The Guardian’s Glenn Greenwald broke a massive story this week by publishing a blatantly unconstitutional, secret court order that directed Verizon Wireless to disclose daily records of every phone call made by every single one of its customers. The administration has responded, claiming that it has the need and the right to monitor everything you do for no reason at all, so long as the government’s own secret courts agree it might prevent a terrorist attack. A lot of the reporting since has focused on the distinction between phone records (“metadata”) and calls themselves (“content”). That distinction isn’t worth the (classified) paper it’s printed on.
The special treatment of “metadata” in U.S. law dates back to Smith v. Maryland, decided by the U.S. Supreme Court in 1979. Smith was the court’s misguided attempt to apply the Katz “reasonable expectation of privacy” standard to phone records. In Smith, the court held that if you voluntarily convey information to a third-party for it to take some action (like connecting and billing a call), that information is not protected by the 4th amendment. Phone call “metadata” like who you call, how long you talk, etc., were treated like the plainly visible routing information on a mail envelope. This decision ignored both the more frequent use of phone calls in everyday communication and the fact that phone record data wasn’t necessarily in plain sight, the way an envelope is during transit. In un-protecting telephone records, Smith decimated the privacy of association between people early on in the electronic era, and critically undermined not only the 4th amendment but the 1st as well.
The order Greenwald unearthed is the logical conclusion of that undermining, and it’s a bigger deal than it seems at first. Metadata is important to governments because it establishes motive and association. It’s important to people to protect, for the exact same reason: in a metadata-driven investigation, you can end up tarred in suspicion merely because you called the wrong person (even as a misdial, perhaps). Murphy’s law is that anything that can go wrong, will. Murphy applies acutely to government investigations and prosecutions, where errors are (sometimes quite literally) life-destroying events for victims.
A corollary of Murphy’s law, in the realm of surveillance:
The need to protect a piece of information from warrantless collection is proportional to the value the government finds in the information, and rises precisely at the moment the collection would be most useful to the government. The need to protect information from government intrusion arises not as a property of what information is being collected, but the very fact the government finds collecting it useful.
The “reasonable expectation of privacy” test from Katz could disappear. If the government is using a class of information to establish reasonable suspicion, reasonable suspicion should be required to collect it. If the government could conceivably take your phone records to a judge for a warrant or a jury for a conviction, collecting those records should require a warrant backed by… you guessed it, probable cause.
Are metadata collection, and the conclusions it could allow the government to draw, dangerous enough to preclude mass collection and require a warrant? You bet they are. To show you how dangerous it really is, I’ll provide you just a portion of the records Verizon has already handed over to the government. These records were collected by the government without a warrant or even a figment of a belief that I’ve done anything illegal. They’re collecting the same records on you too, in all likelihood. From yesterday morning at 9:30 am to now, I have called, been called by, or sent/received text messages to/from from the following people:
- My vet
- Four of my friends
- Two pest control firms
- My realtor
- My home insurer
- A taxi driver
- My dog walker
- My HOA president
- My condo management consultancy
- My dad
- My boss
- Two of my coworkers
It’s a pretty revealing list. Patterns readily emerge:
- The realtor, the HOA President, the home insurer and the condo management consultancy calls establish that I’m likely having some sort of problem at home. The pest control firms establish that it’s a pest problem of some sort.
- The call/texts to the vet and dog walker establish that I had some sort of concern about my dog.
- The call to my boss suggests one or both of these made me miss work.
- I went somewhere by Taxi on a trip that started within the city of San Francisco, because I got a call from a taxi driver who operates there.
Bear in mind, I’ve shared with you only one piece of what the government has collected, and only about 36 hours of years worth of collections. I’ve mingled together calls and texts, presented them out-of-order, and I haven’t shown you how many texts went back and forth between the people I texted. I haven’t shown you call times, or where I was standing when I made those calls/texts. You’d be able to draw further patterns from that data. All these inferences are possible without the government having heard one word of what I wrote or said on any of these calls or messages.
The government can assert that it needs this data, but it owes us the opportunity to challenge that assertion in an open, accountable court. Let’s not pretend that we have any shred of privacy from our bureaucracy in a nation where this much monitoring is routine for every citizen. The fact the President responded personally to the leak (itself, a piece of metadata) is more telling than any of the words he said: it shows he knows the public will be unhappy with the scope of the program. Yet, he reauthorized it anyway, in secret; that piece of metadata is the strongest indicator of the entire controversy that our government is out of control.