem less inclined than ever to sit down with an American. "They see us right now as an angry, reckless giant supporting the bombing of kids in Lebanon," says a top U.S. terrorism official. "If they were to see something troubling nowadays, they'd be more inclined than ever to simply look the other way. It's their inaction -- on a vast scale -- that'll kill us."

Eric uses this excerpt to make a point about the utility of not pissing off the world's Muslims, but I think this also makes an important point about the limitations of electronic intelligence-gathering techniques, and the importance of human intelligence gathering techniques (known as "humint" by the pros). President Bush has emphasized the importance of having the authority to listen to the phone conversations of American citizens without a warrant when those conversations are with with "known" members or associates of Al Qaeda. The legality and constitutionality of that proposition is highly dubious (as discussed here, here, here and here.) But another issue entirely is it's effectiveness not on it's own standing, but as compared to other intelligence gathering techniques. We covered an article in the NY Times in January which reported that FBI officials complained repeatedly about the dead-ends they were being led to as a result of the surveillance program, and this article from the June 30th edition of the Christian Science Monitor includes this quote:

"Nothing that's been revealed in the last few months has had any substantive effect on the war on terror," says Vincent Cannistraro, chief of Central Intelligence Agency (CIA) counterterrorism operations and analysis from 1988 to 1991. "Hardened, organized groups have been aware or assumed that voice, e-mail communications, and bank transfers are monitored at least since 9/11."

Canistraro was referring to the "damage" caused by the leaks to the media of the existence of the NSA surveillance and like programs, but the implication is that if terrorists already assumed that we were listening to their calls into the U.S. (notwithstanding whatever Al Qaeda's legal advisors thought of the subverted warrant requirement) then exactly how many terrorists have we managed to listen to?

Such logic, and the articles above, support the proposition that the NSA domestic surveillance program has not been very effective. But the thwarting of the terrorist plot in Britain is highly suggestive of the fact that old-style intelligence gathered from real human beings by real human beings is very effective at producing leads on terrorist activity. So where do we stand on the ability of the United States to gather intelligence on the activities of terrorists abroad and at home? Here's an example of someone who thinks we're doing a much better job:

We've vastly improved human intelligence. Shockingly, in 2001 way more than half of CIA case officers weren't stationed overseas spying on bad guys, but were manning desks at CIA HQ instead. Today, the ratio has been reversed. The CIA doubled the spy force against terrorism, opening more stations/bases in potential intelligence blind spots, especially in the developing world.

But several others say we haven't learned our lesson.

Time Magazine:

The CIA is larded with Russian specialists left over from the cold war, even as the agency struggles to recruit and train officers with proficiency in other tongues. In last year's graduating class of case officers, just 20% had usable skills in non-Romance languages. When the war in Afghanistan began, the CIA had only one Afghan analyst. As Time reported last month, American intelligence agents in Kabul almost blew the chance to question a top-ranking Taliban minister, who may have had information on the hiding place of Mullah Omar. The spooks had yet to hire a Dari translator.

Sen. Saxby Chambliss:

During the national debate on intelligence reform [in 2004], there was general acknowledgment that HUMINT needed to be improved; however, it was not afforded the primacy in the legislation that I believe it deserved. In fact, HUMINT is not mentioned even once in the 26-page summary of the Intelligence Reform and Terrorism Prevention Act of 2004 prepared by the Congressional Research Service. The reason al Qaeda was able to attack us was because we didn’t have spies to infiltrate their organization. It had nothing to do with intelligence budget execution or the reprogramming of funds.

Michael Scheuer:

We should concentrate on what we do best, which is human intelligence.

...The DNI] has bloated the intelligence system as far as I can tell. It’s just another layer of control, editing, review. And it’s another step in preventing the truth as the intelligence community sees it from getting to the president. It’s a tremendously unfortunate situation.

...I don’t think we’re lacking in resources, at least on the terrorism issue. Among the cheapest operations to run are HUMINT operations. We were never short of money to conduct counterterrorism operations. What we were short of was leadership...Porter Goss really never gave himself a chance to prove his capabilities because he surrounded himself with people who were both untalented and obnoxious.

And while I had to do some serious searching to find someone who said we've made significant improvements in our human intelligence-collecting capabilities, it took far less time to find those who do not think so.

I think the argument can be made that the Bush administration has quietly made important improvements in our non-electronic intelligence capabilities, but at the same time it has invested much in the NSA program, a program whose helpfulness is at best questionable, whose legality is highly in doubt, and the revelation of which has considerably divided public opinion. And there's also no question that human intelligence is only as useful as the means by which you gather and vet it, and the use of illegal means to gather it (see Abu Ghraib and the secret detention program) and the failure to examine it carefully (see the Iraq war) can lead to tremendous policy mistakes, regardless of how much intelligence you gather. Additionally, extraordinarily mis-placed priorities can harm our intelligence-gathering efforts, such as the strict enforcement of "Don't Ask, Don't Tell" even when it results in the discharging of military translators profficient in crucial languages.

So where do we stand in regards to intelligence on terrorist activities? I think you could say without question that we still have a long way to go.