Thinking about contemporary American policing requires thinking about danger: the danger that police officers risk in service to citizens,1×1. See, e.g., Daniel Knowles, What the Cops Say, The Economist (Apr. 27, 2015, 7:13 PM), http://www.economist.com/blogs/democracyinamerica/2015/04/policing-america [http://perma.cc/4XH6-ZHN4]. and the growing fear among citizens that it is dangerous for them to be near police.2×2. See, e.g., Jenna Wortham, Racism’s Psychological Toll, N.Y. Times Mag. (June 24, 2015), http://www.nytimes.com/2015/06/24/magazine/racisms-psychological-toll.html. This toxic combination yields mutual mistrust,3×3. See, e.g., Manny Fernandez, Police Face a Long and Complex Task to Mend Distrust Deepened by Killings, N.Y. Times (Dec. 7, 2014), http://www.nytimes.com/2014/12/08/us/police-face-a-long-and-complex-task-to-mend-distrust-deepened-by-killings.html (reporting community mistrust of police); Knowles, supra note 1 (reporting police mistrust of community members). as well as contentious uses of police force.4×4. See, e.g., Ralph Ellis & Melissa Gray, Tamir Rice Report: No Proof Police Officer Shouted Warning Before Shooting, CNN (June 15, 2015, 10:36 PM), http://www.cnn.com/2015/06/13/us/tamir-rice-report [http://perma.cc/G5RN-47CK]; Alex Horton, Opinion, In Iraq, I Raided Insurgents. In Virginia, the Police Raided Me., Wash. Post (July 24, 2015), https://www.washingtonpost.com/opinions/in-iraq-i-raided-insurgents-in-virginia-the-police-raided-me/2015/07/24/2e114e54-2b02-11e5-bd33-395c05608059_story.html [http://perma.cc/G3NV-V72E]. In 2007, Professor Peter Kraska noted that an “estimated 45,000 SWAT-team deployments [were] conducted yearly among those departments surveyed” — up from an average 3000 annually in the early 1980s. Peter B. Kraska, Militarization and Policing — Its Relevance to 21st Century Police, 1 Policing 501, 506 (2007). Civil claims arising from such force are governed by the Fourth Amendment’s “objective reasonableness” standard;5×5. Graham v. Connor, 490 U.S. 386, 388 (1989). police officers are further protected, however, by the doctrine of qualified immunity, which precludes liability whenever there is not already “clearly established law” prohibiting their conduct.6×6. Pearson v. Callahan, 555 U.S. 223, 243 (2009). This defense, as the Supreme Court has put it, “protects ‘all but the plainly incompetent or those who knowingly violate the law.’” Ashcroft v. al-Kidd, 131 S. Ct. 2074, 2085 (2011) (quoting Malley v. Briggs, 475 U.S. 335, 341 (1986)). Critics have argued that it is in practice even more protective. See, e.g., Susan Bendlin, Qualified Immunity: Protecting “All but the Plainly Incompetent” (and Maybe Some of Them, Too), 45 J. Marshall L. Rev. 1023 (2012). Recently, in Milan v. Bolin,7×7. 795 F.3d 726 (7th Cir. 2015), cert. denied, 2016 WL 763262 (U.S. Feb. 29, 2016). the Seventh Circuit held that qualified immunity did not protect Evansville, Indiana, police from liability for a SWAT raid on a woman’s home following online threats traced to her IP address.8×8. Id. at 730. The court’s emphasis on the officers’ failure to investigate sufficiently could point future courts toward recognizing a new rule as part of clearly established qualified-immunity law: that police act unreasonably when they launch a SWAT raid9×9. That is, a “dynamic entry” raid, see Holland ex rel. Overdorff v. Harrington, 268 F.3d 1179, 1190 (10th Cir. 2001), using the military-type “Special Weapons and Tactics” to which the acronym refers, see Radley Balko, Overkill: The Rise of Paramilitary Police Raids in America, Cato Inst. 1 (2006), http://object.cato.org/sites/cato.org/files/pubs/pdf/balko_whitepaper_2006.pdf [http://perma.cc/AJ96-9UDP]. without further investigation in the face of a nonurgent, ambiguous threat.
On June 20, 2012, Evansville police learned of menacing posts on an online message board.10×10. Milan v. City of Evansville, No. 3:13-cv-1, 2015 WL 71036, at *1 (S.D. Ind. Jan. 6, 2015), aff’d in part sub nom. Milan v. Bolin, 795 F.3d 726 (7th Cir. 2015), cert. denied, 2016 WL 763262 (U.S. Feb. 29, 2016). The posts spoke of “hat[ing] police” and of “hav[ing] explosives” and “some new artillery . . . to test on there thin a$ $ vest.”11×11. Id. One stated: “[Chief] Bolin lives behind parkside.”12×12. Id. Another read: “4th of July a cops house gonna got hit.dont care about your kids or btchs lives.I dnt even care about my own life.”13×13. Id.
The police investigated right away. They discovered that the posts came from an IP address registered to the home that sixty-eight-year-old Louise Milan shared with her eighteen-year-old daughter Stephanie, and that Louise was related to three male Milans — each of whom had a criminal record involving violence or weapons, one of whom had lived at her house several years earlier, and another of whom held a gun in a Facebook photo.14×14. Id. at *1–2. They saw that one of the wireless networks that could have been Louise’s was unsecured.15×15. Id. at *1. And while surveilling Milan’s house, they spotted Derrick Murray, a man known to the investigating officers as a longtime “thorn to the police department,” two doors down.16×16. Id. at *2. Murray had already been convicted of intimidating a police officer, and had once spray-painted a menacing symbol on an officer’s property. Id. at *2 & n.4. The police did not observe any of the three male Milans at Louise’s house.17×17. Id. at *2.
At least two officers saw Murray as the probable culprit; nevertheless, the department executed a search warrant on Milan’s house on June 21 — the day after they learned of the posts.18×18. Milan, 795 F.3d at 728. They did so with “an eleven-man SWAT team,” which “broke open the front door and a nearby window, and . . . hurled two ‘flash bang’ grenades.”19×19. Id. at 729. The SWAT team then “rushed into the house, searched it from top to bottom (finding no males, and also no evidence of any criminal activity), handcuffed mother and daughter, led them out of the house, and questioned them briefly,” before releasing Louise and Stephanie Milan “to return to their damaged and smoking abode.”20×20. Id. For video of the raid, see SWAT Raid Video at Center of EPD Lawsuit Released, WFIE (Aug. 13, 2014, 7:02 PM), http://www.14news.com/story/26274555/swat-raid-video-at-center-of-epd-lawsuit-released [http://perma.cc/5JA2-XM7U]. A day later, the police discovered that Murray was in fact responsible for the posts; they then “politely requested that he come to police headquarters, which he did, where he was arrested without incident.”21×21. Milan, 795 F.3d at 729.
Milan filed suit in the U.S. District Court for the Southern District of Indiana,22×22. Milan, 2015 WL 71036. alleging three Fourth Amendment violations: (1) that the search warrant was defective and the search thus unreasonable; (2) that she was “falsely arrest[ed] and/or wrongly detain[ed]” during the search; and (3) that the police used “unreasonable and/or excessive force.”23×23. Id. at *3. Milan also leveled a “Monell” claim (after Monell v. Department of Social Services, 436 U.S. 658 (1978)), alleging that the underlying violations occurred “pursuant to an established policy, procedure or custom” of the Evansville police. Milan, 2015 WL 71036, at *3 (quoting Third Amended Complaint & Demand for Jury Trial ¶ 28, Milan, No. 3:13-cv-1 (S.D. Ind. July 11, 2014)). The court denied summary judgment on Milan’s Monell claim. Id. at *8. The defendants moved for summary judgment, and the district court granted it with respect to the first two claims.24×24. See id. at *4. With respect to the excessive force claim, however, the district court denied summary judgment, citing an array of factors25×25. The court treated the use of (and lack of care in using) the flash bangs as an important factor, see id. at *5–6, while also noting that the crime was “fairly low on the severity scale” and that the existence of any immediate threat was “questionable,” id. at *6. and concluding that the officers could not claim qualified immunity.26×26. Id. at *7–8. The defendants appealed the denial of qualified immunity, arguing that “[i]t was not clearly established that the circumstances with which the Officers were confronted did not constitute exigent circumstances and a threat to their safety.”27×27. Appellants’ Reply Brief at 8, Milan, 795 F.3d 726 (No. 15-1207).
The Seventh Circuit affirmed. Writing for the panel, Judge Posner28×28. Judge Posner was joined by Chief Judge Wood and Judge Williams. held that performing the raid “with flash bangs, yet without any but the most perfunctory, indeed radically incomplete, preliminary investigation” was too unreasonable to be covered by qualified immunity.29×29. Milan, 795 F.3d at 730. Judge Posner took notice of several investigative failings, emphasizing officers’ failure “to find out whose network” was unsecured, as well as their failure to “delay the search for a day or so to try to get a better understanding” of the situation.30×30. Id. at 728. He also expressed incredulity at the raid’s heavy-handed nature, contrasting the SWAT team’s “body armor[,] big helmets and . . . formidable rifles pointed forward” with the image of Milan’s daughter, whom he described, per the video footage, as “so small, frail, utterly harmless looking, and completely unresisting that the sight of her being led away in handcuffs is disturbing.”31×31. Id. at 729. For a link to the video, see supra note 20. In summing up the officers’ cumulatively unreasonable mistakes, Judge Posner castigated the officers’ use of flash bangs, which are permissible under Seventh Circuit precedent only when there is heightened danger and after police have taken special precautions.32×32. See Milan, 795 F.3d at 730 (citing Estate of Escobedo v. Bender, 600 F.3d 770, 784–85 (7th Cir. 2010)).
Judge Posner’s emphasis on the officers’ failure to make “a minimally responsible investigation”33×33. Id. could point a doctrinal way forward for judges concerned about preventing excessive force and restoring trust in the post-Ferguson era.34×34. Such concerns would be jurisprudentially relevant since, as Professor Alan Chen has noted, “qualified immunity has essentially become a doctrine based entirely on public policy considerations.” See Alan K. Chen, The Facts About Qualified Immunity, 55 Emory L.J. 229, 267 (2006). The case could prompt such judges to define as unreasonable, as part of clearly established qualified-immunity law, a SWAT raid launched without further investigation in the face of nonexigent, ambiguous danger.
Judge Posner’s opinion does not state where exactly the Evansville police crossed the unreasonableness line, but three aspects of the flash-bang raid seem important: the police executed the raid (1) without “a more extensive investigation,”35×35. Milan, 795 F.3d at 730. (2) while there was still plenty of unresolved ambiguity,36×36. See id. at 728 (“[T]he threats might have come from a person (or persons) inside the Milan home who might moreover be armed and dangerous . . . .”); id. at 730 (“The open network expanded the number of possible threateners . . . .”). and (3) while there was still time to investigate further.37×37. See id. at 728 (“Prudence counseled delaying the search for a day or so to try to get a better understanding . . . .”). And though courts must “consider [qualified immunity] question[s] in ‘the specific context of the case,’”38×38. Mullenix v. Luna, 136 S. Ct. 305, 311 (2015) (quoting Brosseau v. Haugen, 543 U.S. 194, 198 (2004) (per curiam)). a rule governing nonurgent SWAT raids does not seem overly broad. Rather, it would be analogous to the rule that the Seventh Circuit already imposes for flash bangs, which treats them as unreasonable absent “a dangerous suspect,” “a dangerous entry point,” and special precautions.39×39. Milan, 795 F.3d at 730 (quoting Estate of Escobedo v. Bender, 600 F.3d 770, 784–85 (7th Cir. 2010)).
Could such a rule — effectively a duty to investigate ambiguous, nonurgent threats further before launching a SWAT raid — avoid falling back into the vagueness of general “reasonableness”? In line with the Supreme Court’s admonitions against “defin[ing] clearly established law at a high level of generality,”40×40. Ashcroft v. al-Kidd, 131 S. Ct. 2074, 2084 (2011). the rule would supply a focused test to apply to a specific police activity: the nonemergency SWAT raid. And while terms like “ambiguity” and “nonurgent” are facially vague, courts have already given them substance in other Fourth Amendment contexts. For example, courts have imposed an analogous “duty to investigate further” in situations “where an officer is presented with ambiguous facts related to authority” to consent to a search,41×41. United States v. Kimoana, 383 F.3d 1215, 1222 (10th Cir. 2004). and they have assessed whether ambiguity exists by asking whether “the facts available . . . [would] ‘warrant a man of reasonable caution in the belief’” that the consent was authorized.42×42. Illinois v. Rodriguez, 497 U.S. 177, 188 (1990) (quoting Terry v. Ohio, 392 U.S. 1, 21–22 (1968)). For investigatory stops, meanwhile, the Court has suggested that ambiguity exists when something is “susceptible of an innocent explanation.”43×43. Illinois v. Wardlow, 528 U.S. 119, 125 (2000). The Milan facts would count as ambiguous on either of these definitions: the unsecured wireless network, Murray’s presence two doors down, and the failure to turn up evidence of a dangerous person at Milan’s home all created substantial ambiguity, both suggesting a highly plausible alternate explanation and precluding a reasonably cautious person from concluding that Milan’s house harbored the culprit.
Courts likewise have ample experience defining “urgent need,” or “exigent circumstances,” in the search-and-seizure context. The D.C. Circuit, for example, has indicated that an urgent need is “a need that could not brook the delay” that otherwise-required procedure would entail.44×44. Dorman v. United States, 435 F.2d 385, 392 (D.C. Cir. 1970) (en banc); see also id. at 392–93 (listing factors for making the determination). On that definition, the Milan facts would fall short of urgency: not because the culprit named a date,45×45. Although the trial judge found that fact significant in determining that there was no “emergency situation,” see Milan v. City of Evansville, No. 3:13-cv-1, 2015 WL 71036, at *6 (S.D. Ind. Jan. 6, 2015), aff’d in part sub nom. Milan v. Bolin, 795 F.3d 726 (7th Cir. 2015) (“[T]here was no real ‘emergency situation’ . . . ; the suspect alleged that he would take action two weeks later on July 4.”), Judge Posner omitted that justification from his opinion. Judge Posner’s choice seems wise; there is no reason to take a person who threatens the lives of police at his word. but because the police already had means available — ongoing surveillance — to continue investigating without fear of being caught unaware by an attack from Milan’s house.46×46. Though Judge Posner did not state exactly why there was time to “delay the search for a day or so,” Milan, 795 F.3d at 728, his observation that “just one extra day of surveillance” would have helped “reassure the police that there were no dangerous men lurking in the house,” id. at 730, could imply that existing surveillance decreased the situation’s urgency. They could keep an eye out while continuing to piece together the largely unfinished puzzle.
A rule that police act unreasonably if they fail to further investigate ambiguous, nonurgent threats before launching a SWAT raid seems likely to yield three benefits: fewer constitutional violations, more police training, and greater police-community trust.
First, the rule would help safeguard a constitutional right that could use additional safeguarding. Currently, courts’ principal means of deterring Fourth Amendment violations is suppression of inculpatory evidence.47×47. See, e.g., James v. Illinois, 493 U.S. 307, 319 (1990) (noting the exclusionary rule’s “ability ‘to compel respect for the constitutional guaranty in the only effectively available way — by removing the incentive to disregard it’” (quoting Elkins v. United States, 364 U.S. 206, 217 (1960))). But suppression “gives no direct remedy to the innocent woman wrongly searched.”48×48. Akhil Reed Amar, The Constitution and Criminal Procedure 91 (1997). Tort, however, does, and the rule proposed here would make it easier for people who can’t benefit from suppression to successfully sue — bolstering the tort-based remedies “presupposed by the Framers of the Fourth Amendment”49×49. Id. at 21; see also id. at 20–22, 31–45, 106–07. but inhibited by current qualified-immunity law. With roughly 45,000 SWAT raids now being executed across the country each year,50×50. Kraska, supra note 4, at 506. such bolstering seems due.
Second, increasing access to such remedies — and thereby shifting the cost of breaches from citizens to the government — may prompt cities to further invest in what many agree is crucial to avoiding future problems: training. Such cost-shifting is the virtue that many thinkers — not least of all Judge Posner — have recognized in tort law, which need not encode the moral blame of criminal law.51×51. See, e.g., William M. Landes & Richard A. Posner, The Economic Structure of Tort Law 187 (1987) (“The economic purpose of tort law . . . is to internalize the costs of accidents.”). See generally Guido Calabresi, The Cost of Accidents (1970). From this perspective, the raid on Milan’s house presents a less charged question: the police made a decision that imposed costs — on Milan’s property and inner peace52×52. Judge Posner noted that the City “replaced the broken door and window, and the burned rug,” but that “[t]here was doubtless other damage” and that the court was “guessing that the principal harm for which compensation [was] sought [was] emotional.” Milan, 795 F.3d at 729. — and the law decides whether those costs are spread among Evansville taxpayers or whether Milan must bear them herself. If the city pays whenever its agents conduct a SWAT raid too hastily, it will be more inclined to train them to investigate more thoroughly and deploy force more carefully.53×53. It is possible, of course, that such a rule would also hinder police from protecting citizens from dangerous crime. Cf., e.g., Julian Grace, Full-Time SWAT Making Indy Streets Safer, WISH-TV (Jan. 14, 2015, 7:18 PM), http://wishtv.com/2015/01/14/full-time-swat-making-indy-streets-safer [http://perma.cc/V9XR-SP8W]. But cf., e.g., Radley Balko, Paramilitary Police Don’t Make Us Safer, Cato Inst. (Apr. 15, 2010), http://www.cato.org/publications/commentary/paramilitary-police-dont-make-us-safer [http://perma.cc/XP4L-SDVC]. This empirical debate cannot be resolved here. Prominent departments are already betting on similar training;54×54. See Timothy Williams, Long Taught to Use Force, Police Warily Learn to De-escalate, N.Y. Times (June 27, 2015), http://www.nytimes.com/2015/06/28/us/long-taught-to-use-force-police-warily-learn-to-de-escalate.html; see also Gene Johnson, Inside the DOJ’s $1.5M Plan to Train Cops to Use Their “Wits Rather than Their Weapons,” Talking Points Memo (May 22, 2015, 3:25 PM), http://talkingpointsmemo.com/news/police-demilitarization-training-blue-courage [http://perma.cc/9V68-JGXP]. a firmer rule should boost that trend.
Third, to the extent citizens are likely to see SWAT raids on their neighbors’ homes as strong prophylaxis — to be employed only in cases of great exigency or clear danger — a city’s paying restitution where such raids fall below that bar can restore trust in law enforcement and, by extension, the law.55×55. Judge Posner may have been gesturing at this point when he observed that the broadcasting of a video of the raid — executed by an apparently all-white SWAT team on a black family’s home — “cannot have helped race relations in Evansville.” Milan, 795 F.3d at 730. As behavioral-science scholars have demonstrated, trust is no idle perk, but rather a prerequisite for a healthy system: “If the public generally view the police as legitimate, much of their everyday behavior will conform to the law . . . . Further, the efforts of the police to manage . . . problematic people and situations will be aided by cooperation from the public.”56×56. Jason Sunshine & Tom R. Tyler, The Role of Procedural Justice and Legitimacy in Shaping Public Support for Policing, 37 Law & Soc’y Rev. 513, 535 (2003); see also Wesley G. Skogan & Tracey L. Meares, Lawful Policing, Annals Am. Acad. Pol. & Soc. Sci., May 2004, at 66; Tom R. Tyler, Procedural Justice, Legitimacy, and the Effective Rule of Law, 30 Crime & Just. 283 (2003). At the same time, there is evidence that police officers who perceive themselves and their departments as more procedurally fair are themselves more likely to engage with the community.57×57. See Scott E. Wolfe & Justin Nix, The Alleged “Ferguson Effect” and Police Willingness to Engage in Community Partnership, Law & Hum. Behav., Oct. 12, 2015, at 7–8, http://www.apa.org/pubs/journals/releases/lhb-lhb0000164.pdf [http://perma.cc/96W7-LZDP]. The cycle, in other words, can be virtuous.
Despite these potential benefits, courts would likely (and justifiably) reject a definition of reasonableness that seemed prone to endanger police, either by keeping them from tackling threats head on or by pushing them to enter dangerous situations unprotected. Those broader safety concerns are palpable in Milan,58×58. See, e.g., Appellants’ Reply Brief at 3–9, Milan, 795 F.3d 726 (No. 15-1207). and ought to carry substantial weight coming from public servants who risk their lives.59×59. The safety concerns for police in Milan were especially weighty, of course, given the threats quoted above.
Fortunately, the rule proposed here is unlikely to force police as a whole to bear extra risk; in fact, it could make them safer. The rule’s deterrence, after all, is mild: clearly dangerous or exigent situations fall outside its limited scope, and police facing a borderline case would still have every reason to err on the side of self-protection in choosing their methods, since the worst sanction they would virtually ever face for an improper SWAT raid is civil liability ultimately borne by city taxpayers.60×60. See generally Joanna C. Schwartz, Police Indemnification, 89 N.Y.U. L. Rev. 885 (2014) (reporting that “governments paid approximately 99.98% of the dollars that plaintiffs recovered” in civil rights suits against law enforcement, id. at 885). Moreover, it’s possible that fewer SWAT raids in ambiguous situations would increase police safety by averting tragedies in which innocent people respond with putative self-defense, assuming themselves to be subject to a home invasion rather than a police search.61×61. See, e.g., Ellen Tumposky, Arizona SWAT Team Defends Shooting Iraq Vet 60 Times, ABC News (May 20, 2011), http://abcnews.go.com/US/tucson-swat-team-defends-shooting-iraq-marine-veteran/story?id=13640112 [http://perma.cc/PNU4-H5ZT]; see also Radley Balko, Rise of the Warrior Cop 278 (2013) (“[I]n raids for nonviolent offenses, sowing confusion only increases the potential for violence. On numerous occasions, police have detonated a flash-bang grenade in the course of a raid, then claimed that a suspect who subsequently grabbed a gun or a knife or who physically attacked one of the officers should have known that it was the police raiding the home.”); cf. id. at 214 (quoting a former federal law enforcement agent as observing: “[I]t’s so much safer to wait the suspect out. . . . [Y]ou wait until the guy leaves, and you do a routine traffic stop . . . . That’s the safest way to do it.”).
Even though the rule discussed here should not visit moral censure or added systemic risk on police, imposing it would entail calling some self-protective measures unreasonable. When weighing such a rule, then, we should still ask how much general danger the law can expect individual officers to reasonably bear without resorting to whatever self-protective force they think best.62×62. Cf. Graham v. Connor, 490 U.S. 386, 396–97 (1989). One case on which Milan partly relied seems to set a high bar: there, the Seventh Circuit imposed liability on police for deploying flash bangs in confronting someone who was “not an unusually dangerous individual,” even though that person was high on cocaine, armed with a gun he refused to surrender, and suffering from extreme emotional distress.63×63. Estate of Escobedo v. Bender, 600 F.3d 770, 786 (7th Cir. 2010); see also id. at 773–74. Viewed in that light, the reasonable officer seems to be one who shows uncommon restraint in a stressful situation. Milan could foretell a comparable shift: defining the reasonable officer as one who, before deploying the publicly entrusted power of a SWAT raid, continues to investigate peacefully so long as there is both ambiguity and time to resolve that ambiguity. That definition, though it certainly could expose cities to more liability, should not increase risks to police. And any taxpayer cost would be an investment — in trust, legitimacy, and “[t]he right of the people to be secure . . . against unreasonable searches and seizures”64×64. U.S. Const. amend. IV. — that judges could deem well worth the price.