For whatever reason, this has not been shared on any of the blogs I frequent despite it being instructive on the future of the surveillance state.
The Wall Street Journal notes the gobsmacking ineffectiveness of our efforts to stop terrorism:
Per life saved, the Food and Drug Administration maintains society should be prepared to spend $7.9 million. The Environmental Protection Agency has its own figure: $9.1 million.
The Homeland Security Department has shied away from any such discipline, though it once suggested that a life saved from terrorism was worth twice as much as any other life, bizarrely based on the public’s exaggerated dread of terrorism. Even so, the Federal Air Marshal Service has been estimated to cost $180 million per life saved. Airport security—don’t ask. One British study estimated that security efforts would have to stop 30 attacks per year on the scale of the 2005 London transit bombing to justify Britain’s spending under normal metrics.
Rather than concluding, however, that maybe we should spend somewhat less on things that don’t work…
Maybe the best outcome, then, would be to take metadata surveillance away from the spooks and apply it more broadly and openly to regular law enforcement.
Absent an improbable Bernanke-sized bailout of civil libertarians by Scalia and his newfound friends, this seems to me that it will eventually be the winning argument.
Pew says that 56% of Americans found acceptable the “NSA getting secret court orders to track calls of millions of Americans to investigate terrorism.” This was somewhat unsurprising because asking some one whether something should be done to “investigate terrorism” is basically like asking whether or not they hate America.
This same public would surely support proactive surveillance to “identify sex offenders” or “identify school shooters”. (There’s no need to include the word “potential” in either of those…unless you’re a sex-offender defender.) Do you think it would be considered inappropriate to troll through the database if it might assist with an AMBER alert?
Whether you support this development or not is immaterial because it is going to happen. If you still doubt this, consider that Julian Sanchez of Cato concedes that even the most sweeping of surveillance measures are probably legal under the Supreme Court’s current interpretations of the Fourth Amendment. Sanchez alludes to the need for a new constitutional amendment, which tells you just how lost a cause it is. Godspeed, Cato. [EDIT: OK, not really. See Sanchez’s piece and the edit below.]
For those of us who accept the constraints of reality, however, we can work on adjusting to our rapidly approaching futures. Here is my humble starter suggestion: a digital mirror.
The government and corporations have tools to help understand everything we do. Google and the government know every search I’ve made using it and how it compares to other users, but I don’t. Costco and Amazon know everything I buy (to the point of being able to contact me about product recalls), but I don’t. The government has apparently been tracking our credit card purchases, something I wish I could get some more help with.
A functioning digital mirror would consolidate whatever data you give it, process it into some intelligible points about what you might look like to an outsider based on that data, and tell you those points and how they compare to other users. Have your Facebook posts been particularly serial-killer-ish lately? Check your digital mirror to find out!
That’s all I have so far. Can you detail how such a service would work and what insights it could provide to a customer? Can you think of other tools that might be useful?
EDIT: Julian Sanchez notes in the comments that he makes no mention of the need for a constitutional amendment. Well, this is awkward. The parts of his piece that made me leap to the need for an amendment was that “When you used technology that left traces of your activity in the phone company’s files, the court reasoned, you ‘assumed the risk’ that the company would reveal that information — even if they had explicitly promised not to — and waived your ‘reasonable expectation of privacy’ under the Fourth Amendment.” He later notes that presently “a vast array of private information that would previously have required a physical search — and therefore a search warrant — to obtain is now available under a far lower standard.” This is the current protection offered by the Fourth Amendment, which suggested to me the need for another amendment if one really expected to protect data held by companies as opposed to just physical property.
Here’s what it would look like. The Collusion plugin for Firefox. You’ll be horrified and amazed how much information you leak, even just coming here.
That’s a nice visualization of who sees (though not what exactly is seen). I’d like to see some content-processing as well though and something more comprehensive than just websites visited.
That’s not websites visited, Vikram. That’s websites talking to other websites about you.
Am I really the best judge of whether my Facebook posts are serial-killer-like? As the author, I think I’ve been perfectly reasonable and appropriate at all times.
A digital mirror might be the wrong metaphor. I’d like to know whether my posts seem serial-killer-like to people who are not me.
Note that according to the testimony at the house hearing on Tuesday, the call records are not looked at when gathered, but rather go into a database. Then when a foreign actor is found making phonecalls, the database is queried to find out what calls were made. The testimony said that there are about 300 queries per year of the database. So they have to have an idea that calls from some number related to a terrorist, and then the database is queried to find out what calls were made from the number. If a US number is found the FBI is brought in to look at it.
Further it appears that before making a decision the FISA courts read the submissions, which are lengthy and come back with things that need to be fixed before the decision is made, the things are fixed and then the official decision is made. It is not the way usual courts in the US work, but does indicate that if you relie soley on the number of requests granted versus denied, it is not a valid metric, since prior consultation is done (which is not done in standard court cases, but since there is really no second side on the FISA court it does work).
Once a database is created, the evidence suggests that it will get used more over time. Notice the recent revelations on the growth of business-records requests under the PATRIOT act.
What the Wall Street Journal is suggesting, however, is that the database be used for additional purposes than whatever limited purposes it is being maintained for now. This is an inevitability.
“It is not the way usual courts in the US work, but does indicate that if you relie soley on the number of requests granted versus denied, it is not a valid metric, since prior consultation is done (which is not done in standard court cases, but since there is really no second side on the FISA court it does work).”
However, this assumes that virtually all requests are in the end valid, and just need some rework done on them. Which I for one don’t accept.
I’ve been considering the non-adversarial nature of the FISA court a lot, and the more I think about it the more I’d be happier with the FISA court’s superficial record if there were some sort of public privacy advocate, an equivalent of a public defender, practicing in opposition to the government. As it stands, the FISA court seems to be more like a grand jury than a trial court issuing search warrants — the threshold appears to be lower than probable cause, although we can’t really know because most of the decisions, and all of the back-and-forth between bench and bar, are still classified.
Courts work well when presented with adversarial arguments. (As I understand it, search warrants are sometimes granted unilaterally and sometimes after adversarial hearings with the court assuming the role of constitutional safeguard.) The grand jury system only works in the way it does because the threshold for an indictment is so low. FISA requests probably ought not to be treated the same way as indictments; adversarial setups are how courts work best. And with a public advocate, the worry that the attorney himself or herself would become a point of communication to the bad guys is minimized; the public advocate can have the same kind of clearances as do the attorneys applying for the FISA search authorization.
Burt, I hope you’ll do a post (or series) on these different processes/courts/systems in the judiciary; The Judiciary for Dummies. Who does what, how, the benefits, and perhaps a bit about the problems you see, what someone ‘trapped’ within a particular system might experience.
No one else seems much bothered by this, but:
I worry about PRISM data being used for commercial purposes. Not so much individual information about you mined and sold to the nearest Russian spammer, but anonymous data mined for statistical information about groups of people who hold certain things in common. Seems it would be valuable for targeted advertising, insurance, employment screening; anything where people’s revealed preferences can be used as a window into their lives as predicted by statistical models.
I want to know if the data being collected can be mined for commercial purposes in any way. I remind you that doing this would not necessarily violate anybody’s individual privacy.
Companies saying stuff like “48% of our customers who bought silverware also needed dishtowels, so we should have a page between buying silverware and checkout asking ‘DO YOU NEED DISHTOWELS?'”
Something like that?
I wonder if that becomes self-fulfilling.
“Well, I don’t need dishtowels, but what if those people know something I don’t about owning silverware?!?!”
That sort of thing freaks some people out, but not me.
I do start getting concerned when I read about targeted pricing and see some very discomforting things there. “A person with this browsing history will pay, on average, 10% more for a laptop than the average person.” Or “A person with this browsing history is statistically a worse driver than average, so we would need to raise his or her auto premiums to reflect that.”
I read an article about that a while back, how this was the future. I have my doubts whether it is, but I was discomforted.
It doesn’t necessarily freak me out. In fact, I’m glad that Amazon tells me what other customers who looked at this item bought.
That said, I’m not a big fan of advertising in general. It’s usually gauche and often intrusive.
There is a company called Acxiom, in Arkansas of all places, that does all this. It aggregates data from web browsing cookies, sites you sign up for, DoubleClick advertising, and the like, and provides very specific data about who your customer is, and their likely consumer and lifestyle choices.
They will make pretty educated guesses about your income, sexual orientation, political predilections, consumer habits and preferences, and the like. I worked with a company that used Acxiom services, and the kind of data they can return is scary: we sent them a list of our customer names / addresses / identifying characteristics, and Acxiom came back with lists that included the kinds of data hinted at above.
I don’t think people really understand yet: privacy is really, truly dead.
I didn’t say anything about the need for a constitutional amendment, nor do I think one is either necessary or realistic. The courts have already begun pulling back dramatically from a line of interpretation condemned by legal scholars for decades, and just need to continue doing so.
Hey! Nice to see you here!
Julian, sorry about stuffing those words down your virtual face. I reached too far in saying you “alluded” to the need for an amendment. Rather, I inferred the need given that you detailed how the Fourth Amendment does not protect business records.
You mention that “the courts have already begun pulling back dramatically” on this view. I didn’t see evidence for that in this particular piece. I will try to look around some other work.
How many lives would be saved if this money were put towards feeding people? Or health care? Or medical research?
Which money?
Oh, I take it you mean the terror-prevention money. The implication is that more people would be saved if the money were spent in almost any other area. The fact that the DHS does not use a cost-per-life metric is not by accident because many of the things they do probably don’t save any.
So the denominator is zero? They are LITERALLY being irrational.
But, yes, you got my implication. If we took the money used by DHS on terror-prevention and used it in other “life-saving” areas, I think you’d see far, far more lives saved. Same with Air Marshalls.
I’m tempted to even make the same argument about the EPA and the FDA, but my understanding is that those organizations are about more than just saving lives/securing our safety, so evaluating them only along that metric seems unfair.
But groups put in place to explicitly save lives and protect us? That aren’t doing much saving or protecting? Well, yea… irrational…
The denominator is quite possibly zero, but the implications of that are not obvious. The TSA has yet to stop any attack that has been attempted. Then again, surely the presence of the TSA has been a deterrent to attempted attacks. The shoe bombers put bombs in their shoes because of the need to conceal explosives to get past security.
So, it might be unwise to reduce spending to zero.
But that doesn’t mean we couldn’t cut it in half without ill effects.
Well said.
Overall, it’d be wise of the government to work smarter, not more expensively. They tend to assume the two are synonymous.
$7.9 million per life saved? That’s best argument for libertarianism I’ve ever heard.