Nota bene, New Republic: the whole “counterintuitive” shtick has its merits, but it’s sort of cheating when the argument is counterintuitive because obviously wrong. The trick is to run pieces with premises as superficially implausible as Ben Wittes’ lukewarm endorsement of the Protect America Act, but which actually turn out to be at least somewhat compelling once you get into them.
Wittes’ central point is that, after all, foreign intelligence agencies could in the past run surveillance operations overseas, with no court order required if they happened to pick up a target’s communications with an American—post hoc retention rules covered that. This is either disingenuous or obtuse. Citizen privacy is a multivariate function: A product not only of statutory limits on surveillance, but of physical and technological constraints as well. That’s why a property-line rule for defining Fourth Amendment searches worked in 1867, but not in 1967. So yes, it was possible for intel operations abroad to pick up some incidental American’s communications, but it does not follow that nothing of interest has changed if the law now licenses the establishment of a massive domestic wiretapping architecture capable of sweeping up (for instance) every international call by an American believed to be in contact with an intelligence target believed to be abroad. (This is assuming DoJ lawyers decide to interpret “targeted” in a way that restricts surveillance to international calls.) The same changes in the telecom architecture that lead to strictly foreign-to-foreign calls being routed through U.S. switches also facilitate far broader dragnet searches. It’s true that if I’m corresponding with a suspect, an authorized search of his home may turn up the letter I sent him. But it would be an especially ridiculous species of sophistry to claim that my privacy is no more compromised when police start rifling through the outgoing letters in my mailbox, so long as their target is still my correspondent.
Moreover, as Wittes seems bizarrely prepared to assume that crucial terms of the law will be interpreted in in narrow and deferential ways:
To know whether the new law represents a strong long-term policy response to the technological changes now challenging FISA, I would have to know a lot more about the NSA’s surveillance technologies both in the 1970s and now than is public. I would want to know also how the NSA interprets phrases like “reasonably believed to be located outside of the United States” and how it means to handle situations in which such people turn out, notwithstanding the agency’s reasonable belief, to be running around Cleveland.
But that’s precisely the problem: We would have to know all these things to be confident that privacy is being suitably respected, and we won’t. And we won’t because this bill removes all but the most nominal vestiges of oversight from an already incredibly secret process. Still, as Matt Yglesias points out, we have enough experience with this administration that it’s not wholly mysterious how unclear provisions might be interpreted:
[H]ere’s a wild guess as to how the NSA is going to interpret the phrase “reasonably believed to be located outside of the United States” — they’ll interpret it so as to give themselves as broad a mandate as possible. Other ambiguous phrases, likewise, will be interpreted so as to give themselves as broad a mandate as possible. What’s going to happen when they mess up: as little as possible. This is why, in the real world, we look not at administrative guidelines but rather at enforcement mechanisms.
Exactly right. If we were prepared to give the benefit of the doubt to government agents whose use of great power was cloaked in secretly, this would all be rather moot, because we wouldn’t need a Fourth Amendment, we wouldn’t need FISA, and we wouldn’t need judicial oversight. Since we have wisely chosen not to be quite so trusting with such power, on the basis of evidence both ample and recent demonstrating how misplaced such trust would be, we do need these things.
Update: Yglesias is a better libertarian than I am today, as he makes the necessary point that nothing about the posit that intelligence agencies will interpret the law to give themselves the broadest possible mandate turns on one’s distrust—however thoroughly justified—of this administration.