![Sample our Humanities journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5939/TOC-Subject-Sample-2021-Humanities.jpg"})
Abstract
Policing in many parts of the world—the United States in particular—has embraced an archetypal model: a conception of the police based on the tenets of individuated archetypes, such as the heroic police “warrior” or “guardian.” Such policing has in part motivated moves to (1) a reallocative model: reallocating societal resources such that the police are no longer needed in society (defunding and abolishing) because reform strategies cannot fix the way societal problems become manifest in (archetypal) policing; and (2) an algorithmic model: subsuming policing into technocratic judgements encoded in algorithms through strategies such as predictive policing (mitigating archetypal bias). This paper begins by considering the normative basis of the relationship between political community and policing. It then examines the justification of reallocative and algorithmic models in light of the relationship between political community and police. Given commitments to the depth and distribution of security—and proscriptions against dehumanizing strategies—the paper concludes that a nonideal-theory priority rule promoting respect for personhood (manifest in community and dignity-promoting policing strategies) is a necessary condition for the justification of the above models.
[For their thoughtful comments, I am grateful to Jonathan Jacobs, Alexander Schlutz, anonymous reviewers, the philosophy faculty at the University of Alabama, and the participants in the 2021 Workshop on Ethics in Criminal Justice AI at the University of Florida.]
[Disclosure Statement: No potential conflict of interest was reported by the author(s).]
Notes
1 See Grossman and Christensen, On Combat.
2 At the outset, I want to be clear that archetypal conceptions of policing (e.g., “warrior” and “guardian”) describe a policing ethos, both at the individual and collective levels. By contrast, the reallocative and algorithmic models do not prescribe an ethos in the same way, though they may impose constraints on them. The archetypal model is thus more about people and the character of their institutions, while the reallocative and algorithmic models are more about methods—though there is much grey area here. I discuss archetypal policing extensively in Hunt, The Police Identity Crisis, and will say little about it in this paper.
3 There are good reasons to think that political discretion is justified and legitimate, not only because of the practical issue of limited resources, but also due to a deep background norm of discretion in liberal polities. This background includes Locke’s arguments in favor of an executive prerogative power, suggesting that discretion is justified and legitimate given certain principled reasons for use of that discretion (e.g., national security situations). See Hunt, The Retrieval of Liberalism in Policing, chapter 5. Police discretion is in some sense derivative of this sort of executive discretion, though there are important boundaries of authority. See Monaghan, “Boundary Policing.” In short, the problem is not discretion itself, but rather exceeding the principled limits of discretion—which is a function of constraining rule of law principles. See Fuller, The Morality of Law; Raz, The Authority of Law. These points raise practical policy questions about how institutions can be reformed such that they operate in a way that is consistent with background norms regarding the rule of law (and mitigate unjustified and illegitimate uses of discretion).
4 By “archetypal bias” I mean bias (ethnic, gender, class, and so on) stemming from prerogative, discretionary power used by police conceived and trained as heroic “warriors” and “guardians” (compared with bias associated with algorithms).
5 Perry et al., Predictive Policing, xiii.
6 The methodology sketched here—a form of transitional nonideal theory—is derived from Rawls. See Simmons, “Ideal and Nonideal Theory,” for a comprehensive account of the method, and Hunt, The Retrieval of Liberalism, chapter 2, for an account of how the method might be applied to policing. This paper proceeds based on the assumption that we know enough about a broad outline of an ideal theory of justice to permit coherent nonideal theorizing (i.e., we know enough about the target ideal such that we have something for which to aim in our nonideal theorizing).
7 See Simmons, “Ideal and Nonideal Theory,” which explicates Rawlsian transitional nonideal theory.
8 Loader and Walker, “Policing as a Public Good,” 27.
9 See Rawls, A Theory of Justice, 3.
10 See ibid. 7.
11 Loader and Walker, “Policing as a Public Good,” 26.
12 See Waldron, Torture, Terror, and Trade-Offs, 116–22.
13 See ibid., 148.
14 See Adegbile, “Policing Through an American Prism,” 2232.
15 See Hunt, The Retrieval of Liberalism in Policing, chapter 3; see generally Jacobs, The Liberal State.
16 See Hunt, The Police Identity Crisis.
17 See Waldron, Torture, Terror, and Trade-Offs. Waldron discusses how legal archetypes become “a sort of emblem, token, or icon of the whole area of law … the spirit of the area of law in question” (227).
18 See, e.g., Stoughton, “Law Enforcement’s ‘Warrior’ Problem”; Stoughton, “Principled Policing”; and Hunt, The Police Identity Crisis.
19 In this sense, the reallocative approach is less a model of policing and more a model of public safety generally. The underlying idea seems to be that the less the police are empowered to do, the less harm they will cause and the better off society will be.
20 See Ray, “What does ‘defund the police’ mean?”
21 See McHarris and McHarris, “No More Money for the Police.”
22 See Perry et al., “To add value.”
23 According to a report by Forbes, at least thirteen cities cut police funding or decreased officer numbers (as of August 2020) in response to systemic racism and police brutality. See McEvoy, “At Least 13 Cities.”
24 See, e.g., Uggen and Shannon, “Productive Addicts and Harm Reduction.” Uggen and Shannon argue that work opportunity reduces crime but not drug use.
25 See, e.g., Wertheimer, “Are the Police Necessary”; Vitale, The End of Policing; McDowell and Fernandez, “Disband, Disempower, and Disarm.”
26 See, e.g., Vitale, The End of Policing, “How Much Do We Need the Police?”
27 McDowell and Fernandez, “Disband, Disempower, and Disarm,” 388.
28 Wertheimer, “Are the Police Necessary?”, 50.
29 Ibid.
30 See Vladeck and Wittes, “DHS Authorizes Domestic Surveillance.”
31 Wertheimer, “Are the Police Necessary?,” 49, 51.
32 Ingraham, “There are more guns.”
33 See Solomon, “Six years after”; Zook, “Lessons of Jordan Davis’s Murder”; Fausset, “What We Know.”
34 See Vitale, The End of Policing; Bayley, Police for the Future; Wertheimer, “Are the Police Necessary?,” 53–55.
35 Wertheimer, “Are the Police Necessary?” 55.
36 See Braga, et al., “Hot spots policing and crime reduction.”
37 See Wertheimer, “Are the Police Necessary?” 57.
38 Vitale, “Policing Is Fundamentally a Tool.”
39 See Bump, “Over the past 60 years.”
40 See Vitale, “The Best Way.”
41 McDowell and Fernandez, “Disband, Disempower, and Disarm,” 387.
42 Ibid. (quoting Spirithouse Executive Director Nia Wilson’s 2016 TED talk).
43 Sharkey, “Why do we need the police?” See also Braga, et al., “Hot spots policing,” for comprehensive empirical work showing that hot spots policing is an effective crime reduction strategy.
44 To be clear, I am not suggesting that algorithmic policing is necessarily a comprehensive model of policing. It is an approach that is used to allocate police resources for a relatively narrow range of crime based on statistically defensible predictions; it competes for resources with other needs (i.e., needs beyond a narrow range of property and violent crime), and in this sense may be construed as a narrow strategy for a subset of police responsibilities. On the other hand, it may be construed as a model inasmuch as it is considered an emerging strategy that could revolutionize our overall approach to policing.
45 See Fuller, The Morality of Law, and Raz, The Authority of Law, for discussions of the connections between dignity and the rule of law.
46 Brayne, Predict and Surveil, 3
47 I discuss this tactic more fully in Hunt, The Police Identity Crisis, chapter 4.
48 Santos, “Predictive Policing,” 372.
49 See ibid; see also Mohler et al., “Self-Exciting Point Process.”
50 See Santos, Crime Analysis; Mohler, et al., “Randomized controlled field trials.”
51 See Bond-Graham and Winston, “All Tomorrow’s Crimes.”
52 See Schauer, Profiles, Probabilities, and Stereotypes; Schauer, “Statistical (and Non-Statistical) Discrimination,” 47.
53 See, e.g., Rawls, A Theory of Justice, 24; Rubel, Castro, Pham, “Algorithms, Agency, and Respect for Persons”; Hunt, The Retrieval of Liberalism in Policing, 93–103 (on the rule of law), 104–33 (on liberal personhood), and 176–86 (on entrapment and probabilistic offenders); Hunt, The Police Identity Crisis, chapter 4 (on human rights). In one sense, the problem is that the algorithmic model dehumanizes people and communities not through the algorithms themselves, but rather by subjecting people to inherently dehumanizing interventions of policing.
54 See Ferguson, “Crime Mapping”’ and Ferguson, “Predictive Policing,” on this worry. It is constitutional for the police to “stop and frisk” a person (detain and pat them down briefly) if the stop and frisk is based upon reasonable suspicion (defined as “specific and articulable facts,” a level of proof less than “probable cause”) that the person is involved in criminal activity and the police have a reasonable belief that the person “may be armed and presently dangerous” (Terry v. Ohio, 392 U.S. 1 (1968)). However, if there is evidence that a police department implements aggressive stop and frisk initiatives amounting to an unconstitutional application of the tactic, then such violations may be exacerbated by the algorithmic model’s lack of qualitative and legal context.
55 See Shue, Basic Rights, as well as the discussion in Waldron, Torture, Terror, and Trade-Offs on these points. One need not commit to Shue’s claim about the basicness of security to think that sacrificing one group’s security for the benefit of another group’s security can be dehumanizing to the first group.
56 Waldron, Political Political Theory, 290–91.
57 Jenkins and Purves, “A Dilemma,” 328.
58 Ibid.
59 See Waldron, Political Political Theory, 56, for a discussion of Locke’s point about “separation in thought.”
60 Evans, “Google Strikes Deal With Hospital.”
61 To be sure, people are centrally concerned with their health when they seek medical professionals, and medical professionals have important relations with individuals. Medical professionals are also central agents in the overall maintenance of public health and are—in a sense—tied to democratic (or at least social) concerns. It thus seems right to say that—though they are often viewed as experts—the expertise of medical professionals has a humane-values orientation inasmuch as their primary consideration is the wellbeing and welfare of other human beings (and inasmuch as the first principle of medicine is do no harm). Arguably, then, both patients and doctors may in a sense be dehumanized in large, profit-driven hospital systems—especially when doctors have little or no discretion to disregard algorithmic guidance. This point suggests that there are a variety of ways that policies and practices can fall on the spectrum of dehumanization.
62 The underlying idea is simply that the algorithmic model may shape—and degrade—the character of the police-to-citizen interaction.
63 See Rubel, Castro, and Pham, “Agency Laundering.”
64 Ibid., 1033–34.
65 For example, “a handbook on data-informed community policing … was published on the LAPD’s website” just the day before the police chief announced an end to its use of PredPol. Miller, “LAPD will end controversial program.”
66 Thomas, “Why Oakland Police Turned Down.”
67 Rubel, Castro, & Pham, “Agency Laundering,” 1020.
68 Howell, “Google Morals,” 412.
69 See Rawls, Law of Peoples, 89.
70 See ibid.; see also Rawls, A Theory of Justice, 267.
71 In The Retrieval of Liberalism in Policing, I set forth a priority rule regarding respect for personhood based on a tripartite conception of liberal personhood; and in The Police Identity Crisis, I discuss justified policing in terms of procedurally just community policing steeped in a move expansive understanding of public reason (one that serves as a unifying rationale and moral foundation for a justified police role, given that public reason promotes legitimacy and security of person by bolstering human dignity within diverse societies). Here I explicitly marry those two ideas: procedurally just community policing as a way to pursue a priority rule based upon personhood. See also, Hunt, “Ice Cube.”
72 See, e.g., Owens, Weisburd, Amendola, and Alpert, “Can You Build a Better Cop?”; and Wood, Tyler, and Papachristos, “Procedural Justice Training Reduces.”
73 Community-oriented policing (COP) is an organizational strategy promoting community empowerment through community partnership and development, as well as collective efficacy. COP may be augmented by procedural-justice policing (PJP), which focuses specifically on legitimacy and “giving citizens police decision processes that manifest demonstrations of police fairness and regard for a person’s dignity.” Sampson, “The Community,” 210.