GPS Surveillance and the Fourth Amendment: Thoughts on United States v. Jones

In United States v. Jones, FBI agents installed a GPS tracking device on Jones’ car and monitored where he drove for a month without a warrant.  Jones challenged the warrantless GPS surveillance as a violation of the Fourth Amendment.  The D.C. Circuit agreed with Jones.

On its face, the D.C. Circuit opinion appears to clash with the Supreme Court’s decision in United States v. Knotts, 460 U.S. 276 (1983), where the police installed a much simpler tracking device (referred to as a “beeper”) to a person’s car.  The Court concluded that the Fourth Amendment did not apply to the beeper because a “person traveling in an automobile on public thoroughfares has no reasonable expectation of privacy in his movements.”

The D.C. Circuit distinguished Knotts because there, the Supreme Court noted that the surveillance was limited and explicitly noted that more pervasive surveillance might be treated differently.   As the Court in Knotts stated that “if such dragnet-type law enforcement practices as respondent envisions should eventually occur, there will be time enough then to determine whether different constitutional principles may be applicable.”

The Supreme Court will determine if it agrees with this theory.   The Court has long held that there is no expectation of privacy in public.  This view has been frequently criticized as failing to recognize that people enjoy much practical obscurity in public and pervasive monitoring will dramatically undermine this obscurity.  Will the Court revisit its view about the lack of privacy in public given the changing capabilities of technology?

If the Court were to conclude that the Fourth Amendment required a warrant for GPS surveillance, it would have to define a coherent line between when a person in public has an expectation of privacy and when a person lacks such an expectation.  Such a line would be challenging to draw.

If extreme enough, quantitative differences can become qualitative differences.  But how does one articulate this into a workable approach in the law that isn’t too vague and mushy?

I was speaking with Peter Swire recently, who said that in order to convince the Court to draw such a line, it would be helpful for scholars to propose a workable test or approach for the Court to use.  I think he’s right.  So here’s my stab at it.

The Fourth Amendment applies to a surveillance technology used in public if the surveillance technology: (1)  extends significantly beyond human capabilities; and (2) is used in a manner beyond its ordinary use by the general public.

Let’s see how this might work with a few examples:

1. Flashlight.  A flashlight provides illumination where ordinarily humans might not be able to see.  But its ordinary or intended use is to allow people to see things in the dark, so it wouldn’t be covered by the Fourth  Amendment.

2. Thermal Sensors as Used in KylloThe use of thermal sensors to detect heat patterns in the home would be regulated by the Fourth Amendment.  My approach would create a more sensible rationale than that used in Kyllo v. United States, 533 U.S. 27 (2001) (whether the technology is in general public use).  Thermal sensors are in general public use, but they are not ordinarily used to spy into people’s homes.

Applying this approach to GPS, the technology can track people much more pervasively than regular stakeouts or  following them around.   It thus extends significantly beyond human capabilities.  The general public does not use GPS as a way to track people’s movements.   They use it as a way to find places and navigate while driving.  Therefore, the police use of GPS to place a person under surveillance would be covered by the Fourth Amendment.  Knotts could be distinguished because the beeeper wasn’t as pervasive and thus could be argued not to have extended significantly beyond human capabilities.

I don’t agree with the reasonable expectation of privacy approach to the Fourth Amendment, but if the Court is going to stick with this approach and try to fit GPS surveillance within its existing caselaw, then the above test will hopefully be consistent enough with the Court’s caselaw and more coherent than the tests previously articulated by the Court.

You may also like...

18 Responses

  1. Doug C says:

    Nice concise test. Thanks for sharing.

  2. Chris Soghoian says:

    Could you expand a bit on your argument that GPS tracking isn’t used by the general public?

    GPS tracking devices can be purchased on eBay and on surveillance gear websites. Likewise, stalkers and private investigators frequently use them (although, the legality of this is iffy).

    I’m sure you saw the recent Stingray article in the WSJ – those devices are 150k a pop, and only sold to law enforcement. In comparison, GPS trackers can be purchased for a hundred bucks.

  3. Larry Rosenthal says:

    I am afraid that I find the proposed distinction of Knotts in the post quite unpersuasive. In the real world, it turns out that without a beeper, it is quite difficult to track a vehicle, especially if the target is surveillance-conscious. Using a beeper to track the movements of a vehicle, and even more clearly to track the location of a container being transported in a vehicle as in United States v. Karo, accordingly “(1) extends significantly beyond human capabilities; and (2) is used in a manner beyond its ordinary use by the general public.” I suppose I should add that a test that could plausibly support either result in cases like Knotts and Karo probably isn’t much of a test.

    Larry Rosenthal
    Chapman University School of Law

  4. Daniel Solove says:

    Chris — My test isn’t general public availability. Nor is it the illegal use of the technology. It is ordinary public use. The key is to look at the legal and common uses of the technology by the general public, not the availability and not any possible use.

    Larry — I’m not a fan of Knotts, and nor am I a fan of the reasonable expectation of privacy test. But my goal is to come up with a plausible and workable approach that allows the Court to articulate some set of principles to draw a line where surveillance in public with sense-enhancement technology will trigger 4th Amendment protection. I believe there is a significant quantitative difference between the Knotts beeper and GPS technology.

  5. Bryan Gividen says:


    I think you’re right that there is a “significant quantitative difference between the Knotts beeper and GPS technology,” but I don’t think that actually addresses Prof. Rosenthal’s critique. It seems you’re arguing that your approach allows you to uphold Knotts and Karo and avoid peeling back precedent. But the test would seem to squarely confront the Knotts decision. Your response seems to say that GPS is much more than a beeper which is why Jones would fail the test, but Knotts wouldn’t; really the question ought to be, what’s the comparison of the beeper to human capabilities. So while a GPS would certainly be not included, you haven’t shown how the beeper would be treated under the first prong.

    Even given the first prong, I think Knotts would certainly fail on the second prong. Any showing that the general public used a beeper in the way law enforcement did would be a huge surprise to me.

    On the other hand, I could be misreading your post and response. You might be arguing the Supreme Court should explicitly overrule Knotts and then adopt this standard of thinking. In that case, I think you have a workable standard, albeit a mushy one.

  6. Daniel Solove says:

    Bryan — My test is that BOTH prongs must be established, and the distinction between Knotts and Jones would turn on prong 1. If we got to prong 2, I agree that beepers would not be ordinarily used for the public to track people. On prong 1, the question is whether GPS surveillance is different from beepers. Limited tracking of a person’s movements in public would be within human capabilities. The police could follow a person around for a while. But GPS surveillance is much more sophisticated and advanced.

    So the key would be how readily the surveillance via beeper vs. GPS could be replicated by the police doing it without the aid of the technology.

    The goal of the test is to do the following:

    1. The first prong looks to see if the technology is significantly enhancing what the police can do. If it is only amplifying to a moderate degree, it doesn’t trigger the Fourth Amendment (if in public). If the first prong isn’t met, then there’s no proceeding on to prong 2.

    2. The second prong recognizes that there are some technologies that significantly alter human capabilities but nevertheless are really an extension of ordinary life. Flashlights are an example of this. This prong singles out technologies that have a material difference in what would be conduct acceptable by the general public. Instead of general public availability of the technology, the focus here is on the *purpose* of the uses of the technology. When the police use a technology for new uses beyond its commonly-accepted purpose, that’s when Fourth Amendment protection would kick in — but, of course, only if prong 1 were satisfied too.

  7. Bryan Gividen says:

    I follow, and excuse the misstep on my assertions about public use of beepers (in other words, looking at the test as disjunctive). Following up on that point and on your response to Chris, what utilization would rise to “beyond its ordinary use by the general public?” For instance, some parents use GPS trackers to follow their children. (See or apps installed on SmartPhones to track it.) Would Jones be “beyond the original use by the general public” because (a) the most prevalent use of a GPS device is to navigate so that type of use overshadows the amount who use it as a tracking device, (b) not enough of the public use it as a tracking device for it to qualify as used by “the general public”, (c) the qualitative difference between the parent-child relationship and a police-suspect, or (d) some combination of those?

  8. shg says:

    The first prong, “significantly,” is a bit mushy and will lend itself to mischief. The second prong, given the way tech advances and pervades society (at least early adopters), will likely have a shelf life of a couple of years before any particular type of tech either comes into common use or is replaced by something new, rendering it old news. Either way, by the time a court decides, it’s likely the tech will no longer be in the same status as it was when the case arose.

    I would go with all sense-enhancing tools north of a flashlight require a warrant, but then, I might be biased. On the other hand, given the speed and modes of communication available today, no reason why police can’t get a warrant nearly instantaneously under almost any circumstances. So let’s rid ourselves of the facile exceptions and thank technology for allowing us to return to the rule.

  9. Joe says:

    “The Fourth Amendment applies to a surveillance technology used in public if the surveillance technology: (1) extends significantly beyond human capabilities; and (2) is used in a manner beyond its ordinary use by the general public.”

    So, if GPS is used by the general public (“good way to keep track of your teens who drive”), it would not be covered?

    I welcome the discussion but the 4A is the sort of thing that is likely not to be determined by crystal clear rules but some sort of balance (that leads to the test the author doesn’t like) to determine what “reasonable” means.

    And, a beeper on that level is much less so than 24hr GPS surveillance, even if even the beeper was a questionable case. It is hard to do, but you can kinda follow a car, but it takes much more effort (and this is part of what my test would be) to imagine keeping track of the the person around the clock. This also is a matter of reach, including personal information, including stopping at various locations in a layer of detail a beeper won’t do.

  10. Joe says:

    Also, the second prong doesn’t really address the “capabilities” and different interests of the government. The government singling out a specific person for a search is different than a private party. It is different if a parent or even friend tracks someone with GPS than the government. This is why the government has standards in place for things like web searches or visits to churches/mosques or such that the public can do if desired, but the dynamics are different.

    I recognize the test reflects (problematic) case law, but at some point, the reach of the the government matters here. As more and more information is easily obtainable, some sort of private information test should be factored in when allowing government action here.

  11. Jim Maloney says:

    Your test (“The Fourth Amendment applies to a surveillance technology used in public if the surveillance technology: (1) extends significantly beyond human capabilities; and (2) is used in a manner beyond its ordinary use by the general public.”) appears to have come pretty close to having been used by Judge Lippman and his non-dissenting colleagues in deciding People v. Weaver, 12 N.Y.3d 433, 909 N.E.2d 1195, 882 N.Y.S.2d 357 (2009), q.v. A couple of excerpts (“tracking” the test’s subparts):


    [1] Here, we are not presented with the use of a mere beeper to facilitate visual surveillance during a single trip. GPS is a vastly different and exponentially more sophisticated and powerful technology that is easily and cheaply deployed and has virtually unlimited and remarkably precise tracking capability. With the addition of new GPS satellites, the technology is rapidly improving so that any person or object, such as a car, may be tracked with uncanny accuracy to virtually any interior or exterior location, at any time and regardless of atmospheric conditions. Constant, relentless tracking of anything is now not merely possible but entirely practicable, indeed much more practicable than the surveillance conducted in Knotts. GPS is not a mere enhancement of human sensory capacity, it facilitates a new technological perception of the world in which the situation of any object may be followed and exhaustively recorded over, in most cases, a practically unlimited period. The potential for a similar capture of information or “seeing” by law enforcement would require, at a minimum, millions of additional police officers and cameras on every street lamp.

    [2] It would appear clear to us that the great popularity of GPS technology for its many useful applications may not be taken simply as a massive, undifferentiated concession of personal privacy to agents of the state. Indeed, contemporary technology projects our private activities into public space as never before. Cell technology has moved presumptively private phone conversation from the enclosure of Katz’s phone booth to the open sidewalk and the car, and the advent of portable computing devices has resituated transactions of all kinds to relatively public spaces. It is fair to say, and we think consistent with prevalent social views, that this change in venue has not been accompanied by any dramatic diminution in the socially reasonable expectation that our communications and transactions will remain to a large extent private. Here, particularly, where there was no voluntary utilization of the tracking technology, and the technology was surreptitiously installed, there exists no basis to find an expectation of privacy so diminished as to render constitutional concerns de minimis.


    Admittedly, the NY Court of Appeals did not parallel the second part of the test exactly, but instead tied it into the “reasonable expectation of privacy” standard (which I think we’re going to be stuck with given the nearly foundational status of Katz and given further the Fourth Amendment’s use of the logical opposite, “unreasonable.”) Still, they leaned toward part (2) pretty heavily…

    Also, the NY Court “moshed the tech” just a bit: using GPS to find one’s way around is not the same as using it to track someone else. But that raises an interesting point, touched upon by other commentators above: If an eavesdropping/info-gathering technology falls into common use, does that make it OK for police to use? NJ’s Supreme Court said no after the Supremes said yes. In California V. Greenwood, 486 U.S. 35 (1988), it was pointed out that children and small animals get into people’s curbside garbage, so why not the cops? But NJ, in State v. Hempele (1990), said that they held their own state-empowered police to a higher standard than that applicable to children and small animals…

  12. Daniel Solove says:

    Bryan — In response to your point that some parents use GPS to track their children, under the law, parental consent is typically deemed to constitute children’s consent, and the children are thus consenting to the GPS tracking.

    When the case doesn’t involve children or situations with consent, it gets quite creepy and troubling. Suppose a parent used GPS to track a college-aged child without consent. That would seem quite beyond commonly-accepted practices and norms.

  13. discretionary docket says:

    The test should be: “Is the surveillance replicable by traditional means?” If traditional human grunt work — like stakeouts — can replicate the surveillance, I see no reason why there should be a reasonable expectation of privacy. Are sophisticated, elaborate stakeouts different from less meticulous ones for Fourth Amendment purposes? No. The degree of laboriousness doesn’t enter into it. The GPS technology is essentially a labor-saving device in this context.

  14. Orin Kerr says:


    I’m not sure your test means anything, because it rests on a set of arbitrary classifications that anyone can manipulate to reach any result in any case. The two biggest problems are classifying “the surveillance technology” and classifying “the manner of use.”

    The problem is that fact patterns involving technologies do not themselves provide these classification. These are both arbitrary classifications that you are coming up with when you purport to apply your test. But this means that your test isn’t doing the real work, as you can always just draw the category of the “technology” or “manner of use” either broadly or narrowly to either include or exclude any fact pattern and say anything is or is not a search.

    Take the case of a flashlight. Say the police use a police-model LED flashlight to illuminate a public street at night in a search for a serial killer. If you want to say that it’s use is NOT a search, you just say that the technology is a flashlight and its use is to illuminate a street. That’s just what the public does all the time, so no search. If you want to say that it’s use IS a search, you say that the technology is a very special police-model LED flashlight, and its use is to illuminate spaces in searches for serial killers. That’s beyond the ordinary use of the public, so it *is* a search.

  15. Joe says:

    So, we are back to reasonable expectation.

    How about if a spouse uses GPS? Creepy?

    Or, if the “child” lives at home and is supported by the parents but is over 18 and has given the parents some trouble. So, they track them with GPS to ensure they aren’t doing something troubling. Again, creepy?

    I think various things the average public does can be creepy at some point. Is this another factor in the test?

  16. walter james mcintosh says:

    It is worth noting that these rulings apply only to government actors. Let’s take the example someone involved in a minor auto accident who then claims they are not able to work . Is there anything to prevent the loaner car provided by the body shop being fitted with GPS and then having insurance investigators following the person about to check/film activities incompatable with their ‘unable to work’ claim ? Would that evidence be acceptable to a court in criminal proceedings – I would think so .

  17. Orin — In my perfect world, I’d wipe the slate clean of nearly all of the Court’s jurisprudence regarding the reasonable expectation of privacy test. I think it is incoherent and silly. But my goal in this post is not to try to reformulate the Supreme Court’s jurisprudence to make it sound — if that were my goal, I’d demolish it all and start anew!

    Rather, my goal is to articulate an approach, consistent with the Court’s jurisprudence, that provides a plausible way the Court can draw lines. As you and I agree, the public availability test for new technologies is a failure, as thermal sensors in Kyllo are publicly available. So my goal is to jettison this test and find one similar yet a tad bit more plausible. My approach focuses not on availability but on use and norms. It allows the Court to draw lines, which the Court might do in this area if it had a plausible way of doing so.

    My test is no more vague that most Supreme Court tests — indeed, I think you’re holding out a standard that most Supreme Court tests wouldn’t meet. All the time we see tests involving “reasonableness” or “totality of the circumstances” or “community norms” or “significant government interest” or “narrowly tailored.” These aren’t very precise and your critique of my test can (and often has) been applied to these tests.

    The purpose of my endeavor here is not to demolish the caselaw and provide what I think would be the best approach. It is to provide an approach that will allow the Court to draw lines, and will focus the inquiry on something that is more sensible than availability of the technology. And if you don’t like this approach, then what approach would you recommend that: (1) can remain consistent with the Court’s jurisprudence; (2) does not make radical departures from existing caselaw or the reasonable expectation of privacy approach; (3) moves beyond looking to public availability; (4) allows the Court to draw lines (rather than a simplistic rule that ‘if you’re in public, then anything goes no matter what’)?

    Better tests may certainly be formulated, but I’ve yet to see them. Typically I see the following types of proposals: (a) follow the simple ‘no privacy in public ever’ approach and make no attempt to draw any lines; or (b) argue that the Court’s caselaw needs a radical overhaul. I usually argue (b). But in this post, I wanted to advance something that is neither (a) nor (b). Can any of the critics of my approach suggest an approach that is neither (a) nor (b)?

  18. Jim Maloney says:

    “Can any of the critics of my approach suggest an approach that is neither (a) nor (b)?”

    Sure. Recognize that the reasonable expectation of privacy approach to the Fourth Amendment, in the context of emerging technologies, will inevitably invite overreaching by police (i.e., warrantless searches using the new technology) with each technological innovation, and accordingly articulate a standard to the effect that any new surveillance technology that extends significantly beyond human capabilities (part 1 of your test) requires a warrant for its application unless and until its use for surveillance without a warrant is “clearly established” (i.e., by a Circuit or Supreme Court case holding that such use is permitted). (This is just the “flip side” of qualified immunity jurisprudence as it exists today to protect police from citizens, so why not employ it to protect citizens from police? And the Supreme Court articulates such standards all the time, e.g., Miranda warnings.)