Home Slideshow

The Fourth Amendment’s Hidden Intrusion Doctrine

The Fourth Amendment’s concept of probable cause is the linchpin of legal standards governing law enforcement actions such as arrests, searches, and seizures. This Article challenges the assumption that the same quantum of evidence can meet the probable cause standard regardless of whether law enforcement seeks to conduct a search, to seize evidence, or to make an arrest, and regardless of the intrusiveness of such search or seizure. This Article demonstrates that the Supreme Court implicitly considers the degree of intrusion into privacy or liberty, not just the quantum or quality of evidence, when determining whether probable cause exists. In doing so, I bring to light the Supreme Court’s “hidden intrusion doctrine.”

By failing to explicitly state that the degree of intrusion is a factor in the probable cause analysis, the Supreme Court injects ambiguity that has many consequences. Some lower courts and law enforcement agencies already balance the quantum or quality of evidence with the severity of intrusion, even without explicit Supreme Court guidance, but others do not. The ambiguity in the doctrine therefore fosters inconsistency and expands police discretion. Moreover, as technological advancements from facial recognition to digital searches reshape investigative techniques, the need for a clear articulation of the probable cause standard is increasingly urgent.

This Article suggests both doctrinal and policy-based proposals that would bring the Supreme Court’s intrusion doctrine out of the shadows and require deliberate consideration of the degree of intrusion in probable cause determinations. Such an approach would preserve law enforcement flexibility while safeguarding individual rights amidst evolving technological landscapes.

Introduction

In 2022, Alonzo Sawyer was arrested after facial recognition software generated a list of potential matches to surveillance images of a man who had recently assaulted a bus driver.1 1.Eyal Press, Does A.I. Lead Police to Ignore Contradictory Evidence?, New Yorker (Nov. 13, 2023), https://www.newyorker.com/magazine/2023/11/20/does-a-i-lead-police-to-ignore-contradictory-evidence.Show More Using the facial recognition software as the basis for probable cause, police arrested Mr. Sawyer and detained him for approximately a week with hardly any other investigation.2 2.Id.Show More After all, the image and software created a nexus between Mr. Sawyer and the crime. There was no question that the person depicted in the surveillance footage was the assailant.

The problem is that Mr. Sawyer was not the person in the image. After police arrested Mr. Sawyer, officers conducted several less invasive investigatory steps: police visited the house where he had been staying to search the premises for the clothing the assailant had been wearing in the surveillance photo and found nothing.3 3.Id.Show More They interviewed his relatives, who confirmed Mr. Sawyer’s alibi that he had been at home many miles away at the time of the assault.4 4.Id. Police also showed the surveillance photograph from the assault to Mr. Sawyer’s probation officer, who at first said he believed the photo was of Mr. Sawyer. Id. The probation officer, however, was told before he made the identification that Mr. Sawyer had been identified by facial recognition. Id. In addition, he had only met Mr. Sawyer wearing a COVID pandemic-era mask. Id. The probation officer later questioned his own identification and reached out to police to say that he had changed his mind—he no longer thought the surveillance image depicted Mr. Sawyer. Id.Show More These actions were not only far less intrusive than arresting Mr. Sawyer; they also revealed that the facial recognition software had been wrong.

Mr. Sawyer’s arrest turned on the definition of probable cause. Police used a match from new, relatively untested software to justify his arrest and detention for several days. Was that evidence enough to meet the probable cause standard to deprive him of his liberty, arguably the most severe intrusion implicated by the Fourth Amendment? What would have happened if police had been required to investigate further using less intrusive methods before arresting Mr. Sawyer? Could his week in jail have been avoided? Does the law require such an intermediate investigation when the only evidence is from new, untested technology?

This Article examines these questions through the lens of the Fourth Amendment’s probable cause standard, challenging the predominant understanding that the same quality or quantum of evidence—the evidentiary inputs—can satisfy the standard regardless of the degree of intrusion police want to exercise. The term probable cause comes directly from the text of the Fourth Amendment, which requires that “no Warrants shall issue, but upon probable cause, supported by Oath or affirmation, and particularly describing the place to be searched, and the persons or things to be seized.”5 5.U.S. Const. amend. IV (emphasis added).Show More Drawing on this text, probable cause has become the standard6 6.In this Article, I focus only on the Fourth Amendment intrusions that rely on probable cause to demonstrate how that analysis also includes the degree of intrusion, even though the Supreme Court does not explicitly acknowledge it as it does in the reasonable suspicion context. There are other types of search and seizure that do not require probable cause, such as searches incident to a lawful arrest, inventory searches, and sobriety checkpoints, which are not justified by any degree of suspicion that a specific person has committed a crime, see infra notes 16, 69, 104 and accompanying text, and so-called Terry stops, which are supported by reasonable suspicion, a lower standard than probable cause. See infra note 37.Show More by which law enforcement can, while investigating a crime, conduct three separate and distinct actions that implicate individuals’ privacy or liberty interests7 7.Law enforcement may also collect evidence through many different types of actions that require little, if any, individualized suspicion that evidence will be found in a particular place. See, e.g., United States v. Tuggle, 4 F.4th 505, 510–11 (7th Cir. 2021); Smith v. Maryland, 442 U.S. 735, 745–46 (1979); 18 U.S.C. § 2703(d) (allowing disclosure of certain electronic communications based on the government’s representation that records are “relevant and material to an ongoing criminal investigation”). Nevertheless, as discussed elsewhere in this Article, the probable cause standard is required for many of the actions law enforcement conduct that most deeply intrude into individuals’ privacy. See infra Section I.A.Show More: arrest,8 8.See, e.g., Maryland v. Pringle, 540 U.S. 366, 368 (2003). Notably, police do not need a warrant to arrest someone in public, as long as the arrest is based on probable cause. United States v. Watson, 423 U.S. 411, 423–24 (1976).Show More search of a constitutionally protected area,9 9.See, e.g., Illinois v. Gates, 462 U.S. 213, 246 (1983).Show More and seizure of evidence.10 10.See, e.g., Chambers v. Maroney, 399 U.S. 42, 51–52 (1970). Probable cause is also the standard for many other iterations of these basic intrusions. See infra notes 20–25 and accompanying text.Show More

Most scholars would say that the same standard of probable cause applies11 11.See 2 Wayne R. LaFave, Jerold H. Israel, Nancy J. King & Orin S. Kerr, Criminal Procedure § 3.3(a), at 118 n.8 (4th ed. 2015) (“Thus, in Spinelli v. United States, 393 U.S. 410 . . . (1969), where the issue was whether the challenged search warrant had been issued on probable cause, the Court found its earlier decision in Draper v. United States, 358 U.S. 307 . . . (1959), involving grounds to arrest, to be a ‘suitable benchmark’ for resolving that issue. Similarly, Illinois v. Gates, 462 U.S. 213 . . . (1983), which significantly altered the probable cause formula in a search warrant context, has readily been deemed applicable to arrest warrants as well. State v. Schroeder, 450 N.W.2d 423 (N.D. 1990).”). The treatise authors note that the question of “probable cause of what” will differ in various contexts, but the general quantum of evidence and the analysis is the same. Id. § 3.3(a). See also Craig S. Lerner, The Reasonableness of Probable Cause, 81 Tex. L. Rev. 951, 953–54 (2003) (“[P]robable cause is widely viewed in the legal community as a fixed standard.”).Show More regardless of whether police wanted to get a warrant for Mr. Sawyer’s phone records or arrest him. I argue that the law requires more, or stronger, evidence to justify a greater intrusion into an individual’s privacy or liberty than the evidence necessary to justify a lesser intrusion. Under such an approach, for example, the facial recognition match might, alone, provide probable cause to obtain phone records, but not to arrest Mr. Sawyer and deprive him of his liberty.

In this Article I show that, contrary to widespread assumptions, the Supreme Court has implicitly acknowledged that whether probable cause exists to justify any given search or seizure depends in part on the degree of intrusion the particular search or seizure entails. For example, the Court has required more or stronger evidence to justify the search of a home or a custodial arrest than it has for the search of a car, even though “probable cause” is the standard for all three. But the assumption that the degree of intrusion does not factor into the totality of the circumstances12 12.See Gates, 462 U.S. at 246.Show More test used to evaluate probable cause is understandable because the Supreme Court has failed to say so explicitly. This Article identifies and defines these considerations in what I term the Fourth Amendment’s “hidden intrusion” doctrine.

The Supreme Court’s vagueness on its intrusion doctrine and the ensuing failure of lower courts and law enforcement to consistently consider the degree of intrusion in determining whether probable cause exists have injected ambiguity that has many consequences,13 13.See Section I.C.Show More including arrests like Mr. Sawyer’s. It allows police to proceed less cautiously, moving straight to an arrest without any intermediate investigation. It gives cover to police who arrest when they could first search to confirm their suspicions, and it allows them to conduct a more intrusive search when a less intrusive search would suffice. It insulates police decision-making from judicial review and civil liability. It sows confusion for trial courts applying the probable cause standard, for judges reviewing warrant applications, and for officers on the street trying to conduct police work that comports with the Fourth Amendment. It undermines transparency and consistency in the application of the probable cause standard. It generally expands police power.

Though some scholars have argued that balancing the quantum or quality of evidence collected by police with the degree of intrusion involved in a search or seizure should be part of the Fourth Amendment analysis,14 14.See Christopher Slobogin, Let’s Not Bury Terry: A Call for Rejuvenation of the Proportionality Principle, 72 St. John’s L. Rev. 1053, 1053–54 (1998) [hereinafter Slobogin, Let’s Not Bury Terry] (“That framework, which I call the proportionality principle, is very simple: A search or seizure is reasonable if the strength of its justification is roughly proportionate to the level of intrusion associated with the police action.”); Andrew Manuel Crespo, Probable Cause Pluralism, 129 Yale L.J. 1276, 1343 (2020). Over thirty years ago, Christopher Slobogin recognized the value of incorporating the degree of intrusion into search and seizure analysis in his article imagining how the government would consider a search and seizure doctrine if the Fourth Amendment did not exist. See Christopher Slobogin, The World Without a Fourth Amendment, 39 UCLA L. Rev. 1, 49–50 (1991) [hereinafter Slobogin, The World Without a Fourth Amendment]. Slobogin argued that the “proportionality principle” was not part of existing Fourth Amendment doctrine but represented the proper balance of interests between the State and the populace. Id. (“Put another way, because the cost to individual interests of a mistake by the police is greatest when the intrusion is greatest, and diminishes as the intrusion lessens, the tolerance for such mistakes should vary inversely with the level of intrusion. Note further that the cost to the state of a police mistake is also greatest when the intrusion is greatest, because the public is more likely to perceive unnecessary police intrusions as illegitimate when they are significant.” (emphasis omitted)). This Article argues that Slobogin’s “proportionality principle” is already being applied by the courts, even if it has never been explicitly incorporated into Fourth Amendment doctrine.Show More this Article is the first to argue that such balancing is already part of the jurisprudence. The common, unitary understanding of probable cause—that one should entirely ignore the degree of intrusion in deciding whether probable cause exists to justify a particular search or seizure—misapplies important nuances in Supreme Court doctrine.

In addition, I show that there is no need for the Supreme Court’s intrusion doctrine to remain hidden and that bringing it to light would increase transparency and consistency in decisions made by lower courts and law enforcement. There is in fact broad understanding among system actors that the degree of intrusion is relevant in evaluating whether a search or seizure is justified. Some courts and police are already doing such balancing on an ad hoc basis,15 15.See infra Section II.A.Show More requiring more evidence to justify the search of a home than a car, or to justify a search of someone’s crotch than less private parts of their body. But without guidance from the Supreme Court sanctioning such consideration, they are merely acting on the instinct that the degree of intrusion should be part of the search and seizure calculus. The result is a hodgepodge legal doctrine of probable cause that rarely discusses degree of intrusion, even if it is implicitly part of the analysis. Police and judges are left with no coherent limiting principle—some follow their instinct that more or more reliable evidence should be necessary to justify a greater intrusion into individuals’ privacy, but others do not.

An examination of how the police apply the probable cause standard in two areas of emerging technology provides useful case studies for the necessity of a clear intrusion doctrine. As untested technologies, like the facial recognition technology used in Mr. Sawyer’s case, are used to develop leads that then justify searches, seizures, and arrests, courts and police need guidance to help understand how new technologies should be used in the probable cause analysis. A clear intrusion doctrine would guide courts and policy-makers as they respond to an ever-changing technological law enforcement landscape, providing guardrails that protect citizens from undiscovered flaws in new technology while still allowing law enforcement agencies to use new technology. An explicit understanding of how the degree of intrusion applies in the probable cause analysis would encourage police to tread carefully with untested technology. It would encourage them to search before they arrest and to minimize harm from unanticipated errors.

Likewise, bringing the Supreme Court’s hidden intrusion doctrine into the light—by making clear that whether probable cause exists depends in part on the degree of intrusion of a particular search or seizure—could help solve the vexing problems around seizure and search of cell phones. As with any physical evidence they seize in an investigation, police must generally have probable cause to seize a cell phone that they believe contains evidence of a crime.16 16.See, e.g., Chambers v. Maroney, 399 U.S. 42, 51–52 (1970). There are some limited circumstances in which the seizure of physical evidence need not be based on probable cause, such as in the course of an inventory search. See, e.g., Colorado v. Bertine, 479 U.S. 367, 371 (1987).Show More As the Supreme Court recognized in Riley v. California, “[m]odern cell phones are not just another technological convenience. With all they contain and all they may reveal, they hold for many Americans ‘the privacies of life.’”17 17.Riley v. California, 573 U.S. 373, 403 (2014).Show More Police may have reason to believe that some particular type of evidence may be stored in a cell phone, but does that give them license to search every part of the phone? And if it does, should it? Courts authorizing search warrants have grappled with this question and have come to differing opinions on whether the same evidence to seize a phone or search part of it would justify a search of the entire phone.18 18.See infra notes 239–41.Show More Recognition of the Fourth Amendment’s hidden intrusion doctrine would help resolve this tension, requiring more evidence to justify a search of a phone than a seizure of a phone.

In a world in which emerging technologies like facial recognition technology are implemented quickly, often before any meaningful analysis of their reliability or risks, or where such technologies allow police to glean an entirely different degree of information about individuals’ private lives as done through phone searches, courts and policy-makers should be willing to move more slowly. This Article argues that in light of the Supreme Court’s hidden intrusion doctrine, they must move more slowly, and it proposes an approach that I call “graduated probable cause.” Under this approach, to comply with the (previously hidden, now explicit) intrusion doctrine, police would have to collect more or stronger evidence to justify greater intrusions like an arrest or a search of a home. If they had less evidence, or less reliable evidence, they could conduct a lesser intrusion, such as the search of a car or seizure of a cell phone, as an intermediate step. Such a process would maintain law enforcement flexibility, but also protect individuals’ privacy and liberty interests in the face of a constantly changing technological landscape.

This Article proceeds in four Parts. In Part I, I provide an overview of probable cause, examining both how the Supreme Court has defined the standard and how previous scholars have attempted to untangle that messy doctrine, to illustrate that Supreme Court probable cause precedent already implicitly considers the degree of intrusion a particular search or seizure makes into an individual’s privacy or liberty. In Part II, I examine case law from lower federal courts and state courts to demonstrate how, despite broad statements about a unitary standard, courts have already drawn distinctions in practice about the quantum of evidence necessary to provide probable cause based on the degree of intrusion. I also demonstrate how police trainings and policies can either reinforce the idea of a unitary standard to preserve police discretion or provide a model demonstrating the workability of a probable cause analysis that does include consideration of the degree of intrusion. In Part III, I examine in depth the case studies of facial recognition software and phone searches to demonstrate the confusion created by a unitary probable cause standard and the ways in which explicit consideration of the degree of intrusion could improve probable cause analysis in a rapidly changing digital environment. In Part IV, I articulate possible doctrinal and policy solutions for rearticulating the probable cause standard based on the degree of intrusion. Acknowledging that the degree of intrusion is already implicitly part of the analysis, I argue for explicitly incorporating it into the totality of the circumstances analysis that courts already apply. In other words, the courts do not need a new framework; they just need to honestly articulate the interest balancing that already drives decisions. I also set forth policy proposals for police, based on a graduated approach to investigation already endorsed by some law enforcement agencies, that would direct police to first conduct less intrusive searches and seizures in circumstances where they are relying on less, or less reliable, evidence.

  1.  Eyal Press, Does A.I. Lead Police to Ignore Contradictory Evidence?, New Yorker (Nov. 13, 2023), https://www.newyorker.com/magazine/2023/11/20/does-a-i-lead-police-to-ignore-contradictory-evidence. ↑
  2.  Id. ↑
  3.  Id. ↑
  4.  Id. Police also showed the surveillance photograph from the assault to Mr. Sawyer’s probation officer, who at first said he believed the photo was of Mr. Sawyer. Id. The probation officer, however, was told before he made the identification that Mr. Sawyer had been identified by facial recognition. Id. In addition, he had only met Mr. Sawyer wearing a COVID pandemic-era mask. Id. The probation officer later questioned his own identification and reached out to police to say that he had changed his mind—he no longer thought the surveillance image depicted Mr. Sawyer. Id. ↑
  5.  U.S. Const. amend. IV (emphasis added). ↑
  6.  In this Article, I focus only on the Fourth Amendment intrusions that rely on probable cause to demonstrate how that analysis also includes the degree of intrusion, even though the Supreme Court does not explicitly acknowledge it as it does in the reasonable suspicion context. There are other types of search and seizure that do not require probable cause, such as searches incident to a lawful arrest, inventory searches, and sobriety checkpoints, which are not justified by any degree of suspicion that a specific person has committed a crime, see infra notes 16, 69, 104 and accompanying text, and so-called Terry stops, which are supported by reasonable suspicion, a lower standard than probable cause. See infra note 37. ↑
  7.  Law enforcement may also collect evidence through many different types of actions that require little, if any, individualized suspicion that evidence will be found in a particular place. See, e.g., United States v. Tuggle, 4 F.4th 505, 510–11 (7th Cir. 2021); Smith v. Maryland, 442 U.S. 735, 745–46 (1979); 18 U.S.C. § 2703(d) (allowing disclosure of certain electronic communications based on the government’s representation that records are “relevant and material to an ongoing criminal investigation”). Nevertheless, as discussed elsewhere in this Article, the probable cause standard is required for many of the actions law enforcement conduct that most deeply intrude into individuals’ privacy. See infra Section I.A. ↑
  8.  See, e.g., Maryland v. Pringle, 540 U.S. 366, 368 (2003). Notably, police do not need a warrant to arrest someone in public, as long as the arrest is based on probable cause. United States v. Watson, 423 U.S. 411, 423–24 (1976). ↑
  9.  See, e.g., Illinois v. Gates, 462 U.S. 213, 246 (1983). ↑
  10.  See, e.g., Chambers v. Maroney, 399 U.S. 42, 51–52 (1970). Probable cause is also the standard for many other iterations of these basic intrusions. See infra notes 20–25 and accompanying text. ↑
  11.  See 2 Wayne R. LaFave, Jerold H. Israel, Nancy J. King & Orin S. Kerr, Criminal Procedure § 3.3(a), at 118 n.8 (4th ed. 2015) (“Thus, in Spinelli v. United States, 393 U.S. 410 . . . (1969), where the issue was whether the challenged search warrant had been issued on probable cause, the Court found its earlier decision in Draper v. United States, 358 U.S. 307 . . . (1959), involving grounds to arrest, to be a ‘suitable benchmark’ for resolving that issue. Similarly, Illinois v. Gates, 462 U.S. 213 . . . (1983), which significantly altered the probable cause formula in a search warrant context, has readily been deemed applicable to arrest warrants as well. State v. Schroeder, 450 N.W.2d 423 (N.D. 1990).”). The treatise authors note that the question of “probable cause of what” will differ in various contexts, but the general quantum of evidence and the analysis is the same. Id. § 3.3(a). See also Craig S. Lerner, The Reasonableness of Probable Cause, 81 Tex. L. Rev. 951, 953–54 (2003) (“[P]robable cause is widely viewed in the legal community as a fixed standard.”). ↑
  12.  See Gates, 462 U.S. at 246. ↑
  13.  See Section I.C. ↑
  14.  See Christopher Slobogin, Let’s Not Bury Terry: A Call for Rejuvenation of the Proportionality Principle, 72 St. John’s L. Rev. 1053, 1053–54 (1998) [hereinafter Slobogin, Let’s Not Bury Terry] (“That framework, which I call the proportionality principle, is very simple: A search or seizure is reasonable if the strength of its justification is roughly proportionate to the level of intrusion associated with the police action.”); Andrew Manuel Crespo, Probable Cause Pluralism, 129 Yale L.J. 1276, 1343 (2020). Over thirty years ago, Christopher Slobogin recognized the value of incorporating the degree of intrusion into search and seizure analysis in his article imagining how the government would consider a search and seizure doctrine if the Fourth Amendment did not exist. See Christopher Slobogin, The World Without a Fourth Amendment, 39 UCLA L. Rev. 1, 49–50 (1991) [hereinafter Slobogin, The World Without a Fourth Amendment]. Slobogin argued that the “proportionality principle” was not part of existing Fourth Amendment doctrine but represented the proper balance of interests between the State and the populace. Id. (“Put another way, because the cost to individual interests of a mistake by the police is greatest when the intrusion is greatest, and diminishes as the intrusion lessens, the tolerance for such mistakes should vary inversely with the level of intrusion. Note further that the cost to the state of a police mistake is also greatest when the intrusion is greatest, because the public is more likely to perceive unnecessary police intrusions as illegitimate when they are significant.” (emphasis omitted)). This Article argues that Slobogin’s “proportionality principle” is already being applied by the courts, even if it has never been explicitly incorporated into Fourth Amendment doctrine. ↑
  15.  See infra Section II.A. ↑
  16.  See, e.g., Chambers v. Maroney, 399 U.S. 42, 51–52 (1970). There are some limited circumstances in which the seizure of physical evidence need not be based on probable cause, such as in the course of an inventory search. See, e.g., Colorado v. Bertine, 479 U.S. 367, 371 (1987). ↑
  17.  Riley v. California, 573 U.S. 373, 403 (2014). ↑
  18.  See infra notes 239–41. ↑

Free Speech, Breathing Space, and Liability Insurance

An important piece of the “speech-tort” picture has been almost completely missing from doctrinal and policy analysis: the role played by liability insurance in protecting speech. In New York Times Co. v. Sullivan, the Supreme Court began adopting First Amendment restrictions on liability for defamation and the other speech torts—false light, intentional infliction of emotional distress, and public disclosure of private information. The Court’s purpose was to create “breathing space” for valuable speech by precluding liability for some speech that has no constitutional value in itself.

However, there is a little-known but highly important liability insurance regime that also affords breathing space, more broadly than the constitutional rules, by insuring against liability for unprotected speech and the costs of defending virtually all speech-tort suits, regardless of their validity. There have been decades of extensive legal scholarship about the First Amendment’s restrictions on speech-tort liability. Yet this scholarship has largely ignored the fact that all the liability for the speech torts that the First Amendment does permit can be, and often is, covered by liability insurance. In addition, Supreme Court Justices Thomas and Gorsuch recently have separately criticized existing constitutional limitations on liability for defamation as too broad without any mention of the widespread existence and availability of insurance protecting against liability for defamation. The Justices’ criticisms of defamation law have garnered a lot of attention and a barrage of responses, which have also omitted any reference to the possible relevance and significance of liability insurance to the debate about the proper scope of liability for defamation.

This Article takes insurance against speech-tort liability out of the shadows, bringing First Amendment theory and doctrine into the orbit of thinking about liability insurance and its operation in practice. The Article identifies and analyzes the sources and scope of the coverage that insurance provides against speech-tort liability, combining insights about the complex and intertwined consequences of the threat of speech-tort liability with what we know about how liability insurance both creates breathing space and attempts to mitigate excess risk-taking by those who are insured. The Article argues that, whether the end result is to change the law or simply to provide a firmer and more knowledgeable foundation for maintaining the law as it now stands, proponents of reform should either invoke the availability of liability insurance in support of their position or explain why their analyses ignore it. And opponents of reform should explain why they maintain their support of the status quo in spite of the availability of liability insurance. Finally, the Article considers the relevance of liability insurance to different theories of tort liability and analyzes the principal possible alternative to the current constitutional limits on liability, a negligence standard, concluding that such a standard would have considerable deficiencies.

Introduction

The proper scope of liability for defamation—impugning the reputation of an individual or organization—has become a salient public issue. The First Amendment provides considerable protection against liability for defamation (and certain other speech-related torts)1 1.As I explain below, these torts are false light, intentional infliction of emotional distress (“IIED”), and public disclosure of private facts. See infra Section II.A.Show More in order to provide “breathing space”2 2.See infra Section II.B.Show More—an extra margin of protection designed to ensure that free speech is not unduly deterred by the threat of liability.3 3.See Leslie Kendrick, Speech, Intent, and the Chilling Effect, 54 Wm. & Mary L.Rev. 1633, 1636–37 (2013) (explaining that protecting negligent false statements from liability provides constitutional “breathing space”); Frederick Schauer, Fear, Risk and the First Amendment: Unraveling the “Chilling Effect,” 58 B.U. L. Rev. 685, 710–18 (1978) (describing the role of a constitutional “buffer zone” in minimizing the chilling of protected activities).Show More Regardless of these protections, there are still suits for defamation, brought not only by private individuals, but also by, and against, public officials and public figures.4 4.See, e.g., US Dominion, Inc. v. Fox News Network, LLC, 293 A.3d 1002 (Del. Super. Ct. 2023) (voting machine manufacturer against a television broadcaster based on conduct of various news anchors). The parties later reached a settlement for $787.5 million. David Folkenflik & Mary Yang, Fox News Settles Blockbuster Defamation Lawsuit with Dominion Voting Systems, NPR (Apr. 18, 2023, 4:35 PM), https://www.npr.org/2023/04/18/117033911‌4/fox-news-settles-blockbuster-defamation-lawsuit-with-dominion-voting-systems [https://p‌erma.cc/DTD5-TLCE]. See generally Freeman v. Giuliani, 691 F. Supp. 3d 32 (D.D.C. 2023) (election workers against a media personality), judgment entered, No. 21-cv-03354, 2023 WL 9783148 (D.D.C. Dec. 18, 2023), appeal dismissed, No. 24-7021, 2025 WL 634800 (D.C. Cir. Feb. 26, 2025); Carroll v. Trump (Carroll I), No. 20-cv-07311, 2024 WL 97359 (S.D.N.Y. Jan. 9, 2024) (writer against Donald Trump); Carroll v. Trump (Carroll II), No. 22-cv-10016, 2023 WL 3000562 (S.D.N.Y. Mar. 20, 2023) (same); Depp v. Heard, No. CL-2019-0002911, 2022 WL 2342058 (Va. Cir. June 24, 2022) (actor against another actor).Show More These suits receive high-profile attention. Such suits often not only seek personal vindication and compensation for reputational harm, but today, they also figure in the political process; they are moves in a larger set of thrusts and parries occurring outside the direct confines of tort litigation. But the law of defamation should be a law for all seasons, not one bent to suit transient political passions.

Into this picture have come two Justices of the Supreme Court. In judicial opinions, Justice Thomas and Justice Gorsuch have separately criticized existing constitutional limitations on liability for defamation as too broad, Justice Thomas on originalist grounds5 5.McKee v. Cosby, 139 S. Ct. 675, 675–76 (2019) (Thomas, J., concurring in the denial of certiorari).Show More and Justice Gorsuch because of the quick and easy harm to reputation that can occur in the world of digital media.6 6.Berisha v. Lawson, 141 S. Ct. 2424, 2427–28 (2021) (Gorsuch, J., dissenting from the denial of certiorari).Show More Both Justices contend, in effect, that the Constitution should authorize more liability for defamation than it currently permits. In response, a number of established First Amendment scholars have defended existing precedent,7 7.See, e.g., Matthew L. Schafer, In Defense: New York Times v. Sullivan, 82 La. L. Rev. 81, 96–97, 158–60 (2021) (critiquing Justice Thomas’s and Justice Gorsuch’s historical analyses and arguing that overturning Sullivan would have a chilling effect on the speech of news organizations); Lili Levi, Disinformation and the Defamation Renaissance: A Misleading Promise of “Truth,” 57 U. Rich. L. Rev. 1235, 1285–87 (2023) (arguing that Justice Thomas’s historical justification for dispensing with Sullivan is inadequate and that Justice Gorsuch’s concerns about disinformation online are narrow compared to the sweeping effects that reconsidering Sullivan could have on free speech); Matthew L. Schafer & Jeff Kosseff, Protecting Free Speech in a Post-Sullivan World, 75 Fed. Commc’ns L.J. 1, 35–36 (2022) (questioning Justice Thomas’s use of the “PizzaGate” conspiracy theory as an example in favor of loosening liability standards, noting that the theory was spread by anonymous posters on message boards, actors for whom there would likely be little effective legal consequence); Lyrissa Lidsky, Untangling Defamation Law: Guideposts for Reform, 88 Mo. L. Rev. 663, 676 (2023) (arguing that reconsidering Sullivan would likely not have a substantial effect on the disinformation Justice Gorsuch is most concerned with, which does not usually come from mainstream news organizations).Show More and one of the country’s leading organizations supporting media rights has issued a 196-page white paper seeking to demonstrate that reform of existing limits on liability is unnecessary and ill-advised.8 8.See generally Media L. Res. Ctr., New York Times v. Sullivan: The Case for Preserving an Essential Precedent (2022) [hereinafter MLRC], https://medialaw.org/wp-content/uploads/20‌23/01/nytsullivanwhitepaper.pdf [https://perma.cc/B7L4-UH3T].Show More

The two Supreme Court Justices, as well as the First Amendment scholars and media-support organizations opposing them, however, are guilty of a significant analytical and practical omission. They have failed to recognize that the media, other kinds of organizations, and most individuals have, or can obtain,9 9.It is probably the case that some small media organizations cannot easily obtain, or cannot easily afford, Media Liability insurance, although I have found no quantitative data confirming this. See, e.g., Christina Koningisor & Lyrissa Lidsky, First Amendment Disequilibrium, 110 Va. L. Rev. 1, 29–31 (2024) (discussing the financial pressures that media organizations face); Elise Czajkowski, Why Media Liability Insurance Is Key to Making Sure Your Newsroom Continues to Exist, Inst. for Nonprofit News, https://inn.org/research/case-studies/why-media‌-liability-insurance-is-key-to-making-sure-your-newsroom-continues-to-exist/ [https://perma‌.cc/5D47-K9Z9] (last visited Apr. 3, 2025) (recounting difficulties and costs small media organizations face in connection with defamation suits and insurance).Show More insurance against liability for defamation and other speech-related torts.10 10.The MLRC white paper does include a half-sentence reference to “rising libel insurance rates.” Richard Tofel & Jeremy Kutner, A Response to Justice Gorsuch, in MLRC, supra note 8, at 79, 83.Show More Alongside the constitutional regime affording breathing space for the exercise of protected free-speech rights, then, there is a little-known liability insurance regime that also affords breathing space for speech and provides broader protection than what the Constitution provides. It does so by indemnifying speakers against liability not only for harm caused by unprotected speech but also for the cost of defending against suits alleging liability for speech that turns out to have been protected or not to have been tortious.11 11.See infra Section II.A.Show More That is, liability insurance provides protection against liability that the First Amendment permits, as well as the cost of defending against alleged liability. It also supplements the protection the Constitution provides by covering the legal costs that speakers would otherwise incur when they are sued and the speakers win.

What if you analyzed the impact of tort liability on the quality of medical care without taking into account the fact that health care providers typically are protected against liability by malpractice insurance?12 12.For work that addresses this issue, see generally Tom Baker, The Medical Malpractice Myth (2005); Frank A. Sloan, Randall R. Bovbjerg & Penny B. Githens, Insuring Medical Malpractice (1991); Paul C. Weiler et al., A Measure of Malpractice: Medical Injury, Malpractice Litigation, and Patient Compensation (1993).Show More What if you analyzed how the threat of liability for causing an automobile accident affects driving behavior without ever considering that liability insurance pays for virtually all individuals’ automobile liability?13 13.See Kenneth S. Abraham, The Liability Century: Insurance and Tort Law from the Progressive Era to 9/11, at 102–03 (2008) (illustrating how the ubiquity of automobile insurance has slowed automobile liability reform);Tom Baker, Liability Insurance as Tort Regulation: Six Ways That Liability Insurance Shapes Tort Law in Action,12 Conn. Ins. L.J. 1, 11–12 (2005) (explaining the role of liability insurance in shaping tort doctrine); Nora Freeman Engstrom,Sunlight and Settlement Mills, 86 N.Y.U. L. Rev. 805, 815–16 (2011) (discussing how no-fault liability insurance could reshape tort law).Show More You would properly be thought to have ignored an important ingredient of any sensible analysis.14 14.See Kenneth S. Abraham & Catherine M. Sharkey, The Glaring Gap in Tort Theory, 133 Yale L.J. 2165, 2169 (2024).Show More

Yet the constitutional theory of breathing space takes no account of the role that insurance against liability for defamation and other speech-related torts plays in the operation of free speech in theory and in practice. In fact, there have been decades of extensive legal scholarship and constitutional litigation about the First Amendment’s restrictions on speech-tort liability. Most of this scholarship, and the courts, have wholly ignored the fact that all of the liability for the speech torts that the First Amendment does permit can be and often is covered by liability insurance.15 15.There have been at most a few dozen passing references to libel insurance in the law review literature over the decades, but almost no actual discussion of its significance. It would take too much detail to systematically prove this negative. A glance at two of the major First Amendment law casebooks, however, provides some confirmation. There is nothing in the table of contents or the index of either that refers to “insurance” or “liability insurance,” and nothing in the text of either that I can find that mentions these subjects. See generally Noah R. Feldman & Kathleen M. Sullivan, First Amendment Law (8th ed. 2023); Geoffrey R. Stone, Louis M. Seidman, Cass R. Sunstein, Mark V. Tushnet & Pamela S. Karlan, The First Amendment (4th ed. 2012). I have been able to find only two brief but actual discussions of defamation liability insurance in the law review literature, the most recent of them more than thirty years old. See Frederick Schauer, Uncoupling Free Speech, 92 Colum. L. Rev. 1321, 1339–43 (1992); Marc A. Franklin, Good Names and Bad Law: A Critique of Libel Law and a Proposal, 18 U.S.F. L. Rev. 1, 18–22 (1983). And roughly thirty-five years ago, part of one chapter in a book of essays made reference to and analyzed “trends in insurance premiums.” See Henry R. Kaufman, Trends in Damage Awards, Insurance Premiums, and the Cost of Media Libel Litigation, in The Cost of Libel: Economic and Policy Implications 1, 11–15 (Everette E. Dennis & Eli M. Noam eds., 1989).Show More As we will see below, how this all takes place is not an entirely simple matter, but it is not rocket science either. Courts and commentators should be able to understand it. In any event, its technicality and complexity are not a justification for ignoring it.

Whether the existing constitutional limits on liability for defamation and the other speech torts should be maintained is a multilayered issue that I do not claim, or even aim, to definitively resolve here.16 16.For a recent argument that the protections provided to public figures should be reduced, see generally G. Edward White, Reconsidering the Legacy of New York Times v. Sullivan, 23 First Amend. L. Rev. 1 (2024).Show More The principal purpose of this Article is to give analysts and policymakers more sophisticated tools to employ in addressing that issue, though I also make some points regarding the difficulties that would be encountered if liability based on negligence were permitted across the board. My contention is not that the newly realized availability of liability insurance should necessarily make a difference in speech-tort law at the retail level, doctrine by doctrine. Rather, I contend that the invocation of breathing-space concerns in constitutional law, in common law adjudication, and in public debate should be influenced by the recognition that not only constitutional protections against liability, but also liability insurance, provide breathing space for speech.

My message, then, is that those who subscribe to Justice Gorsuch’s position that proving actual malice, in practice, is too high a hurdle for defamation plaintiffs to overcome should either be invoking the availability of liability insurance in favor of their position or explaining why they are not doing so. And those who oppose that position and favor maintaining current First Amendment protections against liability should be explaining why they take this position, notwithstanding the breathing space that liability insurance provides.

In short, in light of the findings about liability insurance that I set out below, it seems clear that any analysis of First Amendment or tort liability issues that turns at least in part on breathing-space concerns should take the operation of liability insurance into account. I show how that can be done. Although liability insurance does not provide blanket protection, it is still a source of substantial breathing space. In addition, I explain why, at this point, we could use a much more complete understanding of the quantitative side of the issues that are relevant to the kind of analysis I am proposing. But we should not wait for perfect information about the incidence of insurance protection against speech-tort liability before taking the availability of this insurance into account. We should go with what data we have now. That is what I do here.

Part I identifies and analyzes the sources and scope of the coverage that insurance provides against speech-tort liability. Such insurance may be known to exist by a few judges and First Amendment scholars, but based on my own informal surveys, the availability, scope, and distribution of such insurance are only dimly understood at most. Perhaps even more surprising, insurance law scholars themselves (again, based on my own informal surveys) also have little awareness or understanding of the scope of speech-tort liability insurance. It is little wonder that there has been no scholarship integrating the two fields; the subject is largely opaque to most scholars in both.

Part II briefly identifies the elements of the traditional common law rules governing four speech torts—defamation, the main such tort, as well as false light, intentional infliction of emotional distress (“IIED”), and public disclosure of private facts—and also examines the particular restrictions on the scope of these liabilities that the Supreme Court has ruled apply to them as a matter of First Amendment law. The focus of this Part is not only the two sets of liability rules, but also what we know (and do not know) about suits and outcomes in speech-tort litigation, for most suits do not go all the way to verdict and judgment. This Part delineates the forms of speech-tort liability that are now precluded and those that remain constitutionally valid, which is an important step in the analysis, because the distinction between the two categories marks the boundary between the domains of constitutional protection from speech-tort liability and potential liability insurance indemnity against speech-tort liability.

Part III constitutes an effort to intervene in the two fields by combining insights about the complex and intertwined consequences of the threat of speech-tort liability with what we know about how liability insurance both creates moral hazard and attempts to neutralize it. This Part thereby identifies for the first time the critical issues for the integrated field. The results can be considered a benchmark in the development of an integrated theory of the relation between free speech, breathing space, and liability insurance. It also provides a sharpened understanding of the manner in which these fields operate together in practice.

Finally, now that liability insurance has come out of the shadows, Part IV first considers the normative relevance of liability insurance to analysis of the speech torts. This Part argues that, regardless of the other debates in torts scholarship about “rights theory” versus consequentialism, in the field of speech torts, the consequentialist conception must prevail. This is because ensuring breathing space for speech is the most consequentialist of purposes. Second, this Part takes a look at what standard could then replace actual malice. It turns out that the alternatives—especially a negligence standard—have significant disadvantages. This insight suggests that evaluating the full mix of considerations relevant to the issue, including the availability of liability insurance, does not lead to a definitive conclusion, but only to a more sophisticated analysis of the advantages and disadvantages of the approach that has now been in place for over fifty years.

  1.  As I explain below, these torts are false light, intentional infliction of emotional distress (“IIED”), and public disclosure of private facts. See infra Section II.A. ↑
  2.  See infra Section II.B. ↑
  3.  See Leslie Kendrick, Speech, Intent, and the Chilling Effect, 54 Wm. & Mary L.

    Rev. 1633, 1636–37 (2013) (explaining that protecting negligent false statements from liability provides constitutional “breathing space”); Frederick Schauer, Fear, Risk and the First Amendment: Unraveling the “Chilling Effect,” 58 B.U. L. Rev. 685, 710–18 (1978) (describing the role of a constitutional “buffer zone” in minimizing the chilling of protected activities). ↑

  4.  See, e.g., US Dominion, Inc. v. Fox News Network, LLC, 293 A.3d 1002 (Del. Super. Ct. 2023) (voting machine manufacturer against a television broadcaster based on conduct of various news anchors). The parties later reached a settlement for $787.5 million. David Folkenflik & Mary Yang, Fox News Settles Blockbuster Defamation Lawsuit with Dominion Voting Systems, NPR (Apr. 18, 2023, 4:35 PM), https://www.npr.org/2023/04/18/117033911‌4/fox-news-settles-blockbuster-defamation-lawsuit-with-dominion-voting-systems [https://p‌erma.cc/DTD5-TLCE]. See generally Freeman v. Giuliani, 691 F. Supp. 3d 32 (D.D.C. 2023) (election workers against a media personality), judgment entered, No. 21-cv-03354, 2023 WL 9783148 (D.D.C. Dec. 18, 2023), appeal dismissed, No. 24-7021, 2025 WL 634800 (D.C. Cir. Feb. 26, 2025); Carroll v. Trump (Carroll I), No. 20-cv-07311, 2024 WL 97359 (S.D.N.Y. Jan. 9, 2024) (writer against Donald Trump); Carroll v. Trump (Carroll II), No. 22-cv-10016, 2023 WL 3000562 (S.D.N.Y. Mar. 20, 2023) (same); Depp v. Heard, No. CL-2019-0002911, 2022 WL 2342058 (Va. Cir. June 24, 2022) (actor against another actor). ↑
  5.  McKee v. Cosby, 139 S. Ct. 675, 675–76 (2019) (Thomas, J., concurring in the denial of certiorari). ↑
  6.  Berisha v. Lawson, 141 S. Ct. 2424, 2427–28 (2021) (Gorsuch, J., dissenting from the denial of certiorari). ↑
  7.  See, e.g., Matthew L. Schafer, In Defense: New York Times v. Sullivan, 82 La. L. Rev. 81, 96–97, 158–60 (2021) (critiquing Justice Thomas’s and Justice Gorsuch’s historical analyses and arguing that overturning Sullivan would have a chilling effect on the speech of news organizations); Lili Levi, Disinformation and the Defamation Renaissance: A Misleading Promise of “Truth,” 57 U. Rich. L. Rev. 1235, 1285–87 (2023) (arguing that Justice Thomas’s historical justification for dispensing with Sullivan is inadequate and that Justice Gorsuch’s concerns about disinformation online are narrow compared to the sweeping effects that reconsidering Sullivan could have on free speech); Matthew L. Schafer & Jeff Kosseff, Protecting Free Speech in a Post-Sullivan World, 75 Fed. Commc’ns L.J. 1, 35–36 (2022) (questioning Justice Thomas’s use of the “PizzaGate” conspiracy theory as an example in favor of loosening liability standards, noting that the theory was spread by anonymous posters on message boards, actors for whom there would likely be little effective legal consequence); Lyrissa Lidsky, Untangling Defamation Law: Guideposts for Reform, 88 Mo. L. Rev. 663, 676 (2023) (arguing that reconsidering Sullivan would likely not have a substantial effect on the disinformation Justice Gorsuch is most concerned with, which does not usually come from mainstream news organizations). ↑
  8.  See generally Media L. Res. Ctr., New York Times v. Sullivan: The Case for Preserving an Essential Precedent (2022) [hereinafter MLRC], https://medialaw.org/wp-content/uploads/20‌23/01/nytsullivanwhitepaper.pdf [https://perma.cc/B7L4-UH3T]. ↑
  9.  It is probably the case that some small media organizations cannot easily obtain, or cannot easily afford, Media Liability insurance, although I have found no quantitative data confirming this. See, e.g., Christina Koningisor & Lyrissa Lidsky, First Amendment Disequilibrium, 110 Va. L. Rev. 1, 29–31 (2024) (discussing the financial pressures that media organizations face); Elise Czajkowski, Why Media Liability Insurance Is Key to Making Sure Your Newsroom Continues to Exist, Inst. for Nonprofit News, https://inn.org/research/case-studies/why-media‌-liability-insurance-is-key-to-making-sure-your-newsroom-continues-to-exist/ [https://perma‌.cc/5D47-K9Z9] (last visited Apr. 3, 2025) (recounting difficulties and costs small media organizations face in connection with defamation suits and insurance). ↑
  10.  The MLRC white paper does include a half-sentence reference to “rising libel insurance rates.” Richard Tofel & Jeremy Kutner, A Response to Justice Gorsuch, in MLRC, supra note 8, at 79, 83. ↑
  11.  See infra Section II.A. ↑
  12.  For work that addresses this issue, see generally Tom Baker, The Medical Malpractice Myth (2005); Frank A. Sloan, Randall R. Bovbjerg & Penny B. Githens, Insuring Medical Malpractice (1991); Paul C. Weiler et al., A Measure of Malpractice: Medical Injury, Malpractice Litigation, and Patient Compensation (1993). ↑
  13.  See Kenneth S. Abraham, The Liability Century: Insurance and Tort Law from the Progressive Era to 9/11, at 102–03 (2008) (illustrating how the ubiquity of automobile insurance has slowed automobile liability reform); Tom Baker, Liability Insurance as Tort Regulation: Six Ways That Liability Insurance Shapes Tort Law in Action, 12 Conn. Ins. L.J. 1, 11–12 (2005) (explaining the role of liability insurance in shaping tort doctrine); Nora Freeman Engstrom, Sunlight and Settlement Mills, 86 N.Y.U. L. Rev. 805, 815–16 (2011) (discussing how no-fault liability insurance could reshape tort law). ↑
  14.  See Kenneth S. Abraham & Catherine M. Sharkey, The Glaring Gap in Tort Theory, 133 Yale L.J. 2165, 2169 (2024). ↑
  15.  There have been at most a few dozen passing references to libel insurance in the law review literature over the decades, but almost no actual discussion of its significance. It would take too much detail to systematically prove this negative. A glance at two of the major First Amendment law casebooks, however, provides some confirmation. There is nothing in the table of contents or the index of either that refers to “insurance” or “liability insurance,” and nothing in the text of either that I can find that mentions these subjects. See generally Noah R. Feldman & Kathleen M. Sullivan, First Amendment Law (8th ed. 2023); Geoffrey R. Stone, Louis M. Seidman, Cass R. Sunstein, Mark V. Tushnet & Pamela S. Karlan, The First Amendment (4th ed. 2012). I have been able to find only two brief but actual discussions of defamation liability insurance in the law review literature, the most recent of them more than thirty years old. See Frederick Schauer, Uncoupling Free Speech, 92 Colum. L. Rev. 1321, 1339–43 (1992); Marc A. Franklin, Good Names and Bad Law: A Critique of Libel Law and a Proposal, 18 U.S.F. L. Rev. 1, 18–22 (1983). And roughly thirty-five years ago, part of one chapter in a book of essays made reference to and analyzed “trends in insurance premiums.” See Henry R. Kaufman, Trends in Damage Awards, Insurance Premiums, and the Cost of Media Libel Litigation, in The Cost of Libel: Economic and Policy Implications 1, 11–15 (Everette E. Dennis & Eli M. Noam eds., 1989). ↑
  16.  For a recent argument that the protections provided to public figures should be reduced, see generally G. Edward White, Reconsidering the Legacy of New York Times v. Sullivan, 23 First Amend. L. Rev. 1 (2024). ↑

Deterring Unenforceable Terms

Contract law doesn’t work the way most people—that is, most nonlawyers—think it works. People think that if they agree to a contract, they are bound by its terms—no matter if those terms are unfair or legally unenforceable. But that’s not correct. Although there is a default presumption that the law will enforce terms that parties agree to, courts can and do decline to enforce terms when they are contrary to statute, regulation, or common law.

This is a bad arrangement. Because people do not understand how enforceability works, contract drafters can include unenforceable terms and benefit from them even when they are contrary to law. Clearly unenforceable terms are used in a wide range of cases, and those terms impose costs on consumers and employees despite being formally toothless.

This Article argues for a change. The problems of unenforceable terms arise from the burden of determining whether a contractual provision is enforceable. The current law makes little effort to allocate or mitigate that burden. But in a common scenario—where a sophisticated actor drafts mass contracts for many unsophisticated counterparties—the drafter is much better positioned to determine the contract’s enforceability. The law should therefore penalize such drafters for including clearly unenforceable terms in their contracts. This Article describes the basic normative case for such a penalty, considers how it might best be designed, and assesses the opportunities and limitations in existing law for applying a penalty to deter the use of unenforceable terms in mass contracts.

Introduction

Contract law has a process problem.

The law governing contracts is full of protections for the people who are party to a contract. Over the centuries, judges at common law have determined that a variety of potential provisions are too unfair to permit. In the modern era, legislators have passed statutes that prohibit a wide range of terms. And regulators have promulgated rules to limit what companies can put in a contract.

But to enforce these protections, contract law often relies on a flawed procedure. The background norm in contract law is that a provision that is contrary to law will not be enforced in court. And that makes sense—where the terms of an agreement violate doctrine or public policy, there is a strong argument that the state should not enforce those terms. But, other than nonenforcement, there is no general default penalty for using unenforceable terms. Individual statutes or regulations may attach a penalty for using a particular term, such as when a statute creates a penalty for including contractual provisions purporting to waive the statute’s protections.1 1.See, e.g., 15 U.S.C. §§ 1693l, 1693m.Show More But absent such a specific regulation, the law does not penalize a party’s choice to include an unenforceable term in their contracts.

The problem is that this arrangement does little to prevent the use of unenforceable terms in the first place.2 2.See Margaret Jane Radin, Boilerplate: The Fine Print, Vanishing Rights, and the Rule of Law 147 (2013).Show More Most contracts do not get litigated. Most people, meanwhile, take the terms of the contracts they sign at face value, not realizing that a term may be unenforceable and carry no legal weight.3 3.See infra Section I.B.Show More As a result, a sophisticated drafter who puts unenforceable terms in their contracts may reap the benefit of those terms by influencing their counterparties’ behavior.4 4.Radin, supra note 2.Show More An employee who thinks they are bound by an unenforceable noncompete clause may decline to look for a better job; a consumer who thinks they are bound by an unenforceable liability waiver may not file a lawsuit even when they are entitled to damages.

And if the contract does end up in litigation, the term at issue is just rendered a nullity, making the drafter no worse off than if the term had not been included. On balance, then, contract drafters often are incentivized to include unenforceable terms and try to get whatever value out of them that they can, short of relying on them in court.5 5.Id.Show More

This arrangement is the result of a legal system that typically does not recognize a particular type of burden: the burden of learning what the law is and how it applies in a given context. To the contrary, courts embrace a clear legal fiction: that parties to a contract “are presumed to know the law.”6 6.E.g., R.L. Polk Printing Co. v. Smedley, 118 N.W. 984, 984 (Mich. 1908); BPP069, LLC v. Lindfield Holdings, LLC, 816 S.E.2d 755, 761 (Ga. Ct. App. 2018) (“[A]ll persons are presumed to know the law and therefore cannot be deceived by erroneous statements of law.” (quoting Lakeside Invs. Grp., Inc. v. Allen, 559 S.E.2d 491, 493 (Ga. Ct. App. 2002))).Show More As a result, although the law puts the onus on a contract drafter to write clear and intelligible terms,7 7.See, e.g., David Horton, Flipping the Script: Contra Proferentem and Standard Form Contracts, 80 U. Colo. L. Rev. 431, 437–38 (2009) (noting that the contra proferentem doctrine “deters imprecision” and describing its deterrent effects in the context of standard form contracts).Show More it does not typically penalize a drafter for writing unenforceable terms. The burden of understanding the enforceability of a term “lies where it falls,”8 8.Cf. Filosa v. Courtois Sand & Gravel Co., 590 A.2d 100, 102 (R.I. 1991) (“[W]here there is no negligence, the aggrieved party is no longer a plaintiff but is a victim of accidental misfortune, and one of the clearest and probably most draconian principles to evolve out of centuries of tort law is that accidental harm lies where it falls.” (citing Brown v. Kendall, 60 Mass. (6 Cush.) 292, 298 (1850))).Show More and if a tenant or an employee mistakenly thinks they are bound by a term, they generally just bear the cost of compliance.

But most people do not know the law. Contracts mediate huge portions of our lives, from our employment and our housing, to our communications and correspondence, to our leisure and entertainment. And the law makes little to no effort to ensure that those contracts’ terms actually create enforceable obligations. Instead, the law leaves it up to everyone to figure out for themselves what parts of their contracts they must listen to and what parts they can ignore. And what’s more, access to those who can help figure that out—i.e., lawyers—is incredibly unequal, with marginalized groups much less able to access legal assistance than those with more resources.9 9.See infra Section II.C.Show More

This arrangement should change. This Article contends that in the contemporary world of mass contracting, the law should reallocate the burden of learning and applying the law. In many private-law contexts, the law plays a role both in the efficient allocation of costs and in the protection of less sophisticated parties. But, I argue, in the world of mass contracting, the burdens of unenforceable terms are not efficiently allocated, and the result is harm to everyday consumers and employees who sign mass contracts.

The problematic incentives posed by unenforceable terms have long been recognized.10 10.See, e.g., Radin, supra note 2; Lee A. Pizzimenti, Prohibiting Lawyers from Assisting in Unconscionable Transactions: Using an Overt Tool, 72 Marq. L. Rev. 151, 158 (1989); Bailey Kuklin, On the Knowing Inclusion of Unenforceable Contract and Lease Terms, 56 U. Cin. L. Rev. 845, 846–47 (1988); Warren Mueller, Residential Tenants and Their Leases: An Empirical Study, 69 Mich. L. Rev. 247, 248 (1970).Show More But the time is ripe for a reconsideration of this basic feature of how our legal system handles contracts. Over the last decade, new empirical research has illustrated the breadth and depth of the problem of unenforceable terms—a problem that remained mostly anecdotal until recently.11 11.See infra Part I. On the lack of robust evidence before the last decade or so, see Meirav Furth-Matzkin, On the Unexpected Use of Unenforceable Contract Terms: Evidence from the Residential Rental Market, 9 J. Legal Analysis 1, 5 (2017) [hereinafter Furth-Matzkin, Unexpected Use] (“[T]here has been so far very little empirical investigation into the prevalence of unenforceable terms in consumer contracts.”).Show More Unenforceable terms are widespread, with studies confirming their ubiquity across economic sectors including housing,12 12.David A. Hoffman & Anton Strezhnev, Leases as Forms, 19 J. Empirical Legal Stud. 90, 90–91 (2022); see also Furth-Matzkin, Unexpected Use, supra note 11, at 17–23 (detailing contract provisions in lease housing contracts that courts have found to be unenforceable or that are prohibited by state statutes).Show More employment,13 13.Evan P. Starr, J.J. Prescott & Norman D. Bishara, Noncompete Agreements in the US Labor Force, 64 J.L. & Econ. 53, 60, 81 (2021) [hereinafter Starr et al., Noncompete Agreements].Show More and recreation.14 14.Edward K. Cheng, Ehud Guttel & Yuval Procaccia, Unenforceable Waivers, 76 Vand. L. Rev. 571, 577 (2023).Show More And new empirical research also demonstrates the effect of those terms.15 15.See Tess Wilkinson-Ryan, The Perverse Consequences of Disclosing Standard Terms, 103 Cornell L. Rev. 117, 139–49, 139 nn.105–10 (2017) (surveying research); Meirav Furth-Matzkin & Roseanna Sommers, Consumer Psychology and the Problem of Fine-Print Fraud, 72 Stan. L. Rev. 503, 516 (2020) (explaining that laypeople may be discouraged from breaking a contract, even if they suspect the contract is unfair, because they assume written contracts are binding); Meirav Furth-Matzkin, The Harmful Effects of Unenforceable Contract Terms: Experimental Evidence, 70 Ala. L. Rev. 1031, 1044–51, 1053–56 (2019) [hereinafter Furth-Matzkin, Harmful Effects] (showing through experiments that unenforceable terms adversely affect tenants’ behavioral intentions and legal predictions); Evan Starr, J.J. Prescott & Norman Bishara, The Behavioral Effects of (Unenforceable) Contracts, 36 J.L. Econ. & Org. 633, 651–55, 659–66 (2020) [hereinafter Starr et al., Behavioral Effects] (suggesting that noncompetes have an effect on behavior regardless of their enforceability, and that noncompetes are associated with longer employee tenure and reduced likelihood of leaving for a competitor). For examples of older evidence in this vein, see also Dennis P. Stolle & Andrew J. Slain, Standard Form Contracts and Contract Schemas: A Preliminary Investigation of the Effects of Exculpatory Clauses on Consumers’ Propensity to Sue, 15 Behav. Scis. & L. 83, 91–93 (1997) (finding that consumers tend to believe that all contract terms are enforceable and that exculpatory language in form contracts appears to deter consumers’ propensity to seek compensation); Curtis J. Berger, Hard Leases Make Bad Law, 74 Colum. L. Rev. 791, 815–16 (1974) (explaining that “[u]ninformed or misinformed parties to a contract are easily terrorized or disarmed into foregoing their rights and remedies”); Mueller, supra note 10, at 248, 272–74 (suggesting that “the bulk of tenants [do] not appear to question the validity of terms found in their leases”).Show More Both consumers and employees are likely to feel bound by contract terms that they have assented to, regardless of those terms’ legality, and even if they have not read the terms before signing.16 16.See infra Part I (surveying research).Show More There is thus now an established, increasingly robust literature documenting that unenforceable terms pose a real problem, and one that is common in the contemporary economy.

Unenforceable terms have also been highlighted by the recent actions of agencies and advocates.17 17.See infra Section I.C.Show More Terms that are frequently unenforceable, like noncompete agreements and liability waivers, are at the center of recent actions by the White House,18 18.Exec. Order No. 14036, 86 Fed. Reg. 36987, 36987, 36992 (July 9, 2021) (discussing noncompete agreements).Show More Federal Trade Commission (“FTC”),19 19.Non-Compete Clause Rule, 89 Fed. Reg. 38342, 38342 (May 7, 2024) (to be codified at 16 C.F.R. pts. 910, 912). But see Ryan, LLC v. Fed. Trade Comm’n, 746 F. Supp. 3d 369, 390 (N.D. Tex. 2024) (“The [Non-Compete] Rule shall not be enforced or otherwise take effect on its effective date of September 4, 2024, or thereafter.”), appeal docketed, No. 24-10951 (5th Cir. Oct. 24, 2024).Show More National Labor Relations Board (“NLRB”),20 20.Memorandum GC 23-08 from Jennifer A. Abruzzo, Gen. Couns., Nat’l Lab. Rels. Bd., to All Reg’l Dirs., Officers-in-Charge, and Resident Officers, Nat’l Lab. Rels. Bd. (May 30, 2023), https://nlrbresearch.com/pdfs/09031d4583a87168.pdf [https://perma.cc/D6QA-2‌RPG].Show More and Consumer Financial Protection Bureau (“CFPB”).21 21.Registry of Supervised Nonbanks That Use Form Contracts to Impose Terms and Conditions That Seek to Waive or Limit Consumer Legal Protections, 88 Fed. Reg. 6906, 6906 (proposed Feb. 1, 2023) [hereinafter Registry of Supervised Nonbanks That Use Form Contracts].Show More And significantly, policymakers and advocates are beginning to focus not only on rendering bad contract terms unenforceable, but also on penalizing drafters for including those provisions in the contracts to begin with.22 22.See infra Section I.C.Show More

Between this empirical work and recent policy developments, unenforceable terms are now in the spotlight. But that spotlight’s focus has often been somewhat granular, examining one particular term or context rather than the problem of unenforceable terms writ large.23 23.The FTC’s recent policy actions, for instance, focus only on noncompetes in employment contracts, while the CFPB’s actions focus on waivers in consumer finance. CompareNon-Compete Clause Rule, 89 Fed. Reg. at 38342 (providing that it is an unfair method of competition to enter into noncompete clauses with workers after the Rule’s effective date), with Registry of Supervised Nonbanks That Use Form Contracts, supra note 21 (proposing that nonbanks be required to register with the CFPB if they use contract terms designed to waive consumers’ legal protections or limit how consumers enforce their rights).Show More The idea of a general penalty for using unenforceable terms, meanwhile, has come up before, but typically only in passing.24 24.See, e.g., Furth-Matzkin & Sommers, supra note 15, at 544–45 (suggesting “statutory damages for fine-print fraud” as part of a set of policy solutions); Radin, supra note 2, at 147–48 (mentioning the possibility of fines as a component of a public regulatory regime for boilerplate terms). A more thorough consideration of an affirmative cause of action can be found in Brady Williams’s Unconscionability as a Sword: The Case for an Affirmative Cause of Action, 107 Calif. L. Rev. 2015, 2041, 2043–45, 2047 (2019), which argues for developing an affirmative remedy in the context of unconscionable contract provisions.Show More

This Article picks up where those conversations leave off and considers the merits of a penalty for using unenforceable terms in mass contracts of adhesion. It argues for a general, affirmative prohibition on clearly unenforceable terms in contracts offered by a sophisticated drafter to large numbers of unsophisticated counterparties.

In doing so, the Article brings to bear normative concerns from both within and outside of traditional private-law theory. In particular, the problem of unenforceable terms implicates both the traditional private-law goal of cost minimization as well as the public-law goal of access to justice. That is because the question of how to approach unenforceable terms can be thought of as a question of how the law ought to distribute the costs of acquiring and applying legal knowledge. The law under the status quo makes no effort to allocate these costs, which is why there is a problem: the signers of mass adhesive contracts are unlikely to know that unenforceable terms carry no legal weight, and so may change their behavior to accommodate those terms even if doing so causes them loss or injury.

There are two basic paths that could address that problem: the signers of mass contracts can acquire and apply the legal knowledge necessary to understand terms’ enforceability, or the drafters of those contracts can acquire and apply the legal knowledge necessary to prevent unenforceable terms from being included in the first place. Comparing those options, it is clear that the party who can more cheaply manage the costs of legal knowledge is the drafter. The drafter in this scenario is both a sophisticated actor (who likely already has counsel) and one who is able to amortize the cost of legal analysis over many transactions.

Placing the burden on the drafter to issue only binding, valid terms also mitigates serious inequities under the status quo. Access to legal knowledge and legal institutions in the United States is not equally distributed.25 25.See infra Part II.Show More Marginalized groups in the United States face the double bind of higher-than-average legal needs and lower-than-average income and wealth to use to manage those needs.26 26.See infra Part II.Show More The inequities that these individuals and communities experience when it comes to the civil justice system both reflect and reinforce racial and gender inequality.27 27.See, e.g., Kathryn A. Sabbeth & Jessica K. Steinberg, The Gender of Gideon, 69 UCLA L. Rev. 1130, 1143–48, 1150–61 (2023); Tonya L. Brito, Kathryn A. Sabbeth, Jessica K. Steinberg & Lauren Sudeall, Racial Capitalism in the Civil Courts, 122 Colum. L. Rev. 1243, 1268–77 (2022).Show More These legal problems will often center around transactions and relationships that are mediated by contract and involve a power imbalance—with a landlord, employer, or health care provider, for instance—and so may implicate many substantive contract law doctrines that are designed to protect less powerful individuals.28 28.See infra Part II.Show More Restricting unenforceable terms thus avoids compounding injustice, in which a person’s existing marginalization prevents them from obtaining the assistance necessary to prevent further injury.

A general penalty for unenforceable terms is the natural development of the “contract as thing” perspective introduced by Arthur Leff more than fifty years ago.29 29.See Arthur Allen Leff, Contract as Thing, 19 Am. U. L. Rev. 131, 131–32, 147–52, 155 (1970); see also Douglas G. Baird, The Boilerplate Puzzle, 104 Mich. L. Rev. 933, 933–37 (2006) (comparing the legal treatment of boilerplate and fine print contract terms to “hidden” attributes of products).Show More Mass contracts of adhesion, ubiquitous in the modern world, are more like off-the-shelf purchased products than the bespoke negotiated instruments that contract doctrine developed around. As Leff wrote, “[i]f . . . a particular contract is a mass-produced inalterable thing, then the words that make it up are just elements of the thing, like wheels and carburetors.”30 30.Leff, supra note 29, at 153.Show More But the doctrine of unenforceability does not treat unenforceable words like wheels or carburetors, or even like other words that a company may utter about its products. A sports equipment company may face liability if its advertisement falsely touts “the highest-rated safety features on the market,” but if its contract says “the company is not liable for any damages resulting from your use of our products,” the standard approach of non-enforceability provides no penalty—even if that statement is, legally speaking, false.

In this way, the world of unenforceable contract terms is one of the last vestiges of the “caveat emptor” doctrine that has long been excised from many other areas of the law.31 31.See infra Part II.Show More Faced with unenforceable terms in a contract, consumers are simply left to their own recognizance. Affirmatively prohibiting clearly unenforceable terms in mass contracts would allow signers to rely on the bindingness of the terms they assent to without being lawyers—just as we can rely on the functionality of the cars we buy without being engineers or the safety of the medicines we purchase without being physicians.

The Article proceeds as follows. First, Part I surveys the research regarding the use of unenforceable terms and their effects on the general public. Part II then builds out the normative argument for penalizing the use of unenforceable terms in mass contracts. Part III considers questions of how such a penalty would be designed, such as how to construct a liability rule and which parties it should cover. Part IV then considers resources in existing law that could be used to combat unenforceable terms short of passing new legislation.

  1.  See, e.g., 15 U.S.C. §§ 1693l, 1693m. ↑
  2.  See Margaret Jane Radin, Boilerplate: The Fine Print, Vanishing Rights, and the Rule of Law 147 (2013). ↑
  3.  See infra Section I.B. ↑
  4.  Radin, supra note 2. ↑
  5.  Id. ↑
  6.  E.g., R.L. Polk Printing Co. v. Smedley, 118 N.W. 984, 984 (Mich. 1908); BPP069, LLC v. Lindfield Holdings, LLC, 816 S.E.2d 755, 761 (Ga. Ct. App. 2018) (“[A]ll persons are presumed to know the law and therefore cannot be deceived by erroneous statements of law.” (quoting Lakeside Invs. Grp., Inc. v. Allen, 559 S.E.2d 491, 493 (Ga. Ct. App. 2002))). ↑
  7.  See, e.g., David Horton, Flipping the Script: Contra Proferentem and Standard Form Contracts, 80 U. Colo. L. Rev. 431, 437–38 (2009) (noting that the contra proferentem doctrine “deters imprecision” and describing its deterrent effects in the context of standard form contracts). ↑
  8.  Cf. Filosa v. Courtois Sand & Gravel Co., 590 A.2d 100, 102 (R.I. 1991) (“[W]here there is no negligence, the aggrieved party is no longer a plaintiff but is a victim of accidental misfortune, and one of the clearest and probably most draconian principles to evolve out of centuries of tort law is that accidental harm lies where it falls.” (citing Brown v. Kendall, 60 Mass. (6 Cush.) 292, 298 (1850))). ↑
  9.  See infra Section II.C. ↑
  10.  See, e.g., Radin, supra note 2; Lee A. Pizzimenti, Prohibiting Lawyers from Assisting in Unconscionable Transactions: Using an Overt Tool, 72 Marq. L. Rev. 151, 158 (1989); Bailey Kuklin, On the Knowing Inclusion of Unenforceable Contract and Lease Terms, 56 U. Cin. L. Rev. 845, 846–47 (1988); Warren Mueller, Residential Tenants and Their Leases: An Empirical Study, 69 Mich. L. Rev. 247, 248 (1970). ↑
  11.  See infra Part I. On the lack of robust evidence before the last decade or so, see Meirav Furth-Matzkin, On the Unexpected Use of Unenforceable Contract Terms: Evidence from the Residential Rental Market, 9 J. Legal Analysis 1, 5 (2017) [hereinafter Furth-Matzkin, Unexpected Use] (“[T]here has been so far very little empirical investigation into the prevalence of unenforceable terms in consumer contracts.”). ↑
  12.  David A. Hoffman & Anton Strezhnev, Leases as Forms, 19 J. Empirical Legal Stud. 90, 90–91 (2022); see also Furth-Matzkin, Unexpected Use, supra note 11, at 17–23 (detailing contract provisions in lease housing contracts that courts have found to be unenforceable or that are prohibited by state statutes). ↑
  13.  Evan P. Starr, J.J. Prescott & Norman D. Bishara, Noncompete Agreements in the US Labor Force, 64 J.L. & Econ. 53, 60, 81 (2021) [hereinafter Starr et al., Noncompete Agreements]. ↑
  14.  Edward K. Cheng, Ehud Guttel & Yuval Procaccia, Unenforceable Waivers, 76 Vand. L. Rev. 571, 577 (2023). ↑
  15.  See Tess Wilkinson-Ryan, The Perverse Consequences of Disclosing Standard Terms, 103 Cornell L. Rev. 117, 139–49, 139 nn.105–10 (2017) (surveying research); Meirav Furth-Matzkin & Roseanna Sommers, Consumer Psychology and the Problem of Fine-Print Fraud, 72 Stan. L. Rev. 503, 516 (2020) (explaining that laypeople may be discouraged from breaking a contract, even if they suspect the contract is unfair, because they assume written contracts are binding); Meirav Furth-Matzkin, The Harmful Effects of Unenforceable Contract Terms: Experimental Evidence, 70 Ala. L. Rev. 1031, 1044–51, 1053–56 (2019) [hereinafter Furth-Matzkin, Harmful Effects] (showing through experiments that unenforceable terms adversely affect tenants’ behavioral intentions and legal predictions); Evan Starr, J.J. Prescott & Norman Bishara, The Behavioral Effects of (Unenforceable) Contracts, 36 J.L. Econ. & Org. 633, 651–55, 659–66 (2020) [hereinafter Starr et al., Behavioral Effects] (suggesting that noncompetes have an effect on behavior regardless of their enforceability, and that noncompetes are associated with longer employee tenure and reduced likelihood of leaving for a competitor). For examples of older evidence in this vein, see also Dennis P. Stolle & Andrew J. Slain, Standard Form Contracts and Contract Schemas: A Preliminary Investigation of the Effects of Exculpatory Clauses on Consumers’ Propensity to Sue, 15 Behav. Scis. & L. 83, 91–93 (1997) (finding that consumers tend to believe that all contract terms are enforceable and that exculpatory language in form contracts appears to deter consumers’ propensity to seek compensation); Curtis J. Berger, Hard Leases Make Bad Law, 74 Colum. L. Rev. 791, 815–16 (1974) (explaining that “[u]ninformed or misinformed parties to a contract are easily terrorized or disarmed into foregoing their rights and remedies”); Mueller, supra note 10, at 248, 272–74 (suggesting that “the bulk of tenants [do] not appear to question the validity of terms found in their leases”). ↑
  16.  See infra Part I (surveying research). ↑
  17.  See infra Section I.C. ↑
  18.  Exec. Order No. 14036, 86 Fed. Reg. 36987, 36987, 36992 (July 9, 2021) (discussing noncompete agreements). ↑
  19.  Non-Compete Clause Rule, 89 Fed. Reg. 38342, 38342 (May 7, 2024) (to be codified at 16 C.F.R. pts. 910, 912). But see Ryan, LLC v. Fed. Trade Comm’n, 746 F. Supp. 3d 369, 390 (N.D. Tex. 2024) (“The [Non-Compete] Rule shall not be enforced or otherwise take effect on its effective date of September 4, 2024, or thereafter.”), appeal docketed, No. 24-10951 (5th Cir. Oct. 24, 2024). ↑
  20.  Memorandum GC 23-08 from Jennifer A. Abruzzo, Gen. Couns., Nat’l Lab. Rels. Bd., to All Reg’l Dirs., Officers-in-Charge, and Resident Officers, Nat’l Lab. Rels. Bd. (May 30, 2023), https://nlrbresearch.com/pdfs/09031d4583a87168.pdf [https://perma.cc/D6QA-2‌RPG]. ↑
  21.  Registry of Supervised Nonbanks That Use Form Contracts to Impose Terms and Conditions That Seek to Waive or Limit Consumer Legal Protections, 88 Fed. Reg. 6906, 6906 (proposed Feb. 1, 2023) [hereinafter Registry of Supervised Nonbanks That Use Form Contracts]. ↑
  22.  See infra Section I.C. ↑
  23.  The FTC’s recent policy actions, for instance, focus only on noncompetes in employment contracts, while the CFPB’s actions focus on waivers in consumer finance. Compare Non-Compete Clause Rule, 89 Fed. Reg. at 38342 (providing that it is an unfair method of competition to enter into noncompete clauses with workers after the Rule’s effective date), with Registry of Supervised Nonbanks That Use Form Contracts, supra note 21 (proposing that nonbanks be required to register with the CFPB if they use contract terms designed to waive consumers’ legal protections or limit how consumers enforce their rights). ↑
  24.  See, e.g., Furth-Matzkin & Sommers, supra note 15, at 544–45 (suggesting “statutory damages for fine-print fraud” as part of a set of policy solutions); Radin, supra note 2, at 147–48 (mentioning the possibility of fines as a component of a public regulatory regime for boilerplate terms). A more thorough consideration of an affirmative cause of action can be found in Brady Williams’s Unconscionability as a Sword: The Case for an Affirmative Cause of Action, 107 Calif. L. Rev. 2015, 2041, 2043–45, 2047 (2019), which argues for developing an affirmative remedy in the context of unconscionable contract provisions. ↑
  25.  See infra Part II. ↑
  26.  See infra Part II. ↑
  27.  See, e.g., Kathryn A. Sabbeth & Jessica K. Steinberg, The Gender of Gideon, 69 UCLA L. Rev. 1130, 1143–48, 1150–61 (2023); Tonya L. Brito, Kathryn A. Sabbeth, Jessica K. Steinberg & Lauren Sudeall, Racial Capitalism in the Civil Courts, 122 Colum. L. Rev. 1243, 1268–77 (2022). ↑
  28.  See infra Part II. ↑
  29.  See Arthur Allen Leff, Contract as Thing,
    19

    Am. U. L. Rev. 131, 131–32, 147–52, 155 (1970); see also Douglas G. Baird, The Boilerplate Puzzle, 104 Mich. L. Rev. 933, 933–37 (2006) (comparing the legal treatment of boilerplate and fine print contract terms to “hidden” attributes of products). ↑

  30.  Leff, supra note 29, at 153. ↑
  31.  See infra Part II. ↑
©Virginia Law Review Association. All rights reserved.