Harvard Law Review Harvard Law Review Harvard Law Review

Harvard Law Review Forum

The Positive Law Floor

Responding to William Baude & James Y. Stern, The Positive Law Model of the Fourth Amendment

The positive law model maintains that a Fourth Amendment search or seizure occurs if, but only if, a private party could not lawfully perform the conduct that the government actually engaged in. The positive law model thus treats laws applicable to private parties as a ceiling on Fourth Amendment protections. But government action is fundamentally different — and often more deserving of regulation — than similar conduct by private parties. Due to its distinctive capabilities, incentives, and social role, the government often threatens the people’s security in ways that private parties simply do not. Still, we can learn from analogies to private parties without being limited by them. The way to do that, I briefly suggest, is to view privacy-related measures applicable to private parties as presumptively triggering the Fourth Amendment’s prohibition on unreasonable searches. I call this alternative approach the “positive law floor.”


In The Positive Law Model of the Fourth Amendment, Professors Will Baude and James Stern have given us a terrific piece of scholarship: that uncommon kind of article that is provocative, thoughtful, and sure to change the way we think about a familiar area of law.1×1. William Baude & James Y. Stern, The Positive Law Model of the Fourth Amendment, 129 Harv. L. Rev. 1821 (2016). What makes Baude and Stern’s piece distinctive is its “categorical” claim that the positive law imposes a hard cap on Fourth Amendment protections. Id. at 1888; cf. Orin S. Kerr, Four Models of Fourth Amendment Protection, 60 Stan. L. Rev. 503, 516–19, 532–34 (2007) (discussing and criticizing exclusive reliance on the positive law model).

But that doesn’t mean it’s right. In giving the positive law model its most thorough elaboration and defense, Baude and Stern push its logic too far. In short, the “positive law model” works via analogy: a Fourth Amendment search or seizure occurs if, but only if, a private party could not lawfully perform the conduct that the government actually engaged in.2×2. See Baude & Stern, supra note 1, at 1825–26; see also id. at 1849 (“The more analogous the government’s position is to that of an ordinary private actor, the less its behavior generates constitutional concern.”). Because the positive law model disregards exemptions for police, it does not “incorporate” laws applicable to private parties, id. at 1833, so much as transmute them. As a result, the positive law model treats laws applicable to private parties as a ceiling on what the Fourth Amendment can do. While the positive law model has several shortcomings, its most fundamental problem is this: government action is different — and often more deserving of regulation — than similar conduct by private parties. Due to its distinctive capabilities, incentives, and social role, the government often threatens the people’s security in ways that private parties simply do not. So in focusing on how the government is more formally empowered than private parties, the positive law model overlooks all the other ways in which the government is special.

The shortcomings in the positive law model point toward another possibility: learning from analogies to private parties without being limited by them. The way to do that, I briefly suggest, is to view privacy-related measures applicable to private parties as presumptively triggering the Fourth Amendment’s prohibition on unreasonable searches. In other words, when the law has made a deliberate choice to protect against certain intrusions on privacy and security by private parties, then police should have to adduce some kind of justification for undertaking a similar intrusion. This alternative approach, which I call the “positive law floor,” has several advantages over Baude and Stern’s positive law model and finds support in recent Fourth Amendment jurisprudence.

I. The Structure of the Positive Law Model

This Part critically explores the positive law model’s analytical structure. In short, the positive law model maintains that (i) whether a Fourth Amendment “search” has occurred (ii) should be distinguished from the question of reasonableness and (iii) resolved instead based on laws applicable to private parties. Each of these component parts is problematic.

A. Distorting the Meaning of “Search”

The positive law model purports to be a necessary and sufficient test for what qualifies as a “search” within the meaning of the Fourth Amendment.3×3. See id. at 1825–29. But under any recognizable use of language, whether conduct is illegal is irrelevant to whether that conduct is a “search.”4×4. See, e.g., Search, n., Oxford Eng. Dictionary (2016), http://www.oed.com/view/Entry/174307 [http://perma.cc/LTE3-CMBW]. Rather, a search is an effort to discover information,5×5. See id. (defining “search” as “examination or scrutiny for the purpose of finding a person or thing” and as “investigation of a question; effort to ascertain something”). and an “unreasonable” search under the Fourth Amendment is naturally read as an effort to discover information that unjustifiably undermines “[t]he right of the people to be secure in their persons, houses, papers, and effects.”6×6. U.S. Const. amend. IV.

To illustrate the oddity of the positive law model, imagine that, “in the course of”7×7. Baude & Stern, supra note 1, at 1833 (emphasis omitted). following a suspect by car, three police officers take steps to avoid losing track of their quarry:

  • The first officer quickly executes a lane change without signaling, as required by state law.
  • The second guns the engine, creating noise exceeding the decibel level set by a local ordinance.
  • And the third activates a GPS navigational system whose operation infringes a patent.

Once apprehended, could the suspect seek evidentiary exclusion or civil damages based on the unsignaled lane change, the revved engine, or the use of the navigational system?

The positive law model seems to answer “Yes,” but only because each officer violated laws applicable to private parties. This conclusion is counterintuitive, to say the least. As a matter of English language, whether each of the examples involves a “search” is unrelated to whether a legal violation occurred. A search is a search, whether it is illegal or legal — reasonable or unreasonable.

What’s more, it would be equally counterintuitive to say that the police actions described above are constitutionally “unreasonable,” either with or without a warrant. Laws involving signaling, noise pollution, and intellectual property may all have been written for good reasons, but those reasons have little or nothing to do with the values of privacy and security that have been thought to underlie the Fourth Amendment and its prohibition on unreasonable searches.

Remarkably, Baude and Stern adduce no evidence that anyone ever defined a “search” with reference to legal demands, much less legal demands applicable to private parties.8×8. See, e.g., id. at 1841 (“[W]e don’t claim that the positive law model was ever articulated in exactly these terms at the time.”). Nor do Baude and Stern make any such claim regarding “seizures.” See id. at 1840–41.

Instead, Baude and Stern point to the Founders’ desire to protect the home from arbitrary searches.9×9. See id. at 1837–39; see also Akhil Reed Amar, The Fourth Amendment, Boston, and the Writs of Assistance, 30 Suffolk U. L. Rev. 53, 69 (1996); Thomas Y. Davies, Recovering the Original Fourth Amendment, 98 Mich. L. Rev. 547, 590 (1999); Geoffrey R. Stone, The Scope of the Fourth Amendment: Privacy and the Police Use of Spies, Secret Agents, and Informers, 1976 Am. B. Found. Res. J. 1193, 1198–99, 1211. According to Baude and Stern, “the most straightforward extrapolation” from that history is that Fourth Amendment violations were “marked by violations of positive law.”10×10. Baude & Stern, supra note 1, at 1839–40. In effect, Baude and Stern imagine that the Founders’ bold declarations about the home’s inviolability from governmental intrusion all had asterisks to allow for the home’s violation, if legal rights against private parties relevantly changed. But when James Otis insisted on “the sacred Rights of the Domicil” against intrusions by “Officers” acting “under color of Law,”11×11. Id. at 1837 (quoting Report of the Committee of Boston on Rights of Colonists (1772), as published by Order of the Town, 15–17, reprinted in Josiah Quincy, Jr., Reports of Cases Argued and Adjudged in the Superior Court of Judicature of the Province of Massachusetts Bay, Between 1761 and 1772, app. 1, at 466–67 (Boston, Little, Brown & Co. 1865)). did he really mean to allow for the possibility that changes in the positive law might annihilate those rights? A more plausible view is that the Fourth Amendment was meant to bar Congress from using its lawmaking powers to erode the people’s security in the home, including by authorizing general warrants.12×12. See, e.g., Davies, supra note 9, at 590 (arguing that the Framers “were concerned about a specific vulnerability in the protections provided by the common law; they were concerned that legislation might make general warrants legal in the future”); Laura K. Donohue, The Original Fourth Amendment, 83 U. Chi. L. Rev. (forthcoming 2016) (manuscript at 98), http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2726148 [http://perma.cc/69E5-LH83] (arguing that because “[a]t least some common law rules could be altered by statute,” the Fourth Amendment “sought to ensure that no further encroachments could occur”). Consistent with that view, violations of positive law may have been regarded as a sufficient reason to enforce the Fourth Amendment. See infra Part IV (discussing the positive law floor). And Fourteenth Amendment incorporation later extended that safeguard against states. So in tolerating legislative subversion of Fourth Amendment rights, the positive law model would seem to turn the Founders’ purpose on its head. Perhaps this alternative reading of history is mistaken, but if so, Baude and Stern have not shown why.

Baude and Stern also point out that Founding-era courts generally vindicated Fourth Amendment rights within the context of private tort suits.13×13. See Baude & Stern, supra note 1, at 1840 (citing Richard M. Re, The Due Process Exclusionary Rule, 127 Harv. L. Rev. 1885, 1919–20 (2014)). But why should Founding-era litigation practices be treated as part of the Fourth Amendment’s meaning at all, much less as part of the meaning of a “search”? The Fourth Amendment clearly allows for the creation of new vehicles for judicial remediation, such as modern-day civil rights lawsuits, and Baude and Stern expressly disclaim any desire to recreate the Founding-era system of Fourth Amendment remediation.14×14. See id. at 1839–40. So the remedial backdrop to the early Fourth Amendment, like many other features of the eighteenth-century world, is simply the changeable factual context in which the Fourth Amendment happened to operate. As that factual context changed, so too might the Fourth Amendment’s means of achieving its broad, stated purpose: to secure the people from “unreasonable searches.”15×15. See Re, supra note 13, at 1918–29 (arguing that the Fourth Amendment’s implications might change in light of the historical development of professional police forces). As discussed below, that broad purpose is undermined by focusing only on whether the government has acted in a way that a private person lawfully couldn’t.

B. Deferring the Key Question of Reasonableness

Because it takes a strong position on the meaning of a Fourth Amendment “search” but no clear position on what qualifies as “unreasonable,”16×16. E.g., Baude & Stern, supra note 1, at 1887 (noting that “the positive law model does not answer” “what the necessary reasonableness entails”). the positive law model operates primarily as a threshold barrier to any Fourth Amendment claim.17×17. E.g., id. at 1829, 1832, 1840. To be sure, the positive law model might sometimes find a “search” where current doctrine would not, but in those cases the positive law model yields only hollow results: satisfying the demands of the positive law model is little more than a prelude to the deeper and more difficult question of reasonableness. Moreover, the reasonableness inquiry that the positive law model leaves unanswered will have to consider the underlying interests that the Fourth Amendment should protect, and those interests presumably include privacy. It therefore seems inevitable that consideration of something resembling the Katz v. United States18×18. 389 U.S. 347 (1967). “reasonable expectation of privacy” test19×19. See id. at 360–61 (Harlan, J., concurring). is simply being deferred but one step in every case where the positive law model is triggered.

It may be tempting to respond that the positive law model should be paired with a stringent standard of reasonableness review, such as a categorical warrant requirement. But Baude and Stern wisely avoid that more comprehensive position. To see why, consider the following scenarios, each of which would seem to trigger the positive law model:

  • Government physicians fail to comply with FDA regulations while they conduct pharmaceutical trials to search for a cure for cancer.
  • Administrators at a government installation transgress workplace-safety laws while building a telescope to search for a distant planet.
  • A state university professor of philosophy searches for the meaning of life by holding onto her copy of Fear and Trembling well past its due date.

At risk of stating the obvious, the activities and laws at issue in the above examples are far afield from the interests in privacy and security that are thought to underlie the Fourth Amendment. In order to acknowledge as much, the positive law model would have to be supplemented with some additional principle about what qualifies as an unreasonable search. And that principle would seem to reintroduce at least some consideration of privacy, à la Katz.

Reasonableness considerations would also have to come into play to prevent the positive law model from allowing localities to hamstring superior governments. Imagine that a city is hostile toward state or federal law enforcement officers and so enacts an ordinance prohibiting all investigatory conduct — such as interviewing witnesses, patrolling the streets, and following individuals on open roads. Further imagine that the ordinance triggers only nominal damages and that private parties feel free to ignore it. Even so, the positive law model would demand that courts treat all banned investigatory conduct in the city as a Fourth Amendment “search.” Would state and federal investigators then have to obtain a warrant, or make some other showing, to perform even the most basic aspects of police work? If yes, then the positive law model would have allowed a city to stymie superior state and federal governments, contrary to the federal structure. If not, then the real work would once again occur as part of reasonableness review — outside the bounds of the positive law model.

Finally, consider Baude and Stern’s claim that the positive law model would improve on current law by replacing the third-party doctrine.20×20. See Baude & Stern, supra note 1, at 1873–74. Given extant privacy legislation, the positive law model would view government orders to disclose metadata regarding electronic communications as a “search.”21×21. See id. But so does current Fourth Amendment case law, which scrutinizes subpoenas and similar coercive orders for reasonableness.22×22. See, e.g., See v. City of Seattle, 387 U.S. 541, 544 (1967); Big Ridge, Inc. v. Fed. Mine Safety & Health Review Comm’n, 715 F.3d 631, 646 (7th Cir. 2013) (“A subpoena also implicates the Fourth Amendment, but only to the extent of requiring that the demand for information be ‘sufficiently limited in scope, relevant in purpose, and specific in directive so that compliance will not be unreasonably burdensome.’” (quoting See, 387 U.S. at 544)). The hard question under current law is whether the applicable reasonableness standard is too lax, given relevant privacy interests23×23. Compare United States v. Graham, 796 F.3d 332, 360–61 (4th Cir. 2015) (holding that a warrant is required for ordering disclosure of cell-site locational data), with id. at 378 (Motz, J., dissenting in part and concurring in the judgment) (maintaining that no warrant is required given the third-party doctrine). — but that is a subject on which the positive law model takes no clear position. Given extant laws, the positive law model would also prohibit the government from directing or excusing unconsented disclosures of electronic content information, such as the text of emails.24×24. See Baude & Stern, supra note 1, at 1874–75. But, again, so too does current doctrine. In fact, current doctrine does one better in concluding that the government generally may not demand content information without a warrant.25×25. See United States v. Warshak, 631 F.3d 266 (6th Cir. 2010); see also infra notes 71, 104 and accompanying text. Once again, the positive law model does not address the issue of reasonableness and so offers only half of a solution.

There is a broader point here, one that extends beyond the positive law model. The “reasonable expectation of privacy” test for whether a search has taken place often blurs with the textually similar inquiry into whether the search was “unreasonable.”26×26. See Akhil Reed Amar, Fourth Amendment First Principles, 107 Harv. L. Rev. 757, 769 (1994). As a result, many cases that purport not to find a search are actually best understood as resting on an implicit — and sometimes not so implicit — finding that the police acted reasonably, even though they didn’t have reasonable suspicion. In other words, the threshold Katz test for whether a “search” has occurred is best viewed as a test for whether it is “unreasonable” for police to undertake the investigative technique at issue based on a mere hunch. And when a Fourth Amendment claim survives that threshold inquiry, courts have to go on to fashion a doctrinal rule to govern the police conduct at issue.

Once the Katz test is understood as addressing the ultimate issue of reasonableness, it becomes clear that the positive law model alone cannot fill Katz’s shoes: even after finding a search, courts applying the positive law model must still figure out whether the search was reasonable without suspicion. And that additional inquiry would reintroduce all the judgment and indeterminacy that Baude and Stern seek to avoid. So when Baude and Stern avoid hard “reasonable expectation of privacy” questions while deferring consideration of “unreasonableness,” they are really just relabeling and postponing the hard questions that arise under Katz, without answering them.

C. Generating Administrability Problems

Baude and Stern celebrate the positive law model for its purported administrability relative to current doctrine,27×27. See Baude & Stern, supra note 1, at 1833, 1876. but their proposal would generate serious problems of its own.

As an initial matter, key components of the positive law model remain underspecified. For instance, the positive law model is triggered only when a legal violation occurs “in the course” of a search28×28. Id. at 1833. — but is that requirement temporal, causal, or purposive? The positive law model also declines to take a comprehensive position on who is actionably searched in any given instance, or on when consent should be viewed as tainted by unconstitutional conditions.29×29. See id. at 1864, 1878–80. Notably, one of the more attractive features of the positive law model resembles the positive law floor by maintaining that private causes of action should establish “a minimum” of Fourth Amendment standing — without specifying a maximum. Id. at 1880. And, as explained below, the positive law model does not specify which private-party analogy is appropriate in any given situation.

Problems persist when it comes time to assess whether a positive law violation occurred. In the kinds of scenarios where current Fourth Amendment doctrine yields hard cases, laws applicable to private parties are often indeterminate.30×30. See, e.g., id. at 1877 n.283 (“[T]he ‘civil trespass by deception cases themselves are largely a mess . . . .’” (quoting Laurent Sacharoff, Trespass and Deception, 2015 BYU L. Rev. 359, 363)). Part of the reason for this indeterminacy is that many recurring situations involving police, like the use of drug-sniffing dogs, only rarely arise in connection with private parties. And even widely litigated laws applicable to private parties frequently rest on open-ended normative standards like “reasonableness,” as illustrated by trespass and many privacy torts.31×31. See, e.g., Sanders v. ABC, Inc., 978 P.2d 67, 72 (Cal. 1999). Indeed, the indeterminacy of many privacy laws — including the Fourth Amendment — is what allows them to apply to unforeseeable questions. So to invoke laws applicable to private parties is often simply to move from one hard question to another one. Confirming as much, Baude and Stern repeatedly reserve judgment as to how the positive law model would apply in commonly arising situations.32×32. See, e.g., Baude & Stern, supra note 1, at 1835; see also id. at 1874 (“[A]pplication of the positive law model will frequently be context-specific and even jurisdiction-specific, so generalizations are perilous.”).

Finally, there are systemic considerations. Today, many Fourth Amendment rules are well settled and nationally applicable, but the positive law model would make all of those rules contingent and jurisdictionally variable, depending on changes in all other areas of federal and state law.33×33. See id. at 1851. For example, a federal officer might have to change her policing methods multiple times a day as she transited the Holland Tunnel between New Jersey and New York. Further, the positive law model would prompt new changes in laws applicable to private parties. As discussed below,34×34. See infra section II.B. for instance, the positive law model would create an incentive for lawmakers to adjust privacy protections for private parties so as to expand the power of law enforcement. And even if it were desirable as a matter of first principles, the positive law model would generate considerable transition and implementation costs.

Ultimately, administrability alone is not a sufficient basis for choosing a Fourth Amendment theory: we should start by asking what is the right theory — the one that aims to achieve the right goals — before trying to maximize administrability. That question leads to the next Part.

II. Analogy and the Positive Law Model

The positive law model maintains that, if the people lose their legal security against private intrusions, then they should also lose their constitutional security against police. But private parties and the police are different, and those differences often provide good reason to think that police should be subject to greater regulation than private parties. This Part shows that the private-party analogy underlying the positive law model is highly imperfect.35×35. See supra note 2.

A. The Limits of Analogy

Baude and Stern defend the positive law model on the ground that it guards against governmental power. But that argument is unpersuasive, even assuming that the Fourth Amendment’s purposes are limited to power-based concerns.36×36. See Baude & Stern, supra note 1, at 1828, 1844. Despite their focus on nonexceptionalism, Baude and Stern acknowledge that the law does and should afford the government many search-and-seizure powers denied to private parties. See, e.g., id. at 1832, 1844. A Rottweiler needs a stronger leash than a Chihuahua, and the government needs legal restrictions where private parties do not.

In general, the government poses special risks of abuse because of its distinctive capabilities, incentives, and social role.37×37. See Bivens v. Six Unknown Named Agents of Fed. Bureau of Narcotics, 403 U.S. 388, 392 (1971) (“An agent acting — albeit unconstitutionally — in the name of the United States possesses a far greater capacity for harm than an individual trespasser exercising no authority other than his own.”). The government’s special capabilities derive from its tax-financed budget, its vast infrastructure and resources for conducting surveillance, and its special ability to pursue criminal prosecutions. As a political institution, the government’s incentives are also distinctive, since it has an uncommon interest in investigating crime and, potentially, in suppressing dissent against the incumbent regime. And the government’s social role as a representative of the people means that it has special moral authority and prestige, as well as a unique capacity to inflict injury by betraying the people’s trust in it. For these reasons, analogies to private parties are necessarily limited in their usefulness — yet the positive law model often affords those analogies dispositive weight.

Start by imagining that the United States government has launched a $10 billion satellite capable of monitoring the movements of everyone in the United States at all times.38×38. This scenario is not so farfetched. See, e.g., Ryan Gallagher, Could the Pentagon’s 1.8 Gigapixel Drone Camera Be Used for Domestic Surveillance?, Slate: Future Tense (Feb. 6, 2013, 10:14 AM), http://www.slate.com/blogs/future_tense/2013/02/06/argus_is_could_the_pentagon_s_1_8_gigapixel_drone_camera_be_used_for_domestic.html [http://perma.cc/99NV-DTNH]. It tracks people everywhere they go, whether they use planes, trains, or automobiles. We can even imagine that it uses X-ray technology to see inside every home in the country and monitor people’s activities within those spaces as well. It seems quite obvious that this device should be subject to Fourth Amendment scrutiny, since it would obliterate privacy from government. But what if existing privacy laws did not apply to orbital observations or the government revised those laws to preclude their applicability to such observations? Under the positive law model, the satellite would then also be exempt from the Fourth Amendment. And the positive law model would stick to that conclusion even if no private party had ever built such a satellite. All that would matter, under the positive law model, is that a private version of the satellite would — as a purely hypothetical matter — comport with the law as it then existed.

For a more familiar example, consider facts like those in Fernandez v. California.39×39. 134 S. Ct. 1126 (2014). When police arrived at a home, one of its inhabitants rejected their request to enter, at which point the police removed him from the scene and placed him under arrest.40×40. Id. at 1130. Police then returned to the home and again asked to enter the house. This time, the police received consent.41×41. Id. Should the recently arrested occupant’s prior denial of consent trump the later consent by the nonarrested inhabitant? Applying the positive law model, Baude and Stern propose asking whether a private person could have lawfully obtained consent to enter.42×42. See Baude & Stern, supra note 1, at 1877. But it would seem to matter that the police had recently removed the individual who had attempted to deny them consent — something that police often do but that a private person is quite unlikely to attempt. And, just before obtaining consent, the police told the consenting party that they had detained the nonconsenting party for prosecution — something that a private party could not do.43×43. See Fernandez, 134 S. Ct. at 1140 (Ginsburg, J., dissenting) (“Police, after all, have power no private person enjoys. They can, as this case illustrates, put a tenant in handcuffs and remove him from the premises.”). Does the positive law model nonetheless require us to imagine a private person who has previously cuffed and removed the nonconsenting party and then returned to explain that the nonconsenting party might be prosecuted? If Baude and Stern answered no, then they would omit critical features of government power; but if they answered yes, then they would have to adjust the positive law to fit the police.

Finally, take the most banal scenario possible: a uniformed officer approaches an individual on a street and asks to talk for a moment. In determining whether a search or seizure has taken place, is the proper private-party analogy to someone in casual clothes who might be exercising their First Amendment rights,44×44. E.g., McCullen v. Coakley, 134 S. Ct. 2518, 2528–29 (2014) (noting similar First Amendment interests). or do we have to imagine a private person who is visually indistinguishable from a police officer, conjures the same emotional reactions as a police officer, and operates as part of a law enforcement system? If we take the latter approach and make the private-party analogy closely match the actual governmental conduct at issue, then we must engage in an imaginative exercise involving something quite unrealistic. But if we settle for a private-party analogy that more closely resembles recognizable private parties, then the positive law model would require us to overlook and distort critical features of the governmental conduct in question. After all, the potential coerciveness of a police officer’s actions stems in part from the police officer’s distinctive capabilities, incentives, and social role; and a police officer, unlike a private party, would not have the same First Amendment interest in addressing a bystander. Current doctrine strives to get at these police-specific concerns by inquiring into what a “reasonable person” would feel free to do under the actual facts at hand.45×45. See Brendlin v. California, 551 U.S. 249, 255 (2007) (discussing the test for when a “seizure” has taken place); United States v. Mendenhall, 446 U.S. 544, 554 (1980) (opinion of Stewart, J.) (same). By contrast, the positive law model assumes that those government-specific considerations do not exist.

These examples point toward a deep tension in the positive law model. Baude and Stern propose imagining that police officers are “stripped of official authority,”46×46. Baude & Stern, supra note 1, at 1825, 1831. but they also recommend focusing on what an “ordinary citizen” could lawfully do.47×47. See id. at 1831, 1861. The problem is that, even when stripped of official authority, a police officer is still very different from an ordinary citizen.

At one juncture, Baude and Stern “imagine a slightly less formalistic refinement of the positive law model.”48×48. See id. at 1865. Without endorsing this refinement, Baude and Stern raise the possibility that “[o]ne might want to look especially hard to see whether laws that are formally neutral with respect to the government in fact smuggle in some kind of hidden legal privilege for government behavior.”49×49. Id. at 1865–66. But if taken seriously, this revision would undo the positive law model. Again, the government’s distinctive capabilities, incentives, and social role regularly affect Fourth Amendment scenarios. Because laws that formally apply to both private parties and the police often have the practical effect of favoring the police, automatic reliance on those laws would systematically underprotect Fourth Amendment values — including not just privacy, but also Baude and Stern’s own concern with limiting governmental power. And if the positive law model were altered so as to reject legal rules that significantly favored the police, then it would quickly cease to be a positive law model at all.

B. How to Choose the Analogy

The positive law model requires that we first define the relevant search and then imagine a private party to engage in it. But how do we define the governmental conduct at issue, and which type of private-party analogue should we imagine? These tasks are often underdetermined and thus call for judgment, despite the positive law model’s claim to easy administrability.

First, the positive law model encounters a level-of-generality problem when defining the governmental conduct at issue. Consider the Sixth Circuit’s recent decision in United States v. Houston,50×50. 813 F.3d 282 (6th Cir. 2016). which found no Fourth Amendment violation where police had placed a camera on a public utility pole and monitored the defendant’s farm for ten weeks.51×51. Id. at 285. The court employed something like the positive law model when it noted that “the camera captured only views that were plainly visible to any member of the public who drove down the roads bordering the farm.”52×52. Id. at 288. Baude and Stern might prefer a closer analogy, such as a private party placing a camera on a public utility pole. But the government had placed the camera on its own property. Should we therefore imagine that a private party has lawfully procured her own utility pole? Maybe that too is the wrong framing since the government may have used tax dollars or eminent domain to acquire the land where the utility pole stands. So should we try to imagine that the private party had exacted funds or lands from other private parties?

Second, the positive law model assumes a sharp public-private dichotomy, but some private parties resemble government actors and vice versa.53×53. See generally David A. Sklansky, The Private Police, 46 UCLA L. Rev. 1165 (1999) (discussing the rise of private security forces). Just think of an executive in a “company town,” or an Amtrak ticket agent whose corporation may be a state actor for some purposes but not others.54×54. See, e.g., Dep’t of Transp. v. Ass’n of Am. R.Rs., 135 S. Ct. 1225, 1231–33 (2015) (holding that Amtrak is a governmental entity for the purposes of determining metrics and standards); Marsh v. Alabama, 326 U.S. 501, 509 (1946) (holding that a state cannot impose criminal punishments for the distribution of religious literature in a company-owned town). See generally Jon D. Michaels, Running Government Like a Business . . . Then and Now, 128 Harv. L. Rev. 1152, 1178–81 (2015) (reviewing Nicholas R. Parrillo, Against the Profit Motive: The Salary Revolution in America, 1780–1940 (2013)) (discussing the malleable boundary between public and private functions). The positive law model creates the possibility that courts will have to confront these hard questions of state action even in routine cases, since even the most humdrum Fourth Amendment claim could demand choosing among competing analogies. And these scenarios will likely multiply as private-public collaborations become ever more routine.

For a basic example of this difficulty, imagine that a police officer temporarily detailed to mall security detains someone on suspicion of shoplifting. Nearby are a private security guard and a passing shopper. Which private party offers the proper analogy? Choosing the security guard seems consistent with Baude and Stern’s suggestion that courts should imagine a “similarly situated” private party,55×55. Baude & Stern, supra note 1, at 1825, 1831. since the security guard behaves, thinks, and looks very much like a police officer. But Baude and Stern also recommend focusing on what an “ordinary citizen” could lawfully do,56×56. Id. at 1831. and the shopper there for a day better fits that description. Neither choice is perfect because no private person is, or could be, quite the same as the police officer. Yet the choice of analogy is critical: as a store employee, a security guard might have the “merchant’s privilege” of investigative detention, whereas a disinterested customer would have only the power of citizen arrest.57×57. See Sklansky, supra note 53, at 1183–84. So the underdetermined choice of analogy could entirely determine how the positive law model applies.

In sum, no analogy to private parties can fully capture what makes the government the special object of the Fourth Amendment. The government’s special capabilities, incentives, and social role critically affect most — and perhaps all — Fourth Amendment scenarios.

III. The Politics of the Positive Law Model

Because it overlooks much of what makes police different from private parties, the positive law model would have harmful political effects. This Part discusses problems related to newly emerging technologies, the rights of politically disempowered groups, and the risk that lawmakers might exploit the positive law model to advantage police.

A. Responsiveness to New Technologies

Baude and Stern argue that the “big changes worked by the positive law model” would pertain to new technologies.58×58. Baude & Stern, supra note 1, at 1861. But those changes would tend to be undesirable.

New technologies are most obviously problematic for the positive law model when the government alone has access to them. Law directed toward private parties might not apply to the new technology and so could leave it unregulated even as the government makes use of it, to the detriment of privacy and other values. Or, as in the NSA satellite example above, lawmakers may be tempted to adjust the laws applicable to private parties so as to enable the government to use the technology without encountering Fourth Amendment scrutiny.

But new technologies can be problematic for the positive law model even when private parties do use them. Lawmakers and the public often allow emerging technologies to go unregulated for significant periods of time without passing any meaningful judgments on them.59×59. See, e.g., Earl Warren, Science and the Law: Change and the Constitution, 12 J. Pub. L. 3, 5 (1963) (“The law lags behind until crisis stirs it into action.”). The new technology might simply be too rare, obscure, or ill-understood to warrant any kind of opinion. And lawmakers may have other priorities that command their attention, creating a regulatory lag. As a result, a lack of regulation directed toward a new technology is not reliable evidence that lawmakers or the public have condoned the technology’s use.60×60. After a period of indeterminacy, preexisting laws may eventually be interpreted to apply to unforeseen technologies. See Kristen E. Eichensehr, Cyberwar & International Law Step Zero, 50 Tex. Int’l L.J. 357, 359 (2015).

Take thermal imagers, which the Court addressed in Kyllo v. United States.61×61. 533 U.S. 27 (2001). According to Baude and Stern, “[a] decision that, say, using a thermal-imaging camera is or is not proscribed by a state’s privacy statute” — and therefore is or is not a “search” under the positive law model — “is likely to elicit some response from the state if it is off-base because that misreading may have wider implications” for private parties.62×62. See Baude & Stern, supra note 1, at 1861. But a legislative response is not “likely” when the technology at issue is only rarely used by private parties. Anticipating that point, Baude and Stern show that thermal scanners are now affordable for many private persons.63×63. See id. at 1864. But if that response is comforting, it is so only because it looks well beyond the positive law model. Indeed, Baude and Stern seem to be engaged in precisely the kind of inquiry suggested in Kyllo itself, which considered whether the technology at issue was “in general public use.”64×64. Kyllo, 533 U.S. at 40.

To make this point obvious, imagine two stages in the rise of thermal-imaging technology. In Stage 1, thermal imagers are large and expensive, requiring a van for transport and costing $1 million each. Later, in Stage 2, thermal imagers come standard on every personal electronic device and so are ubiquitous. Rulings on thermal imagers would have very different political implications in these two contexts. In Stage 1, such rulings are unlikely to affect anyone outside the context of law enforcement. In Stage 2, by contrast, an analogous decision would affect nearly everyone, and not just as a potential user of the technology but also as a potential object of its use. These contextual points control whether a narrow ruling under the positive law model would have “wider implications” and be “likely to elicit some response from the state,” as Baude and Stern promise. Yet the positive law model maintains that these critical contextual points should be legally irrelevant.

Or consider another of Baude and Stern’s principal examples: surreptitious DNA testing, which occurs when someone leaves behind or discards personal genetic material that is then tested by a third party. Baude and Stern report that ten states have enacted statutes banning such testing and that it is “unlikely” that such testing would be a common law larceny in the remaining forty states.65×65. See Baude & Stern, supra note 1, at 1883. According to Baude and Stern, the positive law model provides a good “framework” here: when and only when states prohibit surreptitious DNA testing by private parties, similar police testing becomes a Fourth Amendment search.66×66. See id. at 1882–83. However, Baude and Stern wrongly infer that the forty states “have so far made the judgment that DNA testing of one’s discarded skin cells does not warrant legal prohibition.”67×67. Id. at 1883. Instead of arriving at that conclusion, the states in question may simply be uninformed about the issue or struggling to overcome legislative inertia. Or surreptitious DNA testing by private parties may simply be of low public import because it is relatively rare or tends to be harmless. Given these possibilities, there is no good reason to prevent the Fourth Amendment from regulating police use of surreptitious DNA testing, even where private persons have been left unregulated.

Does the positive law model at least fare better than status quo doctrine, problematic as it is? Even that modest claim seems dubious. While generally upholding surreptitious DNA testing as compliant with the Fourth Amendment, courts have also limited their holdings to testing of noncoding DNA for purposes of identification.68×68. This is true, for instance, of the main case that Baude and Stern cite on this point. See Raynor v. State, 99 A.3d 753, 761 (Md. 2014) (“[T]he character of the information specifically sought and obtained from the DNA testing of Petitioner’s genetic material — whether it revealed only identifying physical characteristics — is paramount in assessing the objective reasonableness of his asserted privacy interest.”). Current doctrine therefore seems to be providing at least some privacy protection against police in the forty jurisdictions that lack specific prohibitions on surreptitious DNA testing by private parties. Perhaps the Fourth Amendment should provide greater protection in this area — but if so, the positive law model is unlikely to help. The regulation of surreptitious DNA testing, even if popular overall or among informed persons, would likely become more difficult if the positive law model were adopted, for law enforcement interests would then lobby against it. The positive law model could thus drag down not just Fourth Amendment protections, but also privacy protections directed against private parties. And even in the ten states where Baude and Stern are confident that their approach would currently deem surreptitious DNA testing a search, the positive law model does not give an answer to the ultimate issue of reasonableness.

B. Exacerbating Democratic Pathologies

Baude and Stern also emphasize that the positive law model takes advantage of the democratic lawmaking process to glean insights into privacy and other values. But when democratic pathologies arise, the positive law model would have perverse effects, causing defects in regular lawmaking to curb the Fourth Amendment.

Democratic lawmaking is often distorted by private interest groups with focused concerns.69×69. See, e.g., Bruce A. Ackerman, Beyond Carolene Products, 98 Harv. L. Rev. 713, 728 (1985). For example, data brokers have an interest in the sale of personal data, so there is a well-organized and financed minority in favor of allowing private parties to share or sell that information — including to the NSA or other government investigators — even if such sales are opposed by a diffuse majority. Indeed, voluntary, informal partnerships between government and the private sector have been a major feature of the war on terror.70×70. See Jon D. Michaels, All the President’s Spies: Private-Public Intelligence Partnerships in the War on Terror, 96 Calif. L. Rev. 901 (2008); Julia Angwin et al., AT&T Helped U.S. Spy on Internet on a Vast Scale, N.Y. Times (Aug. 15, 2015), http://www.nytimes.com/2015/08/16/us/politics/att-helped-nsa-spy-on-an-array-of-internet-traffic.html. The positive law model would approve of these voluntary disclosures of noncontent information and so would insulate such arrangements from Fourth Amendment scrutiny — even though, as Baude and Stern note, recent Fourth Amendment case law seems to recognize the need to regulate those disclosures.71×71. Baude & Stern, supra note 1, at 1874–75.

Further, many laws foreseeably fall especially hard on politically disempowered groups, like the poor or racial minorities, yet the positive law model does not account for those democratic distortions. For example, private businesses have an interest in protecting the legality of closed-circuit televisions, even if the use of those systems primarily affects people who live in dense urban areas. Or take Florida v. Jardines,72×72. 133 S. Ct. 1409 (2013). the case involving a drug-sniffing dog being walked up to the front door of a house. A legal rule that allowed private parties to engage in similar conduct would have no effect on people who live in gated communities. But, under the positive law model, such a rule would give police carte blanche to use drug-sniffing dogs in the less well-to-do locales where police may already be inclined to focus attention. So in curtailing generally applicable privacy laws, politically empowered groups would know that they have special ability to secure their own privacy.

Baude and Stern might respond that the positive law model increases the overall odds that courts would defend Fourth Amendment rights, since the positive law model provides a sure, objective basis for judges to enforce those rights. But as we have already seen, the positive law model doesn’t actually offer an automatic mechanism for the judicial protection of rights. Instead, it imposes a new threshold barrier to Fourth Amendment relief while deferring the hard question of reasonableness. And implementation of the positive law model turns out to require discretionary judgment, both because the choice of private-party analogue is underdetermined and because laws applicable to private parties are hardly more determinate than current Fourth Amendment doctrine. So, on net, the positive law model would seem to preclude important opportunities for courts to live up to their rights-protecting traditional role, without increasing the odds of their doing so.

“If people want to live in fishbowls,” Baude and Stern argue, “the Fourth Amendment should not be what stops them, so long as the government swims alongside them.”73×73. Baude & Stern, supra note 1, at 1866. This statement assumes a homogeneous “people” whose interests are all aligned. In reality, however, the United States is made up of heterogeneous groups, some of which are more powerful than others. In that context, the Fourth Amendment should be a bulwark for unpopular persons and minority groups — a source of countermajoritarian rights. Yet the positive law model would allow politically empowered people to reduce everyone’s privacy from police, perhaps because the police are on their side.

C. The Problem of Exploitation

Baude and Stern celebrate the positive law model for keeping pace with changes in the laws applicable to private parties.74×74. Id. at 1851. But because the positive law model doesn’t consider the nature of those changes, it can be exploited to the advantage of police.

At an extreme, exploitation of the positive law model could negate the paradigm cases that gave rise to the Fourth Amendment and that form the core of current Fourth Amendment protections.75×75. See supra text accompany note 9 (discussing the Fourth Amendment’s history of protecting the home). For instance, a law-and-order state legislature might enact a law providing:

Any private person who reasonably suspects that a residence contains evidence of crime may lawfully enter that residence for the purpose of searching for the evidence.

Given such a statute, the positive law model holds that the Fourth Amendment would no longer play its historical role in constraining police searches of homes. Warrants would no longer be required for police to search homes, and the probable cause standard long applicable to investigative residential searches would be replaced with the reasonable suspicion standard.

A defender of the positive law model might respond that the imagined statute would be politically untenable even in a law-and-order jurisdiction, since it would authorize undesirable private searches.76×76. Baude and Stern respond to this type of concern by doubting that it would arise “regularly” and positing that it would at any rate occur only when “justified.” Baude & Stern, supra note 1, at 1866–67. Yet it is doubtful that even this extreme statute would greatly increase the prevalence of undesirable private-party searches. The basic reasons, as with many examples above, have to do with the government’s distinctive capabilities, incentives, and social role. Private parties do not have access to the government’s sophisticated intelligence apparatus, which allows the police to learn about whether evidence is present in other people’s homes. Further, most private persons are not paid, trained, equipped, and encouraged to ferret out crime and so have little interest in risking a fight or social disapprobation to search one another’s homes for incriminating evidence. What’s more, the statute would likely be applied differently to private parties, as judges and juries might more readily believe that a uniformed officer had “reasonable suspicion” and entered the home “for the purpose of searching it for that evidence,” as opposed to a malicious purpose such as theft or blackmail.

For a less extreme example of exploitation, consider Jardines, which held that the use of a drug-sniffing dog on the porch of a house violated the Fourth Amendment because it had “physically intrud[ed] on [a] constitutionally protected area[],” namely the home.77×77. 133 S. Ct. at 1417. Baude and Stern argue that Jardines should have mined Florida tort law to answer the otherwise insignificant question of whether, when, and precisely how a private party might lawfully walk a drug-sniffing dog up to someone’s house.78×78. See Baude & Stern, supra note 1, at 1834–36. To condition Fourth Amendment protection on such an obscure question is odd for the reasons already discussed. But even if we assume both that this is the right question and that it results in a clear answer in favor of the homeowner, the state legislature (or an elected state common law court) could still reverse that result, including by recognizing an implied easement for dog walkers. The practical effects of that innocent-seeming tweak would be focused on police and their suspects.

IV. An Alternative: The Positive Law Floor

In the limited space available, I would like to gesture toward a better way for Fourth Amendment jurisprudence to learn from the private-party analogy. I call this alternative “the positive law floor.” Whereas the positive law model would treat laws directed toward private parties as a hard ceiling on the meaning of the word “search,” my suggestion is that courts might treat privacy laws directed toward private parties as a presumptive floor on the Fourth Amendment’s prohibition against “unreasonable searches.” On this view, when lawmakers guard against privacy intrusions by private parties, then similar intrusions by the government would be presumptively unreasonable.

A. Repurposing the Private-Party Analogy

In recent years, technological change has transformed law enforcement, prompting some judges and scholars to claim that the legislative process holds the key to police regulation.79×79. See United States v. Jones, 132 S. Ct. 945, 964 (2012) (Alito, J., concurring in the judgment) (“In circumstances involving dramatic technological change, the best solution to privacy concerns may be legislative. A legislative body is well situated to gauge changing public attitudes, to draw detailed lines, and to balance privacy and public safety in a comprehensive way.” (citing Orin S. Kerr, The Fourth Amendment and New Technologies: Constitutional Myths and the Case for Caution, 102 Mich. L. Rev. 801, 805–06 (2004))). After all, rules that spring from the legislative process are plausibly thought to be more detailed, legitimate, and expertly drafted than similar rules promulgated by judges. So the Fourth Amendment’s primary role might be to prompt legislative regulation of police or simply to absorb legislative conclusions about how police ought to be regulated.

Yet there is a way of preserving a distinctive role for the Fourth Amendment while learning from the fruits of the legislative process. When the law forbids private parties from engaging in certain conduct while fashioning a legal exemption for police, the positive law model correctly detects a risk of unjustified governmental exceptionalism. If we assume that private and governmental searches both implicate personal security, a legislative choice to prohibit a private search is powerful evidence that a similar governmental search also implicates personal security. And when lawmakers discriminate in favor of governmental searches, it is fair to infer that lawmakers are asserting an exceptional power for their own agents, thereby justifying probing review for reasonableness under the Fourth Amendment. Thus, courts can draw their own conclusions about privacy and security by leveraging the deliberative work of legislators and other lawmakers. Under this reasoning, laws applicable to private parties go to the issue of reasonableness and supply just one good reason why governmental conduct might warrant more searching review under the Fourth Amendment.

In short, the intuition underlying the positive law floor is this: if it’s objectionable for a private party to encroach on personal privacy or security in a certain way, then it’s at least as objectionable — and probably much more objectionable — for the government to do the same. This approach captures the best aspect of the positive law model: its awareness that police can obtain objectionable exemptions from privacy laws applicable to private parties.80×80. See Baude & Stern, supra note 1, at 1854 (citing Erin Murphy, The Politics of Privacy in the Criminal Justice System: Information Disclosure, the Fourth Amendment, and Statutory Law Enforcement Exemptions, 111 Mich. L. Rev. 485, 503 (2013)). So the positive law floor is an improvement on Fourth Amendment proposals that would simply defer to legislative regulation of police. At the same time, the positive law floor is prepared to look beyond the private-party analogy and so acknowledges what the positive law model overlooks: because of its distinctive capabilities, incentives, and social role, the government poses special threats to the people’s security.

B. Assessing the Positive Law Floor

The positive law floor would have several advantages over the positive law model. As an initial matter, the positive law floor would view laws applicable to private parties as a practical tool for understanding when police have engaged in an “unreasonable” search or seizure, rather than implausibly trying to squeeze the private-party analogy into the word “search.”81×81. See supra section I.A.

More practically, the positive law floor would support the following rule: police searches that would violate privacy-protecting laws if performed by a private party are presumptively “unreasonable” under the Fourth Amendment, with the presumption being overcome, for instance, if the police have an appropriate degree of individualized suspicion.82×82. There is no space here for a full assessment of when the presumption should be overcome. To provide just one example, one might argue that adherence to a transparent search policy should sometimes suffice to justify governmental actions forbidden to private parties since officials are subject to political checks. So when the positive law floor is triggered, police searches justified solely by a hunch or whimsy would be unreasonable. This approach would not categorically deny Fourth Amendment protection when private parties could lawfully engage in the government’s conduct. Thus, Fourth Amendment protection would still be available when laws applicable to private parties are tilted toward the politically powerful or are too feeble to constrain the government’s special incentives and capabilities.83×83. See supra Part III.

The positive law floor would also avoid or mitigate several problems associated with the more rigid positive law model. For one thing, the positive law floor focuses on deliberately adopted laws; by contrast, the positive law model assumes that legislative inaction toward private parties represents tacit approval of similar governmental conduct — a dubious inference given the other explanations for legislative inaction and legal lacunae.84×84. See supra text accompanying note 67. Further, the positive law floor would allow courts to learn from inevitably imperfect private-party analogies, without making the exact choice of analogy critical to Fourth Amendment protection. So courts could vary the strength of the positive law floor’s presumption, depending on the strength of the best private-party analogy and the clarity of the relevant positive law. The positive law floor would also reduce law enforcement’s incentive to lobby against general privacy protections, since police can be regulated even if those protections are not adopted. (The possibility that this incentive would remain significant may be the strongest point against even the positive law floor’s use of the private-party analogy.)

And then there’s the question of whether to fashion local or national Fourth Amendment rules. The positive law model prides itself on incorporating the laws of each jurisdiction, which would yield a contingent and heterogeneous legal regime, as well as empower localities to hamper superior state or federal governments.85×85. See Baude & Stern, supra note 1, at 1861–62; supra section I.C. While the positive law floor could operate in a similar way, a more appealing approach would consider the laws of multiple jurisdictions when fashioning nationwide Fourth Amendment rules, thereby promoting uniformity and stability. So, in general, the positive law floor would be triggered: (i) when the government asserts new exemptions from widespread privacy laws applicable to private parties; and (ii) when new privacy laws extend widespread protections against intrusions by private parties. The positive law floor would thus have both a traditionalist aspect and a modernizing aspect. Notably, this approach would make Fourth Amendment doctrine resemble aspects of Eighth Amendment and substantive due process doctrines, which already consider trends in state and federal laws when fashioning national rules.86×86. Cf. Corinna Barrett Lain, The Unexceptionalism of “Evolving Standards,” 57 UCLA L. Rev. 365, 368–69 (2009) (discussing contexts in which the Supreme Court considers patterns of state law when construing the Constitution); David A. Strauss, The Modernizing Mission of Judicial Review, 76 U. Chi. L. Rev. 859 (2009) (discussing the “modernization” of several areas of law via their anticipation of trends in public opinion).

Baude and Stern might respond that the foregoing advantages come at the price of clarity, a purported hallmark of the positive law model. But as we’ve seen, the positive law model isn’t really so clear and must embrace unspecified exceptions in order to avoid untenable outcomes.87×87. See supra sections I.B, II.A, II.B, III.B.

Ultimately, the biggest difference between the positive law floor and the positive law model has to do with the kinds of answers they provide. The only definitive answers offered by the positive law model favor the government, whereas the definitive answers offered by the positive law floor are favorable to Fourth Amendment claimants.88×88. See supra section I.B.

C. Using the Positive Law Floor

While the Supreme Court has never adopted the positive law floor in so many words, it has long maintained that Fourth Amendment doctrine is attentive to — but not controlled by — laws applicable to private parties.

Take Florida v. Riley,89×89. 488 U.S. 445 (1989). where a plurality approved police use of a helicopter to observe marijuana growing in a partially open greenhouse.90×90. Id. at 447–52 (plurality opinion). The plurality felt that “it is of obvious importance that the helicopter in this case was not violating the law,” apparently because the law provided evidence of customary or frequent behavior.91×91. Id. at 451. As the plurality immediately observed: “there is nothing in the record or before us to suggest that helicopters flying at 400 feet are sufficiently rare in this country to lend substance to respondent’s claim that he reasonably anticipated that his greenhouse would not be subject to observation from that altitude.”92×92. Id. at 451–52.

Further, the plurality denied that “an inspection of the curtilage of a house from an aircraft will always pass muster under the Fourth Amendment simply because the plane is within the navigable airspace specified by law.”93×93. Id. at 451. Confirming as much, the plurality went on to find, for example, that “no intimate details connected with the use of the home or curtilage were observed.”94×94. Id. at 452. In short, the plurality would have viewed the government’s conduct much more skeptically if the police had deviated from laws that protect privacy from private intrusion. And even after concluding that the government complied with those laws, the plurality checked to see whether there was any other reason to be concerned with the reasonableness of the government’s conduct.95×95. See id. While its conclusions on that score may be questioned, the plurality applied something like the positive law floor.

For an analysis that even more closely approaches the positive law floor, consider Justice Scalia’s solo concurrence in Fernandez.96×96. 134 S. Ct. 1126, 1137 (2014) (Scalia, J., concurring). While voting to reject the Fourth Amendment claim, Justice Scalia explained that he would “find this a more difficult case” if “property law did not give [the defendant’s] cotenant the right to admit visitors over [defendant’s] objection.”97×97. Id. at 1138. Justice Scalia then cited treatises and cases from several jurisdictions.98×98. See id. Contra Baude and Stern, Justice Scalia’s attention to broad principles of property law made sense. Property law is avowedly attuned to the kind of privacy factors that matter under the Fourth Amendment — namely, when one person’s consent should authorize third parties to invade another person’s privacy. And Justice Scalia correctly observed that property law should yield a presumptive Fourth Amendment minimum. That approach recognizes that it can be valuable to regulate police even where private actors are unregulated: if the law wouldn’t find consent even for a private person, then there is good reason to think that the law likewise shouldn’t find consent for police.

Justice Scalia’s recent majority opinions have likewise come close to embracing the positive law floor, as both Jardines and United States v. Jones99×99. 132 S. Ct. 945 (2012). emphasized that property law concepts might establish sufficient conditions for Fourth Amendment protection.100×100. See Florida v. Jardines, 133 S. Ct. 1409, 1414 (2013); Jones, 132 S. Ct. at 952. For example, Jardines emphasized that Katz’s “reasonable-expectations test ‘has been added to, not substituted for,’ the traditional property-based understanding of the Fourth Amendment.”101×101. See Jardines, 133 S. Ct. at 1417 (quoting Jones, 132 S. Ct. at 952). The positive law floor would adjust and supplement that statement, most significantly by considering not just “property-based” laws, but also other laws relating to personal privacy.

Yet the positive law floor would do much more than clarify current doctrine. For instance, the positive law floor would allow customers to negate the third-party doctrine and enhance their Fourth Amendment protections by contracting for privacy protections.102×102. See supra text accompanying notes 20–25 (offering a critical assessment of the positive law model’s approach to the third-party doctrine). By choosing to do business with telecoms or other companies that contractually commit to keeping customer information confidential, consumer arrangements would support expanded Fourth Amendment rights: if the government trumped those contractual duties by ordering or excusing disclosure of customer information without a warrant, then it would trigger the positive law floor’s presumptive rule of unreasonableness.103×103. See Randy Barnett, Why the NSA Data Seizures Are Unconstitutional, 38 Harv. J.L. & Pub. Pol’y 3, 13 (2015); cf. Donald A. Dripps, “Perspectives on the Fourth Amendment” Forty Years Later: Toward the Realization of an Inclusive Regulatory Model, Minn. L. Rev. (forthcoming 2016), http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2692296 [http://perma.cc/6GY5-PNUP] (proposing that the Fourth Amendment third-party doctrine depend on the availability of an individual opt-out). This approach not only captures the best feature of the positive law model but also improves on it by addressing the ultimate question of reasonableness. And, just as important, it does all this without closing the door to similar protections for persons who lack the wherewithal to choose privacy-conscious businesses.104×104. But see, e.g., Baude & Stern, supra note 1, at 1875 n.264 (noting “that existing Terms of Service agreements between users and their providers could contain provisions that satisfy the ‘consent’ requirement of positive law”).


In relying much more heavily on the private-party analogy than Justice Scalia ever did, Baude and Stern neglect the valuable lessons underlying Justice Scalia’s admittedly imperfect opinions in Kyllo, Jones, Jardines, and Fernandez. But Baude and Stern also exemplify one of Justice Scalia’s greatest talents: by brilliantly arguing that the Court go several steps too far, they may successfully persuade the Court to take one or two steps in the right direction.


* Assistant Professor of Law, UCLA School of Law. I am grateful to Will Baude and James Stern for many conversations about the merits of the positive law model and the positive law floor. I also thank Donald Dripps, Kristen Eichensehr, Orin Kerr, Jon Michaels, Alicia Solow-Niederman, the Harvard Law Review Forum, and participants in a UCLA faculty workshop.