Remarkable New Fifth Circuit Decision Limiting Cell Phone Searches

Remarkable but wrong, I think -- although, in my view, the correct result in this case for other reasons.


Here's a fascinating new Fourth Amendment case from the Fifth Circuit, United States v. Morton, that sharply limits cell phone searches.  According to the court, warrants must establish probable cause separately for each "category" of information in the phone in order to search it.

In Morton, the Fifth Circuit held that the government properly obtained a warrant for text messages, call logs, and contacts, but that the warrant did not establish probable cause to believe the evidence would be in the form of photographs.  The photographs were therefore suppressed, and the conviction was overturned.

I think the court's reasoning is wrong, although I happen to think the result in this particular case was correct for other reasons. Here's an overview of the case, and my concerns with it.

I. The Case and Its Reasoning

The government arrested Morton with drugs and obtained a search warrant to search his cell phone for evidence of drug possession. A search of the phone revealed photographs of child pornography.  This wasn't exactly a surprise, as the evidence found with Morton suggested he may be a pedophile.  But the warrant was for records of drug possession, not child pornography.

Having discovered the child pornography, the police stopped and obtained a second warrant—this time for child pornography.  The search pursuant to the second warrant led the police to discover over 19,000 images of child pornography on the phone. Child pornography charges followed, and Morton was convicted and sentence to a nine-year prison term.

In the new decision, the Fifth Circuit overturned the conviction because there was not enough information that the evidence of drug possession would be in photograph form to justify searching the photographs when the government searched the phone under the first warrant.  The child pornography had to be suppressed from the first warrant, and that left no basis for the second warrant, so the conviction was overturned.

According to the court, probable cause has to be established for each individual category of information on the phone.  The first warrant was proper in authorizing the government to search three categories of information —  text messages, contacts, and call logs—because there was probable cause to believe these kinds of records would be evidence on the phone.  Morton got his drugs from somewhere, and it made sense that there would be evidence about the buying of drugs in his call logs, contacts, and text messages.  So that part was fine.

On the other hand, the government could not search the photographs because there wasn't any particular reason to think that the evidence of drug possession would be in photographic form.  The evidence against Morton was of personal use, and there wasn't a particular reason to think that a user of drugs would have evidence of it on the phone.

Indeed, there wasn't even enough evidence to get over the Leon good-faith hurdle, the court rules. As a result, all the evidence had to be suppressed.

II.  The Court's Key Move—Probable Cause by Information Category

What makes Morton so important, I think, is its particular solution to the plain view problem in computer searches.  As regular readers know, I have been writing on this issue for 15 years.  The problem is this: Computer warrant searches are extremely invasive, and the government can use warrants pretextually to search for evidence without probable cause.  If everything on a computer can be searched and can come into plain view, how can you limit computer warrant searches to make sure they don't turn out to be general warrants in practice?

The Fifth Circuit's answer is to limit where the government can look.  The government can only look in places where probable cause has been established for that category of record.  Because there wasn't probable cause that the evidence would be in photograph form, the government couldn't look in the photographs.

Where did that holding come from?  Its first source was rather remarkable: At oral argument, the government seems to have conceded the point.

The Court: Do you say you're entitled to everything inside that phone so long as you can look at anything inside the phone?

The Government: No, your Honor.

The Court: Or do you need probable cause for each individual sort of category of information that could be found there?

The Government: That's correct.

But the court goes out of its way to make this point as a holding, not just a conceded point that it assumes to be true.  According to the Fifth Circuit, the probable-cause-by-category holding is required by Riley v. California:

Riley made clear that these distinct types of information, often stored in different components of the phone, should be analyzed separately. This requirement is imposed because "a cell phone's capacity allows even just one type of information to convey far more than previously possible." Id. at 394. Just by looking at one category of information—for example, "a thousand photographs labeled with dates, locations, and descriptions" or "a record of all [a defendant's] communications . . . as would routinely be kept on a phone"—"the sum of an individual's private life can be reconstructed."4 Id. at 394–95. In short, Riley rejected the premise that permitting a search of all content on a cellphone is "materially indistinguishable" from other types of searches. Id. at 393. Absent unusual circumstances, probable cause is required to search each category of content.

The court also suggests that each kind of record is its own constitutional "place," so that probable cause for each "place" must be established to satisfy the particularity requirement.  The court draws this in part from law review articles:

This requirement is especially important in the context of searches of digital devices that contain so much content. See, e.g., Adam M. Gershowitz, The Post-Riley Search Warrant: Search Protocols and Particularity in Cell Phone Searches, 69 VAND. L. REV. 585, 597–600 (2016); id. at 609 (noting that in drug cases, warrants frequently "authorize searches for photos and videos [on phones] . . . for which there is typically no probable cause"); Andrew D. Huynh, Note, What Comes After "Get A Warrant": Balancing Particularity and Practicality in Mobile Device Search Warrants Post-Riley, 101 CORNELL L. REV. 187, 190 (2015) ("The Court's lengthy discussion about the amount of personal information accessible on a modern mobile device suggests that a search warrant's particularity may be the next subject for scrutiny."); William Clark, Protecting the Privacies of Digital Life: Riley v. California, the Fourth Amendment's Particularity Requirement, and Search Protocols for Cell Phone Search Warrants, 56 B.C. L. REV. 1981, 1984 (2015) ("As the U.S. Supreme Court held in Riley, to allow the police unguided review of the entire contents of a cell phone when executing a search warrant would authorize the exact type of general warrants that the Fourth Amendment forbids.").

III.  Why I Think the Reasoning of Morton is Wrong

Is the court's reasoning correct?  I don't think so.  The Fourth Amendment requires probable cause as to each substantive kind of evidence, but not each form of evidence.  You can't know ahead of time what form a particular kind of evidence will take.  But the Fourth Amendment doesn't require probable cause as to form, only to substance.

To see this, imagine I told you that I have put evidence of my crimes—a secret code—somewhere in my cell phone.  At the same time, I pointedly refuse to tell you the form of the evidence.  Maybe it's the text of an e-mail, but maybe it's in a text message. Maybe it's in a note, or a comment attached to a stored contact.  Maybe I took a picture of it and it's stored as a photograph.  Maybe it's stored as a .pdf file.  Maybe it's a mix, like a photograph sent via text.  There are endless forms in which the evidence could be on the phone.

Do you have probable cause to get a warrant to search my phone in the above example? I would say absolutely, yes. The place where the phone is located—or maybe the phone itself—is the place to be searched.  And you have probable cause to believe the evidence is in the phone. That's all the government needs, I think.  It doesn't make sense to limit the government's search within the place to be searched to particular forms of evidence any more than it makes sense to limit the search of tax records in a house to  "documents stored on 8.5 X 11 paper" or "items in boxes designed to store records." As the Supreme Court stressed in Ross v. United States, "[w]hen a legitimate search is under way, and when its purpose and its limits have been precisely defined, nice distinctions between closets, drawers, and containers . . .  must give way to the interest in the prompt and efficient completion of the task at hand." Everything in the place to be searched can be searched.

Under the Fifth Circuit's approach in Morton, though, I gather there wouldn't be probable cause to search the phone.  Sure, you know the evidence is in the phone.  But without probable cause as to any particular "category" of information on the phone, the phone is unsearchable.  And of course this requires courts to develop a Fourth Amendment jurisprudence of "information categories," whereby judges announce the inherent nature of different kinds of constitutional data that is on a phone.  (Apparently, photographs are one category, text messages are another, and call logs a third.)

I don't think the Fourth Amendment can work this way.  And nothing in Riley suggests it, it seems to me.  Riley required a warrant to search a phone given the invasiveness of cell phone searches, but it doesn't say anything about needing separate probable cause based on the "category" of information on the phone.

Granted, I realize the government in Morton seems to have wrongly conceded the point. I have no problem with the Fifth Circuit accepting the concession in this case.  But I think it should have just accepted the point and ruled accordingly, rather than gone on to make the point what certainly looks like the court's holding.  It was fair to accept the concession, but as a holding, it seems to me quite wrong.

IV.  But the Right Result In This Case, In My View

At this point some of you are thinking, but Kerr, this case stinks! It sure looks like the government was using the warrant pretextually.  They suspected that Morton had child pornography on the computer.  And they obtained a warrant for something else but probably planned to look for child pornography with the hope that it all came into plain view.  Doesn't that raise Fourth Amendment problems?

Yes, I think it does. But I think it raises different Fourth Amendment problems than the Fifth Circuit suggested in Morton. In my view, the Fourth Amendment limit is not in where in the computer the government can search, but what evidence the government can use.  As I argued in my 2015 article, Executing Warrants for Digital Evidence: The Case for Use Restrictions on Nonresponsive Data, 48 Texas Tech L. Rev. 1 (2015), the Fourth Amendment should be interpreted to impose a use restriction on nonresponsive data in digital search warrants.  Evidence outside the scope of probable cause named in the warrant should not be usable, and thus should be suppressed.  (As regular readers may remember, the Oregon Supreme Court adopted this view under the state constitution in State v. Mansour.)

What difference would this make?  Under my test, if the government had probable cause to believe there was evidence relating to Morton's drug possession on his phone, it could search the entire phone for that evidence—including the photographs.  However, the government could not use any child pornography it came across during the search based on the warrant it had obtained for evidence of drug possession.  If it came across child pornography in the text messages, they couldn't use that, either.  Nor could they use the discovery of the child pornography to get a new warrant for that—that is a prohibited use, as my article explains.

Under my approach, then, the Fourth Amendment doesn't stop the government from looking for all the evidence in the phone.  But my approach eliminates the incentive to conduct a pretextual search by eliminating the possible benefit of a pretextual search. If the government wants to recover child pornography from the phone, under my test, the government needs to gather probable cause for that and obtain a warrant to search the phone for evidence of child pornography.

In the end, then, I think the result in this case is right. In my view, the Fourth Amendment should not permit the government to use the search for one kind of evidence to get a different kind of evidence.  But the problem isn't where the government looked, or what kind of file was at issue, but rather whether the evidence recovered was within the scope of probable cause established to obtain the warrant.

NEXT: Today in Supreme Court History: January 10, 1842

Editor's Note: We invite comments and request that they be civil and on-topic. We do not moderate or assume any responsibility for comments, which are owned by the readers who post them. Comments do not represent the views of or Reason Foundation. We reserve the right to delete any comment for any reason at any time. Report abuses.

  1. The suppression of evidence is a “solution” imposed by the courts to discourage the government from infringing on someone’s rights enshrined in the Fourth Amendment. Its primary purpose is not to inoculate a criminal from prosecution – that’s just a, probably undesirable, side effect.

    It seems like the problem is now the courts are in the awkward position of “refining” their “solution” to deal with a new world when the original solution was flawed.

    Suppressing evidence actually isn’t a very strong deterrent as it doesn’t impose a substantial penalty on police who infringe on a innocent person’s Fourth Amendment rights – it mostly just lets a criminal go free. There are other deterrents that likely would work much better — such as the legislature making it a capital crime for police to violate the Fourth Amendment unless evidence of a crime was found (no, I’m not seriously suggesting that extreme solution either).

    To be consistent, the solution proposed here for searches of digital devices seemingly should also be imposed on physical searches as well. If, when searching for financial records per a search warrant, police open a shoe box and out fall thousands of pictures showing the owner of the house raping children, the police should also be barred from using that evidence – even to get a search warrant for a more exhaustive search for child pornography.

    It seems, imperfect as it is, that the court’s decision (and I’m working only from the post having not read the decision as of yet) is reasonable. If a search warrant specifies that police are searching for a stolen army tank, they can’t open the salt shaker on the dining room table looking for the tank. If police are looking for evidence of contacts in a drug case and don’t specify that they have probable cause that contact information may be in the form of photographs, they have no business looking at photographs. Call logs, text messages, and contact lists are stored in specific places in a cell phone and those are the only places the police should be looking. Such contact information is not stored as photographs with the possible exception of the images associated with a contact (such as their picture) or a multimedia attachment to a text message — and those could be accessed by following a link to the one image each such link references.

    If, it turns out, that the text of a text message a multimedia attachment (such as “The courier will have the drugs, see him in the attached picture”) is relevant to the drug issue, then police should be able to view the image to see if the courier is, actually, Doug Drugdealer. If it turned out that the image was of the defendant raping a child, then police should be able to use that to get an expanded search warrant for child pornography (which would almost certainly include any and all images stored on the phone — or any computing device in the suspect’s home).

    1. nie c’est une crack pipe

    2. I disagree about the shoe box full of child rape photos, but only because I think it is reasonable to believe documents related to the tax case might be in the shoe box.

      A closer equivalent to the Morton case might be if the judge signed a warrant giving the police authority to only search the defendant’s file cabinets and desk drawers, and they went looking in shoe boxes, under the bed, in the salt shaker, etc, and found evidence of separate crimes.

      As much as I hate seeing a pedophile go free, the Morton decision seems correct. The police had a warrant for text message, contacts and call logs only, but searched photos as well.

      Now, if they had found incriminating photos within the text messages they were lawfully searching through, I think that would be a different story.

  2. Excellent post about a serious topic.

  3. My understanding is police will add something small to a physical search warrant if they want to go fishing. It is not implausible to me that a drug user could take a selfie with incriminating evidence. It’s as likely as a tiny baggie (indicia of drug use or dealing) being hidden somewhere in a house. Maybe the ruling will simply inspire more creativity in search warrant applications. “Based on extensive experience and training, your affiant expects the phone to contain audio, video, and text evidence of drug use, dealing, and conspiring to vote for your opponent in the last election.” (OK, strike that last one in my area where we don’t have judicial elections. The magistrates who sign warrants are patronage hires or other people with connections.)

  4. So how should the court go about deciding the proper level of generality in which to formulate warrants. Presumably the government can’t just issue a warrant for violations of ‘federal law’. So is it by statutory offense? If heroin and fentanyl were outlawed by different laws would you need to show probable cause for each one you search for (eg you though he was a heroin dealer but can’t use any evidence because he’s a fentanyl dealer)? I presume not (if for not other reason so states don’t collapse all laws into one statute).

    If not won’t this be *almost* as difficult as fleshing out kinds/types of info.

    1. I don’t think this case was about what was being searched for, as much as it is about what and where the police were searching on the phone.

      I assume that if they found evidence of drug possession or distribution, in the photos, that evidence would also have been ruled inadmissible.

      Likewise, if they had found evidence of child pornography or the sexual abuse of children in the text messages, contacts and phone logs, that evidence would have been admissible and could have been used to obtain warrants to search his photos, other electronic devices and his home.

      1. I understand that this case wasn’t about that but on Kerr’s preferred approach the courts would have to break up wrongdoing into kinds of things when issuing warrants because that controls what kinds of evidence is responsive/non-responsive and thus what can be used.

        I mean you surely can’t say the warrant should be limited to the exact circumstances the police expect, e.g., if they think John Doe has text messages conspiring with Jane Doe to rob a bank and it turns out they are really with Mary Smith presumably those should still be considered responsive. OTOH presumably he would think that if they are looking for records of text messages planning to rob a bank and they find evidence of drug use that’s non-responsive.

        So I’m wondering how the courts are supposed to figure that out. It kinda didn’t matter when only the place to be searched was really relevant to whether the evidence could be used at trial but if you want to start suppressing evidence that’s found in the place to be searched because it’s not the kind of evidence to be searched for you need to deal with that and this seems super hard.

  5. I don’t think that your solution works. Once the government has full access to the phone, they can collect all sorts of side information that can direct them elsewhere without the need of a warrant.

    For example, if I find that you sent Joe Schmoe some child pornography in an e-mail, what prevents me from going to Joe and asking him for any e-mails from you (perhaps agreeing not to prosecute him)? You’d have to start questioning the motivation for me talking to each and every possible witness.

    I can built a very detailed profile of a person, his movements, his friends, his entire life out of the forensic information on the cellphone.

    1. Yes, would all parallel evidence also have to be prohibited? How about communicating what is know to other LEOs so they can put the perp under a microscope? Where is the line?

  6. Prof. Kerr,

    Can you elaborate on what derivative uses of non-responsive evidence you think are permissible? If I’m following the article, you’re saying that if the police, executing a warrant for (say) child pornography according to its precise terms in perfect good faith, find a video of the phone owner abducting and dismembering dozens of hitchhikers, they can’t “use” it. What exactly can they do (if anything) under your theory?

  7. “On the other hand, the government could not search the photographs because there wasn’t any particular reason to think that the evidence of drug possession would be in photographic form. The evidence against Morton was of personal use, and there wasn’t a particular reason to think that a user of drugs would have evidence of it on the phone.”

    This is such a Boomer ruling (or, given Jolly having been born in 1937, pre-Boomer ruling).

    Seriously. I don’t know how old Morton is, but people take pictures of everything, post everything to social media. One of the drugs in question was marijuana- I suspect that a majority of people who use pot have pictures of them using it somewhere, especially those under 50.

    To be clear here- I don’t think marijuana should be illegal, and it’s not clear to me why, given that he had little enough in the way of drugs, there should be probably cause to do any kind of searching of his phone. But John Roberts’ correct jurisprudence- that a smartphone is many things and that indiscriminate searches are not ok- cuts both ways. Yes, someone who smokes pot and has only a film camera may be unlikely to snap pictures of it regularly, but a smartphone is not merely a camera- it is a camera that can communicate with friends, and through which people share visual media, whether or not it’s necessarily a good idea to do so.

  8. Would it be fair to say Prof. Kerr’s position is the plain-view exception should not apply in computer searches?

    I appreciate that cell-phone searches provide difficult Fourth issue

  9. (Not sure what happened there. Give us an “edit” feature)

    Cell phone searches raise difficult Fourth Amendment issues given the sheer volume of personal information they contain. But I don’t think that asking law enforcement to ignore evidence of crimes outside the scope of a warrant is realistic. Imagine a case where officers find photos of a possible murder scene or of a missing person. Just pretend they didn’t see it?

  10. No argument with the legal analysis. But I can’t imagine many will weep if Mr. 20,000 Child Porn Images is t-boned by a big rig going 50mph.

    1. With side airbags, that may not be fast enough…

  11. First of all, I think you misunderstand the court. If the evidence they have a probable cause to search for could legitimately be found anywhere on the phone, then that would be cause for a warrant to search every category of item on the phone. The individualized analysis of types of data on the phone would only lead to a restriction of the warrant if there is reason to believe the data is somewhere specific on the phone. Secondly, I am not a civil libertarian but your solution embodies all the worst aspects of civil libertarianism for me. It doesn’t protect any non-criminal privacy of the suspect, for instance if they were secretly a member of the Communist party, and that could be found in their photos. It only protects unrelated criminal activity found on the phone. Normally, civil libertarians claim we need to disallow evidence of criminal activity to protect the privacy of everyone, but here the police would be allowed to invade that privacy anyway, and the only thing being done is protecting criminal activity.

  12. With side airbags, that may not be fast enough…

    Of course, I make a distinction between someone who merely is looking at pictures and some schmuck who is actually molesting children. Yes, I understand the “demand” issue, but as we now can (or soon will be able) to computer generate these images (or film a sexbot), I think we need to make a distinction between someone who actually harms living children, and someone who is a creep but who doesn’t.

    If for no other reason than priorities….

  13. “because there was not enough information that the evidence of drug possession would be in photograph form to justify searching the photographs “

    Those judges clearly don’t understand the “selfie” culture.

    The pictures file is the first place I would look because the idiot probably has pictures that will be more damning than any call log or anything else. Remember that all the pictures we have so far by those who broke into the Capital were taken and published by thew perps themselves.

    Question I have is why couldn’t they get a kiddieporn warrant at the getgo?

    1. I tend to agree. There is almost certainly a greater likelihood of finding evidence of personal drug use in the photos than in the contacts and call logs. It is probably also true of the text messages, but that is a closer call, IMO.

      This case might be more about a lack of imagination and understanding of where evidence of drug use is likely to be on a phone, on the part of the judge and possibly the officers who obtained the warrant.

      Contacts and call logs would seem to be a rather weak source of compelling evidence, especially for personal possession, as opposed to drug use.

      I would imagine and large percentage of non-drug users have contacts in their phones of people who have been involved in distributing drugs, on some level.

      A pattern of many calls with a known drug dealer might be evidence of personal drug possession. But, then again, drug dealers also have friends who don’t buy drugs from them.

      On the other hand, these days, I think it is fairly common for people to take photos of themselves using illegal drugs.

      Also, I don’t think it is unreasonable to think that a drug user or drug dealer might use photos to hide drug contacts.

  14. I’m wondering about Kerr’s hypothetical where the police know a piece of information is on the phone but don’t know the location more precisely than that. Suppose that the police have probable cause to believe that the suspect is engaging in illegal activity and that the suspect’s business records will contain all the juicy details, but they don’t know where the suspect keeps his business records. Can they get a warrant for all of the locations that the suspect visits regularly: his office, his house, and his safe deposit box? Kerr seems to be saying that the answer is no.

  15. We cannot be sure it was child pornography. They may claim that it was, but if the people and the press are prohibited by law from making an independent determination, we should give the benefit of the doubt to the defendant.

  16. This blog doesn’t deserve you, Prof. Kerr. Please consider finding better collaborators.

    Everyone makes a mistake now and again.

  17. My entirely normative reactions:

    Leon‘s “good-faith” hurdle entirely swallows the control that the Fourth Amendment is supposed to provide. Leon should be overturned. The courts grant us no good-faith exception when we violate the law. Cops shouldn’t get one, either.

    – Prof Kerr’s criticism of Morton is based on a distinction between the “kind” of evidence and the “form” of evidence. That strikes me as a distinction without a difference.

    The Fifth Circuit’s approach here makes a lot of sense to me. If you’re looking for call logs, you’re not going to find them among the pictures unless the person keeping them has gone to extraordinary lengths to hide them. If you have evidence of such extraordinary lengths, make your case when applying for the warrant. Without such evidence, it’s the equivalent of saying that just because you have a warrant for the surveillance footage of my business, you can also seize all my financial papers just in case there’s a printed image mixed in. That’s not how warrants are supposed to work.

    That said, I am sympathetic to Prof Kerr’s “use” interpretation. I fear, however, that once the information is discovered, it’s use cannot be fully restricted. The suspect’s privacy has been breached and the genie cannot be stuffed back into the bottle. Without a lot more discussion and court control on the uses (in court and out) of that information, I question the practicality of that approach. In the meantime, the “you can’t look there” approach seems simpler and more reliable.

  18. “there wasn’t a particular reason to think that a user of drugs would have evidence of it on the phone.”

    Drug users post pictures online of drugs all the time. Are we to believe they don’t have them on their phones?

  19. From a data science perspective, this seems correct.

    The warrant is for data supporting the claim that defendant is a drug dealer, in whatever form that might be.

    If you are running a separate investigation for child pornography, you need an independent warrant to run that query.

    Does the case explain how they got access to his phone in the first place? Was he dumb enough to give them the passcode?

Please to post comments