Harvard Law Review Harvard Law Review Harvard Law Review

Harvard Law Review Forum

Process Costs and Police Discretion

Cities across the country are debating police discretion. New York, for example, has recently endeavored to end its controversial practice of stopping and frisking citizens as a matter of course.1×1. E.g., Joseph Ax, NYC Police to Reform Public Housing Stop-And-Frisk in Settlement, Reuters (Jan. 8, 2015, 4:38 PM), http://www.reuters.com/article/2015/01/08/us-usa-newyork-stopandfrisk-idUSKBN0KH25R20150108 [http://perma.cc/L633-4ABD]. The debate over police discretion implicates fundamental questions about the role of police in American society, racial discrimination in the criminal system, and the disproportionate use of violence by the police against young black men. Much of the debate over the proper scope of police discretion centers on reforming the criminal code to decriminalize or eliminate minor crimes. This Essay argues that the debate over the proper scope of police discretion should instead focus on the real source of that discretion: the process costs of low-level adjudication.

Minor crimes are a big problem. In 2006 alone, Americans were charged with and detained on misdemeanor offenses approximately 10.5 million times.2×2. Robert C. Boruchowitz et al., Nat’l Ass’n of Criminal Def. Lawyers, Minor Crimes, Massive Waste: The Terrible Toll of America’s Broken Misdemeanor Courts 11 (2009), https://www.nacdl.org/WorkArea/DownloadAsset.aspx?id=20808 [https://perma.cc/8H2H-7E74?type=pdf]. These cases have serious long-term consequences for defendants, their families, and our criminal justice institutions.3×3. See id. at 11–13; see also Marc Santora, City’s Annual Cost Per Inmate Is Nearly $168,000, Study Finds, N.Y. Times (Aug. 23, 2013), www.nytimes.com/2013/08/24/nyregion/citys-annual-cost-per-inmate-is-nearly-168000-study-says.html; Facilities Overview, City of N.Y. Dep’t of Corr., http://www.nyc.gov/html/doc/html/about/facilities-overview.shtml (last visited Apr. 9, 2015) [http://perma.cc/9KJ3-XDS4] (noting that on an average day, there are approximately 11,400 inmates in New York City’s jails). They create criminal convictions and criminal records.4×4. E.g., Alexandra Natapoff, Misdemeanors, 85 S. Cal. L. Rev. 1313, 1313 (2012). They crowd our jails.5×5. See Boruchowitz, supra note 2, at 7. And minor convictions are usually imposed with little process, without counsel, and often regardless of factual guilt or innocence.6×6. Josh Bowers, Punishing the Innocent, 156 U. Pa. L. Rev. 1117, 1119 (2008). Worse, these crimes and convictions arguably form the core of our criminal justice system: while most people incarcerated in the United States were convicted of a felony, a large majority of criminal sentences imposed come from misdemeanor and violation convictions.7×7. Natapoff, supra note 4, at 1320–21.

Many of these minor convictions result from what are often called “public order” offenses.8×8. See Debra Livingston, Police Discretion and the Quality of Life in Public Places: Courts, Communities, and the New Policing, 97 Colum. L. Rev. 551, 556 n.14 (1997) (citing Robert C. Ellickson, Controlling Chronic Misconduct in City Spaces: Of Panhandlers, Skid Rows, and Public-Space Zoning, 105 Yale L.J. 1165, 1168, 1217–19 (1996)). These offenses are relatively petty to be sure, but their more important defining feature is that the actual sentence violators receive for their transgressions — usually time already served in detention via a guilty plea — is not the “punishment” that ought to matter to policymakers. In practice, our criminal justice system primarily enforces public order prohibitions prior to any conviction by subjecting the accused to arrest, detention, and other legal process.9×9. See Malcolm M. Feeley, The Process Is the Punishment: Handling Cases in a Lower Criminal Court (1979). For rough estimates of how many defendants plead guilty to avoid prolonged detention, see generally Charlie Gerstein, Note, Plea Bargaining and the Right to Counsel at Bail Hearings, 111 Mich. L. Rev. 1513, 1515 (2013) (“In 2010, in New York City alone, 16,649 defendants were unable to make bail set at one thousand dollars or less,” id. at 1515 n.3, and “[d]efendants who are required to post bail that they cannot afford . . . end up pleading guilty to avoid waiting in jail,” id. at 1515 (citing Douglas L. Colbert, Prosecution Without Representation, 59 Buff. L. Rev. 333, 348–52 (2011))). These “process costs” are significant; they include not just pre-trial detention, but also the hassle of pre-trial and trial proceedings and the risk and uncertainty that those proceedings necessarily entail.10×10. See Issa Kohler-Hausmann, Misdemeanor Justice: Control Without Conviction, 119 Am. J. Soc. 351, 374 (2013) (discussing “procedural hassle” as a technique for control of vulnerable populations); see also Josh Bowers, Grassroots Plea Bargaining, 91 Marq. L. Rev. 85, 86 (2007). These costs distort plea bargaining so much that the substantive law behind the bargained-for conviction becomes irrelevant.11×11. See William J. Stuntz, Plea Bargaining and Criminal Law’s Disappearing Shadow, 117 Harv. L. Rev. 2548, 2563–64 (2004) (arguing that as the severity of the criminal offense decreases, the influence of the substantive law wanes). Defendants are likely to spend more time in jail if they contest the charges than if they plead guilty. Not surprisingly, they almost always plead guilty, whether or not they committed the charged offense and despite the fact that the criminal conviction may result in serious consequences down the road.12×12. Bowers, supra note 6, at 1119 (discussing the “innocence problem” in plea bargaining); Natapoff, supra note 4, at 1323–27 (discussing collateral consequences); Gerstein, supra note 9, at 1526 (citing Kevin C. McMunigal, Disclosure and Accuracy in the Guilty Plea Process, 40 Hastings L.J. 957, 987 (1989)).

Maintaining public order is nevertheless an important civic function. Many of these offenses — disorderly conduct, minor trespassing, loitering — attempt to serve this function by giving police discretion that allows them to disrupt, to isolate, and to sober.13×13. See Livingston, supra note 8, at 591–92; see also Adriaan Lanni, The Future of Community Justice, 40 Harv. C.R.-C.L. L. Rev. 359, 366–67 (2005). The use of vague terms and broad standards in drafting statutory language can deliver such discretion. In the minds of some, however, discretion leads to abuse, a conclusion that has engendered a heated debate about how much statutory discretion the law should make available to police.14×14. Compare, e.g., Tracey L. Meares & Dan M. Kahan, The Wages of Antiquated Procedural Thinking: A Critique of Chicago v Morales, 1998 U. Chi. Legal F. 197, and Livingston, supra note 8, with, e.g., Albert W. Alschuler & Steven J. Schulhofer, Antiquated Procedures or Bedrock Rights?: A Response to Professors Meares and Kahan, 1998 U. Chi. Legal F. 215.

We do not join this battle. Instead, we suggest that criminal law process costs essentially decouple statutory discretion from actual police behavior, rendering the debate about statutory language largely moot. In other words, in the minor crimes context, process costs — not vague statutory terms — produce police discretion. Abuse is thus better addressed by first recognizing that, in the context of public order crimes, discretion has little to do with substantive criminal law and that, instead, focus is much better placed on mitigating the harmful consequences that discretion can generate and on limiting police discretion through other means. To this end, we propose providing the police with new civil enforcement tools that will be equally effective at preserving order but that will in all likelihood cause significantly less unnecessary harm. We believe — counterintuitively, perhaps — that giving the police an additional power (noncriminal arrest) might encourage them to use the power they currently have (criminal arrest) less frequently and, therefore, might reduce the harm that this latter power causes.

Our argument begins with the fact that when the police feel they need to arrest someone to keep people safe or to prevent property destruction, the police will, in most cases, arrest that person — regardless of how specific or general a given city’s criminal code may be.15×15. See Bowers, supra note 10, at 85–86; David Cole, Foreword: Discretion and Discrimination Reconsidered: A Response to the New Criminal Justice Scholarship, 87 Geo. L.J. 1059, 1071–73 (1999). Why? Because the specificity of the criminal code has little relationship to the costs imposed by an arrest, and it is the ability to impose at least some of these costs (for example, temporary removal and detention) that allow the police to achieve certain ends — disruption, isolation, and others.

American police have an extraordinarily diverse set of responsibilities,16×16. See infra notes 110–114 and accompanying text. and they approach their work with a wide variety of goals in mind. We do not mean to address these goals comprehensively. Rather, we focus only on the goal of “maintaining order” — chiefly, controlling or interrupting low-level misconduct and disrupting potential short-term violence. To achieve this goal, police sometimes have no choice other than to arrest people in order to temporarily isolate them for a few hours or to remove them from a particular location. In such a scenario, police may care about what happens before (and only some of what happens before) any conviction, but not the conviction itself or its consequences for the defendant.

If the police in certain circumstances are indifferent to whether a defendant is convicted of a crime, a realized conviction is likely to be a social waste. And in any event, code reforms are unlikely to control police discretion. When discretionary arrests turn on considerations other than the substantive law that underlies public order criminal offenses, police ought to have tools that do not trigger unnecessary collateral consequences, including criminal records, meaningless pleas, unnecessary risk and uncertainty, and useless (from a police officer’s perspective) process costs. Cities should adopt civil ordinances that free the police to make discretionary arrests for low-level violations, but limit the tendency of those arrests to inflict socially useless harm on defendants.

This Essay proceeds in three parts. Part I quickly recounts the realities of low-level criminal punishment in big cities and shows that low-level arrests are untethered to substantive law, rendering solutions that work within the criminal law unlikely to be effective at controlling police discretion. Part II outlines the debate over discretion to police public order, and argues that it neglects the reality that substantive law is mostly irrelevant to the matter of police discretion in this domain. Part III proposes a solution that comes from a long line of police practice: civil laws with strictly limited periods of detention and other features designed to reduce or eliminate those process costs that have no connection to what police are supposed to be trying to do — maintain order.

Before moving on, we note that the purpose of this Essay is not to discuss whether the police should arrest people as often as they apparently do. Rather, operating on the assumption that the police do feel the need to arrest people, we seek to ameliorate the consequences of those arrests by reforming the law in a particular way. As we explain below, we believe that our proposal (or something like it) can reduce the negative effects of many public order policing arrests without increasing the total number or consequent burdens of such arrests.

I. Arrests for Public Order Offenses

In very low-level misdemeanor prosecutions, the substantive criminal law that generates the punishment is largely irrelevant. Instead, a conviction is the near-certain result of the arrest, and the punishment is the process of criminal arrest, pretrial detention, and adjudication.17×17. See Feeley, supra note 9, at 199; Bowers, supra note 10, at 86; Bowers, supra note 6, at 1119; Natapoff, supra note 4, at 1328; Stuntz, supra note 11, at 2568.

Consider New York City today. The police see (or learn of) someone doing something they do not like.18×18. Sometimes, that can be wearing your pants too low. People v. Martinez, 905 N.Y.S.2d 847, 847 (Crim. Ct. 2010). That person is arrested for a minor offense,19×19. New York law characterizes many of these minor offenses as noncriminal violations. Nonetheless, the evidence suggests that these offenses create serious long-term problems. E.g., Kohler-Hausmann, supra note 10, at 383. usually disorderly conduct,20×20. N.Y. Penal Law § 240.20 (McKinney 2008). trespassing,21×21. N.Y. Penal Law § 140.10 (McKinney 2010 & Supp. 2015). loitering,22×22. N.Y. Penal Law § 240.35 (McKinney 2008). possession of marijuana,23×23. N.Y. Penal Law § 221.10 (McKinney 2008). or drinking on the street.24×24. N.Y.C. Admin. Code § 10-125[b] (Westlaw through 2014 legislation). This arrestee is supposed to be arraigned by a judge within twenty-four hours,25×25. See People ex rel. Maxian v. Brown, 570 N.E.2d 223 (N.Y. 1991) (per curiam). but the process often takes much longer.26×26. Joseph Goldstein, After Budget Cuts, Defendants’ Wait to See a Judge Often Exceeds 24 Hours, N.Y. Times (July 19, 2011), http://www.nytimes.com/2011/07/20/nyregion/new-york-budget-cuts-lead-to-longer-waits-for-arraignment.html. In the interim, the arrestee spends roughly four to six hours in a precinct holding cell before being transferred to courthouse lockup.27×27. What Can I Expect If I’m Arrested?, Legal Aid Soc’y, http://www.legal-aid.org/en/ineedhelp/ineedhelp/criminalproblem/faq/whatcaniexpectifiamarrested.aspx (last visited Apr. 9, 2015) [http://perma.cc/K4D8-AZDB]. When he finally sees a judge, if he has a record, he is likely to be held on bail that he cannot afford.28×28. Mosi Secret, N.Y.C. Misdemeanor Defendants Lack Bail Money, N.Y. Times (Dec. 2, 2010), http://www.nytimes.com/2010/12/03/nyregion/03bail.html. But even if he is released on his own recognizance, which, for defendants with a criminal record, is unlikely,29×29. See Human Rights Watch, The Price of Freedom: Bail and Pretrial Detention of Low Income Nonfelony Defendants in New York City 28 n.75, 29 tbl.3 (2010), http://www.hrw.org/sites/default/files/reports/us1210webwcover_0.pdf [http://perma.cc/NE8R-2G58] (noting that those with a misdemeanor conviction comprised 52.6% of those arrested on misdemeanors in New York City, but only 10.0% of those subsequently released on their own recognizance). the hassle of a trial — with its many courthouse trips, where there might be long lines at secured entrances30×30. See Kohler-Hausmann, supra note 10, at 386–87; William Glaberson, Faltering Courts, Mired in Delays, N.Y. Times (Apr. 13, 2013), http://www.nytimes.com/2013/04/14/nyregion/justice-denied-bronx-court-system-mired-in-delays.html. — starts to look unmanageable. He is offered a plea deal in which the twenty-four hours he just spent in lockup will in effect serve as his sentence. If he does not take it, he will either remain in jail until his trial — which could be a rather long time31×31. Glaberson, supra note 30; see also Human Rights Watch, supra note 29, at 2 (reporting that the average length of pretrial detention for someone who cannot make bail is 15.7 days). — or be forced to attend a series of time-consuming and meaningless court appearances.32×32. See Kohler-Hausmann, supra note 10, at 375 (“[Defendants] must then sit patiently in a crowded courtroom, sometimes all day, watching the seemingly inscrutable logic of other cases being called and courtroom lulls, waiting for their 60–120 seconds in front of the judge. When the lunch break is called at 1 p.m., the crowd of defendants who have been waiting since 9 a.m. for their case to be called invariably express what could be understated as discontent. . . . If defendants fail to return for their case call after lunch a warrant will [likely] be issued.”). If the defendant works full-time, these court appearances are nearly impossible for him to attend. And so at arraignment he does not contest whatever low-level offense is available and goes home.33×33. See Issa Kohler-Hausmann, Managerial Justice and Mass Misdemeanors, 66 Stan. L. Rev. 611 (2014). In New York City, 78.2% of all misdemeanor arrests result in either a conviction for a noncriminal violation (28.7%), a conviction for a misdemeanor (19.6%), or an adjournment in contemplation of dismissal (ACD) (29.9%), where the charge stays on a defendant’s record for a year and is reactivated if the defendant is rearrested. Id. at 647 fig.10. In this Essay, we occasionally refer to “guilty pleas” so as to include the ACD. This is because only a straight dismissal gets you out of court without any record that can come back to haunt you, so agreeing to an ACD is tantamount to pleading guilty for our purposes. Statutory law has no role in this type of prosecution.

Many have noted the startling lack of process in misdemeanor and violation prosecutions generally, as well as the extent to which those prosecutions are driven by process costs.34×34. E.g., Bowers, supra note 10, at 86; Samuel R. Gross, Convicting the Innocent, 4 Ann. Rev. L. & Soc. Sci. 173, 180 (2008); Samuel R. Gross, Pretrial Incentives, Post-Conviction Review, and Sorting Criminal Prosecutions by Guilt or Innocence, 56 N.Y.L. Sch. L. Rev. 1009, 1014 n.15 (2011/12) (discussing “innocent defendants who plead guilty to avoid the process costs of a criminal prosecution, in particular those who have been held long enough in pretrial detention that they will get to go home if they accept the prosecutor’s offer to plead guilty in return for a sentence of imprisonment that they have already served”); Kohler-Hausmann, supra note 33, at 670; Natapoff, supra note 4, at 1328–29. The picture is bleak in New York, to be sure — but at least in New York defendants plead guilty one at a time. In some jurisdictions, defendants are read their rights and enter their guilty pleas en masse.35×35. Natapoff, supra note 4, at 1329 (citing Feeley, supra note 9, at 10). Guilty pleas are a near certainty.36×36. Id. at 1336. Adjudication, in the sense of determining, say, the factual basis of guilt, is absent.37×37. Id. at 1317. This world of low-level criminal processing does not remotely approach the criminal process taught in law school classrooms. At least one scholar suggests that the misdemeanor system in New York is no longer principally concerned with adjudication at all — rather, she claims, its goal is to mark defendants with records so that they can be effectively sorted in future encounters with the system.38×38. See Kohler-Hausmann, supra note 33, at 614.

As a matter of legal doctrine, New York’s disorderly conduct offense is limited in scope and difficult to prove.39×39. See, e.g., People v. Jones, 878 N.E.2d 1016, 1018 (N.Y. 2007) (“The conduct sought to be deterred under the statute is ‘considerably more serious than the apparently innocent’ conduct of defendant here.” (quoting People v. Carcel, 144 N.E.2d 81, 84 (N.Y. 1957))); People v. Richardson, 913 N.Y.S.2d 549, 554 (Crim. Ct. 2010) (dismissing complaint for failure to allege mens rea of “intent to cause public inconvenience, annoyance or alarm,” id. at 552 (quoting N.Y. Penal Law § 240.20 (McKinney 2008)), where a police officer “observed the defendant shouting obscene language to wit: ‘f**k off n[**]ga, stop f**king with me’ in a public area,” id. at 551); People v. Stephen, 581 N.Y.S.2d. 981, 982 (Crim. Ct. 1992) (holding that defendant had not engaged in “violent, tumultuous or threatening behavior” within the meaning of Section 240.20 when he screamed at a police officer “Fuck you . . . If you were in jail, I’d fuck you, you’d be my bitch . . . If you didn’t have that gun and badge, I’d kick your ass, I’d kill you,” id. at 982 (internal quotation marks omitted), and where “a crowd of approximately 15–20 people gathered who joined the defendant yelling, ‘Yeah, fuck the police,’” id. at 982–83). Same with open container violations.40×40. See, e.g., People v. Figueroa, 948 N.Y.S.2d 539, 541 (Crim. Ct. 2012) (dismissing open container violation because “[w]hile the arresting officer’s professional training and sense of smell may be sufficient to support his conclusion that defendant was drinking beer, such does not support the conclusion that the beer contained more than one-half of one percent (.005) of alcohol by volume because the beverage could have very well been non-alcoholic beer”). In the tiny minority of cases that do receive actual judicial scrutiny, the New York Court of Appeals has espoused a common law of disorderly conduct violations that sharply circumscribes the extent to which police can use these laws to intrude on individual liberties. But these laws routinely underlie convictions of defendants who did not violate, and could not have violated, them.41×41. See Bowers, supra note 10, at 85–86; Kohler-Hausmann, supra note 33, at 650 (describing disorderly conduct as “an all-purpose generic charge” that does not indicate “that the defendant is guilty of any specific illegal conduct”).

There are particularly stark examples of this phenomenon: people often plead guilty to crimes that, by virtue of either repeal or unconstitutionality, the police can no longer legally enforce.42×42. See Bowers, supra note 10, at 85–86. In 1993, the Second Circuit struck down New York’s loitering statute because it violated the First Amendment on its face and enjoined the City from prosecuting charges under the statute.43×43. See Loper v. N.Y.C. Police Dep’t, 999 F.2d 699, 701, 705–06 (2d Cir. 1993). Yet between 1992 and 2004, the New York City Police Department (NYPD) arrested 1876 people for violating that very statute.44×44. See Bowers, supra note 10, at 85–86. Eddie Wise, one of those defendants, was convicted of violating the unconstitutional statute seven times after it was declared unconstitutional.45×45. Elva Rodriguez et al., Beggar Gets Change, Wins Suit Forcing City to Lay Off Panhandlers, N.Y. Daily News (June 11, 2005, 12:00 AM), http://www.nydailynews.com/archives/news/beggar-change-wins-suit-forcing-city-lay-panhandlers-article-1.624527 [http://perma.cc/557A-MYP3]. Professor Josh Bowers was Eddie Wise’s attorney, and describes Wise’s story in greater detail in Bowers, supra note 10, at 85. In 2005, local lawyers again sued to enjoin the NYPD from enforcing the statute.46×46. Bowers, supra note 10, at 86; Rodriguez et al., supra note 45. (They won.47×47. Rodriguez et al., supra note 45. ) But the reminder didn’t stop the NYPD from issuing 641 summonses and arresting 58 people for loitering even after the suit was filed.48×48. Kati Cornell Smith, Beggar Buster Blues — Judge Blasts NYPD, N.Y. Post (Nov. 30, 2006, 5:00 AM), http://nypost.com/2006/11/30/beggar-buster-blues-judge-blasts-nypd [http://perma.cc/NFE4-8NEG].

Marijuana arrests present an equally stark example. In New York, possession of marijuana in public view is a misdemeanor.49×49. See N.Y. Penal Law § 221.10 (McKinney 2008). But, since 1977, having marijuana in your pocket is a noncriminal, nonarrestable violation.50×50. 1977 N.Y. Laws 478–83 (codified at N.Y. Penal Law § 221.05 (McKinney 2008)). Between 1996 and 2011, New York City alone made 586,320 arrests for possession of marijuana in public view.51×51. Human Rights Watch, A Red Herring: Marijuana Arrestees Do Not Become Violent Felons 1 (2012), http://www.hrw.org/sites/default/files/reports/us_mj1112webwcover.pdf [http://perma.cc/X99Z-N7C4]. We thank Issa Kohler-Hausmann, a coauthor of this report, for her helpful comments on this point. In most of these arrests, the marijuana “becomes ‘open to public view’ only after the police stop individuals and either ask them to empty their pockets or conduct a frisk.”52×52. Id. at 11 (quoting N.Y. Penal Law § 221.10). Local public defenders, because they were concerned that these arrests “present[ed] clear constitutional and evidentiary problems,”53×53. Marijuana Arrest Project, Bronx Defenders, http://www.bronxdefenders.org/programs/the-marijuana-arrest-project (last visited Apr. 9, 2015) [http://perma.cc/2CFC-AXBR]. began trying to take these cases to trial. They were unable to try a single case.54×54. William Glaberson, In Misdemeanor Cases, Long Waits for Elusive Trials, N.Y. Times, (Apr. 30, 2013), http://www.nytimes.com/2013/05/01/nyregion/justice-denied-for-misdemeanor-cases-trials-are-elusive.html.

How this happens is no mystery. The process costs so outweigh the defendant’s perceived costs of pleading guilty that it seems to make very little sense to contest even patently invalid charges. Almost everyone pleads guilty,55×55. Or, in New York, accepts an ACD that stays on his record for a year. Kohler-Hausmann, supra note 33, at 648–50. even though many did not commit (or could not have committed) the charged offense. This is because successfully fighting the charge is worse for the defendant, at least in the short run, than pleading guilty. The Fourth Amendment, then, imposes no restrictions on police behavior in this realm of criminal punishment, beyond the distant possibility of a § 1983 civil rights suit.56×56. See 42 U.S.C. § 1983 (2012). Because the police can be confident that a trial on these charges is at worst a remote possibility, the exclusion remedy for Fourth Amendment violations is meaningless.

These public order arrests create a cascade of problems for those defendants who are frequently stopped by the police. In well-studied New York, a defendant’s first misdemeanor arrest often results in an adjournment in contemplation of dismissal (ACD), where the charge is dismissed after a year if the defendant stays out of trouble.57×57. See supra note 33. But if the defendant gets rearrested within the next year, the ACD usually results in a worse offer from the prosecution, and often a formal conviction for the offense on which he was rearrested.58×58. Kohler-Hausmann, supra note 33, at 668 (“If a defendant with [an ACD] from a prior arrest is brought back to criminal court on a new arrest, the offer on the new case will go up along one vector or another — the seriousness of the mark, the conditions he must satisfy to be granted the disposition, or the formal sentence.”). And, during the year the ACD is pending, potential employers can see (and make decisions on the basis of) the arrestee’s record.59×59. See id. at 648. Indeed, the purpose of New York’s misdemeanor system may be to mark defendants so that they can be treated differently when they are subsequently arrested. See id. at 649.

But while prosecutors may aim for criminal convictions, the police have much less reason to care about dispositions for loitering, disorderly conduct, or open-container arrests. At least in theory, in some circumstances, they ought to care only about the arrest and pre-arraignment detention.60×60. See generally Feeley, supra note 9. With rare exceptions, once the very low-level defendant is arrested, the police have accomplished their immediate goal of maintaining order. Of course, the defendant is also prosecuted, convicted, and permanently marked by the system, but these fallen dominos are hard to pin on the police. The public expects the police to maintain order, but when an arrest is necessary, the law often arms officers — at least officially — with only the powerful and blunt tools of criminal law. This is a destructive mismatch: public order or “quality of life” policing is conducted almost entirely outside the shadow of substantive criminal law and almost entirely within the discretion of the police.61×61. See generally Livingston, supra note 8.

The problem stems from the misalignment of purposes between the police, who primarily (and optimistically) seek to prevent crime and keep streets safe, and district attorneys, who focus more immediately on pursuing chargeable offenses.62×62. See generally Kohler-Hausmann, supra note 33 (arguing that prosecutors’ principal goal in misdemeanor and violation cases — in New York, at least — is to mark defendants for future encounters). Prosecutorial involvement in a case typically begins when someone has already been arrested. At least according to some, prosecutors are interested in minimizing the risk that a defendant emerges from the system without being “marked” so that, in the event the person reoffends, the prosecutor is not blamed.63×63. Id. at 667–68 (citing an interview with a New York public defender). Prosecutors are not well positioned to weed out those public order arrests that should never have led to a criminal conviction. The police, on the other hand, are expected to enforce public order. They likely care less about the escalating penalties of the criminal system than prosecutors do. But when the police make public order arrests, they (perhaps inadvertently) start a process of escalating punishment that is ill suited to the task of order maintenance.64×64. See Kohler-Hausmann, supra note 33, at 668–70.

“Criminal” public order enforcement is counterproductive in other ways. For one, it erodes the label “crime.” When we ask the police to maintain public order, we do not ask them to focus on crime or to arrest criminals as the typical person uses those terms. We ask them to regulate behavior that may inadvertently create some risk to the public; to deter chronic low-level misconduct that doesn’t rise to the level of criminality;65×65. See generally Ellickson, supra note 8. and even to be our primary — and maybe exclusive — agent for dealing with people with substance abuse problems, the mentally ill, and the homeless.66×66. See Herman Goldstein, Policing a Free Society 9 (1977) (“The police have come to be viewed as capable of handling every emergency.”); Peter C. Patch & Bruce A. Arrigo, Police Officer Attitudes and Use of Discretion in Situations Involving the Mentally Ill, 22 Int’l J. L. & Psychiatry 23, 23 (1999). Calling this sort of policing “criminal” makes the term mean less,67×67. See Daniel C. Richman & William J. Stuntz, Essay, Al Capone’s Revenge: An Essay on the Political Economy of Pretextual Prosecution, 105 Colum. L. Rev. 583, 586 (2005) (arguing that prosecutors must consider the signals their prosecutions send to the public). and therefore makes it less powerful, eroding any deterrent or expressive value of a criminal sanction.68×68. See Glanville Williams, The Aims of the Law of Tort, 4 Current Legal Probs. 137, 155 (1951) (“To stigmatise the ordinary person by the epithets ‘criminal,’ ‘offender,’ and ‘conviction,’ is itself a punishment, and, from a deterrent point of view, it is important that the emotion invoked by these words should be kept at full strength and not weakened by their indiscriminate application.”); see also Stuntz, supra note 11, at 2550 (identifying “a basic irony about criminal law: the more it expands, the less it matters”); William J. Stuntz, The Uneasy Relationship Between Criminal Procedure and Criminal Justice, 107 Yale L.J. 1, 7 (1997). Worse still, this approach brands as “criminals” many who have merely offended other people’s sensibilities or who have engaged in what almost everyone agrees is very minor misconduct that in reality very rarely poses a risk to physical safety.69×69. Cf. Emile Durkheim, The Division of Labor In Society (W.D. Halls trans., Free Press 1984) (1893) (defining “criminal” acts as those that shock the collective conscience of society).

Because defendants cannot (realistically) contest the charges against them, policing outside the substantive law also leaves no account of what happened — or why.70×70. Indeed, New York City has refused to disclose (and may not even have kept track of) how many convictions were generated in non-felony cases between 2002 to 2010. Ray Rivera & Al Baker, Data Elusive on Low-Level Crime in New York City, N.Y. Times (Nov. 1, 2010), http://www.nytimes.com/2010/11/02/nyregion/02secrecy.html. The sentence imposed is in effect subverted by the process, which ought to be administrative and incident to punishment, not the punishment itself. Cases are often resolved at arraignment,71×71. Kohler-Hausmann, supra note 33, at 654 (“In New York City over 57% of all misdemeanor and violation cases reach a disposition at arraignment.”). and very rarely at trial,72×72. See id. at 650 (noting that fewer than 0.5% of misdemeanor cases go to trial). so there is no record of why the system punished someone. All we’ll ever know is that someone was convicted of “disorderly conduct.”73×73. Consider Michigan law, which criminalizes being a “disorderly person” but “provides no standards as to what is a public disturbance.” People v. Gagnon, 341 N.W.2d 867, 869 (Mich. Ct. App. 1983) (per curiam). Those who read the record might think the worst.74×74. See Kohler-Hausmann, supra note 33, at 650 (“In practice, a ‘dis con’ serves as an all-purpose generic charge to mark the defendant for a specific length of time, not to indicate that the defendant is guilty of any specific illegal conduct.”); cf. Old Chief v. United States, 519 U.S. 172, 189 (1997) (expressing concern that when a prior conviction is an element of an offense, withholding the facts of the earlier case “may be like saying, ‘never mind what’s behind the door,’ and jurors may well wonder what they are being kept from knowing”). A criminal record is chief among the unintended and unnecessary costs generated by relying on criminal law to maintain public order. A person arrested for an essentially noncriminal public order offense becomes part of the criminal system alongside those guilty of genuinely transgressive conduct and about whom society would agree on assigning the label “criminal.” Because of the wide variety of conduct covered by public order prosecutions, employers are unlikely to bother drawing distinctions.

Is there a justification for uniformly marking arrestees with criminal convictions in the context of public order offenses? Certain classes of low-level offenses are apparently poor predictors of serious criminality in the future.75×75. See Kohler-Hausmann, supra note 33, at 674–76 (describing the results of an empirical study of New York City arrest data). In some jurisdictions, the probability of being convicted of a more serious low-level offense, as opposed to a less serious one, is chiefly a function of how long it has been since the individual’s last arrest.76×76. See id. at 690. Because people in highly policed areas are arrested at much higher rates, the combination of these facts likely produces a cascade of arrests and convictions that have little relationship to the goals of public order policing (and much more to do with a person’s neighborhood and race).

Finally, there is simply the matter of how much all of this costs. Public order arrests often result in a lengthy period of pre-arraignment detention77×77. See Goldstein, supra note 26 (describing pre-arraignment detentions lasting up to three days and an average detention length of thirty-one hours). — perhaps well in excess of the sentence any institutional actor would rationally want to impose for the “violation.” The defendant is processed by the court system’s personnel and in its buildings and is provided a court-appointed lawyer. All of this jailing and processing translates into a nontrivial amount of money and, in any event, imposes needless suffering.

The debate about police discretion cannot move forward as long as the police are compelled to use extralegal means to police public order by imposing criminal punishment. Using the process in this way interferes with other institutional actors’ ability to limit police discretion: as it stands today, most discretionary arrests result in a conviction with serious consequences. In these arrests, everything seems to have gone right, so the public — including much of the legal academy — continues to think that the text of the substantive law can meaningfully constrain police behavior.78×78. See, e.g., Livingston, supra note 8, at 561 (“Courts cannot ‘solve’ the problem of police discretion by invalidating reasonably specific public order laws — as some have attempted — without seriously impairing legitimate community efforts to enhance the quality of neighborhood life.”).

II. The False Dichotomy of Police Discretion

There has long been a vigorous debate over how much discretion to give the police in initiating street encounters and making low-level arrests.79×79. Compare, e.g., Dan M. Kahan & Tracey L. Meares, Foreword: The Coming Crisis of Criminal Procedure, 86 Geo. L.J. 1153 (1998), Livingston, supra note 8, and Meares & Kahan, supra note 14, with Alschuler & Schulhofer, supra note 14, Cole, supra note 15, and Natapoff, supra note 4. This debate is alive today in the fight over New York’s controversial stop-and-frisk policy and its practice of arresting people for marijuana possession.80×80. E.g., Joseph Goldstein, Marijuana May Mean Ticket, Not Arrest, in New York City, N.Y. Times (Nov. 9, 2014), http://www.nytimes.com/2014/11/10/nyregion/in-shift-police-dept-to-stop-low-level-marijuana-arrests-officials-say.html; Editorial, The Truth Behind Stop-and-Frisk, N.Y. Times (Sept. 2, 2011), http://www.nytimes.com/2011/09/03/opinion/the-truth-behind-stop-and-frisk.html. Some scholars claim that the density of urban spaces requires new forms of police discretion to maintain “social norms” and to smooth community tensions.81×81. See, e.g., Kahan & Meares, supra note 79, at 1163–64. They argue that the increasing empowerment of black communities means that the Constitution should leave them alone to “protect themselves through the political process.”82×82. Cole, supra note 15, at 1061 (citing Kahan & Meares, supra note 79). Courts should no longer be suspicious that public order laws are designed to keep black people out of community life because, the argument goes, black communities increasingly write those laws themselves.83×83. Id.

The early incarnation of this debate centered on City of Chicago v. Morales,84×84. 527 U.S. 41 (1999). which involved a broad antigang loitering ordinance that allegedly gave police the power to arrest (or harass) whomever they wanted. The ordinance essentially criminalized “remain[ing] in one place with no apparent purpose.”85×85. Id. at 97 (Scalia, J., dissenting) (alteration in original) (internal quotation marks omitted). Morales is the most recent in a line of cases in which the Supreme Court invalidated local quality-of-life ordinances on vagueness grounds. See Kolender v. Lawson, 461 U.S. 352 (1983); Papachristou v. City of Jacksonville, 405 U.S. 156 (1972); Coates v. City of Cincinnati, 402 U.S. 611 (1971). Some maintained that this language was fatally overbroad and gave the police inordinate discretion to arrest people for innocent conduct86×86. Alschuler & Schulhofer, supra note 14, at 225–37. — the Court agreed — while others argued that this broad language was necessary for the police to do their jobs and maintain order for the benefit of minority communities.87×87. Meares & Kahan, supra note 14, at 209–14. This debate implicitly assumes, however, that you can have either specific criminal laws that constrain the police, or very general laws that allow the police broad discretion.88×88. See Dorothy E. Roberts, Foreword: Race, Vagueness, and the Social Meaning of Order-Maintenance Policing, 89 J. Crim. L. & Criminology 775, 777–78 (1999) (“For the last several decades, conservative commentators have called for a relaxation of the vagueness doctrine as well as procedural restraints on police discretion to permit bolder law enforcement efforts to investigate, punish, and prevent crime.”). This is a false dichotomy. What the law says — the specific conduct it defines and criminalizes — does little to constrain police discretion in the enforcement of very low-level violations. The need to control police discretion could hardly be more important, but in the context of public order offenses, it has at best a very weak connection with substantive criminal law.89×89. Contra, e.g., Livingston, supra note 8, at 561 (“Courts cannot ‘solve’ the problem of police discretion by invalidating reasonably specific public order laws — as some have attempted — without seriously impairing legitimate community efforts to enhance the quality of neighborhood life.”).

Regardless of how offenses are defined, the police can still use them to generate convictions by using the process to force guilty pleas. Therefore, by focusing primarily on the content of substantive law, policymakers pay too little attention to the real agent of criminal punishment in this setting: the process costs of a criminal arrest.

To illustrate this disconnect, consider an example. Imagine that drinking on the street — a very specific activity — were no longer prohibited in a particular city. Someone is then arrested for drinking on the street in this city even though drinking on the street no longer violates any criminal or administrative rule. If he is offered a plea bargain at his arraignment, he will probably take it, as the previous Part shows. Thus, if the police encounter someone drinking on the street in a manner that they find disruptive or objectionable, they can (and also clearly know they can) still arrest that person despite the fact that drinking on the street is no longer against the rules.

The disconnect stems from a fact that has been true of lower courts since at least the 1950s: the process is the punishment.90×90. Feeley, supra note 9. Beginning with the “due process revolution,”91×91. E.g., Fred P. Graham, The Due Process Revolution: The Warren Court’s Impact on Criminal Law (1970). when the Bill of Rights was incorporated against the states, lower courts have used the process of adjudication to enforce substantive norms of behavior. Because they are no longer able — at least formally — to enforce order without fairly extensive process, the criminal system evolved to use the costs that the process generates to enforce order.92×92. E.g., Kohler-Hausmann, supra note 10, at 374–81.

But the political system and much of the legal academy continues to believe that code reforms can serve to control police discretion. Indeed, in response to the criticism that marijuana arrests do almost nothing to protect public safety,93×93. See Human Rights Watch, supra note 51, at 19. “New York Governor Andrew Cuomo introduced legislation to make possession of marijuana in public view a non-arrestable, non-criminal violation,” just like possession in your pocket.94×94. Id. at 4. Despite support from all five New York City district attorneys,95×95. Each borough of New York City has its own district attorney. the legislation failed because of opposition from upstate and suburban legislators.96×96. Human Rights Watch, supra note 51, at 5; Thomas Kaplan & John Eligon, Divide in Albany Kills Proposal on Marijuana, N.Y. Times (June 19, 2012), http://www.nytimes.com/2012/06/20/nyregion/cuomo-bill-on-marijuana-doomed-by-republican-opposition.html. Both sides of this debate neglected the reality that code reforms cannot alone control police discretion. In other words, even if the legislation passed, if the NYPD wanted to continue arresting people for marijuana possession in public view — despite the fact that it would have become a nonarrestable offense under the law — there would be nothing to stop them.

Other norms and institutions are much better suited to constrain police discretion. Indeed, the political process that led to the passage of Chicago’s gang-loitering ordinance may have strongly influenced police behavior in favor of aggressive enforcement and vigorous public order policing.97×97. Professors Tracey Meares and Dan Kahan argue rancorously with Professors Albert Alschuler and Stephen Schulhofer about how the political process in Chicago ended up generating the anti-gang loitering ordinance. Compare Meares & Kahan, supra note 14, at 199–200 (claiming that black communities on the South and West Sides of Chicago birthed the ordinance), and Tracey L. Meares & Dan M. Kahan, Black, White, and Gray: a Reply to Alschuler and Schulhofer, 1998 U. Chi. Legal F. 245, 247–51 (same), with Alschuler & Schulhofer, supra note 14, at 217–20 (arguing that aldermen from predominantly white wards were the real movers behind the ordinance). Both sides do agree that the process was loud, open, and unusually prominent in the eyes of citizens and police. Civilian oversight can constrain police discretion.98×98. See generally Merrick Bobb, Civilian Oversight of the Police in the United States, 22 St. Louis U. Pub. L. Rev. 151 (2003). So can consent decrees with the Justice Department.99×99. See, e.g., Human Rights Watch, Shielded from Justice: Police Brutality and Accountability in the United States 140–44 (1998), http://www.hrw.org/reports/pdfs/u/us/uspol986.pdf [http://perma.cc/GTA6-AUCP] (discussing consent decrees pursuant to 42 U.S.C. § 14141); Debra Livingston, Police Reform and the Department of Justice: An Essay on Accountability, 2 Buff. Crim. L. Rev. 815, 844–47 (1999). Perhaps most importantly, law enforcement departmental norms can restrain discretion.100×100. Cf., e.g., Neal Kumar Katyal, Internal Separation of Powers: Checking Today’s Most Dangerous Branch from Within, 115 Yale L.J. 2314 (2006). But in the context of minor crimes — the lowest level of criminal punishment — police discretion appears to be largely immune to substantive criminal law.

Fortunately, there is reason to be hopeful about the possibility for reform in our cities. After the legislation in New York to decriminalize marijuana possession in open view failed, the political movement behind those substantive reforms continued to apply pressure to political actors to change practice, if not the law. Turning their focus away from the criminal code, opponents of marijuana arrests were able to persuade New York City simply to stop making them.101×101. Goldstein, supra note 80.

It is important to recognize the limits of our claim. In prosecutions for serious crimes, the substantive scope of criminal conduct really does matter. As the ratio of the expected sentence to the threatened process costs grows, plea bargaining outcomes increasingly mirror trial outcomes.102×102. See, e.g., Stuntz, supra note 11, at 2550–58. It thus matters whether drugs are illegal. It matters very little, however, whether Chicago criminalizes loitering with “no apparent purpose” or “causing a disturbance.” Similarly, our claim is limited to relatively large jurisdictions, where the process costs are high. Smaller jurisdictions may function quite differently.

Finally, this entire discussion is not to say that the text of the criminal code does nothing, even in the context of low-level crimes. Criminal prohibitions send important signals to the public and to the police about the scope of proper conduct.103×103. For a tiny sample of the vast literature on this subject, see Adil Ahmad Haque, Lawrence v. Texas and the Limits of the Criminal Law, 42 Harv. C.R.-C.L. L. Rev. 1, 33–34 & n.145 (2007) (citing Dan M. Kahan, The Anatomy of Disgust in Criminal Law, 96 Mich. L. Rev. 1621, 1623 (1998) (book review)). They can have a tremendously important expressive value, outlining for the citizenry conduct that is to be encouraged and conduct that ought to be forbidden. They can send important messages to the police about the proper scope of their ability to intrude on individual liberty. Our point here is only that, whatever they do, criminal codes do not meaningfully constrain police discretion in the context of public order offenses.

III. Reducing Process Costs and Collateral Consequences

High process costs of low-level criminal adjudication are the problem. The police can — at their discretion and unconstrained by substantive criminal law in any meaningful way — impose draconian, but often unnecessary, even counterproductive, costs on defendants and their families. The police do not necessarily do this out of spite or incompetence. They simply need tools to police public order (often by making arrests), and criminal law is usually all that they have.

We offer one potential solution: in order to reduce the harm of discretionary, low-level arrests — by limiting the process costs and collateral consequences they can generate — we propose a stripped-down, civil form of arrest, the consequences of which include only the arrest itself and a brief period of noncriminal detention.104×104. We discuss briefly below why we believe this proposal is unlikely to significantly increase the volume of total arrests. Counterintuitively, we believe that the availability of such a tool will cause police to use their power to arrest someone for a crime less often, not more. An important historical analog to this approach is the “drunk tank,” in which officers would lock up dangerously inebriated people to sober up overnight.105×105. See, e.g., Livingston, supra note 8, at 640; see also Joshua Partlow, Holiday Rush at Mexico City’s Hangover Prison, Wash. Post (Dec. 26, 2013), http://www.washingtonpost.com/world/the_americas/holiday-rush-at-mexico-citys-hangover-prison/2013/12/26/4edfcf10-6dcc-11e3-a5d0-6f31cd74f760_story.html [http://perma.cc/2BL7-RAZ4]. On the subject of history, it is also worth noting that the earliest Western police forces were permitted — indeed, required — to effect noncriminal, low-process arrests. Early common law arrest doctrines recognized a distinction between the authority of the police in matters of crime and the authority of police in matters of order. For example, the Statute of Winchester, which established London’s first police force in 1285, provided that watchmen were authorized and charged “‘as . . . in Times [passed]’ to ‘watch the Town continually all Night, from the Sun-setting unto the Sun-rising’ and were directed that ‘if any Stranger do pass by them, he shall be arrested until Morning.’” Atwater v. City of Lago Vista, 532 U.S. 318, 333 (2001) (quoting Statute of Winchester, 1285, 13 Edw. 1, stat. 2, c. 4 (Eng.)); see also Thomas Y. Davies, Correcting Search-and-Seizure History: Now-Forgotten Common-Law Warrantless Arrest Standards and the Original Understanding of “Due Process of Law,” 77 Miss. L.J. 1, 58 (2007) (“‘A watchman may arrest a night walker by a warrant in law.’ . . . In effect, being out after dark in town was so suspicious that it was grounds for a temporary arrest . . . .” (quoting 2 Edward Coke, The Institutes of the Lawes of England 52 (1642))). We thank Josh Bowers for his helpful comment on this point. No formal criminal process need be involved and no criminal record would result. Although the debate about police discretion has centered on the substantive scope of low-level regulation, we focus on the real-world process of arresting people for low-level offenses and seek a way to avoid criminal records and disproportionate, socially wasteful costs. If the police are going to enforce public order through discretionary arrests, society would benefit from providing law enforcement with the legal instruments to do so safely, effectively, and legitimately.

To be more specific, we propose complementing (or, alternatively, replacing) public order crimes with a class of civil ordinances that allow only very brief detentions. First, these ordinances should strictly limit the total time of detention imposed — including the sentence and the period that anyone can be detained on suspicion of a violation — to twenty-four hours at the very longest.106×106. New York currently has serious problems complying with a twenty-four-hour deadline for arraignments. See Goldstein, supra note 26. That said, this deadline should be much easier to comply with, though compliance is by no means a certainty. Ideally, the limit would be even shorter. The ordinances should not allow for the imposition of fines or monetary payments of any kind. Second, the ordinances should permit an arrestee to attack the legitimacy of his detention ex post via mail or telephone and to waive in-person arraignment or any other appearance requirement. Lastly, these ordinances should be noncriminal and should not, under any circumstances, leave the defendant with a recorded violation of any kind. The police should be required to retain reliable records of whom they arrested and why, but those records should be accessible to the public only in a reliably anonymous form.107×107. Cf. Jeffrey M. Skopek, Anonymity, the Production of Goods, and Institutional Design, 82 Fordham L. Rev. 1751 (2014) (discussing various aspects of anonymity in the modern legal structure).

These features serve to reduce the unnecessary harm that low-level criminal arrests ultimately impose on arrestees. No longer will there be any reason for you to plead guilty to time served or to accept an ACD-like outcome — you will already have served the maximum penalty you can receive. No longer will there be any reason to plead guilty to avoid trial. You can conduct a paper adjudication if you want to vindicate your version of events, or just let it go — either way, you will not wind up with a criminal record. Lastly, no matter what happens, you’re back home in twenty-four hours or less.108×108. To again use New York City as an example, these detentions should be at the local station house, rather than in the currently overcrowded jails. Even short periods of time in overcrowded jails can be traumatizing and degrading. Station-house lockups — where police are generally present nearby and periods of detention are very brief — should serve to minimize the cruelty of detention.

Such laws would still allow the police to do all the things public order policing enthusiasts expect them to do. Rarely, if ever, do the police need more than a twenty-four-hour detention to accomplish the goals of public order policing: disrupting and isolating, primarily. If the police believe that more than twenty-four hours of detention is appropriate, then other policing goals are in play, and the police should typically arrest for a more serious crime for which the defendant should be charged and tried. In this situation, the criminal system becomes appropriate, and the plea bargaining process functions better because the sentence the defendant would face upon conviction often exceeds the process costs of fighting the charge.109×109. See Stuntz, supra note 11, at 2563–64. Our goal is simply to provide the police with tools that allow them, in appropriate situations, to avoid high process costs and unnecessary collateral consequences while maintaining public order.

When police arrest people for low-level crimes, they seek a wide variety of ends, depending on the context. Sometimes police want to clear a corner where drug dealers are congregating.110×110. See Natapoff, supra note 4, at 1332. Sometimes they want to send a signal to a neighborhood that they are in control.111×111. See id. at 1333. Often, police are maintaining a sense of order in the community, even manifesting that order through the regulation of physical spaces.112×112. See, e.g., William J. Bratton, The New York City Police Department’s Civil Enforcement of Quality-of-Life Crimes, 3 J.L. & Pol’y 447 (1995) (NYPD Commissioner discussing the city’s efforts at policing quality-of-life crimes); Robert C. Ellickson, supra note 8; Dan M. Kahan, Reciprocity, Collective Action, and Community Policing, 90 Calif. L. Rev. 1513 (2002). They are almost always our front-line responders to mental illness and substance abuse; thus police arrest people to keep them safe or ensure that they receive care.113×113. See Goldstein, supra note 66, at 9. Sometimes the police have illegitimate reasons.114×114. Nicola Lacey, Humanizing the Criminal Justice Machine: Re-animated Justice or Frankenstein’s Monster?, 126 Harv. L. Rev. 1299, 1299–1300 nn.1–7 (2013) (reviewing Stephanos Bibas, The Machinery of Criminal Justice (2012)) (collecting books and articles on the subject of policing and punishment in the United States). But no matter why the police make public order arrests, they should rarely be invested in whether the person they’ve arrested is ultimately convicted. The ordinances we suggest leave the police equally effective at maintaining order, but eliminate entanglement in the criminal process as a near-certain result. And, similarly, by lowering process costs, the proposed ordinances can bring public order policing aboveboard, allowing the debate about how much discretion the police should have to occur on more productive terrain.

We recognize that police departments may want to use low-level arrests to incapacitate or simply to keep track of people they worry may pose a threat. With respect to the former possibility, police may use low-level arrests to keep potentially violent criminals off the street for more than a few days,115×115. See, e.g., Human Rights Watch, supra note 51, at 16 n.47 (“[W]e need[] to be more selective about who [sic] we [are] arresting on quality-of-life infractions. When a team of cops fills up a van with arrestees, the booking process can take those cops out of service for a whole day in some cities. The public can’t afford to lose that much police protection for a bunch of first-time offenders, so the units enforcing quality-of-life laws [must] be sent where the maps show concentrations of crime or criminals, and the rules governing the stops have to be designed to catch the sharks, not the dolphins.” (first alteration in original) (quoting Jack Maple, The Crime Fighter: Putting the Bad Guys Out of Business 155–56 (1999))). although the evidence suggests that this approach is unlikely to be effective.116×116. See id. at 16–18 (disputing public safety benefits of marijuana arrests). For better or worse, our proposal does not eliminate the police’s ability make such arrests, although it may frustrate certain plausible law enforcement aims. The police may be concerned, for instance, about the long-term trajectory of chronic low-level violators. By arresting people for, say, drinking on the street, police can keep track of how many times a person has been caught drinking in public and can escalate his punishment accordingly. By allowing an individual police officer to use an unrecorded, noncriminal arrest, our proposal may interfere with the ability of the police to achieve this goal.

In theory, a system of criminal misdemeanors may serve many purposes: it may seek to punish, to deter, and to mark. It may even serve to incapacitate. But the current system achieves these purposes at significant expense. From our perspective, the issue in the public order policing domain is the disparity between the purposes of the police in some circumstances (short-term incapacitation) and the costs of the formal criminal misdemeanor system. We can accomplish short-term incapacitation in a much more humane and less costly way. Public order policing issues are in large part noncriminal, and diverting low-level violators from the criminal system will provide a fairer and lower-cost alternative to the current practice.

This paper does not take a position on the appropriate amount of discretion to give the police in maintaining public order. Nonetheless, assume for the moment that whatever the optimal level of discretion happens to be, police in many big cities currently have too much; and assume that, as a separate matter, police arrest people too often. You might think that our proposal will make both of these problems worse, not better. In response to the first concern (too much discretion), the current system appears to constrain police minimally in this area of criminal law — if it constrains them at all — so our proposal will not free the police much more than the status quo already does. We have a similar response to those who are concerned that our proposal would weaken defendants’ ability to fight the underlying merits of their claims: they have very little ability to do so at present, so, at worst, our proposal is neutral.

With respect to the concern that our proposal will lead to more arrests, though, we are more cautious. Perhaps the hassle of a formal criminal arrest under the current system provides some disincentive to the police. If our proposal makes it faster and easier for the police to arrest people, the argument goes, they will do it more. But there is no good reason to believe that an arrest leading to civil detention under our proposal is (or has to be) any less difficult than a formal criminal arrest is today for the police. For prosecutors, defense lawyers, court personnel, and judges, our system eliminates a tremendous amount of work. The police, on the other hand, still have to arrest someone, lock him up, and fill out paperwork explaining why.

If these civil arrests are no easier on police, however, one might next wonder: why would the police even bother with these new tools when they can arrest someone for basically any reason without them? To this we have two responses. First, we emphasize that, because police discretion at present is hardly constrained at all, our proposal cannot make the situation worse. Even if only a few police officers use the new tools because they recognize the unnecessary costs and consequences that an arrest for a low-level crime can generate for the offender or his family, a few is better than none. Second, we hope that policymakers will give police officers incentives to use these new tools in appropriate circumstances. We note that any policy or practice that makes a criminal arrest more costly to police in absolute terms (e.g., requiring an additional explanation for why the police officer preferred a criminal arrest to a civil arrest) would in theory induce police to use the civil tools without increasing the total number of arrests. But, more generally, what incentives policymakers, the press, the bar, or police departments might employ, and how they would be implemented, is a matter for another essay.

Our proposal might also prompt someone to ask: Why do police need to arrest people at all? If we are concerned about the current system of meaningless pleas and useless process, why not scrap it altogether? Police, indeed, use a wide array of nonarrest techniques to calm situations and ease tensions — why must they arrest?

In reply, we simply point to the fact that police under the current system arrest people for very low-level crimes all the time.117×117. See, e.g., Kohler-Hausmann, supra note 33, at 645. If we simply removed prohibitions on public order offenses from the statute books, there is actually no solid reason to believe people would not continue to be arrested for violating them anyway — or that police would not arrest people for a more serious crime, perhaps exacerbating the current situation.118×118. See supra notes 42–48 and accompanying text. The importance of the debate about how much discretion to afford the police — and about how much public conduct to prohibit — cannot be understated, but we do not believe significant progress can be made simply by reforming the criminal code. Police almost certainly arrest people too often, but this reality is not driven by the substantive content of criminal prohibitions. If our proposal is adopted and a less destructive form of arrest becomes established, we can turn to other, more productive, means to advance the debate over police discretion to arrest.119×119. See supra notes 97–100 and accompanying text.

* Charlie Gerstein is a Law Clerk to the Honorable J. Paul Oetken, United States District Court for the Southern District of New York, and a Research Fellow at the University of Southern California Gould School of Law. J.J. Prescott is a Professor at the University of Michigan Law School. We would like to thank Josh Bowers, Seth Bowers, Sam Erman, Alec Ewald, Colleen Fitzharris, Nick Frayn, Kate Gilbert, Sam Gross, Jason Harrow, Issa Kohler-Hausmann, Justin Marceau, Joshua Matz, Eve Brensike Primus, John Rappaport, Laurent Sacharoff, Alex Sarch, and Sonja Starr for their helpful comments on early drafts. We are also grateful to participants at the Robina Institute’s annual conference on the Future of the Criminal Law.