Over a decade ago, Nancy Kim and I wrote an article, contrasting our attitudes as consumers when we share our data with private companies with our attitudes as citizens when those companies share that same data with state actors. People were outraged by the Snowden revelations, but there is no similar outcry when private companies engage in data mining, even though the two intrusions on privacy are equally severe and the former is in the service of national security while the latter serves less compelling commercial interests. One of our concerns was that courts were binding consumers to terms of service (ToS) that empowered the Internet giants and telecommunications giants to share customer data with government actors. We argued that consumers did not give meaningful consent to such use of their data and that such data should not be shareable unless the government can overcome Fourth Amendment barriers to unlawful searches.
Once again, listening to Con Law podcasts has paid off. I hadn’t paid much attention to the recent U.S. Supreme Court (SCOTUS) decision in Chatrie v. United States until the latest episode of What Trump Can Teach Us About Con Law, with Elizabeth Joh and Roman Mars. Although SCOTUS did not acknowledge our influence, part of the decision in that case follows from our reasoning. That is, like us, the Court is skeptical that people, in sharing information with service providers, agree that the information should be made available to police agencies without satisfying Fourth Amendment requirements or lawful searches.
Chatrie is about the government’s use of so-called “geo-fence warrants” that enable the government to collect information about cell-phone users’ locations. Simplifying matters a bit, in the case at issue, somebody robbed a bank. Video footage caught the perpetrator using a cell phone just before the robbery. The footage did not enable the police to identify the suspect, so the government asked Google to share location information from all of the cell phones in the area at around the time of the robbery. Once in possession of that information, the police were able to identify the suspect.
SCOTUS determined that the geo-fence warrant was a search, because we have a reasonable expectation of privacy with respect to our cell-phones’ locations. The police thus intrude on a constitutionally protected interest when they request information related to that location, even for a limited time, from a third party. I think that’s a great outcome, and once again, kudos to Justice Kagan (below) for getting four other Justices to sign on to a decision that helped criminal defendants. She also did so in Hunter, where she assembled an even broader coalition of Justices, as we blogged about here. The case may have very broad implications, but as to Mr. Chatrie himself, some tough issues remain. The Court remanded the case for the Fourth Circuit to determine whether the geo-fence warrant satisfied the Fourth Amendment’s requirement of probable cause. Even if it did not, the Fourth Circuit may affirm the District Court’s invocation of the good faith exception to the exclusionary rule. That is, the evidence obtained through the geo-fence warrant could come in, even if the search was constitutionally impermissible.
Don’t get me wrong. I like the outcome of this case, and I think it strikes a powerful blow in the struggle to protect the scraps of privacy that remain to us. I have two criticisms: one empirical; one doctrinal.
First, the empirical problem: Justice Kagan’s test for determining when a reasonable expectation of privacy arises is rather subjective. If you asked anybody under the age of forty, and probably a lot of older people as well, whether they thought that the information that they shared with companies like Google was private, I think they would shrug and concede that not much is really private anymore. Who knows? Perhaps I am lacking in common sense, but I would have more confidence in Justice Kagan’s finger-in-the-wind, common-law constitutional adjudication if she could cite to some empirical research on people’s expectations of privacy in connection with information shared with private service providers.
A colleague who actually teaches in this area pushed back, arguing that the “reasonable expectation of privacy” test was intended to be more normative that empirical. The Court is more or less asking whether we want to live in a world in which the government can intrude on certain activities that human beings ought to be able to do in private. He also conceded that the Court is not entirely consistent in its application of the test. I suppose I think public attitudes towards privacy ought to inform the Court’s normative intuitions.
Now on to the doctrinal problem. What really jumps out at me — and this part is situated at the junction of contracts law and constitutional law where I have been camped out for much of the last decade — is that Justice Kagan too breezily rejects attaching any significance to the fact that turning on Google’s geo-location services is a choice. This is not the first time in recent years that the Court, confronted with a contractual obligation, has wasted little or no time weighing contractual obligations against constitutional rights or interests.
I wrote about this in connection with Fulton v. City of Philadelphia here. There, the Court vindicated Catholic Charities Free Exercise rights to discriminate against married same-sex couples notwithstanding non-discrimination policies embedded in a Philadelphia ordinance and in the contract which Catholic Charities signed when it agreed to provide services to the city. I wrote about it again in connection with Austin v. U.S. Navy Seals here. That case involved Seals who wanted to opt out of vaccines without any effect on their service despite their contracts with the military that specifically provided that they would have to be vaccinated according to the military’s needs. The case was decided on different grounds; the contract never came up. I wrote about it again in a post about Mahanoy Area School District v. B.L. That case was about a cheerleader who had clearly violated her undertaking not to post disparaging things about her school or about cheerleading on social media. The Court gave no weight to that undertaking in finding that the cheerleader could not be disciplined for her protected speech. In those cases, contracts rights were not weighed against constitutional rights. Matters work out differently when the constitutional rights are the Seventh Amendment right to a civil jury and the contract rights relate to arbitration.
I concede that in this case, I’m not entirely sure that opting in to Google’s Location History service gives rise to a contract. It sure seems like Google makes an offer by asking you to turn Location History on, which you accept by doing so. In any case, Justice Kagan’s opinion speaks of other apps, which do require users to agree to terms and conditions, and all indications are that she would treat them the same way.
This case more-or-less follows from the Court’s decision in Carpenter v. United States. That case involved the police accessing seven days of historical data about his location based on data shared with his cell-phone provider. Here, the data shared covered a much shorter period of time, but the Court did not find that distinction to eliminate the privacy concerns.
The government argued that this case is distinguishable from Carpenter. While in Carpenter the police made use of information that is automatically shared with cell-phone providers, here Mr. Chatrie opted into a service. Only one-third of cell-phone users opt in to the Location History service. While a cellphone may be indispensible in modern life, Location History, it seems, is not. It follows that the third-party doctrine, according to which information voluntarily shared with third parties can be turned over to the government without violating the Fourth Amendment, is implicated in Chatrie in ways it was not in Carpenter.
Nonetheless, Justice Kagan focuses on how insistent Google is in trying to get users to sign up for its Location History service and how little information Google provides about the extent to which information is being shared with and retained by Google. Moreover, this is not the only app on phones that gathers and stores personal information. Using those apps also involves choices. Justice Kagan imagines that if the Court found that a geo-fence warrant is not a search, the third-party doctrine could become a pathway to permitting the government to collect information through service providers in myriad ways.
This is all great stuff, and I think it is relevant to determining whether and under what conditions, consistent with the Fourth Amendment, the government should be permitted to access information shared with third parties through apps. I just don’t think a “reasonable expectation of privacy” test helps us sort through the mess. Moreover, this situation seems quite similar to a lot of other situations in which people sign up for services and terms and conditions attach. With most apps that we use, we have to agree to terms of service (ToS), and those ToS often put us on notice (if we were to read them) that by using the service we share information with the provider. Courts usually enforce ToS, absent unconscionability. The Court does not explain to my satisfaction why people are treated as consenting to ToS in most contexts but are not treated as consenting to the consequences of opting in to Location History.
In addition, it’s not just that we are sharing information about our locations with Google. We are sharing information with Google in order to get something in return. That something may be driving directions or the location of the nearest Starbuck’s. It may be an advertisement for a shop we are about to pass or a warning that a potentially dangerous criminal has been spotted nearby. Even if we didn’t know much about the kinds of information that Google was getting about us when we signed up for Location Services, we certainly could deduce it from the kinds of information we get from our phones all the time. And if we found that invasive, we could opt out of Location Services.
It’s great that the Court is willing, in this case, to look beyond objective manifestations of consent when they result in a surrender of a right to privacy. It would be better still if courts were willing to look more carefully at all of the ways in which vendors delete rights through form contracts, as Peggy Radin documented in Boilerplate. Perhaps contractual consent should not be treated as a toggle switch turned on with a click of the mouse. Perhaps more contextual analysis is necessary. I think the Court stumbled on to the right outcome here. It could provide clearer guidance going forward by showing the work. That is, what is it about this situation that makes the Court think that users have not consented to particular uses of their data. Is the problem notice? Is the problem foreseeability? Or is there something particular to the Fourth Amendment that makes the Court think that one simply cannot consent to searches of this kind, even though one can, for example, waive the Seventh Amendment right to a civil jury in like circumstances?
Recent Comments