The Shadow of the Law of the Police

Shielded: How the Police Became Untouchable. By Joanna Schwartz. New York: Viking. 2023. Pp. xxi, 336. $30.

Introduction

Joanna Schwartz’s Shielded: How the Police Became Untouchable¹Professor of Law, UCLA School of Law. is, in many ways, a triumph. It catalogues the laws regulating police with a breadth and depth unexpected from a book targeted to a popular audience, and it does so in gripping style. Although her ability as a researcher is well known to those who have encountered her earlier work, it is Schwartz’s ability as a storyteller that sets Shielded apart and makes the stakes of these doctrinal discussions clear. Admittedly, for those lawyers, judges, and academics who come to Shielded already familiar with Schwartz’s academic work, there is little new analytical ground covered. Most (but importantly not all) of the cases she discusses are those we have read and taught. Likewise, the points raised about the on-the-ground reality and interlocking nature of doctrines like qualified immunity and indemnity are, while always shocking, not unfamiliar to those who have read Schwartz’s pathbreaking law review articles. But even those of us who teach and write about criminal law and procedure will likely learn more from this book about the lives of the plaintiffs who inhabit the cases we teach.

Shielded queries how the police became untouchable. But it rarely asks why. Schwartz can be forgiven for not delving into this complex question.²This is not to say that Schwartz totally evades interrogating why. She suggests reasons that are often raised in the cases themselves, such as a desire to protect the police from frivolous lawsuits, pp. 11–12, and to enable police discretion in the name of safety, even if at a cost to racial equality, pp. 13–14. Instead, I mean that she does not explain why, despite the varying explanations and competing interests underlying each of these doctrinal areas, they all end up pointing in the same direction. That is the why that I focus on here. Describing the interlocking web of doctrine, policy, and practice that regulates police is a daunting task unto itself. But turning the focus from how to why surfaces an unexplained issue within Shielded: what benefit do we get from the current state of the world? This Review suggests that we can begin to resolve this question by turning our attention to the shadow of the law of the police.

One of Shielded’s most thought-provoking discussions centers on how civil rights attorneys have responded to the Court’s doctrines regulating civil rights cases. Part I builds on Schwartz’s observations by explicitly naming the shadow of the law’s role and applying civil rights litigation tools developed in other areas to analyze the law’s shadow.³See generally Robert H. Mnookin & Lewis Kornhauser, Bargaining in the Shadow of the Law: The Case of Divorce, 88 Yale L.J. 950 (1979). This helps to explain both why the Court’s jurisprudence may have been a rational but misguided attempt to encourage civil rights settlements, and why people both in and outside of academia have focused on doctrines, like qualified immunity, that rarely formally decide cases.⁴Joanna C. Schwartz, How Qualified Immunity Fails, 127 Yale L.J. 2, 47 (2017) (finding that qualified immunity decided 38 of over 1000 cases in her dataset) [hereinafter How Qualified Immunity Fails]; Joanna C. Schwartz, Civil Rights Ecosystems, 118 Mich. L. Rev. 1539, 1567 (2020) (“I have never heard of a motion for qualified immunity in just excessive force here. Not just my cases but any cases.” (quoting a Philadelphia attorney)); Adam A. Davidson, Procedural Losses and the Pyrrhic Victory of Abolishing Qualified Immunity, 99 Wash. U. L. Rev. 1459, 1462 (2022) [hereinafter Pyrrhic].

Although Schwartz aptly describes how civil rights attorneys bargain in the law’s shadows, she does not discuss the broader shadow the law casts on policing writ large. Part II attempts to fill this gap by turning to the law of pedestrian stops. This Review posits that the law forces police and “subject[s] of [the] carceral state”⁵Utah v. Strieff, 136 S. Ct. 2056, 2071 (2016) (Sotomayor, J., dissenting). to bargain over safety and carceral involvement in settings shaped by legal and social power imbalances. These “bargains,” in turn, are no bargain at all. Instead, the law’s expectation is a submission to law enforcement that enables subordination, especially racial subordination.⁶This is not to say that the police do not subordinate along other lines, like sexuality or gender. Adam A. Davidson, Managing the Police Emergency, 100 N.C. L. Rev. 1209, 1232–36 (2022).

The conclusion of this Review returns to Shielded. It argues that Schwartz’s failure to recognize and name this why may make her proposed solutions hollow and suggests why abolitionist theory may be necessary to make durable change.

I. The Shadow of the Law of the Police

This Part builds on Schwartz’s work describing civil rights bargaining by applying Mnookin and Kornhauser’s canonical analysis⁷Mnookin & Kornhauser, supra note 3. of the “shadow of the law.” While their article has spawned decades of commentary in numerous fields,⁸See, e.g., Ben Depoorter, The Upside of Losing, 113 Colum. L. Rev. 817, 822 (2013) (discussing the potential mobilizing effects of losing litigation); Shirley Lin, Bargaining for Integration, 96 N.Y.U. L. Rev. 1826, 1834 (2021) (discussing the Americans with Disabilities Act’s interactive process for determining an accommodation). including criminal law,⁹Within criminal law, plea bargaining is one such area. E.g., William J. Stuntz, Plea Bargaining and Criminal Law’s Disappearing Shadow, 117 Harv. L. Rev. 2548 (2004); Stephanos Bibas, Plea Bargaining Outside the Shadow of Trial, 117 Harv. L. Rev. 2463 (2004); Michael M. O’Hear & Andrea Kupfer Schneider, Dispute Resolution in Criminal Law, 91 Marq. L. Rev. 1, 4–5 (2007) (discussing symposium contributions on this subject). Immigration is another. See Katherine Beckett & Heather Evans, Crimmigration at the Local Level: Criminal Justice Processes in the Shadow of Deportation, 49 Law & Soc’y Rev. 241 (2015). it has not yet been applied to the law of the police.

This Part applies a shadow of the law analysis to illuminate why the Court’s jurisprudence regulating civil rights litigation—particularly its fee cases and qualified immunity—could be a rational, but ultimately misguided, shift to push these cases towards negotiated settlements.

A. The Shadow of the Law

I use the phrase “shadow of the law” similarly to Mnookin and Kornhauser. I seek to identify the “negotiations and bargaining that occur outside the courtroom” when the law “provid[es] a framework” for parties to interact within but does not “impos[e] order from above.”¹⁰Mnookin & Kornhauser, supra note 3, at 950. In other words, I define the shadow of the law as existing when the law affects, but fails to formally resolve, an issue—whether for substantive, procedural, or other reasons—so the two counterparties resolve their dispute outside of the formal legal process.

Unlike some scholars,¹¹See, e.g., Stuntz, supra note 9, at 2548, 2548 n.3 (arguing that “law” and “trial outcomes” are substitutes because trial outcomes reflect the broader law and “[i]nternalizing the one means internalizing the other”). I believe it is fruitful to analyze the shadows cast by law at various stages of the criminal legal process separately from the ultimate legal outcome. That is because, particularly in the criminal legal system, the legal process often is the punishment of concern to both the police and the people they stop. As such, any number of “procedural” legal decisions are functionally equivalent to an ultimate outcome for the negotiating parties. The relevant law might be that governing whether someone can be arrested, held pretrial, or stopped at all, all of which might cast long shadows. Assuming this law is subsumed into ultimate legal outcomes would provide a skewed view of the work that the law and its shadow do and thus a skewed view of the possible effects of altering the law.

Relatedly, the law’s role in these “shadow” negotiations is not necessarily neutral. As parties negotiate in the shadow of an expected legal outcome, the framework the law provides may tilt the negotiating table firmly in one party’s direction.¹²See Bibas, supra note 9, at 2465 (“[Numerous] scholars treat plea bargaining as just another case of bargaining in the shadow of expected trial outcomes.”). Indeed, one purpose of evaluating the law’s shadow is learning how to reshape legal frameworks to provide the “procedural or substantive safeguards . . . necessary to protect various social interests.”¹³Mnookin & Kornhauser, supra note 3, at 951.

Mnookin and Kornhauser “identify five factors that” shape negotiations in the law’s shadow.¹⁴Id. at 966. These factors are: (1) party preferences; (2) bargaining endowments that legal rules create to “indicate the particular allocation a court will impose if the parties fail to reach agreement;” (3) risk tolerance and “the degree of uncertainty concerning the legal outcome if the parties go to court . . . ; (4) transaction costs and the parties’ respective abilities to bear them; and (5) strategic behavior.”¹⁵Id. Analyzing plea bargains, then-Professor Stephanos Bibas identified other factors that he argued guided negotiations in the shadow of the law in addition to (or sometimes supplanting) the expected trial outcome: “[U]ncertainty, money, self-interest, and demographic variation.” Bibas, supra note 9, at 2469. To the extent these differ from Mnookin and Kornhauser’s, Bibas’s factors can all be underlying reasons for a party’s risk tolerance, preferences, or the existence of and ability to bear transaction costs. While these factors are likely familiar to legal audiences, it is important to highlight—particularly in the context of the law of the police—that each of these factors is shaped by the various intersections of each party’s legal,¹⁶See, e.g., Beckett & Evans, supra note 9, at 246–47 (discussing the impact of citizenship status on criminal case outcomes). psychological,¹⁷See, e.g., Richard Delgado et al., Fairness and Formality: Minimizing the Risk of Prejudice in Alternative Dispute Resolution, 1985 Wis. L. Rev. 1359 (finding that informality increases the risk of prejudiced actions that disadvantage those of lower social status); Trina Grillo, The Mediation Alternative: Process Dangers for Women, 100 Yale L.J. 1545, 1549–50 (1991) (arguing that societal expectations of women might mean that “mediation has the potential actively to harm women”). social,¹⁸See Lin, supra note 8, at 1858–66 (discussing how “education below a college degree, racial and gender minority status, and physically demanding jobs” correlate with being granted a workplace accommodation under the ADA). and economic status.¹⁹See, e.g., Mnookin & Kornhauser, supra note 3, at 971–72, 971 n.72 (noting the financial costs of litigation and identifying divorce law’s shift from the usual rule that each party pays their own attorney).

Although Schwartz does not engage with this line of scholarship, she commendably often highlights how vital the shadow of the law is to on-the-ground experiences. Perhaps the best example of this is Schwartz’s discussion of the law’s role in shaping the practice of civil rights attorneys. Attorneys are sophisticated legal actors who often engage in negotiation professionally. It is unsurprising that the law expects they could, and pushes them to, reach a better and cheaper resolution for their clients outside of the courts in most cases. As Schwartz persuasively argues, however, it is far from clear that the law gets this balance right (ch. 2).

B. Civil Rights Attorneys

Being a civil rights attorney, though a profession of passion for many, is work²⁰Unlike Schwartz, who begins her chapter on civil rights attorneys by discussing their choices in where to locate their practices, I do not discuss the first step in this process. P. 20. That is not to say, however, that the shadow of the law may not operate on that choice too.—and work requires pay. Unlike lawyers with deep-pocketed clients, however, civil rights attorneys mostly rely on contingency fees and fee-shifting provisions (pp. 17–32). Fortunately, Congress recognized the importance of this work and passed 42 U.S.C. § 1988, which provides that a court may award attorney’s fees to the prevailing party in civil rights litigation. Unfortunately, Congress did not account for the shadow of the law.

As Schwartz has found, most civil rights cases are resolved through settlement or another negotiated outcome.²¹P. 26; How Qualified Immunity Fails, supra note 4, at 46 (finding that 672 out of 1183 cases in her sample were decided by a settlement or voluntary/stipulated dismissal). Defendants thus discovered they could induce civil rights attorneys to forego statutory fees as a condition of settlement. Although this tactic could have been stopped with a firm legal rule, the Supreme Court instead blessed it in Evans v. Jeff D.²²P. 26; Evans v. Jeff D., 475 U.S. 717 (1986). As a result, the settlements offered to civil rights plaintiffs “almost always waive lawyers’ ability to recover attorneys’ fees,” and “the contingency-fee system that Congress intended to avoid by enacting Section 1988 is basically back in place” (p. 26).

Even when attorneys win at trial, they are not guaranteed full compensation. Fee awards are left to the discretion of the trial court, and the Supreme Court warned in Hensley v. Eckerhart that “[c]ases may be overstaffed, and the skill and experience of lawyers vary widely.”²³P. 27 (quoting Hensley v. Eckerhart, 461 U.S. 424, 434 (1983)). Additionally, the Court has interpreted the prevailing party requirement to allow district courts evaluating fee requests to adjust awards based on the “results obtained”—meaning that hours spent on a claim decided against the plaintiff may be excluded if that claim is considered “unrelated” to the plaintiff’s successful claim, despite being part of the same case.²⁴Hensley, 461 U.S. at 434–35. District court judges take cues from the Court and “in the end, [] seem almost always to give plaintiffs’ attorneys’ fee applications a haircut” (p. 27).

Schwartz finds that the result of these legal choices²⁵Both Evans and Hensley featured powerful separate opinions from justices arguing for stronger protections for § 1988 attorneys’ fees. Evans, 475 U.S. at 743–66 (Brennan, J., dissenting); Hensley, 461 U.S. at 441–57 (Brennan, J., concurring in part and dissenting in part). is that many civil rights attorneys have highly selective, and sometimes problematic, case selection practices (p. 27). Without gruesome facts—“ ‘[I]s there blood on the street?’ Because if there isn’t, why are we doing it?” one attorney said to Schwartz)—and a perfect victim—too often meaning those who are not people of color, LGBTQ+, or diagnosed with a mental illness—the case may not be sufficiently profitable (pp. 27–28).

While numerous other doctrines are relevant to fully examining how the shadow of the law of the police operates, here I mention only one: qualified immunity. One of the most surprising findings from Schwartz’s work is that qualified immunity formally decides few cases.²⁶How Qualified Immunity Fails, supra note 4, at 45 (finding that qualified immunity was the basis for dismissal in only 3.9%of the cases in her dataset). Relying on this finding, one might expect qualified immunity to be an unimportant doctrine. But qualified immunity continues to play an enormous role in academic debates and the popular imagination.²⁷See Pyrrhic, supra note 4, at 1475–85. Its role in the shadow of the law illuminates why: Even when it does not decide a civil rights case, qualified immunity almost certainly increases the case’s complexity and length.²⁸Joanna C. Schwartz, After Qualified Immunity, 120 Colum. L. Rev. 309, 338–42 (2020); Alan K. Chen, The Intractability of Qualified Immunity, 93 Notre Dame L. Rev. 1937, 1938 (2018) (explaining that judges and litigators “struggle to implement a doctrine that suffers from serious administrability problems”); John C. Jeffries, Jr., What’s Wrong with Qualified Immunity?, 62 Fla. L. Rev. 851, 852 (2010) (“[D]etermining whether an officer violated ‘clearly established’ law has proved to be a mare’s nest of complexity and confusion.”).

The Court’s decision to allow the law’s shadow to operate makes sense when one views these cases like any other civil litigation: a negotiation between two relatively equally-matched, represented parties. And the Court’s discussions in the fees cases suggest this was the majority’s view. As Chief Justice Burger wrote for the Court in Marek v. Chesny, “[t]here is no evidence . . . that Congress, in considering § 1988, had any thought that civil rights claims were to be on any different footing from other civil claims insofar as settlement is concerned.”²⁹Marek v. Chesny, 473 U.S. 1, 10 (1985); see also Evans, 475 U.S. at 733–35 (quoting this portion of Marek approvingly before going on to discuss the “unpredictability” and “magnitude” of potential attorney’s fees in civil rights cases). The fees cases, then, can be read as an attempt to realign civil rights negotiations with those in other civil suits. Similarly, qualified immunity could be thought of as an inducement to avoid complex, protracted litigation.

C. Miscalculating the Shadow of the Law

Assuming the Court intended only to create an equal field primed for negotiation,³⁰This is a generous assumption. As Schwartz discusses, there are good reasons to think the Court was not attempting to create an equal negotiating field. Pp. 24–26. why did this shift lead to such one-sided outcomes? I argue the Court miscalculated along three dimensions.

First, it undermined civil rights attorneys’ bargaining endowments to a greater extent than it realized. By allowing parties to bargain away Section 1988 fees, the Court forced civil rights attorneys to choose between pursuing the potentially more lucrative path of taking a case through trial to receive full fees or doing what was best for their client by waiving fees in order to get a more sizable merits award.³¹For example, imagine a pretrial settlement offer of 0,000 with 50% for fees in a case with ,000 in expected damages and 0,000 in already-accrued fees. As the Court observed, fees could be the primary economic driver of a case because they might be significantly larger than damages.³²Evans, 475 U.S. at 734–35 (“[P]otential liability for attorney’s fees may overshadow the potential cost of relief on the merits and darken prospects for settlement if fees cannot be negotiated.”); see also City of Riverside v. Rivera, 477 U.S. 561, 565–67 (1986) (approving over 5,000 in fees on ,350 damages). The Court compounded in Evans its earlier decisions encouraging district courts to look skeptically at fee requests in Hensley and further limiting fees in Marek. Once fee waivers were possible in settlement, the rarity of a posttrial fee award meant that for an award to have salience in bargaining, the size of the possible award needed to be large. Instead, the Court limited the possibility of large fees.

Second, the Court increased the level of uncertainty in civil rights proceedings through difficult-to-litigate doctrines like qualified immunity. Combined with unfriendly substantive law and hostile juries,³³Joanna C. Schwartz, Qualified Immunity’s Selection Effects, 114 Nw. U. L. Rev. 1101, 1153 (2020) [hereinafter Selection Effects]. this increased uncertainty made it financially untenable for civil rights attorneys to take anything but the most egregious cases. This rationally led them to litigate only cases with “perfect” victims and horrific facts.³⁴Id. at 1132–33.

Third, and perhaps most importantly, the Court failed to appreciate the transaction costs of litigating civil rights cases and the abilities of attorneys and their clients to bear those costs. The nominal reason for qualified immunity’s allowance of interlocutory appeals is to save government defendants the time and expense of litigation.³⁵Wheatt v. City of East Cleveland, No. 1:17-cv-377, 2017 WL 6031816, at *4 (N.D. Ohio Dec. 6, 2017). But it seems to do the opposite, instead allowing defendants to benefit through delay. Courts of appeals affirm qualified immunity denials at high rates,³⁶How Qualified Immunity Fails, supra note 4, at 44. and the additional round(s) of appeals can add years of delay “in place of a trial that [] could have finished in less than a week.”³⁷Wheatt, 2017 WL 6031816, at *4. Because these cases can last years, the pressure to settle can quickly become too much to resist. This is certainly true for attorneys, who remain aware of the mounting costs of litigating in comparison to the limited fees they may win³⁸Selection Effects, supra note 33, at 1149.—but it seems especially true for their clients.

The people most likely to suffer at the hands of the police are those who are already subordinated in some way, often along multiple intersecting lines. As one example: “Black people are significantly more likely than White people to be stopped and 3.5 times more likely than White people to experience police use of force, even after controlling for their rates of offending and contact with the police.”³⁹Robert O. Motley, Jr. & Sean Joe, Police Use of Force by Ethnicity, Sex, and Socioeconomic Class, 9 J. Soc’y for Soc. Work & Rsch. 49, 61 (2018) (citations omitted). But within that broad disproportionality, Black men with lower incomes “were significantly more likely to report exposure to police use of force during their most recent street stop than Black males with incomes of $50,000 or more.”⁴⁰Id. at 58.

Even if a client can bear the financial cost of litigating a case to completion, the psychological toll may prove too much. Schwartz recounts the story of Robbie Tolan, a promising second-generation professional baseball player shot in his parents’ driveway while defending his mother from a police officer (pp. 117–18). After facing seven years of litigation, multiple appeals, and a hostile trial judge, Tolan decided to settle his claims for $110,000 on the eve of trial. This amount “was a pittance compared to” his near million dollars in medical and legal fees.⁴¹P. 132 (quoting Robbie Tolan & Lawrence Ross, No Justice: One White Police Officer, One Black Family, and How One Bullet Ripped Us Apart 215, 218–19 (2018)). Rather than achieving victory or vindication, Tolan said he reached only “a conclusion to a long personal nightmare.”⁴²P. 132 (quoting Tolan & Ross, supra note 41, at 215, 218–19).

II. The Shadow of the Law of the Police Stops

Schwartz ends her final chapter with an exchange she had with a California legislator. This legislator was considering a bill to limit qualified immunity but was afraid of the possible consequences. Schwartz told the legislator about the reams of evidence that showed his fears about runaway liability for individual police officers was unfounded. The bill’s opponents “had no comparable evidence of civil rights lawsuits’ supposed dangers” (p. 240). And yet, even in the liberal bastion of California, “the bill died on the assembly floor” (p. 240).

Why? As Schwartz notes, this is not an example of dueling experts causing political paralysis. Instead, there appears to be some unspoken fear driving legal actors to support qualified immunity and doctrines like it even in the face of increasingly overwhelming contrary evidence (pp. 139–41). Throughout Shielded, Schwartz hints at what this fear might be. From her first pages to her very last, Schwartz’s description of the plaintiffs in these cases almost always includes their race, their gender, and some signifier of their class⁴³See, e.g., p. x (describing Onree Norris); p. 209 (describing Shawn Schenck).—but Schwartz never explicitly says these markers are driving the ecosystem she discusses. I argue that by looking further into the law’s shadows, we can see they are doing just that.

Much civil rights litigation doctrine can be justified under the “traditional” model—where two equal counterparties engage in a private bargain in the shadow of the law.⁴⁴See supra Section I.A. What cannot be justified under that model is the way doctrine shapes negotiations in police-civilian interactions. Beyond unequal social and economic positions, these counterparties are differently situated as a matter of law. The result of this structural inequality is a “bargain” that is no bargain at all. Instead, it is the expected submission of one party to the other. I posit that a primary interest being served by the choice to regulate policing through shadow negotiations instead of legal mandates is the advancement of subordination, particularly along racial lines.⁴⁵While race often predominates this area, it is of course only one aspect of the broader political economic structure. The rest of this Part excavates this underlying purpose by exploring the example of pedestrian stops.

A. The Law of Stops

First, a brief overview of the law of stops is in order.⁴⁶For a more detailed discussion of the many complications of pedestrian stops, see Devon W. Carbado, Unreasonable: Black Lives, Police, Power, and the Fourth Amendment 41–76, 101–54 (2022). The majority of police stops happen in one of two scenarios: pedestrian stops, which I focus on here, and traffic stops.⁴⁷ Susannah N. Tapp & Elizabeth J. Davis, U.S. Dep’t of Just., Contacts Between Police and the Public, 2020, at 2 fig.1 (2022). The law governing traffic stops differs from the law governing pedestrian stops because most traffic stops begin with probable cause of a traffic violation. Carbado, supra note 46, at 77–100. Pedestrian stops are governed by two standards: consent and reasonable suspicion.

A police officer may approach and engage someone without triggering Fourth Amendment scrutiny if the stop occurs with the individual’s consent.⁴⁸Florida v. Royer, 460 U.S. 491, 497 (1983). Here, the individual is also free to end the encounter at any time.⁴⁹Id. In contrast, an officer might stop someone because the officer has reasonable suspicion that they are committing a crime. In this sort of stop, known as a Terry stop, the person is not free to leave.⁵⁰See Terry v. Ohio, 392 U.S. 1, 19 & n.16 (1968). If an officer additionally has a reasonable and articulable suspicion that the person is armed and dangerous, the officer can conduct a pat down, or “frisk,” for weapons.⁵¹Id. at 27.

From this bare-bones description of the law governing pedestrian stops, three areas emerge where the law’s shadow might play: First, how does an officer decide who to stop, either for a consent encounter or when they have reasonable suspicion of a crime? Second, how does the reasonable suspicion standard operate? If the bar is low enough, an officer could have reasonable suspicion about almost everyone they meet. Third, what is the line between the different types of stops, and how does a stopped person know which side of the line they fall on? In other words, how does a person know whether they can end the encounter?⁵²See Rachel Harmon, Law and Orders, 123 Colum. L. Rev. 943, 969–74 (2023).

The law answers these questions by creating shadows that favor police. Though the law could constrain police discretion about who to engage for consensual or Terry stops, it does not. So long as the requisite level of suspicion is met (which is zero for consent stops), the Fourth Amendment allows an officer to approach and engage someone for any reason, or no reason, at all.⁵³Whren v. United States, 517 U.S. 806, 811–13 (1996) (holding that the subjective motivations of police are irrelevant to the Fourth Amendment analysis). The Equal Protection Clause’s prohibition on disparate and irrational treatment is a supposed constraint.⁵⁴See id. at 813. But courts have made proving equal protection claims extraordinarily difficult.⁵⁵See, e.g., Mario L. Barnes & Erwin Chemerinsky, The Once and Future Equal Protection Doctrine?, 43 Conn. L. Rev. 1059, 1066 (2011) (arguing that the Court’s framework “all but ensures only a narrow group of discrimination claims will be actionable or succeed”). The end result is a long legal shadow where police discretion is essentially unfettered.

Further, the reasonable suspicion standard is an easily satisfied “totality of the circumstances” inquiry.⁵⁶United States v. Arvizu, 534 U.S. 266, 273 (2002). Courts value an officer’s purported ability to identify criminal activity even when a layperson would view the activity as innocuous, thereby allowing otherwise innocent activities to create reasonable suspicion.⁵⁷See, e.g., United States v. Dyson, 639 F.3d 230, 234 (6th Cir. 2011) (finding reasonable suspicion when officer observed defendant removing items from vehicle immediately after auto accident, showing no concern for driver or accident investigation). Those activities can be as inoffensive as driving in the wrong sort of place or being too nervous to interact with the police.⁵⁸United States v. Pack, 612 F.3d 341, 361–62 (5th Cir. 2010). Perhaps the most startling incarnation is that reasonable suspicion exists to stop basically anyone in a “high-crime” area who is sufficiently averse to talking to the police.⁵⁹See, e.g., Illinois v. Wardlow, 528 U.S. 119, 124 (2000) (high-crime area plus flight); United States v. Oglesby, 597 F.3d 891, 894 (7th Cir. 2010) (finding reasonable suspicion in a high-crime area when officer observed defendant “slowly retreat[ing] from” a group being approached by police “while looking from side to side” and having “angled his body away from” officers). Of course, because this is a totality of the circumstances inquiry, sometimes being in a high-crime area and performing activities the police find suspicious isn’t enough for a court to find reasonable suspicion.⁶⁰See, e.g., United States v. Johnson, 620 F.3d 685, 692 (6th Cir. 2010) (finding no reasonable suspicion when Johnson was found “in a high drug-trafficking area” at 4:00 AM, was the only one close to a car identified in a 911 call the officers were responding to, and threw a bag into another car after not stopping when officers called him). Reasonable suspicion, then, is a low but unpredictable bar.

Additionally, a stopped person likely does not know what sort of interaction they are having. The courts have not required officers to affirmatively inform people whether a stop is consensual or whether it is a Terry stop, where an attempt to leave will be stopped with force.⁶¹Instead, the test is whether “in view of all the circumstances surrounding the incident, a reasonable person would have believed that he was not free to leave.” United States v. Mendenhall, 446 U.S. 544, 554–55 (1980). Instead, someone interacting with the police has two options. They can ask the officer if they are free to go, or they can just go and see what happens. The first option can permissibly be met with a truthful-but-less-than-clear answer,⁶²Seth W. Stoughton, Terry v. Ohio and the (Un)forgettable Frisk, 15 Ohio St. J. Crim L. 19, 22 (2017) [hereinafter The (Un)forgettable Frisk] (“My go-to phrase was some variant of ‘Hey, lemme talk at you for a minute.’ I might know that the individual I was approaching was not free to leave . . . but I didn’t want them to realize that.”); see Brandon Garrett & Seth Stoughton, A Tactical Fourth Amendment, 103 Va. L. Rev. 211, 265–66 (2017) (discussing “Verbal Judo” and “positive engagement”). and the second option may be met with violence.

B. The Stakes

There is one final piece of this equation: the stakes. The stakes for police officers in these interactions fall into two buckets. If the officer breaks the law and violates a stopped person’s constitutional rights, any evidence the officer discovers may be suppressed in court.⁶³Mapp v. Ohio, 367 U.S. 683 (1961). But see, e.g., Heien v. North Carolina, 574 U.S. 54 (2014) (holding that an officer’s reasonable mistake of law can still create reasonable suspicion). As Shielded makes clear, the risk of personal, economic, or professional consequences for officers is minuscule in all but the most egregious cases. See, e.g., chs. 4–5, 11 (describing procedural protections for police officers). Likely more important to the officer is the possibility that the stopped person will physically harm the officer.⁶⁴The (Un)forgettable Frisk, supra note 62, at 31 (“For more than 50 years, criminologists have identified a preoccupation with danger as one of the defining characteristics of policing.”).

The stakes for the stopped person appear facially similar but are ultimately categorically different. A stopped person is almost certainly going to be concerned that they, rightly or wrongly, will be subjected to the coercive machinery of the state. An arrest, after all, does not require definitive proof, only probable cause. And because of the criminal code’s breadth, an officer may have, develop, or fabricate (likely without consequence), reasonable suspicion for any number of crimes.⁶⁵I. Bennett Capers, Crime, Legitimacy, and Testilying, 83 Ind. L.J. 835, 860 (2008).

Significant academic, legal, and policy attention has been given to “broken windows” policing, which targets minor offenses.⁶⁶See, e.g., Alexandra Natapoff, The High Stakes of Low-Level Criminal Justice, 128 Yale L.J. 1648 (2019) (reviewing Issa Kohler-Hausmann, Misdemeanorland: Criminal Courts and Social Control in an Age of Broken Windows Policing (2018)); Amanda Geller, The Process Is Still the Punishment: Low-Level Arrests in the Broken Windows Era, 37 Cardozo L. Rev. 1025 (2016); Bernard E. Harcourt & Jens Ludwig, Broken Windows: New Evidence from New York City and a Five-City Social Experiment, 73 U. Chi. L. Rev. 271 (2006); Debra Livingston, Police Discretion and the Quality of Life in Public Places: Courts, Communities, and the New Policing, 97 Colum. L. Rev. 551, 578–91 (1997). These low-level charges, like resisting arrest or disorderly conduct, “which may raise special concerns about trumped up charges because of the vague nature of the offenses,” are dismissed at extraordinarily high rates.⁶⁷Andrew B. Coan, The Legal Ethics of Release-Dismissal Agreements: Theory and Practice, 1 Stan. J.C.R. & C.L. 371, 387 (2005) (finding that 43% of release dismissal cases between 1995 and 2005 involved resisting arrest and disorderly conduct). Indeed, one study found a resisting arrest conviction rate of only 23 percent.⁶⁸Scott Holmes, Resisting Arrest and Racism—The Crime of “Disrespect,” 85 UMKC L. Rev. 625, 632 (2017). Still, our system for processing low-level offenses has a “high tolerance for wrongful petty convictions. . . . As a result, every year the criminal system punishes thousands of petty offenders who are not guilty.”⁶⁹Alexandra Natapoff, Misdemeanors, 85 S. Cal. L. Rev. 1313, 1316 (2012). And even if they are cleared, too often “the process is the punishment” in our criminal system.⁷⁰See generally Malcolm M. Feeley, The Process Is the Punishment: Handling Cases in a Lower Criminal Court (paperback ed. 1992). The Supreme Court has given officers the ability “to effectuate a full custodial arrest for the most minor, nonjailable offenses.”⁷¹Alexandra Natapoff, Atwater and the Misdemeanor Carceral State, 133 Harv. L. Rev. F. 147, 152 (2020). This means every stop carries the risk of pretrial incarceration and disruptions to work, childcare, and other responsibilities as a natural result of detention,⁷²See id. at 153. as well as the numerous consequences of even a “minor” criminal conviction.⁷³Natapoff, supra note 69, at 1316–17.

As such, there is a seeming parallel between the first concerns discussed by both the police and the civilians they stop. Both groups would like to avoid what they might view as “wasted” use of the criminal system’s machinery, but this generalization hides vast differences. The officer is concerned that their police work might be undone because they overstepped the Constitution’s bounds. These stakes are usually low.⁷⁴See, e.g., The (Un)forgettable Frisk, supra note 62, at 21–22 (discussing how stops and frisks “were the background noise of my professional life, the elevator music of my law enforcement career”). But the stopped person is worried that the criminal legal system will be used to detain them, to cage them, to take them away from their family and work—even if only for a few hours or days—or to convict them, despite their innocence.⁷⁵See id. at 21 (“Those frisks . . . were without exception so utterly unremarkable as to be entirely forgettable. At least to me they were. I wonder now . . . whether those stops and frisks were as forgettable to the individuals on the receiving end. I rather doubt it.”).

The second set of stakes seems even more facially similar. Both officers and the stopped person would like to make it home that night. Although policing is certainly not the most dangerous profession,⁷⁶That would be logging. Danielle Kurtzleben, Being a Police Officer Is Dangerous. These Jobs Are More Dangerous., Vox (Aug. 22, 2014, 12:00 PM), https://www.vox.com/2014/8/22/6053627/being-a-police-officer-is-dangerous-these-jobs-are-more-dangerous [perma.cc/DZE9-YNM8]. it is relatively dangerous. Police suffer at a fatal work-injury rate multiple times the national average.⁷⁷Id. (reporting 15 fatal injuries per 100,000 full-time workers for police officers versus 3.4 per 100,000 for all occupations); U.S. Bureau of Lab. Stat., Fatal Work Injuries to Police Officers Fell 20 Percent in 2019, Econ. Daily (May 13, 2021), https://www.bls.gov/opub/ted/2021/fatal-work-injuries-to-police-officers-fell-20-percent-in-2019.htm [perma.cc/9YAV-M8S9] (finding a rate of 11.1 fatal injuries per 100,000 full-time workers for police officers versus 3.5 per 100,000 full-time workers for all occupations). A police officer’s fear that they won’t make it home at the end of their shift is not irrational. But that fear is also not, either by scale or structure, equal to the fear of the people they stop.

Putting it bluntly, police kill far more people than people kill police. While there are about 100 fatal work injuries to police officers in any given year,⁷⁸ U.S. Bureau of Lab. Stat., supra note 77 (reporting “86 fatal work injuries in 2019,” down “from the 108 reported in 2018”). These numbers include all fatal work injuries to police officers, not just those occurring during stops. police shoot and kill about 1,000 civilians each year.⁷⁹Fatal Force, Wash. Post https://www.washingtonpost.com/graphics/investigations/police-shootings-database [perma.cc/G8EK-4VVP] (last updated Oct. 12, 2023) (reporting over 8,000 fatal police shootings between 2015 and 2023). Each year police threaten or use force against about one million people, injure about 250,000 people, and cause 85,000 nonfatal injuries that require hospital treatment.⁸⁰ Law Enf’t Epidemiology Project, U.S. Data on Police Shootings and Violence, Univ. of Ill. Chi. https://policeepi.uic.edu/u-s-data-on-police-shootings-and-violence [perma.cc/RCT7-QTXE]. By contrast, the FBI reported that 60,105 law enforcement officers were assaulted in 2020, and 18,568 sustained injuries from those assaults. Melinda Urbina & Katie Chaumont, FBI Releases Statistics for Law Enforcement Officers Assaulted and Killed in the Line of Duty, FBI Dall. (Oct. 22, 2021), https://www.fbi.gov/contact-us/field-offices/dallas/news/press-releases/fbi-releases-statistics-for-law-enforcement-officers-assaulted-and-killed-in-the-line-of-duty [perma.cc/7BGK-5BBN]. Further, these numbers underestimate the violence—recognized in Terry—inherent to a frisk itself.⁸¹Terry v. Ohio, 392 U.S. 1, 16–17, 17 n.13 (1968) (“It is a serious intrusion upon the sanctity of the person, which may inflict great indignity and arouse strong resentment, and it is not to be undertaken lightly.”); see also The (Un)forgettable Frisk, supra note 62, at 25–30 (describing the process of frisking someone from the perspective of an officer).

Although one might dispute or complicate this basic-but-stark arithmetic,⁸²For example, one might argue that the comparison that really matters is the rate of fatal work injuries (11.1 per 100,000 police officers, as of 2019) versus the rate of fatal police shootings for the entire population (5.7 per one million for African Americans). See U.S. Bureau of Lab. Stat., supra note 77; Statista Rsch. Dep’t, Rate of Fatal Police Shootings in the United States from 2015 to October 2023, by Ethnicity, Statista (Oct. 6, 2023), https://www.statista.com/statistics/1123070/police-shootings-rate-ethnicity-us [perma.cc/YX9R-7RET]. there remain structural differences in the stakes police and civilians face. Police officers are concerned with illegal violence against them. Civilians, by contrast, must reckon with the possibility that they will be subjected not only to illegal excessive force from someone visibly armed, but also to legal force backed by the state that arms their aggressor. As Schwartz carefully catalogues, the law affords ample protections for police officers who meet the possibility of illegal force by a civilian with their own law-empowered force (ch. 5).

C. Negotiating in the Shadow of the Law of Police Stops

With this background, one can begin to analyze how negotiations during a stop in the shadow of the law of police proceed. The rest of this section applies the Mnookin and Kornhauser framework of preferences, bargaining endowments, strategic behavior, transaction costs, and uncertainty.⁸³See supra Section I.0

1. Preferences

The major preferences of the parties seem relatively simple.⁸⁴These are almost certainly not the only preferences either police or civilians are likely to have, but given the stakes of the encounter, they are the preferences likely to dominate. See supra Section II.B. Both parties would primarily like to avoid physical harm to themselves. The police have a secondary preference for constitutional policing—they would like to make a legal arrest to avoid having their time and effort invalidated by the rest of the criminal legal system. Civilians, by contrast, have a secondary preference to avoid the various harms that stem from forced engagement with the criminal legal system.⁸⁵See supra Section II.0.

What differs between the officer and civilian are (1) the likelihood of their primary preferences being violated and (2) the respective harms they would suffer from engaging with the criminal legal system. As I discussed earlier, the likelihood that a civilian will suffer violence from an officer seems far greater than the likelihood that a civilian will physically harm an officer.⁸⁶See supra notes 74–80 and accompanying text. However, given the grievousness of the potential harm, as well as legal and social acculturation, this seeming empirical difference may make no difference to the strength of the officer’s preference to avoid physical harm.⁸⁷See Seth W. Stoughton, Principled Policing: Warrior Cops and Guardian Officers, 51 Wake Forest L. Rev. 611, 631–67 (2016). Nobody is doing math when they think they might die.

But the secondary preferences are distinctly different. The officer has a desire to do a good job. Assuming police care about crime prevention and apprehending people who commit crime—an assumption that undergirds virtually all exclusionary rule jurisprudence⁸⁸See United States v. Leon, 468 U.S. 897, 900 (1984) (describing the exclusionary rule’s goal as “deterring official misconduct and removing inducements to unreasonable invasions of privacy”).—they want to ensure that any arrest they make is “good” (i.e., constitutional). This preference might rationally lead them to blur the lines between whether an encounter was consensual or required, hoping that the stopped person would expressly consent, thereby removing the question of reasonable suspicion; or, that a court would validate their action ex post, thereby turning an action taken in the shadow of the law into one explicitly blessed as legal.⁸⁹See The (Un)forgettable Frisk, supra note 62, at 22–23.

For the civilian, the primary and secondary preferences seemingly collapse into one crucial concept: liberty. To a stopped person, both the officer and the criminal legal system pose threats to their liberty. This threat to liberty can take the form of physical safety or dignity violations by a police officer or the broader system.⁹⁰Natalie Lima & Susan Nembhard, Pretrial Deaths in Custody Are Prevalent but Preventable, Urb. Inst. (Dec. 14, 2022), https://www.urban.org/urban-wire/pretrial-deaths-custody-are-prevalent-preventable [perma.cc/MBY8-NNYB]; Terry v. Ohio, 392 U.S. 1, 16–17, 17 n.13 (1968) (stating that police action may cause individuals to suffer “great indignity”). Thousands of deaths in pretrial detention, and the indignity with which the deceased may be treated, over the last decade serve as a stark reminder of this physical risk. Lima & Nembhard, supra; Jon Schuppe, A Mother Reported Her Son Missing in March. Police Kept the Truth From Her for Months., NBC News: Lost Rites (Oct. 26, 2023 5:28 PM), https://www.nbcnews.com/news/us-news/bettersten-wade-dexter-jackson-mississippi-police-missing-rcna121697 [perma.cc/KNA5-6VQQ] (reporting the story of a man fatally struck by a police car and buried in an unmarked grave after the police did not notify his next of kin, despite her being in active communication with investigators). Moreover, the disruption caused by an arrest—let alone prosecution or conviction—can lead to cascading negative social and economic consequences for the stopped person and their community.⁹¹See Natapoff, supra note 69, at 1316–17; Natapoff, supra note 71, at 147–49 (describing the lasting effects of a mother’s arrest for a seatbelt violation on her child).

Recognizing this overriding preference for liberty illuminates the tradeoffs a stopped person might be forced to make. A stopped person might believe that by not resisting physical harm caused by an overly aggressive officer, they can better convince the officer not to induce the liberty deprivations caused by an arrest. Alternately, another stopped person may believe the increased risk of extended legal proceedings is worth taking, if an attempt to resist or flee might minimize or avoid the physical or dignitary harms caused during a stop.

2. Bargaining Endowments

Schwartz’s work, culminating in Shielded, fulsomely recounts the bargaining endowments created by both substantive and procedural law that allow police to escape consequences for unconstitutional behavior. These include broad leeway for police decisionmaking, including whether to use force or whether to approach at all, when interacting with civilians, as well as numerous postencounter protections like qualified immunity and indemnification (chs. 4–5, 11). These endowments make the police “untouchable,” to borrow Schwartz’s label.

Civilians are protected by the Fourth Amendment. However, as I have briefly described, that Amendment’s touchstone is reasonableness. Whether filtered through the lens of reasonable suspicion, probable cause, or objectively reasonable use of force, the focus on reasonableness means there are few hard and fast rules that either a police officer or a civilian can rely on during a stop. This lack of clarity, combined with the broad legal protections given to police, reduces the Fourth Amendment’s force as a bargaining endowment for stopped pedestrians.⁹²Cf. Lin, supra note 8, at 1857 (discussing the ADA’s interactive process as a constraint on the bargaining power its accommodations mandate provides).

These endowments, however, are likely secondary to the fundamentally different legal baselines from which the two parties start. Police are legally authorized to use force against civilians. Civilians are not authorized to use force against police. In most places, the prohibition on using force against police includes the inability to resist an unlawful arrest or act in self-defense.⁹³Alice Ristroph, The Constitution of Police Violence, 64 UCLA L. Rev. 1182, 1230–35 (2017). Even in states that have a right to resist an unlawful arrest, courts have found state law does not affect the Fourth Amendment inquiry.⁹⁴Id. at 1234.

3. Strategic Behavior

The role of strategic behavior in these negotiations mirrors the one-sided nature of the other considerations. Police are legally permitted to obfuscate—and at times, outright lie—to gain compliance.⁹⁵See infra note 98 and accompanying text. This is perhaps seen most clearly in cases that blur the line between consent and mandate. Because they are not required to tell someone what sort of encounter they are engaged in,⁹⁶United States v. Mendenhall, 446 U.S. 544, 554 (1980). police are able to use unclear language and exploit naturally coercive circumstances⁹⁷See, e.g., Florida v. Bostick, 501 U.S. 429 (1991). to induce consent when they lack reasonable suspicion. And while there are legal limits on officers’ abilities to outright lie, those limits are themselves limited; in turn, the police’s ability and willingness to lie have seemingly flourished.⁹⁸Melanie D. Wilson, An Exclusionary Rule for Police Lies, 47 Am. Crim. L. Rev. 1, 4–15, 22–33 (2010) (describing evidence that police lie, as well as the Supreme Court’s jurisprudence around police lying).

For civilians, however, the opportunity to engage in strategic behavior is fraught, since lying to the police invites legal jeopardy. States may require you to identify yourself to the police when asked;⁹⁹Know Your Rights Stopped By Police, ACLU, https://www.aclu.org/know-your-rights/stopped-by-police [perma.cc/S5X2-LZNZ]; see also Hiibel v. Sixth Jud. Dist. Ct. of Nev., 542 U.S. 177 (2004). courts have found that an officer’s belief that someone is lying can help establish probable cause;¹⁰⁰See United States v. Ortiz, 422 U.S. 891, 897 (1975). and lying to the police in the course of an investigation can lead to criminal charges.¹⁰¹See, e.g., People v. Simon, 984 N.W.2d 800, 808 (Mich. Ct. App. 2021) (discussing Michigan’s obstruction of justice statute).

4. Transaction Costs

There are some stops in which no bargain can be reached. If a police officer wants to make an arrest, and a stopped person does not want to be arrested, then the two desired outcomes are fundamentally incompatible. The overwhelming majority of stops, however, do not end in arrest or any other enforcement action,¹⁰²Jeffrey Fagan, No Runs, Few Hits, and Many Errors: Street Stops, Bias, and Proactive Policing, 68 UCLA L. Rev. 1584, 1620 (2022) (finding that 12 percent of stops in the study’s sample resulted in an arrest or summons); Margaret A. Hickey, Consultant Report: Progress Update and Data Analysis of Chicago Police Department Stops Between 2018 and 2020, 36 tbl.2 (2023), https://www.aclu-il.org/sites/default/files/2023.06.14_consultant_report_0.pdf [perma.cc/A49W-ZLFM] (reporting that in 2018–19, 23.6 percent of Chicago Police Department stops resulted in any enforcement action). meaning a negotiated outcome occurs.

From one perspective on transaction costs, this makes sense. A police stop often involves relatively few individuals (such as a few officers and a stopped person), who are interacting face-to-face.¹⁰³Stoughton, supra note 62, at 25 (discussing the contact and cover tactic). The tangible transaction costs are low. The parties can communicate in real time with no financial expense or need to operate through intermediaries and little, if any, required formal process. Indeed, the police seemingly use informality to further lower transaction costs in these interactions.¹⁰⁴See Garrett & Stoughton, supra note 62, at 264–65 (describing how “officers can use conflict avoidance techniques to encourage cooperation”).

But the psychological transaction costs can be enormous. I can most succinctly describe them as a battle between fear and power. Both parties fear the other will cause them harm¹⁰⁵ Aaron Stagoff-Belfort, Daniel Bodah & Daniela Gilbert, Vera Inst. of Just., The Social Costs of Policing 2 (2022), https://www.vera.org/downloads/publications/the-social-costs-of-policing.pdf [perma.cc/LG3G-5HED]; see Stoughton, supra note 87, at 652–53 (noting that officers “are trained to consider each civilian they interact with as a potential threat and to approach every civilian interaction as a potential deadly-force encounter”).—but both parties also want to assert some degree of power. The law places the officer in a position of power that allows them to command the civilian’s obedience and dominate through force if necessary.¹⁰⁶Id. at 652–58 (describing how criminal law ultimately leads police to expect deference from civilians and punish those who “must be taught the consequences of ‘contempt of cop’ ” (quoting Jack L. Colwell & Charles “Chip” Huth, Unleasing the Power of Unconditional Respect: Transforming Law Enforcement and Police Training 89 (2010))). For the civilian, the desire is more basic—to have their full dignity recognized.¹⁰⁷See id. at 668–69 (discussing how minimizing harm and indignity to civilians can improve police-civilian interactions). It is, perhaps, a recognition that the power of the state wielded by the officer purportedly stems from the civilian, from “We the People.”¹⁰⁸ U.S. Const. pmbl. The fear of the other and the desire to assert one’s power collides, too often with deadly results.

Note again, however, the asymmetrical nature of fear and power that the counterparties possess. The law guards against the officer’s fear, and their power is legally enforced. The civilian faces the opposite structure: Their fear is legally enforced because the officer is visibly armed and empowered to use force against them, while their power is legally circumscribed. The state’s power may stem from “We the People,” but the state does not give any single person the legal ability to resist with force its monopoly on violence.

5. Uncertainty

If there is a single reason for the lengthy shadow of the law of the police in this arena, it is likely uncertainty. Jurisprudence in this area provides few rules that allow the police or civilians to know ex ante whether a court will decide a given action is legal or illegal.¹⁰⁹See Garrett & Stoughton, supra note 62, at 216–18. Nonsubstantive hurdles further entrench that uncertainty by making formal legal rulings rare.¹¹⁰See generally Aziz Z. Huq, Judicial Independence and the Rationing of Constitutional Remedies, 65 Duke L.J. 1 (2015) (discussing how scarce judicial resources limit the number of formal constitutional rulings and remedies given). This uncertainty casts an enormous shadow over police, who attempt to investigate and make arrests without violating unclear constitutional bounds, and over civilians, who attempt to escape the grip of the carceral state by standing on their rights. Because neither party has firm legal rights to fall back on, both would seem to have a strong incentive to reach a negotiated outcome. The police, for example, might refrain from using force they are legally entitled to use in exchange for a civilian providing information they are legally entitled to withhold.

While this negotiation in the face of uncertainty explains some police stops, it does not explain how the police developed such strong expectations of civilian deference in the face of legal uncertainty.¹¹¹See Stoughton, supra note 87, at 652–58. Shielded helps answer this question. Even if the doctrine is unclear about whether a court will deem any given action legal or illegal, it is clear who the doctrine favors: the police.

Considering each party’s ability to use force illuminates this favoritism. Because the facts of police-civilian interactions are near infinite, it is unlikely that a court has ruled that a particular use of force against a civilian in a particular situation is definitively legal. But the use-of-force test is designed to give officers leeway. It asks whether some use of force was “reasonable” from an (objectively reasonable) officer’s perspective.¹¹²Graham v. Connor, 490 U.S. 386, 397 (1989). And it explicitly recognizes the need for “split-second” decisionmaking by police, while also cautioning courts against Monday-morning quarterbacking.¹¹³Id. at 396–97. This test does not guarantee that using X force in Y situation is legal, but it tilts the field decidedly in the officer’s favor.¹¹⁴Mitch Zamoff, Determining the Perspective of a Reasonable Police Officer: An Evidence-Based Proposal, 65 Vill. L. Rev. 585, 608 (2020).

In contrast, a person has a legal entitlement to not consent to police requests.¹¹⁵Bumper v. North Carolina, 391 U.S. 543, 548 (1967). But they must exercise that entitlement while facing great uncertainty and enormous consequences if they guess incorrectly. They might not know whether an encounter is a consent stop or an investigative stop, and they will almost certainly not know whether the officer has adequate cause to detain them. Are they in a high-crime area? Did their rational fear of force lead them to make furtive movements? A person might be totally innocent, but the consequences of leaning into that uncertainty to stand on their rights are arrest and potentially prosecution. From there, they may be pushed into a guilty plea to avoid a lengthy criminal legal process.¹¹⁶Somil Trivedi, Coercive Plea Bargaining Has Poisoned the Criminal Justice System. It’s Time to Suck the Venom Out., ACLU (Jan. 13, 2020), https://www.aclu.org/news/criminal-law-reform/coercive-plea-bargaining-has-poisoned-the-criminal-justice-system-its-time-to-suck-the-venom-out [perma.cc/ATL9-DC3V]. Even if they are ultimately acquitted or the charges are dropped, significant damage may already be done through the costs and disruption of the process itself.¹¹⁷See generally Eisha Jain, Arrests as Regulation, 67 Stan. L. Rev. 809, 820–26 (2015).

The result of this analysis is that the rational response for a civilian being stopped by the police is not the back-and-forth negotiation that the platonic ideal of bargaining in the shadow of the law suggests. Instead, it is submission. The law of the police both forces civilians to act in its shadow and shapes that shadow, giving police a near-insurmountable bargaining position.

D. Why?

At first glance, it is not clear why the law would take this indirect route to increase police power. After all, if our concern is simply empowering the police to solve crimes or protect themselves, the law could establish clear rules that allow the police to do exactly that. And there are doctrines where this is the case. The police, for example, can rely on warrants issued by judges and magistrates so long as they are not facially deficient (an exceedingly low bar), or procured through a falsified affidavit—a scenario which the officer controls.¹¹⁸Arizona v. Evans, 514 U.S. 1, 9–10 (1995); Franks v. Delaware, 438 U.S. 154, 155–56 (1978).

But the law of the police does not provide such blanket permissions in most areas. To see why, turn again to police stops. Criticisms of pedestrian stops turn on two well-established facts. First, the percentage of stops that lead to arrests, other enforcement actions, or the apprehension of contraband flips Blackstone’s Ratio¹¹⁹“[B]etter that ten guilty persons escape than that one innocent suffer.” William Blackstone, Commentaries, *352. on its head. Only 12 percent of stops the NYPD made between 1998 and 2012 resulted in an arrest or summons,¹²⁰Fagan, supra note 102, at 1620. and 76 percent of Chicago Police Department stops in 2018–19 resulted in no enforcement action at all.¹²¹ Hickey, supra note 102, at 36.

Second, stop regimes are marked by large racial disparities. The prerequisite to being stopped is being in an area where police are. So it is unsurprising that in New York, “[a]t each level of Black and Latinx population concentration, there are more officers controlling for crime than in precincts with the lowest concentrations of Black and Hispanic residents.”¹²²Fagan, supra note 102, at 1632. Once civilians encounter officers, the racial disparities continue. In Chicago, “a Black person was nine times as likely to be stopped as a White person, and a Latino person was about three times as likely to be stopped as a White person.”¹²³ Hickey, supra note 102, at 5. Stops were racially disproportionate in each police district regardless of its crime rate or racial or ethnic makeup.¹²⁴Id. at 5–6. Black and Latino people were also patted down more often than white people, even though pat downs of those Black and Latino people were less likely to result in the discovery of contraband.¹²⁵Id. at 6. Other empirical research has shown that, in response to legal challenges, police often replaced the challenged “problematic practice with another, perhaps equally discriminatory, practice.”¹²⁶Alex Chohlas-Wood et al., Identifying and Measuring Excessive and Discriminatory Policing, 89 U. Chi. L. Rev. 441, 451 (2022).

Here is where we see the impact of allowing the shadow of the law to flourish instead of imposing legal commands. If the law forced the police to operate under strict rules about who to stop, when to make a stop, and how to behave during that stop, it would be much more difficult to maintain a racially disparate regime. The police might have to stop everyone who fell within the ambit of these rules, regardless of race or class, and would be in a much weaker bargaining position when they did.

Ultimately, analyzing the shadow of the law shines a light on how the law shapes negotiations that occur against the background socioeconomic inequalities that exist in our society. By focusing on these negotiations, we can see how the law affects these inequalities.

What happens in the shadow of the law of the police is a terrifying exacerbation of the underlying racial inequality in our society. By empowering “negotiations,” the law enables the police to demand submission from racial minorities under threat of prosecution or death, while simultaneously allowing the racial majority to largely escape this subordinating experience. And though my analyzing the shadow of the law of the police is novel, I am not the first to suggest the law has developed in this way for racially subordinating purposes.¹²⁷See, e.g., Paul Butler, The System Is Working the Way It Is Supposed to: The Limits of Criminal Justice Reform, 104 Geo. L.J. 1419 (2016); Dorothy E. Roberts, The Supreme Court, 2018 Term—Foreword: Abolition Constitutionalism, 133 Harv. L. Rev. 1, 19 (2019) (“[W]e can see the extreme cruelty and degradation that characterize today’s penitentiaries, police forces, and executions as the inevitable result of a racially subordinating system.”); India Thusi, Policing Is Not a Good, 110 Geo. L.J. Online 226 (2022).

The effects of this violent state-enforced subordination are as predictable as they are horrific. The police help maintain racialized borders that encourage racial segregation¹²⁸I. Bennett Capers, Policing, Race, and Place, 44 Harv. C.R.-C.L. L. Rev. 43, 68–69 (2009).—”[s]paces thought to be ‘public’ such as streets, parks, sidewalks, and shopping districts turn out not to be entirely public, but rather color-coded.”¹²⁹Id. at 69; see also Jeffrey Fagan & Elliott Ash, New Policing, New Segregation: From Ferguson to New York, 106 Geo. L.J. Online 33, 36 (2017) (“New Policing has been translated into a web of legal and financial controls that can only be appreciated as the new segregation.”). Policed peoples suffer “legal estrangement,” while policing reifies stereotypes of Black criminality and deals disproportionate harms to Black individuals and communities.¹³⁰Monica C. Bell, Police Reform and the Dismantling of Legal Estrangement, 126 Yale L.J. 2054, 2066–67 (2017); Ekow N. Yankah, Pretext and Justification: Republicanism, Policing, and Race, 40 Cardozo L. Rev. 1543, 1555–64 (2019). As Justice Sotomayor said, the law of the police “implies that you are not a citizen of a democracy but the subject of a carceral state, just waiting to be cataloged.”¹³¹See Utah v. Strieff, 136 S. Ct. 2056, 2071 (2016) (Sotomayor, J., dissenting).

Conclusion

Schwartz’s solutions seem overwhelmingly positive. Qualified immunity is ridiculous on many levels, and there are good reasons to end it (p. 229). Police departments should learn from lawsuits (p. 233). Municipalities should be held liable (p. 230). And we should diversify the bench (p. 237). I agree with Professor Brandon Hasbrouck that if we could implement only Schwartz’s prescriptions, with no backsliding elsewhere, “I’d be tempted to take it.”¹³²Brandon Hasbrouck, The Untouchables and the Stakes of Abolition, Balkinization (Mar. 7, 2023), https://balkin.blogspot.com/2023/03/the-untouchables-and-stakes-of-abolition.html [perma.cc/V2HU-M2P3].

But Hasbrouck’s concession has an enormous caveat. There is no guarantee that courts and legislatures would remain in stasis if Schwartz’s prescriptions became law. Indeed, I have previously argued that qualified immunity is not a dragon to be slain but a hydra: By cutting off its head, you only further empower the courts to rule against civil rights plaintiffs through substantive law.¹³³See Pyrrhic, supra note 4, at 1485–1518.

This returns us to why. If the why of police untouchability is an underlying desire for racial subordination, would these doctrinal changes matter? Or, even if the shadow of the law of the police was eliminated, would the doctrine simply morph to some shadowless shape that still advances our society’s subordinative desires?¹³⁴See id. at 1489–92 (discussing the equilibration thesis). To the extent that Shielded has any overarching flaw, it is this one. Because she focuses on the “how” of police untouchability, Schwartz fails to grapple with whack-a-mole nature of the problem she addresses.

It is here that the solution Schwartz does not suggest is glaring. Despite several references to the Summer 2020 Uprisings and their aftermath (pp. 227, 234–35, 238), Shielded’s final chapter contains only one mention of abolition. That mention does not occur in a discussion of the merits and demerits of abolitionists’ proposals. ¹³⁵See, e.g., Mariame Kaba, Opinion, Yes, We Mean Literally Abolish the Police, N.Y. Times (June 12, 2020), https://www.nytimes.com/2020/06/12/opinion/sunday/floyd-abolish-defund-police.html [perma.cc/L994-KU4J]. Schwartz never speaks the word herself. Instead, she quotes Republican Congressman Jim Banks, who said attempts to end qualified immunity are just “another way of saying abolish the police” (p. 227). In her response to Hasbrouck, Schwartz admits she decided not to address abolition head on.¹³⁶Joanna Schwartz, Not Everything That Is Faced Can Be Changed, but Nothing Can Be Changed Until It Is Faced, Balkinization (Mar. 10, 2023), https://balkin.blogspot.com/2023/03/not-everything-that-is-faced-can-be.html [perma.cc/BCB8-YJXP]. Schwartz, to her credit, begins to grapple with abolition in her most recent work. Joanna C. Schwartz, An Even Better Way, Calif. L. Rev. (forthcoming 2024–5) (on file with author). Given Shielded’s already impressive scope, I cannot fault her for that choice.

Nevertheless, I cannot help but think that Schwartz’s failure to engage with abolitionist theory and activism limited the depth of her solutions. Although abolition envisions change occurring through a bottom-up process,¹³⁷See, e.g., Amna A. Akbar, Demands for A Democratic Political Economy, 134 Harv. L. Rev. F. 90, 104–05 (2020) (arguing that nonreformist reforms “aim to shift power away from elites and toward the masses of people”). Schwartz assumes that change will occur in a top-down fashion: “How we move forward will depend—as it always has—on what those in power believe about the scope and severity of police misconduct and the need for accountability and reform.”¹³⁸P. 238; see also Aziz Z. Huq, A Comment on Shielded: How the Police Become Untouchable, Balkinization (Mar. 2, 2023), https://balkin.blogspot.com/2023/02/a-comment-on-shielded-how-police-become.html [perma.cc/Z885-9NHV]. But recent history has proven this description at best incomplete and at worst untrue. After decades of pro-police legal development, the movement to abolish qualified immunity gained steam after millions of regular people took to the streets.¹³⁹Ed Yohnka, Julia Decker, Emma Andersson & Aamra Ahmad, Ending Qualified Immunity Once and For All Is the Next Step in Holding Police Accountable, ACLU (Mar. 23, 2021), https://www.aclu.org/news/criminal-law-reform/ending-qualified-immunity-once-and-for-all-is-the-next-step-in-holding-police-accountable [perma.cc/4EWG-DE8Q]. Direct democracy has aided other battles against the carceral state.¹⁴⁰See, e.g., Adam Davidson, Administrative Enslavement, 124 Colum. L. Rev. (2024) (on file with author). And abolitionist groups have long led successful charges to build a better future through local organizing and community building.¹⁴¹See Allegra M. McLeod, Envisioning Abolition Democracy, 132 Harv. L. Rev. 1613 (2019).

Further, abolitionist theories of power shifting and contestation suggest how the law might durably alter the negotiations discussed in both Shielded and this piece.¹⁴²See Amna A. Akbar, Non-Reformist Reforms and Struggles Over Life, Death, and Democracy, 132 Yale L.J. 2497, 2511, 2562–77 (2023) (explaining that non-reformist reforms “embrace antagonism and conflict rather than depoliticization and neutrality” and “aim to shift the balance of power”). By ameliorating economic precarity through a more robust social safety net, we would increase the ability of people to bear some of the transaction costs of negotiating both civil rights litigation and police stops. Because many stops are negotiations, the law could empower civilians with information. Turning stops into “Know Your Rights” lessons by requiring police to divulge both the suspicion (if any) underlying a stop and a clear statement about whether a person is free to leave or refuse consent would reduce opportunities for strategic police behavior. Further, abolitionist proposals to replace police, and especially armed police, not only shrink the footprint of the carceral state but also may alter the preferences at play by making the threat of physical harm less salient for civilians—who would face someone tasked with solving problems through means other than violence—and nonpolice actors—who are less acculturated to viewing civilians as threats.

Abolition may not have the only answer. But it provides something that Schwartz does not. While Schwartz suggests an inevitability to the harms that police cause (p. 238), abolition dares us to imagine a future where these harms will not merely be reduced; they will be eliminated. Abolition forces us to propose solutions that address not only how the police are untouchable but why.¹⁴³See Roberts, supra note 127, at 118–19 (arguing for abolitionists to engage in “treating the symptoms while ending the disease” of prisons (cleaned up)).

Assistant Professor of Law, The University of Chicago Law School. I thank Aziz Huq, Richard McAdams, John Rappaport, Joanna Schwartz, and Fred Smith for thoughtful comments and conversations, the Paul H. Leffmann Fund for research support, and the editors at the Michigan Law Review for their excellent editorial work.