The Limits of Signals Intelligence

The outbreak of violence in Iraq is a good reason to think carefully about the NSA, not an excuse to ignore critics' warnings.

A former NSA base in Germany (Reuters)

When scary geopolitical events happen, Americans are particularly likely to speak favorably about the national-security state. This is understandable: The national-security state encompasses a lot of hardworking men and women engaged in legitimate efforts to protect us, sometimes by putting their lives at risk. But it is doubly important to think critically in moments of crisis, for history shows such moments are exactly when abuses, excesses and missteps are enabled.

This permits the dark side of national security to get out of control, as it has so many times in American history, including the period after the attacks of 9/11.

As a new crisis unfolds in Iraq, Andrew Sullivan is mostly heeding this wisdom: He is skeptical about American intervention because he sees that national-security officials are not capable of doing much of anything with predictable results. But he did offer one aside about the role the NSA might play that I want to flag:

If there is something we can do, it should be to ratchet up our ability to monitor these groups—sorry, NSA-haters, but spying is one of our strongest and least disruptive tools in preventing attacks on the homeland—and to provide as much diplomatic and political advice, if asked, as to how to render the situation less volatile. 

This gets a few things wrong that matter very much to the ongoing debate about the NSA. First off, note that not even the staunchest critics of the NSA, from Edward Snowden to Glenn Greenwald to Senator Ron Wyden to Representative Justin Amash, want to eliminate the agency or prevent it from spying on foreign terrorists or soldiers. Without presuming to speak for any individual, the typical "NSA-hater" would love nothing more than for the NSA to focus its intelligence capabilities on war zones where anti-American fighters plausibly threaten the lives of soldiers or diplomatic personnel, and away from Angela Merkel and every cell-phone call Americans make. Spying on ISIS, however intrusively, is fine by me.

That said, events in Iraq seem to have taken us by surprise, despite the fact that the NSA is totally unencumbered, both legally and politically, in the intelligence it can gather there. And even if the seeming surprise is an illusion, even if the NSA anticipated the fall of cities to Islamic militants, knowing didn't stop it. That isn't a knock on the NSA. It's a statement about the limits of signals intelligence. The NSA didn't stop the underwear bomber or the Times Square bomber or the shoe bomber either. That's not a knock on the NSA. They can't know everything. And if they could, that would be a lot more dangerous than terrorism.

Sullivan asserts that "spying is one of our strongest and least disruptive tools in preventing attacks on the homeland." Really? How does he know that? Maybe he's right. I cannot disprove the statement. But can he prove it? Or is he making a huge assumption without any evidence to justify it? I'd like to contest that general statement, pending further evidence. What we do know is that the specific programs that "NSA-haters" complain about most have no track record stopping terrorism.

Take the phone dragnet:

An analysis of 225 terrorism cases inside the United States since the Sept. 11, 2001, attacks has concluded that the bulk collection of phone records by the National Security Agency “has had no discernible impact on preventing acts of terrorism.”

In the majority of cases, traditional law enforcement and investigative methods provided the tip or evidence to initiate the case, according to the study by the New America Foundation, a Washington-based nonprofit group.

Or to move away from the several NSA programs that blatantly violate the Fourth Amendment rights of Americans—how many terrorist attacks do you think have been stopped by spying on the content of virtually every phone call made in the Bahamas?

The reforms that "NSA-haters" advocate are not at all inconsistent with a signals-intelligence agency that uses the most intrusive methods imaginable to target the communications of ISIS fighters. Unfortunately, I think Sullivan overestimates how much the United States will gain by spying on fighters in Iraq, and how important the NSA is to stopping terrorist attacks within the United States. But if there's reason to think a foreigner is plotting with ISIS, or even that an American is doing so, nothing the "NSA-haters" are advocating would stop the national-security state from lawfully investigating those individuals.