WALTON v. MCCARTHY

• Court: District Court, E.D. Pennsylvania
• Decided: September 29, 2025
• Docket: 2:24-cv-02049
• Status: Unknown

Opinion

THE UNITED STATES DISTRICT COURT FOR THE EASTERN DISTRICT OF PENNSYLVANIA

TERRELL WALTON, CIVIL ACTION v. No. 24-2049 JOSEPH McCARTHY, et al.

Henry, J. September 29, 2025 MEMORANDUM Terrell Walton, a Black Philadelphian, was pulled over for a traffic violation and charged with driving under the influence (DUI). After suppression was granted in municipal court, all charges were withdrawn. Walton then filed this suit claiming, among other things, that the traffic stop, arrest, and prosecution were racially discriminatory in violation of the Equal Protection Clause. Although he has limited direct evidence from his own experience to indicate the necessary elements of such a claim, Walton relies on public statistics to argue that the police department can be proven to discriminate racially. In particular, he produces figures from a report put out by the District Attorney showing discrepancies in the rates at which Black drivers are stopped and charged. In response, the arresting officer moves to dismiss, arguing that the complaint fails to allege “a single fact that supports [the] equal protection claim.” The question before me is, at core, whether Walton’s statistical account, in the context of the complaint, is enough to make out a fac- tual basis from which the elements of an Equal Protection Clause violation could reasonably be inferred. Given the posture of this case, I am not called to opine, nor can I, on whether Officer McCarthy or the Philadelphia police racially discriminated against Walton or anyone else. The question is whether the complaint alleges facts sufficient to move on to the discovery phase. Be- cause I hold that the few statistics in the complaint do not sufficiently address how police treat non-Black drivers, despite the law requiring that information, I dismiss this count and allow Walton an opportunity to amend it with more information.

Separately, the arresting officer’s supervisor, who arrived during the latter part of the stop and arrest, moves to partially dismiss the charge against him on the basis of immunity. I. BACKGROUND At this level, I accept the factual content of the complaint as true.1 See infra § II. At about 10:00 pm on March 1, 2023, Terrell Walton was driving his car in the Cobbs Creek section of Philadelphia when he was pulled over by Officer Joseph McCarthy of the Philadelphia Police De-

partment. Officer McCarthy approached Walton’s vehicle and asked for Walton’s paperwork. Wal- ton asked why he had been pulled over, and Officer McCarthy said Walton had failed to use his turn signal twice. Walton denied as much. The complaint is silent as to whether Walton had in fact failed to use his signals. Walton was unable to produce a paper vehicle registration document, despite looking around the car, but he said it was only a matter of having misplaced it. He suggested that Officer McCarthy simply run his information through police databases to show authoritatively that the car was properly registered to him. Officer McCarthy responded, “That’s not how this works” and asked Walton to turn off the car. The officer again asked Walton to produce the physical registration card, and Walton again suggested that he instead use the police database to the same effect. Walton

began looking for the registration anew. He also called his girlfriend Chassidy King on

1 When I refer to “the complaint,” I mean the operative complaint. In this case, that is the second amended complaint (ECF 13). speakerphone so that Officer McCarthy could hear, asking if she would send the registration. King said the registration ought to be in the car. Walton repeated his suggestion to Officer McCarthy. Officer McCarthy repeated his demand for the physical card. Walton asked to speak with Officer McCarthy’s lieutenant. Officer McCarthy said, “No

problem” and requested the same over the radio. He also asked additional questions about whether Walton had any weapons, had any pertinent medical conditions, or had used marijuana or alcohol that night. Walton denied each of these. The two then returned to the prior registration card dia- logue for “several minutes.” Eventually, Walton said he was getting cold and said he was going to roll up his window. Officer McCarthy forbade this and opened Walton’s door, claiming it was now a matter of safety. The registration card dialogue continued. About ten minutes after the traffic stop began, Sergeant Matthew Stankiewicz arrived on the scene. Sgt. Stankiewicz was a supervisor, evidently the “lieutenant” on offer, and Walton re- counted to him the basics of what had happened thus far. Then, Walton asked Officer McCarthy again to run his information through the database.

At this point, Officer McCarthy asked Walton to step out and walk to the back of the car. Walton did that and turned around at Walton’s request. Officer McCarthy put handcuffs on Walton, explaining to him and Sgt. Stankiewicz that Officer McCarthy suspected Walton of illegally driv- ing under the influence of drugs or alcohol (“DUI”).2 Walton had not shown any signs of impair- ment. Officer McCarthy had not attempted to employ any chemical or behavioral field sobriety testing.

2 Walton was charged at arrest under 75 Pa.C.S. §§ 3802(d)(1) and (d)(2). These related to two different versions of Pennsylvania’s DUI statute relating to driving under the influence of drugs, which may be charged in the alternative. Nevertheless, Walton was arrested on suspicion of DUI. When he was brought to the police Crash Investigation Division (CID) and evaluated, he consented to a blood draw, which did not detect the presence of any intoxicants. The CID officer observed that Walton did not display slurred speech, disorientation, or the smell of alcohol, and he concluded that there was no reason to believe

Mr. Walton was impaired. Officer McCarthy did ultimately run Walton’s information through the database at some point after putting him in handcuffs, which confirmed that the car was properly registered to him. No citation was issued regarding the lack of registration and failures to properly signal. Walton was in a holding cell until March 3, 2023, when he was released on bail.3 At a hearing on a motion to suppress on September 7, 2023, a Judge of the Court of Common Pleas found that no evidence supported the allegation that Walton was impaired and therefore that there was no probable cause to support his arrest. On November 6, 2023, the District Attorney withdrew the charges.4 In his complaint, alongside the factual assertions about the arrest itself, Walton included a

short selection of statistics taken from Racial Injustice Report: Disparities in Philadelphia’s Crim- inal Courts from 2015-2022 (“the Report”) a report published by the Philadelphia District Attor- ney’s Office in June 2023. These statistics are pleaded within the claim under the Equal Protection Clause, and include, for instance, “Black people make up 69% of people stopped by Philadelphia

3 Given his arrest late on March 1, he spent between one and two days in a holding cell. 4 The complaint does not attach the suppression court’s order, nor does it characterize what was suppressed or for what reason. I am aware that such orders in Philadelphia Municipal Court are generally issued orally, sometimes with extremely limited findings of fact and conclusions of law, and may be reduced in writing only to a bare order granting or denying the motion. But cf. Pa. R. Crim. P. 581(I) (requiring suppression court to enter findings of fact and conclusions of law at the conclusion of the hearing). The complaint also does not offer any allegation regarding whether the prosecution withdrew the case based on the suppression order, which may have pre- cluded it from proving its case, or for other reasons. police, 62% of arrests, but only 38% of the Philadelphia population.” Compl. ¶ 81. The complaint does not attach the report itself or any of its underlying data, nor does it indicate whether the statistics it alleges—all of which are either percentages or multiples, e.g., “Black individuals in Philadelphia are arrested and charged with DUI at a rate almost twice of that for white individuals,”

id. ¶ 84—are the report’s analysis or Walton’s own. The attributed statistics do not include quota- tions or pin citations. In his supplemental briefing, Walton includes a link to the report, which I preserve in the margin.5 A. Procedural background On May 14, 2024, Walton filed an initial complaint in the present case. On June 14, he amended the complaint. After a motion to dismiss, on August 9, he filed the operative second

amended complaint. The complaint includes eight counts. On October 7, 2024, the named Defend- ants filed the present motion, which only moves to dismiss the Equal Protection Clause claim under § 1983 against Officer McCarthy (“Third Claim for Relief”) and the claims against Sgt. Stankie- wicz (“Fourth Claim for Relief”). On January 10, 2025, the case was administratively reassigned to me. On March 25, 2025, during an initial pretrial conference in chambers, Counsel for Walton moved for leave to provide supplemental briefing, which I granted. Counsel additionally indicated that he would withdraw the charges against three unnamed officers, a stipulation to which effect was entered on May 25, 2025.

5 This permanent link captures the report from the address Walton supplied in briefing, at least as it appeared during the pendency of this motion: https://perma.cc/B7HB-RBFY. II. LEGAL FRAMEWORK Motions to dismiss under Federal Rule of Civil Procedure 12(b)(6) are considered under the standard set forth in Fowler v. UPMC Shadyside: First, the Court separates all well-pleaded facts from the legal elements of a claim. 578 F.3d 203, 210–11 (3d Cir. 2009). The Court must then decide whether those facts, accepted as true, “show that the plaintiff has a plausible claim for

relief.” Id. (quotation marks omitted). The “plausibility determination” is “a context-specific task that requires the reviewing court to draw on its judicial experience and common sense.” Fowler, 578 F.3d at 211 (citing Ashcroft v. Iqbal, 556 U.S. 662, 679 (2009)). The requirement to state a claim “does not impose a probability requirement at the pleading stage, but instead simply calls for enough facts to raise a reasonable expectation that discovery will reveal evidence of the nec- essary element.” Phillips v. Cnty. of Allegheny, 515 F.3d 224, 234 (3d Cir. 2008) (quotation marks and citations omitted). The rule “does not permit dismissal of a well-pleaded complaint simply because it strikes a savvy judge that actual proof of those facts is improbable.” Id. (quotation marks omitted). In considering a motion to dismiss under Rule 12(b)(6), I consider “only the allegations

contained in the pleading to determine its sufficiency.” In re Asbestos Prods. Liab. Litig. (No. VI), 822 F.3d 125, 133 (3d Cir. 2016). In general, where matters outside the pleadings are presented to and not excluded by court, the motion must be treated as requesting summary judgment. Fed. R. Civ. P. 12(d). However, I may consider “matters incorporated by reference or integral to the claim, items subject to judicial notice, matters of public record, orders” Buck v. Hampton Twp. Sch. Dist., 452 F.3d 256, 260 (3d Cir. 2006) (stating standards of appellate review but referring to identical standard for review in the first instance); accord Tellabs, Inc. v. Makor Issues & Rights, Ltd., 551 U.S. 308, 322 (2007) (establishing the “prescription[]” that “courts must consider the complaint in its entirety” including documents incorporated by reference). With regard to assertions of qualified immunity, because it is an affirmative defense, “qual- ified immunity will be upheld on a 12(b)(6) motion only when the immunity is established on the face of the complaint.” Leveto v. Lapina, 258 F.3d 156, 161 (3d Cir. 2001) (quotation marks and citations omitted).

* * *

Although their motions are joined, Officer McCarthy and Sgt. Stankiewicz move to dismiss on separate grounds: Failure to state a claim under the Equal Protection Clause and qualified im- munity, respectively. I address each in turn. III. EQUAL PROTECTION CLAIM AGAINST OFFICER MCCARTHY Under the Fourteenth Amendment, no state actor may “deny to any person within its juris- diction the equal protection of the laws.” U.S. CONST. amend XIV, § 1. The Equal Protection Clause (EPC) “is essentially a direction that all persons similarly situated should be treated alike.” City of Cleburne v. Cleburne Living Ctr., Inc., 473 U.S. 432, 439 (1985). Its “central pur-

pose . . . is the prevention of official conduct discriminating on the basis of race.” Washington v. Davis, 426 U.S. 229, 239 (1976). The Equal Protection Clause prohibits selective law enforcement based on race. See Whren v. United States, 517 U.S. 806, 813 (1996). To prevail, a plaintiff must show both discriminatory effect and discriminatory intent. Davis, 426 U.S. at 241. Officer McCar- thy argues that the complaint alleges neither that Walton was treated differently from similarly situated persons nor that he was subjected to intentional or purposeful discrimination. Instead, he argues briefly that Walton’s complaint essentially pleads legal conclusions that ipso facto insuffi- cient to state a claim under Rule 8(a) and Twombly. To consider McCarthy’s motion, I begin by examining how discriminatory effect can be made out on facts that derive from statistical, rather than directly observed, facts. I consider whether the pleaded figures are adequate to meet the “benchmark” requirement apparently re- quired by the Third Circuit. I then briefly address discriminatory intent.

A. Discriminatory Effect I begin with discriminatory effect under the EPC. Discriminatory effect is proven where the plaintiff is “a member of a protected class and . . . was treated differently from similarly situ- ated individuals in an unprotected class.” Bradley v. U.S., 299 F.3d 197, 206 (3d Cir. 2002). Officer McCarthy concedes that Walton, a Black person, is a member of a protected class, Defs’ Supp. Br. (“Reply”) 2, so the “sole inquiry” in this aspect is whether he has pleaded evidence that he was

treated differently from otherwise similarly situated comparators. Bradley, 299 F.3d at 206; Reply 2.6 As to which actions he is challenging, Walton refers to “initiating the vehicle stop, . . . prolong- ing and extending the stop . . ., arresting and charging Mr. Walton with DUI . . ., and initiating and continuing the prosecution of Mr. Walton[.]” Compl. ¶ 86. To make out the evidentiary basis, Walton offers primarily a statistical account about stops and arrests. As to the allegations from his own experience, he points to the lack of any factual basis off which he was he was stopped, arrested, charged, and prosecuted. He implicitly claims that other people, i.e., white people, are not subjected to such baseless actions, but he does not have any

6 Neither the complaint nor the briefing concerns racial nuance beyond Blackness and whiteness, whether as to other races, mixed races, Hispanic ethnicity, or any number of realities complicat- ing this view of criminal process. Cf. Report 11–12 (noting that “current and historical public data systems do not accurately capture the ethnic, cultural, and racial identities of all people”). I do not mean to suggest that such would be legally necessary at this stage or otherwise. Neverthe- less, my discussion follows the complaint’s and parties’ paradigm and therefore does not engage with the factually deeper complexity. direct evidence to provide on that point. Instead, he provides several allegations sourced from Racial Injustice Report: Disparities in Philadelphia’s Criminal Courts from 2015–2022, published by the Philadelphia District Attorney’s Office in June 2023.7 This includes the following state- ments, which I reproduce from the complaint directly:

7 As mentioned above, Walton did not attach the report to his complaint as an exhibit, instead re- ferring directly to facts derived therefrom. In a footnote to his supplemental brief in opposition, he offers a web address for the report. The complaint does not explicitly incorporate the Report or purport to quote directly from it, except to indicate that the percentages and multipliers cited are taken from it. The complaint itself does not offer a way to access the report directly, although Walton’s supplemental briefing included a website URL. See supra note 5. He does not refer to any part of the report outside the selected facts, and I am unsure whether he seeks to incorporate it in its entirety. Officer McCarthy raises no procedural concerns, and neither side addresses whether the Report is a matter of public record or judicially noticeable. I could treat the complaint as seeking to incorporate the entire Report. Maybe it goes against the spirit of the Federal Rules’ demand for brevity. See Fed. R. Civ. P. 8(a)(2) (requiring com- plaint to provide a “short and plain statement of the claim”). The trouble with the latter is that it leaves the somewhat barren figures without context, and I must construe pleadings so as to “do justice.” Id. Rule 8(e) The first statement, for instance: “From 2015-2022, Black defendants in Philadelphia were subjected to vehicle stops and charged at a disproportionately higher rate rela- tive to other groups.” Compl. ¶ 79. That language is very broad—and it does not appear in the Report itself. The Report is 68 pages, but just three pages concern the rates of stops and arrests. Four more consider charging. Report 36–38; id. at 39–42. Much of the content is concerned with accessible, public-facing explanations of criminal procedure, with little more relevant statistical content than what is pleaded. The Report offers minimal direct statistical analysis and practically none of the underlying data. I do not consider the Report incorporated into the operative complaint, but it does not make much difference. Far from a detailed statistical analysis, the report presents overall figures with- out extensive context, caveats, or explanation. I mean no critique of its authors, analysts, and de- signers. The report is simply more of an executive summary than a detailed analysis of particular questions. Some of the information is not provided textually: Walton’s allegation regarding the neighborhood in which he was arrested relies on a choropleth (a map with colored sections) of the city divided into unnamed tracts that are colored according to a key from “[l]ess than 1/10 the city rate” to “[m]ore than 5x the city rate,” rather than, for instance, given in a table with each neighborhood’s figures and precise boundaries. Compare Report 38 (choropleth) with compl. ¶ 83 (indicating data read from the choropleth). Since this information is not directly available in a table or as raw data, it is hard to say for certain whether the location of the stop and arrest here actually coincides with a tract that has stop and arrest at the maximum relative concentration. To my eye, the report appears to indicate that the area of 61st and Catharine Streets has a somewhat- elevated concentration of stops but only a mildly elevated concentration of arrests. Incidentally, 79. From 2015–2022, Black defendants in Philadelphia were subjected to vehicle stops and charged at a disproportionately higher rate relative to other groups. . . . 81. Black people make up 69% of people stopped by Philadel- phia police, 62% of arrests, but only 38% of the Philadelphia popu- lation. 82. White people account for only 18% of police stops and 21% of arrests, even though Black and white people make up similar shares of Philadelphia’s population. 83. Philadelphia’s Cobbs Creek section (where the subject vehi- cle stop and arrest occurred) is a predominantly Black area and has a police stop and arrest rate more than 5 times the overall city rate. 84. Black individuals in Philadelphia are arrested and charged with DUI at a rate almost twice of that for white individuals. Compl. ¶¶ 79–84 (citations omitted). In his motion to dismiss, Officer McCarthy reads the complaint, including the excerpted paragraphs, as alleging “not . . . a single fact that supports his equal protection claim.” Defs’ Memo. of L. (“Memo.”) 5. He offers very little argument or authority specific to discriminatory effect, including where the complaint relies on statistics rather than a direct comparator. Officer McCarthy ignores the paragraphs excerpted above, making no argument regarding their use other than by passing reference to the underlying facts as “general occurrences” that do not relate to his own discriminatory intent. Id. In response, Walton cites Bradley and Chavez to show that the sta- tistical evidence can be used to prove discriminatory effect. Pl’s Brief in Opposition (“BIO”) 12, ECF 19.8 He grants that, under United States v. Whitfield, the “statistics must compare the racial

although the key on the relevant choropleths says that it shows the rates “in each census tract,” the map outlines combine present U.S. Census tracts into larger groupings. 8 Walton’s citation to United States v. Washington, No. 13-CR-171-2, 2021 WL 120958 (E.D. Pa. Jan 13, 2021) is not persuasive in his favor. The plaintiff’s statistics there, which show a stark ra- cial pattern in local “stash house cases,” were found insufficient to show discriminatory effect because they failed to produce comparators who were not arrested or charged. Id. at *24. The Supreme Court has set a high standard for criminal defendants challenging their prosecu- tion as unlawful selective enforcement to show similarly situated comparators. United States v. Armstrong, 517 U.S. 456, 470 (1996). Although Walton refers to alleged discrimination in prose- cutorial functions, i.e., charging and further criminal process, I consider his claims as to stop and composition of those targeted to some appropriate benchmark” of “similarly situated individuals of another race who could have been targeted but were not,” 29 F. Supp.3d 503, 513–15 (E.D. Pa. 2014), but asserts that “the statistical evidence . . . meets this standard” at this level. BIO 12. Of- ficer McCarthy replies that, even under Walton’s own standard, the argument fails because “[t]here

are no facts that demonstrate [w]hite individuals who committed a traffic violation were not stopped nor are there any facts that white individuals who exhibit signs of [DUI] were not stopped and[/]or not arrested.” Reply 2–3, ECF 19.9 In addition, Officer McCarthy argues that the statistics are insufficient because they cover data from a span of seven years, which is insufficiently partic- ular to the facts of the time of this arrest. Id. at 3. Finally, he argues that there is no reasonable expectation that discovery will bear out the evidence necessary for Walton’s claims.10 1. Lacking a Direct Comparator, the Plaintiff May Rely on Statistical Evidence There is good reason why a litigant like Walton must rely on statistical evidence to prove discriminatory effect rather than something more direct. Some other plaintiffs have the legal

arrest only under the more permissive standards for a selective policing civil rights case. See Chavez, 251 F.3d at 640 (distinguishing affirmative racial profiling equal protection case from motion to dismiss criminal prosecution and therefore not applying Armstrong requirement of a named comparator). 9 Officer McCarthy claims that Whitfield “did not involve an equal protection claim.” Reply 2. Although Whitfield considered federal criminal defendants seeking post-conviction discovery ra- ther than a plaintiff taking on state law enforcement, the defendants did base their claims on ra- cial disparities under the Equal Protection Clause—albeit through the 5th Amendment’s equal protection component. 29 F. Supp.3d at 512 (citing Armstrong, 517 U.S. at 464–65); Whren, 517 U.S. at 813. See supra note 8. 10 Officer McCarthy, as the movant, bears the burden of persuasion, despite the underlying ques- tion being whether Walton has himself alleged sufficient facts to suggest the required elements under Iqbal and Phillips. See Marcure v. Lynn, 992 F.3d 625, 631 (7th Cir. 2021). His initial mo- tion to dismiss simply ignored the statistical facts from the complaint in his original motion, but his very brief supplemental briefing on discriminatory effect at least takes on the best version of Walton’s case. Although these arguments were substantially missing in the original motion, Wal- ton does not argue that they were forfeited or waived, and I will consider them here. advantage of being able to produce obvious comparators. In most situations, however, the odds of a bona fide target of impermissible profiling producing a helpful comparator are low. See Chavez v. Illinois State Police, 251 F.3d 612, 640 (7th Cir. 2001) (“[P]laintiffs who allege that they were stopped due to racial profiling would not, barring some type of test operation, be able to provide

the names of other similarly situated motorists who were not stopped.”). For instance, if there were a funeral procession of twenty identical cars traveling the same speed along the same route at the same time and only one of the cars in the middle were stopped, that driver-plaintiff might not only be able to offer not just general evidence about those circumstances but even specific information about each of the other drivers’ races. Walton presumably could not plead that there was a white driver immediately ahead of him, whose driving he was copying closely but who was not stopped. Yet, in general, when a plaintiff with a factually meritorious claim lacks the good fortune of such exceptional evidence, she is unable to show discriminatory effect. The Third Circuit has acknowledged the problem of producing direct comparators in many situations of alleged profiling, and it has endorsed the use of statistical evidence in their stead “in

some cases.” Bradley, 299 F.3d at 206 & n.11 (noting that “statistical evidence of discrimination may be the only means of proving a discriminatory effect” in case where plaintiff submitted none). The Supreme Court itself “relied on statistics” that proved unlawful discrimination under the Equal Protection Clause in several cases at least as far back as Yick Wo v. Hopkins in 1886. Chavez, 251 F.3d at 638. In that case, San Francisco had banned the operation of laundries under a facially neutral but functionally discriminatory ordinance, denying exemptions from a building code to some 200 Chinese applicants who applied for them, while granting exemptions to all but one of the non-Chinese launderers who applied. 118 U.S. 356, 374 (1886).11 The Supreme Court

11 The Court in Yick Wo characterized its own holding as not needing to rely on statistics: overturned judgments against the Chinese petitioners where “the conclusion cannot be resisted that no reason for [the discrimination] exists except hostility to the race and nationality to which the petitioners belong, and which, in the eye of the law, is not justified.” 118 U.S. at 374. Similarly, in Gomillion v. Lightfoot, Alabama sought to redefine the boundaries of the city of Tuskegee from a

simple square to a twenty-eight-sided shape, with the effect of removing all but four or five of its Black voters while retaining all its white voters. 364 U.S. 339, 341 (1960). On that record, the Court found discriminatory effect was “amply” pleaded. Id. at 342 (principally relying on Fifteenth Amendment rather than Equal Protection Clause). These are somewhat exceptional statistical cases in that they represent nearly one-sided phenomena, but even a far weaker observation may suffice. In one case where lower courts had found that a felon disenfranchisement law in Alabama was plainly motivated at its enactment by the desire to disenfranchise Black Alabamans, the Court upheld the finding of discriminatory effect where in some counties “[B]lacks are by even the most

[W]e are not obliged to reason from the probable to the actual, and pass upon the validity of the ordinances complained of, as tried merely by the opportunities which their terms afford, of unequal and unjust discrimination in their administra- tion; for the cases present the ordinances in actual operation, and the facts shown establish an administration directed so exclusively against a particular class of persons as to warrant and require the conclusion of unlawful discrimination. 118 U.S. at 373. The Chavez court’s (and my) characterization of Yick Wo as nevertheless premised on a statistical account of similar situation draws on the fact of so many Chinese launderers being refused license while so many non-Chinese launderers were granted it. What distinguishes this from the same claim made by a single Chinese plaintiff offer- ing a single non-Chinese comparator is the statistical certainty, after about 280 occurrences, that the determinative factor is race. Cf. Gomillion, 364 U.S., at 341 (characterizing the “stark” evi- dence in Yick Wo as “tantamount for all practical purposes to a mathematical demonstration”); McCleskey v. Kemp, 481 U.S. 279, 294 n.12 (1987) (characterizing Yick Wo and Gomillion as resting on stark statistical showings of discriminatory effect). Notably, the statistics in Yick Wo were hardly needed to show discriminatory effect, i.e., that comparators existed, given the non-Chinese launderers’ being granted licenses and the “fact of this discrimination [being] admitted.” 118 U.S. at 374. The statistics were evidence that discrimi- natory intent drove the plaintiffs’ outcomes. See infra § III.B.2. modest estimates at least 1.7 times as likely as whites to suffer disfranchisement.” Hunter v. Un- derwood, 471 U.S. 222, 227 (1985). This finding was not contested or under review—but the Su- preme Court noted that it found “no evidence in the record below or in the briefs and oral argument in this Court that would undermine” discriminatory effect. Id.

Although these Supreme Court decisions provide an elementary background from which to imply a case for statistical EPC cases, neither concerned itself much with the actual interpreta- tion or operation of statistics, particularly at the early stage of pleadings. The Third Circuit has not addressed the same, but the Seventh Circuit reviewed a very similar racial selective policing claim and considered the use of statistics to show discriminatory effect. In Chavez v. Illinois State Police, that court considered two consolidated appeals. In the first, a Hispanic driver Chavez was pulled over by state troopers for failure to signal on the highway, although the driver testified that he had properly signaled. 251 F.3d at 623. He was then detained for a dog sniff and a search of his car and luggage, neither of which revealed any lawbreaking. Id. at 623–24. In the second case, Gregory Lee claimed to have been repeatedly stopped while driving by troopers presenting and recording

false pretexts. Id. at 624–25. Because neither had direct comparators to offer, Chavez and Lee had the Illinois State Police provide a database containing information on stops to the Temple Univer- sity Center for Public Policy. Experts from that department provided an analysis opining that there was a “systemic overrepresentation” of Hispanic and Black drivers like the plaintiffs. Id. at 626. As the Seventh Circuit concluded, the legal issue that the Chavez plaintiffs faced was not whether statistics could provide a factual underpinning at all. Id. at 637–40 (noting that statistical facts are of frequent use in other civil litigation, such as employment discrimination cases); see also Mayor of City of Philadelphia v. Educ. Equal. League, 415 U.S. 605, 620 (1974) (“Statistical analyses have served and will continue to serve an important role as one indirect indicator of racial discrimination . . .”). Instead, the issue with the statistics is the matter of whether the figures of- fered can sufficiently do the work they are called on to do. For the Chavez court, weaknesses in the statistics themselves were the problem, because they failed to provide “comparative racial in- formation” to “prove that [Hispanic or Black drivers] were stopped, detained, or searched, when

similarly situated whites were not.” Chavez, 251 F.3d at 643. See also Int'l Bhd. of Teamsters v. United States, 431 U.S. 324, 340 (1977) (statistics’ “usefulness depends on all of the surrounding facts and circumstances”). 2. Requirement of a Benchmark I have similar concerns here about the quality and analysis of the statistics provided, in part because the appellate courts have characterized the legal question for a case based on statistics somewhat differently from one with a directly pleaded comparator. The requirement of showing a similarly situated comparator allows that even a single example would suffice. For instance, the Third Circuit recently overturned summary judgment in favor of an officer who had arrested only one of the two men who were in an outrageous physical altercation (one was driving on the high- way with the other on his windshield) because a jury could “easily compare[] side-by-side” the

Black man arrested with the similarly situated white man who was not. Harvard v. Cesnalis, 973 F.3d 190, 206 (3d Cir. 2020). It seems reasonable that a statistical case would simply need to be strong enough to raise the inference that such similarly situated comparators did exist, without being directly observed—similar, in essence, to pleading the comparator on information and belief. For good reason, however, the Third Circuit (following the Seventh in Chavez) appears to apply a different standard to EPC cases built off statistics. The Third Circuit has not directly ad- dressed this issue, but it has offered reasoning, citations, and dicta demonstrating its position. In Bradley, the plaintiff was an African American airplane passenger arriving from Jamaica who was patted down and had her luggage searched. 299 F.3d at 201. Bradley had specifically pleaded the presence of several deplaning white passengers who were not selected to have their luggage searched, despite being asserted as similarly situated along whatever relevant dimensions. Id. at 206 (not suggesting any reason why the men were not similarly situated). The court recited the requirement that Bradley show “evidence that customs officials treated her differently from simi-

larly situated members of an unprotected class.” Id. Yet, the Court nevertheless held that she had failed to show that prong, since, [e]ven if we assume that the white males were similarly situated [be- cause they, like Bradley, were disembarking while wearing hats], an assumption we are somewhat loathe to make, Bradley failed to sub- mit any evidence that she was unfairly singled out. The mere fact that a few unidentified white males on a flight of many passengers were not selected when Bradley was does not, without more, demon- strate a discriminatory effect. Id. Why was the single white co-combatant in Harvard sufficient but not the several white passengers in Bradley? An additional fact in Bradley that would otherwise seem irrelevant to the analysis is the likeliest contender: As the Court noted, and Bradley conceded, the only other pas- senger who was targeted for a search was a white man. Id. The presence of a similarly situated white person who was targeted obviously does not countermand the presence of similarly situated white people who were not targeted. Instead, it potentially undermines whether the effect was actually discriminatory. Although the Third Circuit did not alter its standard for discriminatory effect as recited, i.e., that Bradley must “show that she . . . was treated differently from similarly situated individuals in an unprotected class,” it repeatedly and favorably cited Chavez for that element. Bradley, 299 F.3d at 206 (emphasis added). But the standard in Chavez frames in classes, not individuals. The Seventh Circuit held that statistics must address “the crucial question of whether one class is being treated differently from another class that is otherwise similarly situated.” Chavez, 251 F.3d at 638 (emphasis added). For his part, Walton concedes that this is the proper standard. BIO 12. Or, as the parties agree, the “statistics must compare the racial composition of those targeted to some appropriate benchmark” of “similarly situated individuals of another race,” Whitfield, 29 F. Supp.3d at 513–15. One straightforward way to interpret the Bradley court’s analysis is that it turned on the “racial compo-

sition of those targeted”: ½ Black, ½ white. But, perhaps equivalently and of more use here, I read the opinion as considering some background benchmark targeting of whites to have been demon- strably increased by the presence of the white detainee:12 His detention and search undermined any assertion that white passengers were not also being stopped or searched. In other words, once it was shown that the arithmetic was not as “stark” as the patterns in Yick Wo and Gomillion—not as black and white as all Black detainees and no white ones—the plaintiff needed to make a greater showing. The weakness of the showing provided was only exacerbated by the small amount of data involved: Bradley discussed two arrests and says nothing of how many other passengers were on the plane (or their racial composition). Under such circumstances, the panel in Bradley all but recommended using statistics. 299 F.3d at 206 (citing Chavez).13 The panel also seems to have

incorporated Chavez’s standards. The work to be done by statistical facts therefore goes much farther than providing an inference that one direct comparator individually existed. Instead, it must permit the inference of an appropriate benchmark comparison central to the analysis. Neither the complaint nor the brief- ing identifies what that benchmark is or describes what it should be. Both parties characterize the figures as indicating a “race-based disparity,” Pl’s Supp. Br. 4, Reply 3, but the question and the

12 He was arrested after the search when contraband drugs were found with him. Bradley, 299 F.3d at 206. 13 That is apparently how Bradley saw things too. She “d[id] not contest” the failure to make out discriminatory effect, but instead argued that she should have been permitted discovery. She lost that motion for unrelated procedural reasons. Id. answer must be much more specific: What would constitute “similar situation” to Walton, and is there a disparity in the rate at which Black people are stopped (or have their stops prolonged, or are arrested, etc.) compared to non-Black people who are similarly situated? I do not mean to undermine the gravity of the correlations on their face from a larger social standpoint. However,

without further analysis, context, or explanation, I cannot to infer a dispositive disparity given the details of Walton’s arrest.14 The interpretive work of finding the proper “benchmark” is functionally similar to describ- ing which aspects of Walton’s situation are relevant to considering a “similarly situated” but un- targeted comparator, plus the added analysis of how race affects those comparators. For instance, if I considered all nighttime drivers in Philadelphia to be similarly situated to Walton, an appropri- ate benchmark would address the targeting of white nighttime drivers. But an additional concern critical to the benchmark is that it cannot be inferred from a mere disparity in the group of those targeted by enforcement. As both parties here agree, the focus of the inquiry must not be on the composition of those targeted, but instead on comparison of the

class of those targeted to those not targeted—those “who could have been targeted but were not,” in the case and phrase cited throughout the briefing. Whitfield, 29 F. Supp.3d at 514–15; BIO 12; Reply 2; see Chavez, 251 F.3d at 635, 642–45 (plaintiffs would need reliable data on not only the racial makeup of those stopped and searched but also the racial makeup of local motorists).

14 The parties have not briefed the question of what kind of analysis, if any, the plaintiff must provide within a complaint that relies on statistics, or in his briefing. The Third Circuit has re- quired plaintiffs to carry the burden of providing a statistical analysis to make out a prima facie case at the level of summary judgment. E.g., Meditz v. City of Newark, 658 F.3d 364, 373 (3d Cir. 2011) (disapproving of district court’s merely comparing percentages rather than standard devia- tion analysis); Green v. City of Philadelphia, No. 21-1034, 2022 WL 1165644 at *3 (3d Cir. 2022) (mere percentages insufficient for prima facie case unless stark) (unreported). Far less is demanded at this stage, at which mere pleadings providing for inferences suffice. Nevertheless, the inferences must draw only from the factual content of the pleadings. See supra § II. 3. Similar Situation The aspects of Walton’s situation that are relevant likely differ for the purposes of the stop and the arrest.15 Beginning with the stop, Walton does not plead any facts about his factual inno- cence or guilt of the traffic infraction (other than his own denial on the scene). For the purposes of the stop alone, the similarly situated driver appears to be, at the most general, someone driving

having just committed a minor traffic infraction. To say nothing of what information might be available elsewhere, the figures provided in the complaint provide no factual basis off which to infer that the police are stopping Black drivers who have been observed violating traffic rules at different rates from other drivers who have been observed doing the same.16 As to the arrest, the question of who ought to be “similarly situated” is blurrier. The task of determining who is “similarly situated,” or equivalently of determining which aspects of Wal- ton’s situation were relevant “common features essential to a meaningful comparison” is a “com- mon-sense inquiry.” Chavez, 299 F.3d at 636. On one hand, Walton was evidently sober at the

15 I leave aside Walton’s complaint regarding discriminatory effect through prolongation of the stop, since he provides no information concerning how prolonged stops affect different classes of people. Similarly, I do not address discriminatory effect in charging and court process, since 1) under Armstrong, they require the showing of a direct comparator not pleaded here; and 2) there is no suggestion that Officer McCarthy had any substantive authority in those prosecutorial deci- sions other than giving the charge at arrest, which I find to be inseparable from the arrest itself. 16 Commentators have suggested, apparently without evidence, that the scope of the traffic codes and practicalities of driving mean that “no driver can avoid violating some traffic law during a short drive, even with the most careful attention.” David A. Harris, Driving While Black and All Other Traffic Offenses: The Supreme Court and Pretextual Traffic Stops, 87 J. Crim. L. & Crimi- nology 544, 545 (1997); David Rudovsky, Law Enforcement by Stereotypes and Serendipity: Ra- cial Profiling and Stops and Searches without Cause, 3 U. Pa. J. Const. L. 296, 318 (2001) (cit- ing Harris and stating that “virtually every driver commits violations of the traffic laws on a reg- ular basis”). In one study, more than 98% of highway drivers were observed to exceed the speed limit, State v. Soto, 734 A.2d 350, 352 (N.J. Super. 1996), although contradictory evidence by state officials was excluded on procedural grounds, id. at 355 n.7. point of arrest.17 He had committed a minor traffic infraction in failing to signal. In addition, by the time he arrested, Walton concedes that he had been unable to produce a valid paper registration as required, although he asserts that he was the valid registrant and owner. Compl. ¶ 47. Perhaps a similarly situated driver would be one who was driving in Philadelphia when she was observed

committing a minor traffic violation. Perhaps it would be one who was stopped for a traffic viola- tion and then observed to violate a second minor traffic violation, like illegal tint, outdated inspec- tion, or failure to carry a registration card. Or perhaps violations like these are of such minor impact on the outcome in Walton’s telling that they hardly affected it. What zoom lens should the Court apply when imagining a proper class of similarly situated individuals? One candidate might be Philadelphians between 2017 and 2023. Another candidate might be drivers in Cobbs Creek on the evening of March 1, 2023. There are many levels of gen- erality in between. Levels of generality escalate geographically, from the block on which Walton was arrested to the whole region. They escalate temporally, from the night of March 1, 2023, to all nights or all times for two decades on either side.18 They escalate based on similarity in fact pat-

terns, from all DUI arrests of factually innocent drivers to all traffic stops that led to arrest to all DUI traffic stops to all stops. Similarly, it might be necessary to consider only arrests by this officer

17 As with the signaling violation, Walton does not directly plead his factual innocence of the DUI. Nevertheless, he pleads the facts that a timely blood test produced no evidence of intoxica- tion and that he demonstrated no behaviors indicating intoxication. 18 Officer McCarthy’s argument that the Report data specifically is unusable because it is from the wrong timeline (2015–22) is hard to understand, particularly at this level. The Report relates to the seven years immediately prior to the arrest. Its underlying data apparently ends on Decem- ber 31, 2022. The present arrest occurred two months and a day later. The report was not pub- lished until June of 2023, months after the arrest. In addition, the complaint directly alleges that the statistics in the Report continued to be descriptive. Compl. ¶ 80 (“At the time of the subject incident, this racial disparity persisted with regard to vehicle stops and arrests in Philadelphia.”). Barring a showing of some remarkable change of circumstances, perhaps relating to a change since the COVID-19 closures, there is no reason not to infer that the Report is at least very well matched to the time of the arrest. himself, or only in his police service area or district. It might be valuable to consider only drivers on roads in residential neighborhoods, to include highway drivers, to exclude commercial driv- ers.19 It may be unnecessary to address almost all of this at this stage. As to geography and time

and personnel, I make no decision for now as to where on those spectra of generality a set of people treated differently across a racial classification would be sufficiently similarly situated, or where the data would be too far removed from Walton’s own situation. As to factual similarity, however, I do decide that it is legally insufficient, even at this level, to rely on is statistics that are not confined to traffic stops but instead address the overall rates of all stops and arrests of Black and white people. The nature of traffic stops, including the breadth of reasons for their beginning, the many rules and policies surrounding them, and the ways in which they may skip from one justification to another, have made them a very particular interac- tion. As a matter of “common sense,” and with no argument to the alternative, I find that traffic stops and arrests are sufficiently distinguishable from other stops and arrests so that they are not

similar situations for this purpose. Of the statistical allegations excerpted above, only ¶¶ 79 and 84 confine themselves to traffic stops and arrests, and neither alleges benchmark information about untargeted people. First, “From 2015–2022, Black defendants in Philadelphia were subjected to vehicle stops and charged at a disproportionately higher rate relative to other groups,” Compl. ¶ 79. This sentence is unfor- tunately ambiguous. Its structure indicates that by “other groups” it refers to “non-Black racial groups of defendants.” All defendants are, ipso facto, those who have been charged. In one reading,

19 I discuss below a defendant-created survey of drivers on the New Jersey turnpike which was designed well enough that similar situation was not even a topic the court needed to discuss. See generally State v. Soto, 734 A.2d 350 (N.J. Super. 1996). the allegation compares the proportion of Black defendants who were charged after a vehicle stop (with all Black defendants as the denominator) to the proportion(s) for non-Black defendants. Comparing such proportions would be irrelevant here, however, since they would be as influenced by racial differences in targeting as by racial differences in overall driving or in the number of non-

vehicle stop arrests. In a second reading, at least preferable for relevance, it compares the ratio of Black defendants arrested after a vehicle stop to the total Black population and compares that ratio to that of other racial groups. And again, the other traffic allegation says, “Black individuals in Philadelphia are arrested and charged with DUI at a rate almost twice of that for white individuals.” ¶ 84.20 One might imagine that the alleged disproportions in vehicle stops in ¶ 79 and the wide difference in the number of DUI arrests in ¶ 84 would each permit the inference of a disparate effect. Why not compare the ratio of Black vehicle-stopped defendants to Black share of the total population with the ratio of white vehicle-stopped defendants to the white population share? Why should it not be enough to say, “the rate of Black people arrested after traffic violations is far

greater than the rate of white people”? And similarly with the DUI arrest rates? The hidden presumption is that drivers commit traffic violations and DUI at similar rates across races. Armed with that presumption, the much higher stop and arrest rates for Black drivers

20 Because of the ambiguity of the word “rate” in this context, is possible to read this paragraph as actually alleging what would be required, to wit, that the rate at which Black people in Phila- delphia who could be targeted for DUI stops and arrests are arrested and charged is almost twice that at which similar white people in Philadelphia are. (Technically, the comparison would in- clude one more step, comparing the first group and white people in Philadelphia who could be targeted and are not.) I think this reading is a stretch. Because I grant the plaintiff the opportunity to amend, he has the opportunity to rephrase or research, as the case may be. Separately, I characterize DUI arrests as traffic stops at this point, but some number of DUI arrests do not result from routine traffic stops. In general, those may occur when police arrive at the scene of a vehicle accident, stop a driver based on a tip, or make a stop at a checkpoint. under the influence could at least indicate a benchmark of much lower targeting of white drivers. Yet, Supreme Court case law strongly suggests that I must question whether there could be a dif- ference in the rate of actual violations among people of different races. When a Court of Appeals relied on the presumptions that “all races commit all types of crimes” and that no crime was pri-

marily associated with a particular race, the Supreme Court explicitly doubted the premise and forbade reliance on it—at least, barring further evidentiary development thereon. United States v. Armstrong, 517 U.S. 456, 469 (1996) (citing Sentencing Commission statistics showing that the vast majority of those sentenced for pornography and prostitution were white, whereas the vast majority of those sentenced for crack cocaine trafficking were Black). Armstrong seems to pre- scribe a null hypothesis that people of different races commit different crimes at different rates. There is nothing in the complaint or briefing that would transform the presumption of racial simi- larity into a founded inference, so the null hypothesis must hold. In other words, if Black drivers are stopped more often, I am not supposed to infer that it is not because Black drivers might violate traffic laws more—at least, Walton cannot avail himself of such an inference until he can offer

some evidentiary basis. The additional statistical allegations are also insufficient, including in combination with ¶¶ 79 and 84, to infer disparate effect. For instance, ¶ 81 states, “Black people make up 69% of people stopped by Philadelphia police, 62% of arrests, but only 38% of the Philadelphia popula- tion.” The next paragraph even offers that the Black and white populations of Philadelphia are of similar size. According to the next paragraph, “[w]hite people account for only 18% of police stops and 21% of arrests, even though Black and white people make up similar shares of Philadelphia’s population.” Id. ¶ 82. The implicit allegation—again, the complaint and briefing do not offer their own analyses—is that white people are stopped only about a quarter as often as Black people and arrested about a third as often as Black people. The figure is gravely concerning as a general matter, but it remains a measure of all stops and arrests. Were ¶¶ 81–82 to address traffic stops and arrests rather than the full mass of them, it could be getting somewhere. It would not solve the issue that the present allegations only address those targeted, rather than those not targeted. Yet, combined

with the allegation, perhaps plausible at this level, that all drivers violate traffic laws during all drives, they might form part of a viable basis of facts to infer the proper benchmark.21 4. Leave to Amend On the present complaint, there is no information sufficient to ascertain the benchmark of those drivers who were not of a protected class, and who could have been targeted but were not. Nonetheless, I do not think that Walton is without any hope of finding statistical evidence in his favor. Walton presented in his complaint five allegations drawn from a report that, although cen- trally concerned with race, did not itself focus on stops, or even traffic stops. Walton is not stuck with only the Report. There is substantial information available for analysis that might address the concerns that thwarted a showing of discriminatory effect. In 2011, in a case resulting in part from complaints of policing that violated the equal protection clause,

Philadelphia entered into a consent decree in this Court. As part of that decree, the city agreed to record electronically mountains of data on all its stops.22 This data is available publicly through the City’s online OpenDataPhilly portal, and multiple analyses have already used it to consider the role of race in traffic stops in in Philadelphia. E.g., Defender Association of Philadelphia, Traffic Stops in Philadelphia, http://driving-equality.phillydefenders.org; Hannon, L., Neal, M., & Gus- tafson, A. R., Out-of-Place and In-Place Policing: An Examination of Traffic Stops in Racially

21 See supra note 18. 22 Settlement Agreement, Class Certification, and Consent Decree at 3–7, Bailey v. City of Phila- delphia, No. 10-5952 (E.D. Pa. June 21, 2011). Segregated Philadelphia, 67 Crime & Delinquency 6–7, p. 868 (2020), https://doi.org/10.1177/0011128720926122. In the alternative, some drivers have produced their own data. The Chavez case itself re- sulted from that plaintiff being hired as a comparator for another driver who was arrested after a

vehicle stop he felt was based on skin tone. Chavez, 298 F.3d at 623 (“Chavez, who is Hispanic, emulated the circumstances surrounding [the arrestee’s] stop and arrest, to see if he would be stopped by the Illinois State Police.”). More programmatically, and more successfully, Black crim- inal defendants pursuing suppression under EPC developed their own surveys under a research plan designed by a Temple University expert in State v. Soto, 734 A.2d 350, 352 (N.J. Super. 1996). This involved first a “traffic survey” that confirmed that Black vehicle occupancy on the Turnpike close to the arrests closely tracked the Black share of the overall population in the states of the vehicles’ license plates. Id. Next, the defendants completed a “violator survey” in which they ob- served that more than 98% of all drivers near the same portion of turnpike were driving faster than the speed limit, and that the difference between the rate of all traffic that was Black-occupied and

all traffic violators in cars that were Black-occupied was statistically insignificant. Id. at 352–53. What was exceptionally significant was the disproportion of the rate at which Black-occupied cars were stopped, which was calculated to represent between 16 and 22 standard deviations. It would be galactically improbable to observe such an event occurring randomly,23 which the court found sufficient to support a finding of discriminatory intent and therefore suppression. Id. at 84–85 (de- scribing disparities as “indeed stark” and sufficient under McCleskey).

23 The probability of observing an event at 16 standard deviations from the norm may be in the range of one in 1057, or a billion-trillion-trillion-trillion-trillion, according to my law clerk. I join Justice Stevens’s tradition by relying on a law clerk’s calculations for matters of statistics. Hazel- wood School District v. United States, 433 U.S. 299, 318 n.5 (1977). I offer the preceding paragraphs to justify my ruling on his request for leave to amend, not as recommendations to the plaintiff. I make no ruling as to what kind of statistical evidence would be sufficient to infer discriminatory effect from an amended complaint.

B. Discriminatory Intent Although I have found that the complaint fails to adequately plead the necessary element of discriminatory effect and therefore the equal protection claim fails, I will venture briefly into the matter of discriminatory intent as it concerns Walton’s arrest. Assuming the complaint will be amended and another motion to dismiss require briefing, this analysis may be of use in directing the parties’ arguments. That is particularly the case since, as I discuss below, I find discriminatory intent to have been sufficiently pleaded already, at least as to Officer McCarthy himself.

1. Proving Discriminatory Intent Using Statistics In a case pleading statistics such as this, the statistics might be called in to contribute a second time for the other element. Again, Officer McCarthy argues that Walton has “not alleged a single fact that supports his equal protection claim” including ¶ 86, which he quotes in full because it could at least “remotely support his claim,” and which I include in the margin.24 Defs’ Memo.

24 Here is the paragraph in question: 86. In initiating the vehicle stop, unlawfully and unreasonably prolonging and ex- tending the stop while insisting that Mr. Walton produce a physical registration, arresting and charging Mr. Walton with DUI despite the absence of any reasona- bly reliable information that he was impaired from safely operating a vehicle, and initiating and continuing the prosecution of Mr. Walton after a blood draw re- vealed no evidence that Mr. Walton was under the influence of alcohol or a con- trolled dangerous substance, Officer McCarthy engaged in racial profiling and se- lective enforcement of the law, treating Mr. Walton differently from other simi- larly situated individuals in an unprotected class, and engaging in this selective treatment based on Mr. Walton’s race. Memo. 5 (quoting compl. ¶ 86). Reproducing this paragraph constitutes more than a fifth of Of- ficer McCarthy’s entire argument. 5. Walton, in turn, says that he has given sufficient factual support of McCarthy’s discriminatory intent by quoting portions of the same paragraph, as well as a separate paragraph that states that Officer McCarthy “acted willfully, wantonly, intentionally and/or with deliberate indifference to Mr. Walton’s rights.” BIO 13 (quoting compl. ¶ 67).25 Although he does not provide further au-

thority on discriminatory intent specifically, he presumably relies on his prior citations to Chavez and Bradley. The parties therefore contemplate this element as concerned with whether the statistics are sufficient to plead discriminatory intent.26 In Yick Wo, the Supreme Court not only held that the stark, near complete opposition of the laundry regulation was evidence of discriminatory effect, but also that it was sufficient to find the Equal Protection violation, since “the conclusion cannot be resisted that no reason for [the discrimination] exists except hostility to the race and nationality to which the petitioners belong.” 118 U.S. at 374. If the present statistics are sufficiently extreme as in Yick Wo so that the conclusion “cannot be resisted,” then the claim certainly survives. Yet, many equal protection claims cannot succeed where proof of discriminatory intent

relies on statistics alone. In Washington v. Davis, the Supreme Court considered a suit by two Black police officers challenging the constitutionality of “Test 21,” an examination “designed to test verbal ability, vocabulary, reading and comprehension” administered to police officers seeking promotion. 426 U.S. 229, 234–35 (1976). Upon motions for summary judgment, the trial court determined that Test 21 was not intended to be discriminatory. Id. at 235–36. The plaintiffs ap- pealed the constitutionality of Test 21, and the Court of Appeals overturned, applying Title VII

25 Walton focused in his supplemental briefing on discriminatory effect rather than intent. In his reply, Officer McCarthy points out that “Plaintiff provides no argument on the use of the statisti- cal data cited to prove” the intent element. Defs’ Reply 4. 26 Despite their framing, this question is left very lightly argued by both sides. standards to determine that where “four times as many” Black applicants failed the test than white applicants, the “disproportionate impact” was sufficient to establish a constitutional violation.” Id. at 236–37 (characterizing Davis v. Washington, 512 F.2d 956 (D.C. Cir. 1975)). The Supreme Court disagreed. Id. at 242. It noted that it had rejected claims “based solely on the statistically dispro-

portionate racial impact of various provisions” in the context of Social Security. Id. at 240–41 (1976) (discussing Jackson v. Hackney, 406 U.S. 535, 548 (1972)). The Court allowed that suffi- ciently discriminatory effects might be sufficient to show intent, citing Yick Wo and several other major civil rights cases. Id. at 242. But the Davis majority nevertheless expressed “difficulty un- derstanding how” a law that is racially neutral on its face could violate EPC “simply because a greater proportion” of Black people than people of other races are affected by it. Id. at 245; accord Bradley, 299 F.3d 206 n.11 (“statistics alone rarely state a violation of equal protection” as opposed to mere discriminatory effect (quoting Chavez, 251 F.3d at 640)). Where should a court draw the line between impacts that are mere racial disproportion as in Davis and those that are ipso facto evidence of unconstitutional discrimination as in Yick Wo?

Statistics may make out the evidence, but I am unaware of whether any court has determined a statistical test to decide when disproportion becomes prima facie discrimination. At best, the Su- preme Court counsels looking elsewhere. Where there is no pattern “as stark as that in Gomillion or Yick Wo, impact alone is not determinative, and the Court must look to other evidence.” Vill. of Arlington Heights v. Metro. Hous. Dev. Corp., 429 U.S. 252, 266 (1977) (cleaned up). But see Davis, 426 U.S. at 253 (Stevens, J., concurring) (“Frequently the most probative evidence of intent will be objective evidence of what actually happened rather than evidence describing the subjective state of mind of the actor. For normally the actor is presumed to have intended the natural conse- quences of his deeds.”). First, it is worth recalling the present posture: Whereas Yick Wo and Davis were appeals from final judgment, the present case is only at the pleading stage. If from the complaint there is a reasonable inference that either the statistics alone or the statistics in combination with other evi- dence would rise to the level of showing intent, the complaint should survive. Fowler v. UPMC

Shadyside, 578 F.3d 203, 210–11 (3d Cir. 2009). As the statistics at present are insufficient even to show discriminatory effect, they essen- tially cannot indicate discriminatory intent by the city or police. Bradley, 299 F.3d at 197 n.11. It is hard to imagine that if Philadelphia or its Police Department were intent on discriminating against Black drivers, it could not accomplish that end. 2. Discriminatory Intent as to Officer McCarthy Himself But the complaint files a claim against Officer McCarthy in particular, and the wording of ¶ 86 alleges that it is Officer McCarthy’s own intent that was discriminatory. See compl. ¶ 4 (nam- ing Officer McCarthy in his individual capacity). I believe there is sufficient evidence at this level from the non-statistical allegations to support the inference of discriminatory intent on the part of Officer McCarthy himself.

Officer McCarthy’s actions as alleged are difficult to explain, at this level, for any reason other than intentional discrimination against Walton. There may be important information Walton did not include in his complaint. Regardless, at this level, Officer McCarthy’s prolonged refusal to confirm Walton’s registration through a police system makes little sense. Although the Pennsylva- nia Vehicle Code requires drivers to carry and produce a registration card, the law specifically excludes anyone who travels without it but later presents the card to police within a few days. 75 Pa.C.S. § 1311(c). Further, Officer McCarthy arrested Walton for driving under an intoxication that was detectible to neither a chemical test nor to the specially trained Crash Investigation Divi- sion officer. If Officer McCarthy suspected that Walton was intoxicated, he nevertheless waited more than 15 minutes before removing Walton from the driver’s seat. Finally, Officer McCarthy’s announcement to his supervisor of the reason for the arrest (“He’s getting locked up for DUI.” ¶ 42) could be read as peculiarly defensive under the circumstances and indicates Officer McCar- thy had not offered any suspicion of DUI when he requested a supervisor for backup.

None of these allegations is generally inexplicable, but at the present level, they support the inference that Officer McCarthy was engaged in some combination of especially unprofes- sional and mendacious conduct targeting Walton. There are no surrounding circumstances to sug- gest any alternative reason for this behavior, nor is there any noticeable trait separating Officer McCarthy and Walton besides race, as they apparently share, for instance, gender identity and locality. It could be that Walton had a sports team bumper sticker or Southern-accented speech or something else that set off Officer McCarthy that would not offend the Equal Protection Clause. However, the contradictions of this particular traffic stop and the lack of any plausible alternative will suffice at this level to make reasonable the inference of a discriminatory intention. In the “context-specific task” of considering the adequacy of the complaint’s facts to support its legal

inferences, the Court is charged with “draw[ing] on its judicial experience and common sense.” Fowler, 578 F.3d at 211. I find that the allegations support an inference of discriminatory intent. Should the complaint be amended with statistics regarding Officer McCarthy’s stops spe- cifically, it is clear how this inference will connect with allegations of discriminatory effect. If the complaint relies on statistics more general than Officer McCarthy’s stops alone, I will rely on the parties’ assistance to address any putative inference about how the discriminatory effect and dis- criminatory intent elements connect. IV. QUALIFIED IMMUNITY AS TO SGT. STANKIEWICZ Finally, I address the claim by Sgt. Stankiewicz that he is shielded by qualified immunity from the claims against him. In his Fourth Claim for Relief, Walton claims that Sgt. Stankiewicz is liable first “[a]s a supervisor on the scene, [who] directly and personally oversaw, participated in, and ratified the aforesaid unlawful and unconstitutional acts and omissions of Defendant Of-

ficer McCarthy,” compl. ¶ 91, and second for failing to intervene, id. ¶ 92. Walton conceded in his briefing that the claim against Sgt. Stankiewicz should be dismissed under qualified immunity as to the failure to intervene theory. BIO 13–14. Sgt. Stankiewicz argues that “Plaintiff’s claim against Defendant Stankiewicz contain [sic] nothing but boilerplate allegations without any supportive facts.” Memo. 5. He writes that “both the individual and derivative theories of liability fail as Defendant Stankiewicz is entitled to qual- ified immunity given that such a right was not clearly established.” He then dedicates his entire argument to defeating the theory of liability by failure to intervene. Walton, having conceded dismissal on the failure to intervene theory, argues that this count may nevertheless proceed under the supervisory liability theory. He offers Third Circuit case law

upholding liability where a supervisor participated in a rights violation or had knowledge of the violation and acquiesced. A.M. ex rel. J.M.K. v. Luzerne Cnty. Juv. Det. Ctr., 372 F.3d 572, 586 (3d Cir. 2004). Walton claims that Sgt. Stankiewicz “was personally involved in the deprivation of Walton’s well-established Fourth Amendment rights not to be arrested or prosecuted in the absence of probable cause.” As the movant asserting qualified immunity, Sgt. Stankiewicz bears the burden of persua- sion. Burns v. PA Dep't of Corr., 642 F.3d 163, 176 (3d Cir. 2011) (“The burden of establishing qualified immunity falls to the official claiming it as a defense.”). Inasmuch as his assertion of qualified immunity extends to the supervisory liability theory, it fails for lack of argument. Although he lays out the requirements for qualified immunity (in service of the non-intervention theory), including showing that the right in question was not “clearly established at the time of the challenged conduct,” George v. Rehiel, 738 F.3d 562, 572 (3d Cir. 2013), Sgt. Stankiewicz provides case law exclusively pertinent to non-intervention. Memo. 6–7.

Sgt. Stankiewicz’s motion might still be read as moving to dismiss under Rule 12(b)(6) based alternatively on the inadequacy of the complaint to support its claim against him. Although the motion does not invoke that rule, it alludes to “boilerplate allegations without any supportive facts.” Memo. 5. It does not offer a framework for considering supervisory liability or identify why what is alleged fails to state a claim under that theory or provide any legal analysis. See L. Civ. R. 7.1(c) (requiring “a brief containing a concise statement of the legal contentions and au- thorities relied upon in support of the motion”); Dennis v. DeJong, 867 F. Supp. 2d 588, 622 (E.D. Pa. 2011) (“[F]ailing to address substantive matters raised in a motion may result in the un- addressed issue being granted as uncontested.”).27 His argument certainly does not contend with the allegations that Sgt. Stankiewicz sub-

stantively interacted with Walton while he was still in the driver’s seat and therefore must have observed that there were none of the indicia of DUI (whether in Walton’s behavior or appearance or smell or in the lack of an accident or visible intoxicants), nor even indicia that Walton had been suspected of DUI prior to his arrival (since Officer McCarthy had not removed Walton from the vehicle), not does it contend with Officer McCarthy’s feeling the need to explain to Sgt. Stankie- wicz the charge for which he was arresting Walton. Where a police sergeant arrives on scene and sees the arresting officer holding a driver just because the latter did not have a registration card on him, and sees that the officer has refused for some time to call in the registration to clarify whether

27 The motion’s conclusory claim that the complaint makes conclusory claims succeeds as irony. it is valid, one reasonable inference is that the sergeant knows very well that his inferior officer is in the midst of an unreasonable seizure and making an arrest without probable cause. This case is still not even at the stage of responsive pleadings. Sgt. Stankiewicz will have ample opportunity to deny, contextualize, or otherwise meet these allegations. At the present pos-

ture, however, they can form a sufficient basis for inferring that Sgt. Stankiewicz was both knowl- edgeable of Walton’s arrest without probable cause and that he acquiesced in that action. V. CONCLUSION For the reasons given in this memorandum, I will issue an order granting the motion to dismiss the equal protection claim against Officer McCarthy. Similarly, I will issue an order grant- ing the motion to dismiss the claim against Sgt. Stankiewicz under the non-intervention theory.

Because I believe the plaintiff might be able provide additional allegations to support the claim against Officer McCarthy, the order will grant Walton’s request for leave to amend his com- plaint a second time.