Next week a panel of the Third Circuit will be hearing oral argument in a case that considers whether federal law requires a warrant for the government to obtain historical cell-site records. I blogged a bit about this when the District Court’s decision was handed down, and I thought I would say a bit more now that the issue is before the Court of Appeals.
As I mentioned when I blogged about this issue in 2008, I think the answer to this case is easy: A Terry-stop “specific and articulable facts” court order is required, but a probable cause search warrant is not. As a statutory matter, the Stored Communications Act is clear here. The cell-site records count as a “record concerning an electronic communication service” under 18 U.S.C. 2703(c). Under 18 U.S.C. 2703(c)(2), such records can be compelled with a Terry stop “specific and articulable facts” court order obtained under 18 U.S.C. 2703(d).
But is this standard unconstitutional? That is, does the Fourth Amendment require a warrant backed by the exclusionary rule instead of the lesser standard and lesser remedy Congress has chosen? In her amicus brief, Professor Susan Freiwald argues that the Fourth Amendment protects cell-site info, which would require a warrant. But I think that is pretty clearly wrong under the Supreme Court’s decision in Smith v. Maryland. A cell site signal is closely analogous to numbers dialed in Smith: It’s a signal that the user sends to the phone company that is necessary for the phone company to deliver the user’s calls. It is a necessary part of placing the call, and information that is necessarily transmitted to the phone company. Professor Freiwald relies on various authorities to try to get around Smith v. Maryland, but I don’t think any of them work. Justice Harlan’s dissent in United States v. White was a dissent and has never been the law; Berger v. New York was a case involving content rather than non-content information; and Judge Posner’s opinion in United States v. Torres was on what kind of warrant was required for video surveillance, not whether such surveillance was a “search” in the first place. I don’t think any of these arguments can get around the pretty clear analogy to Smith v. Maryland.
The EFF’s amicus brief argues that cell site data is different than numbers dialed in Smith because people can reasonably not know that their cell phones need to communicate with cell towers to work. On that theory, people reasonably expect that their location information is private because they don’t know how cell phones function. But that seems to me like the kind of “magic box” argument that is inconsistent with Smith. In Smith, the Court presumed a telephone user who had a general understanding of how phones work: The Court presumed that people understand that when they dial a phone number, that phone number is communicated to the phone company and disclosed to them so the phone company can place a call.
Following Smith, I think the Third Circuit needs to assume that people know the basics of how cell phones work. Cell phones don’t work by magic: The phones need to communicate their location with nearby cell towers so the service provider knows where to route the calls. In my view, it’s hard to see why the Fourth Amendment should afford constitutional protection to a user’s failure to have that basic understanding. Such an approach would be especially problematic given that social understandings of how technology works can change quickly. My sense is that the percentage of cell phone users who have a basic understanding of how cell phones work increases every year. And once you know how a technology works, that understanding tends to stick. Given that, basing a rule on the incorrect understandings of a decreasing percentage of the population seems quite short-sighted.
Another argument sometimes made is that the statutory protections of the Electronic Communications Privacy Act itself should make an expectation of privacy in location data constitutionally reasonable. The argument is that Fourth Amendment privacy is sometimes phrased in terms of what “society” is prepared to accept as reasonable. By protecting the location data by statute, “society” has spoken. This argument doesn’t work, either, for a range of reasons. The first problem is the obvious bootstrapping problem. When the Fourth Amendment protects information, it protects that information with a probable cause standard and particularity backed by the exclusionary rule. In contrast, when Congress enacts legislation to protects by statute that which the Fourth Amendment does not cover, it often chooses a lesser standard and lesser remedy. (That is the case here: the standard is lower and there is no exclusionary rule.)
Saying that the legislative creation of lesser protection and a lesser remedy triggers the constitution’s higher protection and higher remedy is just bootstrapping; one might equally read the same clues as evidence that society does not construe the information as private because the legislature specifically rejected the constitutional protection standard. Put another way, if the existence of legislation shows that “society” recognizes that something is private, it is a strange homage to that judgment to strike down the legislation that is allegedly the evidence of society’s judgment. It seems to me that statutory and constitutional protection have to be considered separately.
Finally, some argue that Smith v. Maryland is wrongly decided. I happen to disagree; I think Smith is correct, as I argued in this article. But whether Smith is right or wrong, the Third Circuit is bound to follow it.
Mark N. says:
While I’m not sure if it’s enough of a difference to change the analysis, disclosure of location information seems at least somewhat less obvious to a user than disclosure of phone numbers. The user actively dials a phone number into the phone, in a way that is hard to view as anything other than an affirmative attempt to transmit that number over the telephone line. But a user doesn’t do any similar typing in of their lat/lon coordinates; it happens implicitly, rather than as an active communication from the user to the telephone company.
I think it might also depend on what kind of location information. The closest situation to Smith (though again, I think it’s not quite analogous) is when the location information is solely information about which tower the call was routed through, since that information is at least necessary to route the call. But location logs these days typically contain more accurate location information than that, collected via methods that are not actually necessary to route the call, like triangulation via multiple towers, and in some cases, GPS on the phone.
February 3, 2010, 2:16 amJohn Thacker says:
Have you heard anyone argue, analogous to the dissent in Citizens United, that Corporations aren’t individuals and thus don’t have Fourth Amendment rights? So therefore, while the individual’s personal phone records are secure, anything on a company server has no protection?
February 3, 2010, 2:39 ammethodact says:
Government baits and switches and then ratchets up laws incrementally a step or few at a time. Thus by slippery sloap, our freedoms are eroded a little at a time.
The pretext for wiretap and pen register laws was that of exigency – oh, let’s save this kidnap victim, there is a short window of opportunity – we only need these laws for the rare extreme and exigent circumstance. They then build out from there.
Because life is so important. Yet compare how the government will not pay ransom for hostages – there are higher ideals at play. Yet they can compromise the Peoples’ rights – for ever-pettier and pettier reasons still.
Ironic that right up until Olmstead, in 1928, the police were wiretapping members of the Court, itself. Yet the Court ruled that warrantless wiretapping didn’t violate the Fourth and Fifth Amendments, as taps were planted outside the home. It took all the way up until Katz in 1967, to admit that the Fourth Amendment extended to warrantless wiretaps. In Katz, Justice Stewart wrote, “No less than an individual in a business office, in a friend’s apartment, or in a taxicab, a person in a telephone booth may rely upon the protection of the Fourth Amendment.”
And back in Olmstead, Louis Brandeis had vigorously dissented and is quoted often, such as:
and
Of course now, we have a Hydra of laws and programs like CALEA and CARNIVORE and DIRT and MAGIC LANTERN and international planning at the economic summit in Davos Switzerland to implement driver’s licenses for the Internet and we have naked body scanners and thermal imaging and FLIR and bio thumb scans and iris scans and voice recognition and gait recognition and ubiquitous surveillance cameras and traffic cameras and deep packet inspections and NSA DATA CENTERS and SNEEK AND PEEKS and FBI SELF WARRANTS and DNA BANKS and cell-phone triangulation and RFID’s and oh yeah, your TERRY-STOPS.
February 3, 2010, 4:46 amCareless says:
When the “reasonable person” (or whatever standard is at use here) becomes significantly more intelligent/knowledgeable than the median actual person, there’s a problem. I’d happily bet against more than a third of the population knowing that much about cell phones. I wouldn’t be surprised if it’s half of that.
February 3, 2010, 9:21 ambillb says:
Why would a normal person assume that cell phone companies log the location data long-term? Sure, they need the cell tower data at the time of the call in order to properly make the call, but after the call is over why would someone believe that the data’s been logged? Phone numbers need to be logged because billing is based on where and who you called (e.g. long distance and in- and out-of-network calls). But which tower(s) serviced your call? I don’t see it.
I can see real-time location data being commonly understood as available to the cell companies, but there’s a bigger gray area around historical records in my mind.
February 3, 2010, 10:04 amGarrett says:
One issue that I would like to see discussed more is the interaction between technology and privacy. At what point does a person have to completely distance themselves from all modern technology to have some degree of privacy? Modern cell phones like the iPhone will send and receive data, even when they look like they are turned off. I assume, going forward, that devices are going to be more and more connected. If that’s the case, I’d need to give up all of the new gadgets which make me more productive and safer. This is far different from when a person takes an active, affirmative step to use a piece of technology, either making a call or swiping a credit card.
February 3, 2010, 10:24 amyankee says:
I had absolutely no idea that my phone company was storing data about which cell tower my calls were routed through. Why would they need such information?
February 3, 2010, 10:58 amArthurKirkland says:
A proposal:
Require all elected officials, judges and other important government officials to possess and use government-issued cellular phones when “on the clock” (performing official duties), and make all information (including historical cell-site records, and number and time of calls received and dialed) associated with such telephones available to the public to the extent any government agency is able to obtain similar information with respect to a citizen’s telephone.
This might help lawmakers to reach a proper balance between privacy and disclosure.
A question:
Could a telecommunications provider distinguish its service by refraining from maintaining customer-specific cell-site records? That could enable a customer to maintain the privacy of historic cell-site information by choosing such a customer-friendly vendor (although any issues concerning real-time or prospective information might differ).
February 3, 2010, 11:00 amEric Rasmusen says:
It wasn’t till a few sentences down that I realized the post wasn’t about stem-cell research cell-lines, but about telephones. I was disappointed.
February 3, 2010, 11:09 ampete says:
There was a New York city school employee fired when his employer tracked his movements via his cell phone and found out he was not spending all day at work.
February 3, 2010, 11:20 amenrico says:
Well, here’s what the govt brief says: “No Global Positioning System (“GPS”) data or other more precise location information (such as
February 3, 2010, 11:41 am“triangulation” data) is contained in the historical records requested pursuant to the Application.”
Loren says:
Until recently, cell phones had absolutely no idea where they were. They just “shouted out” periodically via low powered radio pings, and whichever cell towers heard the “shout” noted that they had heard the phone, along with relative signal strength. When a call is made to a cell, the phone company would look up which towers had most recently “heard from” the cell, and have them transmit the call.
The records of which towers have heard a cell phone should have a “life”. If a tower has not heard from a cell in x number of hours or y number of “ping” cycles, the phone company should no longer care if the phone was in range of a tower. It has obviously moved out of range or been disabled. When it comes back on the net, the cell phone will “shout out” again and whatever tower hears it will inform the phone network that the cell phone is back in range.
February 3, 2010, 11:44 amtheobromophile says:
When I use my cell phone’s VZ Navigator, it asks me for permission to access my cell phone’s location for four hours. Now, for those of us who know how cell phones (and E911 and all that) work, this is not very confusing; however, the general population may make an assumption that the approval is needed because cell phone companies don’t otherwise track your location.
Furthermore, most people probably would not think that their cell phone company tracks the specific location of their calls (as opposed to in-network or out-of-network), stores that data (after all, isn’t data storage supposed to be expensive?), and can hand it over to the government without a warrant.
Finally, call me crazy, but I don’t think a reasonable person would know how much information the government can constitutionally collect about them without a warrant. In this day and age, it’s increasingly difficult to interact with society (use a phone, use the internet, get cash from an ATM, etc., without creating the kinds of paper trails that the government can access without a warrant. I doubt that the Framers intended for the protections of the Fourth Amendment to be mooted by technological advances.
February 3, 2010, 12:01 pmdhlii says:
Deciding constitutional issues based on obscure technical details regarding the way something works is insane. There are nearly infinite technical solutions to most any technical problem. we should not be solving technical problems based on the extent to which we surrender our rights. Worse still this allows an end run around our rights – the government can just mandate a technical implementation that surrenders our rights. The expectation of privacy is a reasonable standard, what is unreasonable is the belief that the details of a technological implimentation diminish that expectation.
February 3, 2010, 12:01 pmDJR says:
Wouldn’t the information stored by cell sites be like the beeper in Knotts v. U.S., which police could monitor so long as the information was the same as could have been observed with the naked eye? So long as the police don’t attempt to use the cell info to locate the defendant in a particular house (cf. Karo v. U.S.), which does not seem possible unless GPS info is transmitted to the tower, the after-the-fact monitoring of a person’s movements by tracking cell phone sites would not be a problem.
February 3, 2010, 12:29 pmChris Travers says:
I think there is a significant difference between the pen register issue and cell-site records. Whether or not it creates a fundamental difference under the 4th amendment is an issue for lawyers and law professors.
With a pen register, you can know which numbers were dialed to and from a specific telephone which is at a specific place. This phone is of fixed location and provides some information on who MAY have called who and when. It doesn’t track an individual’s movements, and so the information is of limited investigatory help.
Cell-site information is fairly different. Depending on where this takes place it could be entirely useless (say, one cell tower that covers a large rural area such as where I live) or it could provide detailed information about an individual’s movements (in a large city like New York).
IMO, I don’t think cell-site information should be seen as analogous to pen registers. The fundamental question needs to be at what point tracking devices become a fourth amendment violation. While in general tracking devices are not prohibited categorically under the 4th Amendment, it might be different now that everyone has something that the police could arbitrarily track.
For example, I think it is one thing if the police place a tracking device on, say, a car. A cell phone arguably provides a more intrusive means of tracking than that in two ways:
1) It usually follows the person, and
2) Historical tracking information (from before the time when the investigation was launched) would be available. In essence, police could ask “so, where did Prof. Kerr go on New Years Eve last year?” even if they hadn’t been doing such an investigation at that time.
I am not convinced these questions have been adequately addressed by the Supreme Court. Indeed the only cases I can think of explicitly deny deciding it. So I personally think the lower courts have a fairly free hand.
Let me ask an analogous question: Suppose we put license plat detector cams at every intersection and logged every car’s license plate that went through every intersection. Would that violate the 4th Amendment?
February 3, 2010, 12:43 pmDJR says:
Chris,
Under what theory would it be reasonable to expect that your car’s presence at a particular public intersection at a particular time would remain private?
February 3, 2010, 12:52 pmChris Travers says:
Just as a note: Theoretically you could probably get:
1) Which tower was hit
2) Which antenna on the tower was hit. Each antenna may hit 30 degrees or so of lateral direction and are uni-directional. So what you get are not large circles where a person is travelling between but much smaller triangles.
In a single tower for a larger rural area, if the tower is off to the side of a town, you could go all around town and never change antennas. Each antenna is mostly limited (usually) not so much by range but by maximum subscribers, so the triangles are very large out in rural Kansas, and very small in NYC or San Francisco.
Because of the way cell phones work, there is an actual hand-off between antennas on the same tower (under either CDMA or TDMA, FDMA poses the same issues but isn’t used much anymore).
Also if you can estimate speed, because of the fact that you are looking at triangles rather than something like a circle or a square, it becomes pretty easy to actually track exactly where someone is going. Driving right by a cell tower at even 20 mph will cause a series of very rapid handoffs, for example.
I personally think this data should be protected for the same reason that the plain view exception should be removed from computer searches. It is simply (in urban environments) too detailed and if it is kept around for an extended period of time, poses major privacy risks analogous to CCTV cams on every street corner.
February 3, 2010, 12:57 pmChris Travers says:
So do you think dragnet surveillance of public streets would ever violate the 4th Amendment?
I am asking because in general, there seems to be disagreement about where, not if, a line is drawn regarding public surveillance.
We don’t allow every car to be stopped and searched for drugs even though any given car can be stopped and searched, for example.
February 3, 2010, 12:59 pmArthurKirkland says:
Framers, no.
The moral and intellectual heirs of the Framers, no.
Drug warriors, panicked terror warriors, nanny-staters and freedom-haters, yes.
Also, what dhlii said.
February 3, 2010, 1:04 pmChris Travers says:
You also get which direction from which tower. Cell tower antennas are unidirectional.
February 3, 2010, 1:05 pmOrin Kerr says:
I responded to that argument here.
February 3, 2010, 1:11 pmChris Travers says:
Here is a quick intro:
Cell towers usually have at least six (and sometimes many more) antennas which are unidirectional. Each antenna picks up a range of acceptable frequencies, and several phone calls can occur on each frequency band on each antenna. This is managed using either Time Division Multiple Access (TDMA) or Code Division Multiple Access (CDMA) algorithms. I have seen cell towers with either six or twelve antennas, making each antenna responsible for either 30 degrees or 15 degrees.
If the cell needs to be made bigger, the antennas can be made stronger (to a certain point of course), and if they need to be made smaller, the power applied to the antennas can be reduced. Cells will be made smaller in cities so that more calls can be processed (more cells, more antennas, lower power). Most of the ones in cities will have at most a 15-degree coverage and a fairly limited range.
When you move from one antenna (not tower) to another, TDMA and CDMA require re-allocating the appropriate bandwidth to the new antenna. This requires a hand-off. Given the triangular shape of the site, it isnt hard to figure out where someone is going based on handoff information.
These are logged by the cell phone company for a number of reasons including attempts to prevent fraud (use of the same identifier on multiple cell sites for example). I don’t know what sort of date retention policies the companies have.
February 3, 2010, 1:15 pmChris Travers says:
In case someone wants more info on TDMA and CDMA:
In TDMA, each frequency range is divided into bands, and each band is divided into time slices and each of these time slices is allocated to a phone call. Each time one moves to another antenna, a new frequency band/timeslice combination has to be allocated to the phone.
On CDMA, there is a different way of handling this (which I don’t fully understand) which also involves both timing and frequency, as well as signal modification. If you want to read more about how it works, look here.
February 3, 2010, 1:26 pmA. Criminal says:
“In her amicus brief, Professor Susan Freiwald argues that the Fourth Amendment protects cell-site info, which would require a warrant. But I think that is pretty clearly wrong under the Supreme Court’s decision in Smith v. Maryland.”
…
The EFF’s amicus brief argues that cell site data is different than numbers dialed in Smith because people can reasonably not know that their cell phones need to communicate with cell towers to work.”
There aren’t any exceptions listed in the 4th amendment. So, the supreme court was wrong in Smith v Mayland, which is nothing new since that bunch of government lawyers is wrong most of the time.
I think it must be really frustrating to be a lawyer, at least for normal people, because you have to pretend that false things are true because Somebody Officially Important said they’re true. Come to think of it, maybe normal people, here defined as those who care whether something is true of false, don’t become lawyers…?
The 4th reads somewhat like the 2nd, since both have the form: “Because of A, the gov’t can’t do B.”
The important part isn’t A, which isn’t prescriptive, it’s B, and here B is “no Warrants shall issue, but upon probable cause…”
Some people like to make these issues sound all subtle ‘n’ complicated and therefore ever so intellectually difficult, but they’re not rocket science by a long shot.
Probably for network analysis. Data storage and access are so cheap and fast now that they might as well keep just about every bit of information indefinitely.
February 3, 2010, 2:01 pmChris Travers says:
Just to clarify:
If the police are watching a specific intersection for a known stolen car, that is legal. If in the course of an investigation, they put a tracking device on a specific car that they suspect to be involved, the Supreme Court has said that is legal even without a warrant, but explicitly rejected the notion that they were deciding more intrusive surveillance measures. I bring this up because I think cell-site data is closer to to this case than to a pen register, but is quite a bit more intrusive.
However, what I am less sure of is whether the police are allowed to track every car at every intersection ostensibly looking for stolen cars, and then archive the data for an extended time period, using it in, say, drug dealing investigations later. I personally think this would be well over the line, and while I don’t think cell cite info is quite that intrusive (yet), it is closer than most people suspect.
February 3, 2010, 2:12 pmDJR says:
Chris,
You appear to be arguing that too much of a constitutional practice is unconstitutional. I’m not aware of any case that has adopted such a rule. If an officer standing at an intersection can note the license plate of someone driving by for any reason or no reason at all, that officer can also note the license plate of everyone driving by, and the police can put an officer at every intersection to do the same thing. I see no reason why it would be any different if an electronic device did the same thing. We may not like living with that kind of surveillance, but that does not mean it’s unconstitutional.
To take another example, police can generally follow someone traveling in public without a warrant or any degree of suspicion. Could the police assign someone to follow every non-police officer’s movements in public? I see no reason why not.
The solution in such cases is not to say the conduct is unconstitutional, but to go to the political branches and tell them to stop having police follow everyone around all the time. I would certainly vote for the candidate who supported an end to big brother style surveillance.
What you appear to be contemplating is a kind of “right to privacy by getting lost in the crowd,” but that’s not how the courts have generally approached the Fourth Amendment. Maybe some courageous court would do so if the extreme practices you contemplate came to pass, but if so it would be an entirely new line of reasoning.
February 3, 2010, 3:03 pmzuch says:
Prof. Kerr:
While you might be right on the basis of the “call data”/”call content” distinction WRT the level of proof and judicial involvement involved, I’m not sure why the Terry “protection of the law officers” rationale would apply to such records. Can you explain (or link to an explanation)? To be sure, the “call content”/”call data” distinction seems, to me, in this digital era far beyond pen registers, a bit quaint. Thanks in advance.
Cheers,
February 3, 2010, 3:16 pmDavid M. Nieporent says:
But what “line” is it “over”? While I think such behavior is inappropriate, the Constitution doesn’t limit police behavior for being too effective. It limits searches. What you’re describing isn’t a search; it’s observing and recording what happens in a public place. No matter how creepy that is, it doesn’t implicate the fourth amendment.
February 3, 2010, 3:27 pmCory J says:
Zuch,
I believe the answer is in these bits from Orin’s post:
“The cell-site records count as a “record concerning an electronic communication service” under 18 U.S.C. 2703(c). Under 18 U.S.C. 2703(c)(2), such records can be compelled with a Terry stop “specific and articulable facts” court order obtained under 18 U.S.C. 2703(d).
…
In contrast, when Congress enacts legislation to protects by statute that which the Fourth Amendment does not cover, it often chooses a lesser standard and lesser remedy. (That is the case here: the standard is lower and there is no exclusionary rule.) ”
Since they are not protected by the 4th Amendment under Orin’s view, the law enforcement officials need only follow any applicable statute protecting the records. Since the legislature chose the Terry standard, that’s what the LEOs need to show.
February 3, 2010, 3:27 pmloki13 says:
According to the police-
Every breath you take, every move you make, I’ll be watching you.
/sorry, couldn’t resist. BTW, I went to a wedding were the happy couple played this as “their song.” Creeeeeeeeepy.
February 3, 2010, 3:45 pmOren says:
Assuming I understand what’s involved, it’s hard to articulate a decent 4A argument against it. There are no private facts involved, nor entry into private space nor interference with enjoyment of property. The police violate no positive law in collecting this data (in fact, a civilian traffic engineer would be well within his rights to sit in a public place and collect this information).
This policy seems like a born winner.
February 3, 2010, 3:48 pmOrin Kerr says:
Zuch,
It’s a statutory issue, not a constitutional one. See 18 U.S.C. 2703(c)(2).
February 3, 2010, 3:48 pmOren says:
I would bet instead on the State legislatures curtailing the process.
February 3, 2010, 3:52 pmChris Travers says:
Katz type analysis maybe? There is a general and reasonable expectation that not everyone will have all their movement tracked at all times? that is an expectation of privacy, right?
February 3, 2010, 4:10 pmCory J says:
I don’t think privacy turns on the likelihood of any bit of particular information being “seen.” I wouldn’t expect a pen register to be installed to capture my phone numbers, or a low-flying helicopter to peer into my field, etc.
Think of privileges and how they can be lost if a third party overhears a discussion involving a privileged matter. As far as I know, it isn’t a good defense to say something like, “I didn’t expect someone to hear me…we were in a crowded restaurant and I assumed no one was paying attention to us.”
February 3, 2010, 4:27 pmChris Travers says:
Agreed on both these points. Instead I think a lot of it has to do with the intrusiveness of a given measure.
True. However, does that mean the government can put bugs on the outside of all phone booths? Certainly not….
February 3, 2010, 4:45 pmMark N. says:
It’s dicta and not quite on point, but it’s possible a court could use some version of the “dragnet type law enforcement practices” exception that Knotts hinted at. I’d bet there are at least 3 votes on the current Supreme Court for expanding the Knotts dicta into a general prohibition on certain kinds of dragnet-style practices. No idea if there’s any chance of 5 votes, though.
February 3, 2010, 4:53 pmOren says:
It’s not reasonable to assume that your movements on the public street are not being observed — quite the opposite, in fact, it’s almost certain someone saw you. Nor does using the work “tracked” as opposed to “observed” change anything substantially. It is entirely uncontroversial that an officer is entitled to make notes of anything he lawfully observes. So if observing you on the street is lawful, so to is noting it and the collating those notes into a picture of where you’ve gone.
BTW, as an aside, in my State (MA), it would be legal for a private individual to live in his car and follow me around 24/7 (our stalking/harassment law requires an actual threat) and write that in his notebook. If he did that, I would have no recourse at law — that makes it hard to contend that I have a reasonable expectation of privacy in that information.
February 3, 2010, 4:54 pmChris Travers says:
So, Oren:
If I (and/or a police officer) stand outside a payphone booth and listen (assume I am undetected) to what someone is saying inside, that doesn’t pose a 4th Amendment problem either, right?
Would it be different if a small microphone was attached to the outside of every payphone booth in town?
Constitutionally a police officer, articulating a reasonable suspicion, can pull over a car and search it for drugs, right? Does that mean that a police officer can pull over every car on the highway and search them for drugs?
I guess what I am arguing is that there is a fundamental difference between a small, targetted, gathering of data which may be “reasonable” and a large-scale, automated gathering of data which may be beyond what is reasonable.
In the public sphere, the 4th Amendment requires reasonableness. In the private sphere it clearly requires a warrant. I don;t think easy access to data tracking an individual’s historical movements is “reasonable.”
February 3, 2010, 5:04 pmOren says:
That’s interesting that you would bring it up, because I was going to cite the non-intrusiveness of it as an argument for it. I didn’t, because I didn’t think it was necessary to get into the policy realm of things.
Now that we are into it, what exactly is intrusive here? If anything, passive monitoring of goings-on in public space would seem to be an example of the absolute minimal intrusion. Certainly less than flying over your house with a helicopter, at any rate.
Indeed. That would be new doctrine and, IMO, a quite messy one. How would you would systematically quantify when a sum of otherwise-lawful observations becomes a “dragnet”? Can an officer write down the license plate of 10 cars crossing his intersection but not 20? Does that change based on whether other intersections are monitored? I would hope the Court would shy away from such a doctrine solely out of sympathy for the lower courts.
February 3, 2010, 5:12 pmzuch says:
I didn’t know that Terry stops required court orders.
Cheers,
February 3, 2010, 5:13 pmzuch says:
That I understand. Why that implicated Terry is what has me confused. Terry wasn’t a statutorily enacted standard.
I understand that courts have refused Fourth Amendment protections to call data (as they did to Terry in his circumstances as well). But the rationale is not quite the same.
Wouldn’t it make more sense to refer to it as a “specific and articulable facts” standard rather than as a Terry standard?
Cheers,
February 3, 2010, 5:18 pmOren says:
Of course it does. The conversation is protected because we reasonably don’t think others will listen in. What the officer can do, however, is sit across the street and note who uses that telephone and when. So you used it at 8:30AM, then Joe at 8:37AM, …
I don’t know what I said that would make you think I believe this. I would ask you to kindly reread my posts and articulate where I even address any issue involving actual seizure, let alone that such seizure need not be supported by probable cause.
To reiterate, though, AFAIK there is no standing legal bar to a police officer driving down the highway and observe that Joe is traveling in his car. There is no further bar from him collecting this information in a notepad and then concluding that Joe seems to travel between I95 exits 35 and 38 every morning at around 10AM.
Whether this difference is “fundamental” or not, it does not appear to be one recognized by current 4A doctrine (although Mark’s pointer to the dicta in Knotts is interesting).
As it stands (correct me if I’m wrong), if the police can observe something is independent of how many times they have observed the same or similar data. It is either legal for police to observe the outside of every house from the sidewalk or it is never legal. Perhaps a new doctrine will change that, but that’s where we are now (again, AFAIK).
The 4A makes absolutely no distinctions based on “access to data” except to inquire as to whether that data is properly acquired in the first instance. That is, either the search is legal, in which case the police can do with the data as they please (including writing it down, entering it into a database, …) or the search is illegal and they may do nothing with it (fruit of the poisoned tree).
That is, AFAIK, the 4A cannot act on your “historical movements”, it can only act on the searches (and perhaps seizures) that the police use to acquire information about those movements. The reasonableness requirement attaches to the searches and seizures, not whatever distilled product the police make from it the next day.
February 3, 2010, 5:23 pmOren says:
Yes, but the latter is a lot more catchy.
February 3, 2010, 5:28 pmDJR says:
It’s not fair to extrapolate from one search with reasonable suspicion to many searches without any suspicion. Assuming you’re right about the standard, if there is reasonable suspicion to pull every single car on the highway over and search it then absolutely the police can do that. Again, you’re presuming that it’s reasonable to assume you won’t be searched because the police can’t search everyone. But if what the police are doing is legal, whether observing people on a public street or pulling someone over when they have sufficient suspicion to do so, there’s no constitutional principle against too much of a constitutional thing.
Your phone booth example is a little different, because the expectation of privacy of your conversation there comes from being enclosed in the booth and being able to see that nobody else is around. So a surreptitious listening device would likely upset that expectation in the way that a person standing right outside the telephone booth would not. If there was a person standing close enough, you probably would expect that person could hear you.
February 3, 2010, 5:30 pmChris Travers says:
A lot of this has to do with the fact that the data collected is very comprehensive in many events both in terms of its volume and what is actually collected. This goes well beyond what even a fairly large law enforcement department could expect to do unaided (which then goes into questions of whether a rationale looking at the Knotts dicta might also draw on Kyllo).
I would think that an officer would be limited by typing speed, and an ability to coordinate hands and eyes. I can’t imagine that officers could ever be precluded from manually recording what they observed. Rather I think courts should look at four factors:
1) Comprehensiveness of surveillance
2) Length of retention of data
3) Comprehensiveness of data
4) Rationale for surveillance
Now one argument that could be made in Orin’s favor here is that the conditions to subpoena the data are reasonable as well. I guess I would like to see a developed record of evidence before concluding for or against.
To my mind the question is not whether the Constitution protects the data but whether the law meets the Constitutional requirements.
February 3, 2010, 5:34 pmOren says:
While we are one Knotts:
Again, the fact that the data is comprehensive has nothing to do with the 4A, which is concerned only with the manner in which it is collected. There are no cases in which the Court has ruled that a transformation of the data after the search is enough to reach back in time and cause the initial search to be unconstitutional.
So using a computer to look up a license plate would also be forbidden because it allows the officer to do more in less time? Besides being foreclosed by Knotts, that argument strains credulity.
[ Story: I went on a police "ride around" as part of a citizen education course and I was impressed by how many license plates the officer could run while stopped at a red light. At a long light, he easily cleared a dozen just punching them in. He said they catch a lot of warrants and suspended license/registration cases like that, even though less than 1 in a 1000 (his rough guess) are hits. ]
But the Law is already in excess of the Constitutional requirement!
February 3, 2010, 6:00 pmArthurKirkland says:
I’m convinced. Require every elected official, judge, police officer and prosecutor to carry a functioning mobile telephone while on duty. Make the call and cell-site logs available to the public that pays their salaries. If it advances the public interest without offending the Fourth Amendment, who would object to the availability of such public information?
February 3, 2010, 6:31 pmChris Travers says:
Not necessarily.
Let’s come up with some hypotheticals.
Suppose a large city decides it has a problem with vehicle thefts. So it institutes the following program:
Police cars are outfitted with licence plate/ocr scanners which scan the license plate information. The information is immediately sent to the base station, where it is compared with a database of vehicles reported stolen. Matches are retained and officers notified. Non-matches are immediately discarded.
I see absolutely no fourth amendment problem there.
Suppose this isn’t enough. So the city places, at every major intersection similar scanners and tracks stolen vehicles. Matches are retained and officers notified. Non-matches are immediately discarded.
Still no fourth amendment problem there.
Suppose the city then decides to improve the cameras, look for certain driving behaviors which suggest driving drunk and flag these vehicles. Drunk driving matches are retained for 24 hours and then discarded. Nothing else changes. Still no 4th Amendment problem.
Now, suppose the city decides to crack down on drug dealing. They then reprogram the system to identify suspect patterns of traffic, log ALL license plates at every intersection for later analysis, and begin using these to subpoena other information in investigations. At this point (only) I think this goes beyond what is reasonable under the fourth amendment.
In essence what I am suggesting is that a number of factors really might need to be weighed in order to determine whether or not a surveillance scheme is overreaching in public space. These include how comprehensive the surveillance is (one police officer on one street corner isn’t comprehensive), how the data is retained or discarded, and the purpose of the surveillance. The nature of cell site location information is such that we should expect some sort of reasonable court oversight. The comprehensiveness of the information means that there needs to be a way of limiting inquiry only to data that is necessary.
I think this is the reason that the government spends so much time asserting that the location information is fairly imprecise, and the EFF arguing that this is not the case. Indeed, in a dense city, the information could be quite precise.
Now, reading the EFF’s brief I notice a few things missing in this discussion.
The EFF’s argument is actually that the magistrate is empowered under the statute to determine if the information sought demands 4th Amendment warrant protection and therefore ALLOWS the magistrate to use good judgement in denying subpoenas, forcing an application for a search warrant instead.
So, question one is whether the use of the word “may” in 18 USC 2703(d), and while “shall issue only if” allows magistrates to evaluate possible 4th amendment issues on a case-by-case basis. I think the EFF is right on this because otherwise “only” would be superfluous.
Now, there is one point here that I think Prof. Kerr misrepresents about the EFF’s case.
The EFF is arguing in part that data available NOW from some phone carriers may be precise enough to locate someone within a Constitutionally protected space (a building for example), and that even mere cell site data may be becoming precise enough to warrant concern.
From the EFF’s brief:
In support they cite Karo and Knotts.
Maybe this is why Prof. Kerr is trying to avoid the conclusion that cell phone data of this sort makes the cell phone a tracking device?
February 3, 2010, 7:37 pmFub says:
I agree with this. There has only rarely been a substantial penalty for politicians, judges, prosecutors, or police for riding roughshod over the rights of citizens, even when their methods were later held unconstitutional by higher courts. Let them suffer the same indignities that “the little people” suffer.
February 3, 2010, 7:38 pmOren says:
Chris, your scheme is just plain contrary to current 4A doctrine. The legality of a search cannot be made contingent on what the police do with the evidence after its collection. Nor can it be made contingent on how many other similar, concurrent searches are being made. Rather, each search* is viewed as to whether it is, in isolation, reasonable at inception and execution.
Your attempts to stretch the 4A to include the “back-end” data processing, e.g. what happens long after the search has terminated, is absolutely unprecedented. There is literally not a single ruling by the Court where the reasonableness of a search (or a non-search) has turned on what happened to the fruits of that search. What has started out as a mildly interesting hypothetical has now taken full flight into fantastical new worlds of jurisprudence.
* I should point out that an officer viewing a car on the street isn’t even a search in the first instance, by analogy to looking at the outside of your house from the sidewalk. That is, the Court did not find that to be reasonable but even more damning, that they don’t even meet the predicate requirement for any 4A scrutiny.
Quoting Scalia from Kyllo
So things are only getting worse for your argument, since now you need to leap all the way from “non-search”, into “search” and then on to “unreasonable search”.
Sounds like a policy judgment for the legislature to hash out.
February 3, 2010, 8:51 pmOren says:
Because we can reasonably expect them to divine the future rulings of the Court?
Don’t get me wrong, I’m quite in favor of strict application of the rules and doctrines as they exist but it’s absurd to ask a police officer to obey a ruling that the Court has not yet even handed down.
February 3, 2010, 8:52 pmDavid Nieporent says:
The problem with your argument is that the fourth amendment doesn’t govern “surveillance,” and doesn’t say that the government behavior must be “reasonable.” It governs searches (which must be reasonable). If it isn’t a search, it doesn’t matter how “comprehensive” it is; the fourth amendment doesn’t say anything about it.
Looking at people in a public place isn’t a search.
February 3, 2010, 9:05 pmFub says:
It’s absurd to expect that a public official could ever be expected to divine that some acts might be inherently wrong and immoral, without a Court ruling that they are coincidentally unconstitutional or otherwise unlawful?
February 3, 2010, 10:26 pmOren says:
The finer points of the law, for instance when an officer is entitled to search a vehicle subsequent to the arrest of the driver, are not particularly ethically consequential. That is, when the Court moves the line a little bit forward or a little bit back, it is almost invariably acting inside the range of policies that are ethically sound (it is important to note here that ethics surely condones more than a single point in the spectrum of all possibly policy choices). So it is hyperbole to assert that the Justices that wrote and the police that implemented Belton could see that it’s “inherently wrong” even after it reversal in Gant. One might be better policy or better law than the other, but both are plainly well (very well) within the bounds of ethical policy.
IOW, the point of my comment was not to excuse “inherently wrong and immoral” conduct but rather incredulity at expecting law enforcement to divine the smaller twists and turns of the Supreme Court’s doctrine. Where the behavior “shocks the conscience” (see, e.g. Rochin v. California), the lack of on-point precedent is never an excuse.
February 3, 2010, 11:21 pmFub says:
The CA 2nd DCA’s opinion in the original case reinforces my original point above, that
From People v. Rochin (1950) 101 CA2d 140:
A little clucking and tut-tutting from the bench, admitting that the police and physician committed felonies, suggesting that the victim sue civilly, and suggesting that the sheriff and medical board review qualifications. Any court has capacity to refer the record to the AG for criminal prosecution, and any AG has the capacity to prosecute. But actual prosecutions of officials for egregious violations of citizens’ rights are scarce as hen’s teeth.
Which is why I agree with Arthur Kirkland above. That criminal conduct for which officials are unwilling to actually sanction their own brethren, ordinary citizens should at least be legally empowered to discover and act upon politically as it occurs.
February 4, 2010, 1:30 amOren says:
This is entirely non sequitor — you cite the CA court ruling on Rochin despite the fact that he won his appeal in the Supreme Court. The seizure was ultimately ruled unlawful, the conviction overturned — it stands as a vindication of the rule of law. Rochin did bring a civil suit against the county (under S1983) and settled for a hefty sum. I’m baffled that you consider the case evidence that police can get away with unlawful conduct.
At any rate, if anyone wants to get back on the topic of cell records and/or Chris’ related hypothetical on aggregating traffic data, I’d be happy to continue discussing it. Arguing about police abuses in such vague and general terms (I’d welcome, of course, specific instances relevant to the tracking of cell-site or other location data) might be catharsis for some, but it seems unproductive to me.
February 4, 2010, 2:44 amEdward says:
Some facts about what a major cell phone company collects from your cell phone:
(Source: a friend of mine who has worked a market analysis contract firm and worked with the data in question.)
* They never throw away any piece of data once they collect it.
* They use cell tower triangulation and, where available, phone GPS
to determine the location of the phone.
* They started collecting this data because of government laws–it
is not a technical requirement for providing phone service. The
original government mandate was motivated by the requirement to
find 911 callers who do not give their address.
* Once they had to have the data, they developed ways to monetize it.
* They track how quickly and in what direction you were moving:
Walking or in a car, etc. Speeding?
* This data is collected whenever the phone is on. Not just when
you are making a call. Very few people realize this.
* They track who you spend time with by tracking which phones
move together.
* The marketing folks do a bit of work to de-anonymize
pre-pay anonymous phones to be able to use their data.
From this data they are able to determine where you sleep, where you
work, and where you shop. They track who you are with and for how
long. The developed an application–it was originally created as a
joke–called “Who’s f*ing who?”
All of this data is used for marketing efforts. Some of the possibilities
have not been exploited yet, but they have studied them. And they have
already developed the profiles. (Stop at daycare in the morning: you have
young children. Often pick them up late? You are probably a
workaholic professional.) Shop at Walmart or Neiman Marcus? Who is
your doctor? Do you see your dentist every 6 months or are you lax
in such care? Are you seeing a cardiologist or oncologist? How often do
you use the restroom? (The quants use this to determine menstrual
cycles and stages of pregnancy.)
All of this is from the location data. They develop profiles that they
then track via the SSN from the subscriber info. This allows them
to mix in data from other marketing sources to develop a more
thorough picture of the individual.
One reason they have not used the data even more extensively is
customer privacy concerns. They are aware that most customers don’t
realize that the data is collected even when they are not on a call.
And they don’t want to anger their customers by calling it to
attention.
Can law enforcement legally request these profiles?
February 4, 2010, 9:26 amArthurKirkland says:
A former high-ranking government official currently employed as a high-ranking telecommunications firm officer recounted a similar tale — with an additional layer of revelation about the government side of the equation — at a banquet table.
The conversation that evening is a large part of the reason privacy issues involving cellular telephones trouble me. If Google and [insert your telecommunications provider] ever pooled information . . .
February 4, 2010, 9:49 amCory J says:
Was the government side the pervasiveness of using cell phone microphones to eavesdrop?
February 4, 2010, 10:16 amFub says:
Because government officials who perpetrate crimes yet retire to cushy pensions instead of striped sunshine by definition “get away with unlawful conduct.”
Until government vigorously prosecutes its own officials’ crimes as the ordinary event instead of the rarest exception, mandatory sousveillance of officials as advocated by Arthur Kirkland above is at worst a measure that will reduce some of that criminal behavior. It will at least provide ordinary citizens a few clues about officials’ criminal behavior as it occurs, so that citizens can make political decisions based on that knowledge.
February 4, 2010, 11:40 amOren says:
Yes, but Federal Law requires them to do so through a court order that requires “specific and articulable facts”. That is, this is a case where the Court has declined to protect the data through the Fourth Amendment (see, e.g. Smith v. Maryland) but the legislature stepped in to implement policy. It’s almost (almost) like democracy works or something.
Also, I’m skeptical of how many customers care about “privacy concerns” for (a) roughly anonymous data and (b) data divulged pursuant to a court order. Of course, (b) is not within the power of the cell phone company to decide — compliance with the subpoena is mandatory.
As far as (a) goes, imagine the following experiment — a cell phone company offers you a service in which the company promises not to archive any more data than is necessary to provide service (including billing, emergency services and compliance with Federal Law) and to expunge that data as soon as it is both legal & practical. Let’s also assume for the moment that you trust they carry out this service faithfully. How much would the average consumer pay for such a service? I don’t know the answer, but I’d wager that at $5 a month you would see less than 5% adoption.
February 4, 2010, 12:05 pmChris Travers says:
I am pretty sure cell tower triangulation doesn’t work where I live because there is only one cell tower per provider. In an urban environment, however, if your phone is picked up on by three cell towers when it calls out, you could get a very accurate measurement of where the individual is (perhaps even the location within a closed structure such as a home or an office building). Not only do you get an intersection of three fairly narrow 15-degree spreads, but the relative strength should theoretically allow someone to pinpoint the cell phone quite accurately within that area. There are a few obstacles to doing this with perfect accuracy (some buildings’ reinforced concrete doesn’t pass cell signals well), but this is probably a problem only with a small minority of commercial and residential buildings.
In a lot of ways this is sort of similar to the beepers at issue in Knotts, except that:
1) Nearly everyone has them, and
2) there are sufficient detection mechanisms in large areas to provide very accurate triangulation including inside closed structures.
I am curious to know why folks wouldn’t see this sort of surveillance that reaches inside closed structures as requiring a warrant under the 4th Amendment?
February 4, 2010, 12:12 pmChris Travers says:
Just a note:
I suspect, but cannot prove at the moment, that with aggregated triangulation data, it would be possible to position a phone inside a closed structure where the structure partially interfered with cell signals. There would be a number of ways of doing this. Furthermore, I suspect it might be possible, with enough aggregated data, to place a phone within a few feet.
February 4, 2010, 12:27 pmChris Travers says:
Just a note:
I don’t think that blanket, pervasive surveillance in public space would be beyond the definition of a search. Until recently, there hasn’t been sufficient technology to make such possible and I think the fundamental interest of the 4th amendment would require drawing some limits there. My reasoning here is originalist: I think the fundamental concern was to prevent the government from initiating searches as fishing expeditions in order to find a crime to match the person. In colonial times this was accomplished via writs of assistance. I think the 4th Amendment should be read to preclude clearly overreaching public surveillance connected to domestic law enforcement as well. If the 4th Amendment can’t cover that, we need an amendment that does.
However, in this case, I think there is an easier line that can be drawn. Pen register programs, as Orin has pointed out before, fall into non-content categories, and are generally speaking “outside” protected interests of the 4th amendment where warrants would be required. His argument is that since CSLI is non-content it should also be outside, as a matter of category.
However, CSLI is different. It is used to place individuals inside closed environments where they have legitimate privacy interests under the most restrictive reading of the 4th Amendment. Thus, if CSLI is used to locate a farmer in his open field, that might not require a warrant, but when it is used to locate a farmer in his home, it might. I would therefore argue that Kyllo is a closer precedent than Smith because the nature of the question of what is protected is closer.
The problem is you can’t know in advance what the data will tell you. CSLI therefore should require a warrant. CSLI may be non-content, but it reaches inside protected areas.
February 4, 2010, 1:01 pmOren says:
… in this case the police must go in front of a neutral magistrate and justify themselves. In Knotts, the police did the whole thing without any judicial approval.
This is fanciful. Even in urban areas, the data are good only to a 1/4 block (250 foot radius). Thus, it would reveal in which building you (as in Knotts) are but no private facts about where in that building.
There is no need to speculate about this, by the way, just grab any phone that has a way to disable the GPS and go into Google Maps. I just did so with my phone and got a location to within a few buildings but certainly not enough to reveal any private fact that one could not figure out from outside.
February 4, 2010, 1:03 pmChris Travers says:
You are assuming a few things about the data retrieved by the federal government, and you are assuming certain things about the specifics of the location being detected. Not all environments are equally imprecise. In some office buildings (though only rarely), you even have indoor mini-cell-towers (I worked in an office where this was considered at one point). There is no reason why a high-rise appartment building with cell reception problems couldn’t take a similar approach.
Where I live, 250 feet is far more accurate than one can get from the cell site info.
And, according to the EFF, the magistrate is certainly allowed to say, “apply for a search warrant instead” if there is any legitimate question as to the nature of data retrieved. Do you disagree with the EFF on this?
I think the EFF’s brief poses a workable (though IMO suboptimal) approach in that it allows the magistrate to look at the facts involving the case and decide whether a 4th Amendment interest is potentially violated, and if so, require a search warrant. I could live with that.
February 4, 2010, 1:41 pmOren says:
I don’t think this is factually true — CSLI is 250 feet radius, not enough to discern useful information about where in a protected area you are.
Even if it were, would you object to rewriting the statute to force the provider to “smudge” the data to 250 feet — i.e. to divulge in which buildings the person was but nothing more (recall that information on which building can be observed from the street and is thus not even a search)? I think your claim is much more broad than that (especially reading it in parallel with your opposition to a hypothetical traffic-camera setup that, by construction, does not act on protected areas).
Just curious why an amendment and not statutory provisions. Congress can require a higher standard of proof than it has — all the way up to probable cause if they desire. The fact that they have chosen to protect the records at all, when the 4A does not require it, is at least proof they have some interest in restraining the police.
February 4, 2010, 2:28 pmOren says:
(1) I’m fairly familiar with cell phone technology, so unless the government is supplementing the triangulation data with something else that allows it to be refined (what, I’m not sure), it’s 250 foot radius at best. Even a microcell of the smallest variety (the kind you can get for your home for $50 such as the Sprint Airave) tell you the customer is within 200 feet of the base station.
(2) If the cell phone triangulation data were sufficiently detailed to reveal private information (e.g. which room of a house), I would gladly revise my opinion. In that case, either the data would have to be smudged or the warrant would have to be supported by probable cause.
I do, not because that would be bad policy but because it is not how the statute is written.
I can live with that too, provided his analysis of 4A is within the corners of Smith v. Maryland. In fact, I would require it of him.
It just so happens that I think it is impossible to find a violation of a 4A interest as understood through current doctrine, although, as Orin frequently reminds us, the 4A is a rather fact-intensive thing so perhaps some requests might fall afoul of this. A “normal” request (e.g. what location data do you have for John Doe on between 8AM and 11:30AM 1/3/2009) seems to me too uncontroversial to require any serious analysis.
February 4, 2010, 2:43 pmFurshlugginer says:
That sounds like extortion to me, same as with non-published (unlisted) numbers. Although I would pay the $5/month for that service, and do pay $3/month for an “unlisted” number, I think the hypothetical charge would be just extortionary as the real one, abetted by obviously bogus PUC rulings. The “unlisted number” fee was originally justified as compensation to the telco for the excess 411 calls it handled as people tried to get numbers which weren’t in the white pages. However, since the telco now recovers all costs (and some profit too) by charging for 411 calls (with PUC approval), the old justification for the “unlisted number” monthly fee has evaporated. Since it obviously costs the telco less to print thinner white pages, the “unlisted” fee is pure unjust enrichment. Since keeping unneeded location data obviously costs the telco more than discarding it, any “blackmail” charge you might be asked to pay to get the telco to stop tracking you would be unjust as well.
(Since someone pointed out that the original impetus for cell carriers to install costly and unnecessary location tracking systems was an FCC rule, let me point out that it was understood at the time that the FCC rule’s ostensible justification (911 location) was bogus– that rule was prompted mainly by law enforcement spying desires.)
February 4, 2010, 2:51 pmChris Travers says:
How so? Where do you think the EFF is misparsing the statute?
The EFF is arguing that the use of the word “only” would be surplussage if we read it to require issuing a warrant under those circumstances. Why do you think they are wrong?
February 4, 2010, 4:35 pmChris Travers says:
Also re: accuracy of triangulations and femtocells…
1) I suspect (though at the moment cannot prove) that techniques which roughly parallel synthesized aperture signal processing (the sort used by synthesized aperture radar) might be able to, with a sufficiently large data source, correct location data and make it quite accurate. This would probably depend on a fair bit of processing and a very large set of data (and a lot of data from each of a lot of phones). It would be a lot of work, and it would require a large data set, but I would bet one could get remarkably accurate. Consequently, one key issue might be whether mere raw measurements were provided to law enforcement or whether the cell phone carriers corrected that data in any way.
2) One of the main applications regarding femtocells is where construction of a building is such that it blocks cellular communications. In the event of, say, a high-rise apartment or an office building, I don’t think it would be a safe assumption that all signals get out as far as 200 ft. It might indeed require multiple femtocells in a much smaller area, and installation records might be enough to make that determination.
I used to work in a building which had issues of this sort. One proposal discussed was to bring femtocells into each floor of the building. Given that only DAMPS cell phones worked in the building, the plan was scrapped in favor of requiring the management to switch cell phone carriers….
February 4, 2010, 5:03 pmOren says:
Extortion requires (fundamentally) a predicate unlawful act — the cell companies collecting and selling the data is perfectly lawful.
Except that this data has considerable resale value, as discussed above. If the company can resell it for more than it costs to collect (and given that such collection is lawful)
Accepted as true, for the purposes of argument, but so what? Law enforcement demanded the implementation of a technology to assist law enforcement — dog bites man.
On second reading, they might have a decent argument here. The statute vest the discretion in the magistrate to approve the request based on less than probable cause but does not require him to do so.
On the other hand, there is no indication from Congress on when the magistrate ought to do so (except for voluminous records that would be burdensome for the telco), so it’s a bit of a stretch to say he should do so for particular reasons — especially when it suggests reasons I don’t think are very good e.g. 4A concerns when precedent clearly indicates that it is not even a search, let alone an unreasonable one.
On the whole it’s odd — the magistrate is empowered to refuse the request but not instructed on when and why … Certainly this bears more thought.
February 4, 2010, 6:17 pmOren says:
So, the first time I read it, I skipped to part III assertion that CSLI is a search under the 4A which I believe is incorrect under precedent (or, at the very most, woefully devoid of a factual record to support the claim).
You were quite correct to point me back to part II and the discretion vested in the magistrate to deny the application even if it meets the standards of part (d). As I said before, I don’t know what this gets you because the plaintiff suggests using that discretion to vindicate a 4A right I don’t believe exists.
February 4, 2010, 6:38 pmFurshlugginer says:
Uh, Oren, it’s a criminal offense for a telco to sell or give away location info or call detail info to anyone other than “law enforcement” without the subscriber’s consent. So who cares if the telco could make a lot of money that way? Banks could make a lot of money selling the contents of customer’s safe deposit boxes too, but that doesn’t mean they have a right to do so.
http://assets.en.oreilly.com/1/event/25/Navigating%20Privacy%20Laws%20when%20Developing%20and%20Deploying%20Location%20Tracking%20Applications%20Presentation%201.ppt
February 4, 2010, 9:14 pmOren says:
The document clearly says “Can use customer info in aggregate form“.
Carriers are perfectly free (and already do) sell anonymized bulk information. Federal privacy law only prohibits selling personal or personally identifying information.
February 5, 2010, 12:56 amFurshlugginer says:
Yeah, Oren, but your hypo was telco charging subscriber an extra (blackmail) fee to avert disclosure! You can’t have it both ways.
February 5, 2010, 1:07 amOren says:
I don’t understand your point.
The normal procedure to is to scrub the data, aggregate it and sell it — which is legal so long as the information is not individually identifiable. Why would offering people the option to opt-out change the analysis here?
Would it help if I restructured it to be a ‘discount’ for opting-in? Can a company break the law by offering a discount for consumers that enter an otherwise-legal program?
February 5, 2010, 12:19 pmChris Travers says:
Indeed, maybe the company can offer a discount if you provide them blanket permission to sell identifying data. The current laws do allow telco’s to give out any info with the customer’s consent (though I think the consent has to be in writing).
Better yet, what if the telco gave the customer 5% of the revenues from selling his information…..
February 5, 2010, 2:12 pmOren says:
No, that would be prohibited by law. I was talking about a discount for allowing them to do something that is completely lawful.
This discussion has been quite polite but I really don’t see the point of this comment. I reread my comment and it was quite clear that I was referring to a discount for opting in to a lawful disclosure. I wrote, precisely:
Offering a discount for opting into an already-unlawful program was a straw man of your own invention (or perhaps you misread my statement, but I find that hard to believe).
If you don’t mind, can we get back to my question before Fub — what do you think I should make of the statute if I concur entirely with Part II of the EFF’s brief (that the magistrate may compel disclosure if section (d) of the statute is met) but I reject part III (that doing so would be the way to maintain the 4A protection of CLS).
February 5, 2010, 4:09 pmChris Travers says:
Really? I thought “prior consent” was a safe harbor. Now, that consent has to be in writing, but if a discount was offered and to qualify you had to mail or fax a signed consent form, I don’t see how that would be a problem. Basically there you have a contract, with due consideration, allowing them to sell the information and to compensate you for it.
Would you think this would be different if royalties were paid to the customer on the basis of sold information?
February 6, 2010, 12:31 pmEdward says:
The personal economic value of privacy is an asymmetric problem. The public will not pay for privacy–in most cases they assume their information is kept confidential. They are not examining the issues and consciously choosing to expose information. If they did, we would not constantly see folks releasing private information on Facebook.
Folks will stop doing business with any firm that violates their trust. Customers won’t pay an extra $5 a month to keep their cell phone data private, but will cancel their $80/month service if their privacy is invaded. (And presumably start using another carrier.)
This is why some business marketing data is not fully exploited. Businesses are not sure exactly where the customers believe the public/private line exists. Marketing departments know that they can legally use the data they collect in thousands of ways that place that line firmly in the public. But customers don’t know that. And the firms often don’t want to risk the $80 revenue in exchange for another dollar in profits.
February 7, 2010, 1:55 am