Ninth Circuit on ALPR Surveillance and the Fourth Amendment

A fascinating new case.

|The Volokh Conspiracy |

I blogged recently on the Massachusetts high court's ruling on how the Fourth Amendment should apply to querying automated license plate reader (ALPR) databases.  Here's a new ruling on the topic that deepens the plot: In United States v. Yang, the Ninth Circuit ruled that a driver's Fourth Amendment rights were not violated by querying an ALPR database to learn the car's location when the car was a rental that was been driven beyond the contract and the rental car company had tried to repossess the car.  In those circumstances, the court reasoned, the driver lacked standing in the location of the car.

I'm not sure I agree with the reasoning of the decision.  But it's certainly interesting, so I thought I would blog more about it.

I. The Facts

Yang was spotted committing a crime while driving a GMC Yukon. Investigators traced the Yukon's license plate to a rental car company, Prestige Motors.  The rental car company told the officers that the Yukon had been rented to Yang but that he hadn't returned the car on time.  But by the time investigators contacted the rental car company, it was already six days beyond the return time. Prestige Motors had tried to locate the Yukon through its company-owned GPS, but that the GPS had been disabled.

In an effort to find the Yukon, investigators entered the license plate number of the Yukon into an ALPR database called LEARN.  LEARN is a database run by a private company that is available only to law enforcement.  The LEARN database receives about 35% percent of its images from law enforcement vehicle cameras and the remaining 65% percent of its images from commercial vehicle cameras. It contains more than 5 billion scans of license plates.

The LEARN database was queried on April 13th, and it revealed a single scan in the database that had been taken on April 5th at 11:24pm—slightly more than 12 hours after the Yukon was supposed to be returned to the rental car company.  The scan revealed that, on April 5th at 11:24pm, the Yukon had been spotted on a road near a particular intersection.  Immediately after getting the database hit, also on April 13th, the officer went to that intersection and was able to locate the Yukon in a parking lot nearby.

II.  The Ninth Circuit's Reasoning

Here's the question: Did the querying the ALPR database constitute a "search" that violated Yang's rights?

The Ninth Circuit's answer, in an opinion by District Judge Piersol joined by Judge Ken Lee: No. Yang didn't have standing in the car that was supposed to have been returned to the rental car company, so his rights weren't implicated in the scanning.  Here's the reasoning:

While the mere expiration of the rental period does not automatically end a lessee's expectation of privacy, see United States v. Dorais, 241 F.3d 1124, 1129 (9th Cir. 2001), we conclude that Yang has failed to establish that he has a reasonable expectation of privacy in the historical location information of the Yukon under the facts of this case. There is no evidence in the record that Prestige Motors had a policy or practice of allowing lessees to keep cars beyond the rental period and Prestige had made affirmative attempts to repossess the vehicle by activating the GPS unit to locate and disable the vehicle.

In so holding, we find instructive our decisions in United States v. Dorais and United States v. Henderson, 241 F.3d 638 (9th Cir. 2000) which both analyze a lessee's expectation of privacy in rental property after the expiration of the rental period. [Editor: The gist of these cases is that you eventually lose Fourth Amendment rights in rented property after the rental period has ended.] . . . .

At the outset, we reject Yang's argument that the above-mentioned cases are inapposite because they regard an expectation of privacy in property or premises rather than an expectation of privacy on the whole of one's movements that is at issue in this case. We are simply unwilling to conclude that a person has a reasonable expectation of privacy in his movements as revealed by the historical location data of a rental vehicle after failing to return the vehicle by the contract due date, when there is no policy or practice of the rental company "permitting lessees to keep cars and simply charging them for the extra time." See Dorias, 241 F.3d at 1128 (citing Henderson, 241 F.3d at 647).

In this case, the rental contract provided that vehicles not returned by the due date will be reported as stolen to the proper authorities. Yang contends that Prestige Motors' decision not to immediately file a stolen vehicle report after the rental contract expired is evidence that the company does not strictly follow this policy. However, unlike in the cases discussed above, Yang presented no evidence at the suppression hearing of any other custom or practice by Prestige that led him to believe that rather than adhering to the rental contract terms and reporting the vehicle as stolen, Prestige would, absent any request by him, simply extend the lease term and charge him the additional fees.

While the rental agreement provided that "[a] charge of $20.00 per day will be applied to the rental for every day the vehicle is late," the contract also provided that "[i]f a customer wishes to extend, he or she must notify the company 1 day in advance to make arrangements and additional payments." There is no evidence in the record to suggest that Yang notified Prestige of any intent on his part to extend the rental period. In addition, the rental contract warned lessees that Prestige may repossess a vehicle if not returned by the contract due date and that a $250.00 repossession fee will apply.

And in case there were any lingering doubts about whether Yang had a reasonable expectation of privacy in the location of the Yukon at the time Inspector Steele searched the LEARN database on April 13, 2006, we conclude that Prestige's private attempts to repossess the Yukon by activating the GPS and disabling the vehicle placed Yang, the sole authorized driver, on notice that Prestige did not intend to extend the lease term, but rather sought to repossess the vehicle.

At oral argument, Yang also argued that he had standing to object to the query of the LEARN database because it revealed his location on April 5, 2016, at approximately 11:24 p.m., at which time, he alleges, he still had a reasonable expectation of privacy in his movements. Given the rental agreement provided that vehicles not returned by the "due date" would be reported as stolen, Yang contends that he had a reasonable expectation of privacy in his movements, as revealed by the location data of the Yukon until 11:59 p.m. on April 5th even though the vehicle was due back by 10:48 a.m. that day. Because the ALPR camera captured the Yukon's location information well after the close of Prestige's business hours, as clearly advertised on the rental agreement, we need not determine whether a defendant has standing to object to a "search" of a rental vehicle's historical location information that was captured and uploaded to a database prior to the expiration of the rental agreement.

III.  Judge Bea's Concurrence

Judge Bea concurred in the judgment.  Judge Bea reasoned that Yang had standing because "in Carpenter, the Supreme Court was clear that the relevant inquiry, at least where location data is concerned, is what personal location information is revealed by a search of the records, not what type of data was collected and under what circumstances."

But according to Judge Bea, no search had occurred because this particular query had not revealed the whole of Yang's physical movements:

Looking at the specifics of this case, it's clear that the LEARN database did not contain information that revealed the whole of Yang's physical movements. Despite its 5 billion total records, the LEARN database contained a single entry for the Yukon that Yang had rented. Yang was unlucky that the one observation was recorded when he was in possession of the vehicle and was made near his residence. But even accepting that the search of the LEARN database revealed where Yang lived, it exposed nothing else about his "particular movements" whatsoever. Carpenter, 136 S. Ct. at 2217.

. . . It would be folly to hold that searches of ALPR databases require a warrant without identifying even one case where the "whole of [one's] physical movements" was implicated in an ALPR database search. Id. at 2217. If the technology evolves in the way that amici hypothesize, then perhaps in the future a warrant may be required for the government to access the LEARN database, but this should only be the case if the database evolves to provide comparable location information to the records at issue in Carpenter.

IV.  A Few Thoughts

The issue of Carpenter standing is fascinating.   Traditionally, people have Fourth Amendment rights in a search of a particular place—persons, houses, papers, and effects, as the constitutional text puts it.  To have standing, the search needs to be a search of "their" persons, houses, papers, and effects.  So when we conduct a traditional standing analysis, we ask if the person has enough of a relationship with the property searched for the property to be effectively theirs.

The complication is that Carpenter recognized a new theory of Fourth Amendment searches. Under Carpenter, people now also have a new Fourth Amendment right in the "whole of their physical movements."  This creates a standing puzzle.  Does standing to challenge a Carpenter search hinge on whether a person has enough of a relationship with the property that was surveilled to learn about the person's movements?  Or does standing to challenge a Carpenter search depend only on whether the government did in fact learn the person's movements?

The majority takes the former view, looking to Yang's rights in the car.  Judge Bea's concurrence takes the latter view, looking to whether Yang's location was obtained.

Based on my read of Carpenter, and the opinions on which it is based, I would think that Judge Bea's approach is probably more consistent with those cases.  Carpenter focused on the fact that the location of Carpenter's phone was revealed. The Court didn't dwell on whether Carpenter had paid the bill or was a legitimate user of the phone.

Granted, the idea of standing being based only on the location of property merely thought to be linked to a person is pretty weird.  After all, we don't know that Yang was actually driving the car when the car was spotted. Who is to say that locating the car at that time was even actually locating Yang?  But I take that oddity to be largely brought on by Carpenter. We didn't know Carpenter was using the phone that was located using the cell-site location information (CSLI).  And yet the Supreme Court treated revealing the CSLI of the phone thought to be linked to Carpenter as effectively revealing Carpenter's location.  The same would seem to be true here, I would think. Perhaps the fact that the Yukon was thought to be linked to Yang meant that his rights were implicated in locating the Yukon even if he wasn't in the car at the time. It's odd, but it's plausibly an oddness caused by Carpenter.

Back to Yang, with just one more thought on standing.  Even if majority was right that Carpenter standing is based on rights in the car,  I wonder if I'm alone in finding it somewhat misplaced that the majority focused so much on the specific language of the rental contract.  I would think that this kind of standing would be governed by the Supreme Court's recent decision in Byrd v. United States, which was expressly about standing to search rental cars outside the terms of the contract.  And yet the opinion doesn't cite Byrd, although it was discussed extensively in Yang's reply brief.

Finally, I'm also fascinated by the timing issue, a question I keep mentioning when I write on Carpenter and the mosaic theory that keeps getting more interesting.  The question is, when does a Carpenter search occur:  When the information was first collected?  When it was requested by the government?  When it has given to the government? When it was analyzed by the government?  Some combination of the above?  This matters if standing is based on the property surveilled, as a person can have standing in property at one time but not at another time.

In the Yang case, the panel majority focuses on whether Yang had standing at the time the query was made, on April 13th.  By the time of the April 13th query, the court reasons, Yang lacked standing.  But the panel then also offers the idea that Yang lacked standing when the snapshot of the car was originally taken, on April 5th, because the car was already overdue by a few hours by then.  The court then expressly leaves unresolved whether Yang would have had standing had the April 13th query returned a hit that revealed the car's location from an earlier time when the car was still properly rented to Yang.  If I'm reading the opinion correctly, the court seems to be slightly hedging its bets here on which timing matters.

As always, stay tuned. Carpenter's novel conceptual puzzles are the gift that keeps on giving, at least for academics.  It's like Einstein's theories meeting Newtonian physics.

Advertisement

NEXT: Brickbat: Justice Delayed

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 Reason.com or Reason Foundation. We reserve the right to delete any comment for any reason at any time. Report abuses.

  1. How can private behind-the-scenes conduct that a person doesn’t know about place a person on notice? It seems to stretch the concept of notice well beyond its meaning.

    That said, if the car had been sitting in the same place for more than a week, that’s evidence it was abandoned, making the renter or former renter’s claim even weaker.

  2. It also seems this case could have been decided more narrowly, based more closely on the specific facts of the case. It seems a stretch to say that because a renter returns something a few hours late, the renter has lost all possessory and privacy interest and lacks even standing to assert them. I’m by no means sure that the person who is caught in traffic and late to the airport thereby loses all rights, even if the contract says to notify before extending.

    1. A few minutes is one thing. Six days is entirely different, not just a matter of stuck in traffic. Plenty of time to call and say why he’s late. Presumably he left an address, phone number, credit card, etc, and the rental agency tried to contact him in that six days, in addition to using the GPS locator which had been disabled, lending support to the idea that something nefarious was underway.

      I don’t like ALPR and the government searching them willy nilly, but this seems perfectly cromulent to me. Maybe they should have had a search warrant, but the facts imply to me it should have been easy to get.

      1. A few minutes is one thing. Six days is entirely different, not just a matter of stuck in traffic.

        The majority rejected D’s argument that he still had a reasonable expectation of privacy less than 13 hours after the “time due” on the rental agreement. The opinion squirms on this a bit, basically hiding behind the fact that the plate image returned in the database search was captured after close of business, thus triggering the policy of reporting as stolen cars not returned “by the due date” and defeating D’s reasonable expectation of privacy. Given that it’s very common to return rentals after COB and leave the keys in a drop box, this feels like a bit of a superficial analysis to brush off an otherwise problematic fact.

        1. It’s commonly acceptable to return rentals after COB and leave the keys in a drop box if you are dropping the car off after hours but before the required return time.

          Regardless, “due date” commonly means the specified date/time that the thing is due. If checkout time at the hotel is 11 am, that’s when you have to be out. Where time is not specified, “due date” commonly ends at COB, not at midnight.

          1. Where time is not specified, “due date” commonly ends at COB, not at midnight.

            Ironically enough, in the legal world it’s exactly the opposite — midnight unless otherwise specified.

          2. “It’s commonly acceptable to return rentals after COB and leave the keys in a drop box if you are dropping the car off after hours but before the required return time.”

            assuming the “required return time” isn’t COB. It’s possible to arrange for an after-hours dropoff, but if you haven’t pre-arranged it, the likely result is that they bill your credit card for an extra day or a “late drop-off fee”.Found this out the hard way from renting a U-haul from a place that closed at 4pm on weekends.

  3. Where in this deliberation does the disabling of the GPS fit?

    1. by the time investigators contacted the rental car company, it was already six days beyond the return time

      If he had just re-enabled it and returned it, he could have been in Antarctica by now.

  4. The timing issue seem to me to be somewhat misplaced. By the time the database was “searched” the car had been missing for 6 days. That they located it at some past time seem not really relevant as it is exceedingly unlikely that had the car been returned even a day two two late that “search” would have been made. Supposing the car was abandoned before the return time and six days later the same events had taken place why would the location of the car be a violation of Yang’s rights?

  5. The majority opinion makes much more normative sense. Yang lost his rights to privacy over the car’s location when he failed return it on time.

    In my view, the Carpenter decision ignored payment of the phone bill and rights to the phone because that factor simply wasn’t relevant. Had there even been any question about whether it was Carpenter’s phone, his lawyers presumably would have raised the issue and the Court would have addressed it.

    re: the timing issue – In my opinion, the search conducted on Apr 13 should have been structured to only return hits after the required return date/time. A search for hits before that date/time would still be allowed – but only with a warrant.

  6. Automated license plate readers operate on vehicles that are operated in public. Seems like the “plain sight” doctrine would have some application. This system as described sounds like an unusually thorough and unusually reliable form of canvassing witnesses. Should police obtain a warrant before asking anyone if they saw such-and-such kind of car in the last couple of days?

    1. Automated license plate readers operate on vehicles that are operated in public. Seems like the “plain sight” doctrine would have some application.

      The problem here is what Prof. Kerr calls “equalibrium adjustment”. The “plain sight” rule works perfectly fine when police departments didn’t have the resources to compile KGB dossiers on everyone in the city. So an undercover cop across the street has you under surveillance and sees you leave your house, in plain sight, at 8 am. That’s fine!

      But now, technology permits police departments, and anyone else who wants to, to collect massive amounts of data of about everyone’s movements through license plate readers. Suddenly, you’ve lost the privacy protecting limitations of the plain sight doctrine.

      What Prof. Kerr has suggested in the past is that when this happens, you want to adjust the equilibrium of the Fourth
      Amendment so the police doesn’t get too much power.

      1. Meh. It’s in private hands. And the rule is the same. Don’t do things in public if you object to people (including police) knowing about them…

  7. “It’s like Einstein’s theories meeting Newtonian physics.”

    Now, don’t get it twisted. We all love you, Kerrbot. But I think you need to take a cold shower.

Please to post comments

Comments are closed.