Edward Snowden's Really, Really Bad Logic

Okay, so, in case I have any regular readers other than the people I've been having actual face-to-face conversations with about various interesting topics over the past month, it seems to have been a while since I wrote one of these. I've been just a wee bit busy, graduating from college among other things. Today, however, there's a story that I do want to comment on, the most recent developments in the whole Edward Snowden/NSA thing. There's a lot about Snowden that pisses me off; overall I feel like he isn't really a civil disobedient in the MLK sense or in any way that has a consistent philosophy of how to reconcile disobedience with the general legitimate authority of the law, but rather an anarchist who thinks that he himself is entitled, as a "majority of one," to decide which laws are Bad Laws and then to not follow them and to get away with it. (This, I should say, is completely separate from my feelings about the NSA program itself, which are basically that I don't know enough of the details of what actually happened to know exactly what I think but that I feel fairly sure that this is not a genuine Obama Administration scandal.)

But today there's a whole bunch of new Snowden content, and it contains this nugget that really pisses me off. He did an online Q&A with The Guardian earlier today, and used this same little rhetorical device in two different contexts:

Congress hasn’t declared war on the countries - the majority of them are our allies - but without asking for public permission, NSA is running network operations against them that affect millions of innocent people. And for what? So we can have secret access to a computer in a country we’re not even fighting? So we can potentially reveal a potential terrorist with the potential to kill fewer Americans than our own Police? No, the public needs to know the kinds of things a government does in its name, or the “consent of the governed” is meaningless.

And:

Journalists should ask a specific question: since these programs began operation shortly after September 11th, how many terrorist attacks were prevented SOLELY by information derived from this suspicionless surveillance that could not be gained via any other source? Then ask how many individual communications were ingested to acheive that, and ask yourself if it was worth it. Bathtub falls and police officers kill more Americans than terrorism, yet we’ve been asked to sacrifice our most sacred rights for fear of falling victim to it.

You can see the idea that he's using in here. The surveillance practices he is leaking information about have the purpose of preventing terrorist attacks, but as terrorism lacks even the potential to cause as many American deaths as, say, accidental falls in the bathtub, or, to take a more politically charged example, American police officers, this is an insufficient justification. But this is bad logic. It resembles, in inverted fashion, the thing people sometimes say when accused of having done something bad where they invoke something else that sometimes happens that's worse, as if that excuses their actions. In both cases the argument is that because Y is worse than X, and therefore X is not the worst thing in the world, X is trivial and not worth worrying about. This is very, very wrong, whichever context it's used in.

The correct thing to say is that both X and Y are bad. In Snowden's example, we should note, correctly, that deaths from terrorist attacks are bad, deaths from falls in the bathtub are bad, and deaths at the hands of police officers are bad. We should then try, in principle, to prevent all of them from happening. In the one case that means trying to figure out what can be done to improve bathtub safety; there might not be much, beyond telling people to get mats and to be careful. In another case that means examining the question of why police officers kill so many people. They don't seem to in other countries, so in principle there should be things we could be doing to reduce these fatalities, the details of which are not within the scope of this blog post. And in the central case, it means, probably, conducting a certain amount of intelligence operations. We can argue about the details of how to structure a surveillance program to make it as non-invasive as possible, and theoretically about whether the least-invasive effective program possible is a price worth paying to save the number of actual lives in question. But the fact that other things, that might or might not be preventable through other means, cost more lives is strictly irrelevant to that utility calculus.

And I will say, though I am a committed civil libertarian who tends to be skeptical about the claims made on behalf of the security state, I think it's important to keep in mind that human lives are incredibly precious and that any argument of the form "well yes, Policy X would save a few lives, but it's not worth it" faces an incredibly uphill argument. It may sometimes be the case that something is too damaging, even in ways that can't be measured in terms of Net Lives Saved, to be worth avoiding some deaths; for instance, we don't prohibit driving because doing so would unquestionably avoid deaths in car accidents. Perhaps certain forms of surveillance aren't worth it even if they would save some lives, but you need to make the argument in great detail, not just make vague allusions to quotes from Founding Fathers living in an entirely different political environment.