Can the DC Circuit Use the Mosaic Theory to Invalidate the NSA Telephony Metadata Program?

Here’s a thought experiment. Let’s assume that when DOJ appeals Judge Leon’s opinion in Klayman v. Obama, the DC Circuit does not reverse on the procedural issues. Instead, on appeal the DC Circuit takes precedent seriously and tries to square the NSA surveillance program with United States v. Maynard and its mosaic theory on the merits of the Fourth Amendment. If that happens, how should the court rule? That is, how should the NSA program fare under the Maynard mosaic theory?

In some sense, there is no real answer to the question. Maynard‘s mosaic theory is novel, strange, and yet also strikingly vague. Still, I thought it would be useful to imagine what issues the DC Circuit might encounter if it tries to fairly apply the Maynard precedent on appeal. There are two key issues. First, does the mosaic theory apply to the kind of data collection at issue in the NSA program? And second, if the theory applies generally, how does it apply specifically to the facts of the Klayman case?

(1) Does the Mosaic Theory Apply to the Data at Issue in the NSA Program?

The mosaic theory developed by the DC Circuit in Maynard reasoned that long-term surveillance can allow the government to collect and analyze so much information about a person that eventually the government can develop a complete picture of their lives. When that happens, a Fourth Amendment “search” is deemed to have occurred. In Maynard, the government installed a GPS device on the car that the defendant drove and monitored it for 28 days. The GPS recorded the location of the car every few seconds. On those facts, the DC Circuit concluded that a search had occurred because the GPS surveillance “reveal[ed] an intimate picture of the subject’s life that he expects no one to have — short perhaps of his spouse.” The idea was that knowing the location of a person’s car on public streets is so revealing over time that at some point the surveillance crosses a line and becomes a search.

Applying this reasoning to the NSA surveillance program raises a threshold question: Does the kind of NSA data collected for phone calls raise the same conceptual problem? Under the NSA’s Section 215 program, the government doesn’t know any names of who is making any calls. Instead, they collect phone numbers without names, and have records of which numbers were called at different times (and for how long) by and from other numbers. When they query the database, they query a specific phone number, generally not associated with any known person. They can then find out what numbers that phone number called, collecting information about what numbers were dialed from that number, when, and for how long. The question is, does such information create such an intimate picture of an individual’s person life that it creates a full mosaic, thus constituting a search (at least at some point)?

I have a lot more questions about the mosaic theory than actual answers, but it seems to me somewhat uphill to argue that NSA Section 215 surveillance creates a mosaic about a person. First, because the NSA doesn’t include the names of anyone in their database, the amount of information the program would learn about a specific person would seem to be, well, none, at least based on that database alone. Maybe a mosaic is created when the government pairs that information with other information linking the number to a person — more on that in a minute. But it’s not clear that the database alone constitutes a mosaic search.

Second, the average American is a party to about 10 phone calls a day. At least a significant chunk of those are business calls, which probably aren’t intimate. Given that, I’m not sure that knowing the numbers dialed from a person’s phone creates such an intimate picture of a person’s life that it creates a mosaic about them. I suppose it depends on how that person uses the phone. Some people dial revealing numbers, and others don’t. Everyone is different. But assuming that the mosaic theory requires courts to hand down a rule that applies across the board to all people in a specific era regardless of variations — as in Jones, where the court had no information about what the monitoring of Jones’s location actually revealed about him — then I’m not sure that telephone numbers dialed and incoming numbers dialed are so intimate alone in some generic sense as to be capable of constituting an intimate mosaic of a person’s life.

(2) If NSA Surveillance Triggers the Mosaic, then When is the Mosaic Triggered?

Assuming that the Section 215 program triggers the mosaic theory, the next question is when the mosaic is triggered. This is a tricky issue. Most expressions of the mosaic theory focus on the combination of the raw acquisition of the data and the subsequent analysis of information in ways that create an intimate picture of a person’s life. Justice Alito’s opinion in Jones looked to whether a person reasonably expects others to “secretly monitor and catalog” a person’s movements. Justice Sotomayor asked “whether people reasonably expect that their movements will be recorded and aggregated” in a manner that creates the mosaic. Taking this language seriously, something more than mere data acquisition is required. As I wrote in my article in the mosaic theory:

[C]ourts will need to determine what kinds of postacquisition conduct are required to create a mosaic. Imagine the government collects a great deal of information but never combines it into a single database. Has a mosaic been created? Or imagine the evidence is collected into a database but never analyzed. Does that cross the line? If some analysis of the evidence is required to trigger the mosaic, what kind of analysis counts? Does any analysis suffice, or is there some threshold of sophistication or computational complexity before the mosaic line has been crossed?

If you take the mosaic theory seriously, I would think that some kind of post-acquisition data analysis is required. The mosaic is presumably created when the government actually analyzes the data in order to create the mosaic. In other words, merely having a number in the database somewhere, unknown to any person, in raw form, doesn’t itself create a mosaic. The government likely creates a mosaic of numbers only when those numbers are actually queried.

If that’s right, that seems like a problem for relief in the Klayman case. Although all calls enter the database, extremely few are actually queried. It seems exceedingly unlikely that the government had reasonable suspicion about Klayman’s number and queried it. It’s also quite unlikely that his number was picked up in a hop analysis, especially in any kind of detail needed to create an intimate mosaic of Klayman’s life — the kind of intimate detail that only Klayman’s spouse would know, as would be needed to satisfy the mosaic standard. If that’s right, it may be that the NSA surveillance program constitutes a search only when the government queries a person’s number, which the government asserts happens only when there is reasonable suspicion to think that the number is likely to reveal foreign intelligence information.

In that case, Klayman would lack standing to challenge the program because there is no reason to think that his number was searched under Clapper. And more broadly, the reasonableness of the mosaic search would balance the privacy invasion only of those whose numbers actually queried sufficiently to create an intimate picture of their lives counterbalanced by the furthering of the government interest in the cases where reasonable suspicion exists, which is a very different reasonableness balancing than Judge Leon assumed would be the case.