§ 5.04. Nondiscrimination in Determining the Targeted Group

Any suspicionless policing policy that targets a particular group of persons or entities rather than the entire relevant population should be justified by a sound basis in fact.

Comment:

a. Nondiscrimination principle. In a suspicion-based regime, the requirement that there be cause for suspicion provides a basis for distinguishing between those who may be subject to a policing action and those who may not. It keeps government action from being arbitrary or discriminatory. This Section provides an analogue to the cause requirement for suspicionless policing activity targeted at a subgroup of the population. It requires that there be a demonstrable basis for distinguishing between the target group and the rest of the population.

Although, by definition, suspicionless policing activities are not based on individualized suspicion directed at a particular person or entity, if § 5.03 is followed there always will be some evidence-based reason for conducting them. To the extent that a suspicionless regime is directed at the entire population, the justification required by § 5.03 is sufficient. Many suspicionless policing policies apply on their faces to every person or every entity that triggers the need for the policies in the first instance. Policies that authorize airport checkpoints affect all plane passengers; health-and-safety-inspection laws govern all houses and businesses within the municipality; a surveillance program might be aimed at everyone in the jurisdiction. The added advantage of such population-wide programs is that they are likely to have been enacted only after careful consideration. Legislators generally will be reluctant to establish a program that affects influential constituents or large numbers of constituents unless it is aimed at achieving a substantial public-safety objective.

In contrast, a policy that singles out a subgroup that is no more likely than others in the population to pose the particular public-safety or regulatory risk is likely to be both ill-considered and discriminatory. Thus, to the extent that a suspicionless policing policy targets a subgroup of the population, there must be a basis in fact for treating that group differently. As one obvious example, a policy that explicitly authorizes inspections of only Muslim-owned businesses, or one that authorizes sobriety checkpoints only in Black neighborhoods, would be irrational and presumably unconstitutional in virtually all contexts.

But suspicionless policing also can work more subtle discrimination. To take examples from four U.S. Supreme Court cases: a drug-testing policy might explicitly target students engaged in extracurricular activities rather than all students; a DNA-collection program might target arrestees rather than the general population; a parolee-monitoring policy might authorize searching a parolee at any time; and a drug-interdiction program might focus solely on buses. If the rationale for the first policy is to detect or deter drug use in the schools, the rationale for the second policy is to obtain evidence from arrestees that later can be used to identify perpetrators of past crimes, the basis of the third policy is to minimize criminal behavior, and the rationale for the fourth policy is curtailing drug trafficking, it is not clear, at least without more data, that any of the policies justify targeting the groups at which they are aimed. If drug use is fairly consistent throughout the entire student population, or those who are arrested but not convicted are no more likely to have been involved in “cold cases” than the average citizen, or parolees on the street are no more likely than others to be committing a crime, or those who use buses are no more likely than those who use cars to transport drugs, the rationales targeting the particular subgroup of the population lose force.

b. Comparison groups. In applying this Section, agencies and legislatures should consider whether the suspicionless program is either under- or over-inclusive, both of which are of concern. A program is underinclusive if it focuses too narrowly on a particular group to the exclusion of other groups that may be just as likely to be engaged in the sort of wrongdoing that the program is designed to prevent. If a suspicionless program is underinclusive, it may fail to address the harms posed by those who fall outside the target group. It also may undercut agency legitimacy by making people feel that they have been unfairly singled out.

A program is overinclusive if it casts its net too broadly and includes individuals or entities who do not in fact pose the sorts of risks the program is meant to address. An overinclusive suspicionless program wastes agency resources by taking up officer time to conduct unnecessary or unproductive policing actions. It also imposes an unnecessary burden on those who need not have been included among the target group. Moreover, the legitimacy of an entire program may be undermined if it is overinclusive. Although some degree of under- or over-inclusiveness is to be expected in any government endeavor, agencies should strive to minimize those sorts of costs both in designing the program and in evaluating its effectiveness.

Reporters’ Notes

The nondiscrimination principle endorsed in this Section has not been adopted explicitly by the U.S. Supreme Court in the search-and-seizure context, and the outcomes in some cases are inconsistent with it. However, the principle is consistent with the U.S. Supreme Court’s repeated assertion that the government must demonstrate a “special need” to justify regulating conduct through a suspicionless search regime. For example, in Skinner v. Ry. Labor Execs.’ Ass’n, 489 U.S. 602, 620 (1989), the Court emphasized that suspicionless drug testing of railway employees was permissible due to the “safety-sensitive tasks” in which they were engaged, signaling that drug testing of other types of government employees may not be permissible. In Vernonia School Dist. 47J v. Acton, 515 U.S. 646, 661 (1995), the Court permitted drug testing of student athletes in part because “the risk of immediate physical harm to the drug user or those with whom he is playing his sport is particularly high.” Applying this principle, the Eleventh Circuit invalidated a suspicionless drug-testing program of public-assistance recipients under the Temporary Aid for Needy Families (TANF) program because the government had failed to show that drug use was especially prevalent among TANF applicants, and because the rationales that the government offered would apply with equal force to all residents in the state. Lebron v. Sec. of Florida Dep’t of Children & Families, 772 F.3d 1352 (11th Cir. 2014).

The nondiscrimination principle also resonates with, although it is not equivalent to, the Equal Protection Clause’s requirement that the government have a rationale for singling out some individuals or groups as opposed to others. See, e.g., City of Cleburne v. Cleburne Living Ctr., 473 U.S. 432, 439 (1985) (“The Equal Protection Clause of the Fourteenth Amendment commands that no State shall ‘deny to any person within its jurisdiction the equal protection of the laws,’ which is essentially a direction that all persons similarly situated should be treated alike.”); Vill. of Willowbook v. Olech, 528 U.S. 562, 564 (2000) (“Our cases have recognized successful equal protection claims brought by a ‘class of one,’ where the plaintiff alleges that she has been intentionally treated differently from others similarly situated and that there is no rational basis for the difference in treatment.”). It should be unacceptable that some individuals are subjected to government information gathering or encounters while others are not, without a basis in evidence for distinguishing the two. Although courts have not decided whether the decision to search some individuals but not others should be subject to heightened scrutiny, “heightened scrutiny regularly is applied when enumerated rights are at stake.” Barry Friedman & Cynthia Benin Stein, Redefining What’s “Reasonable”, 84 Geo. Wash. L. Rev. 281, 337 (2016); see also Dist. of Columbia v. Heller, 554 U.S. 570, 628 (2008) (heighted scrutiny for statutes that infringe on the right to bear arms); Foucha v. Louisiana, 504 U.S. 71, 84-86 (1992) (heightened scrutiny when classification implicates freedom from physical restraint). At the least, a rational basis for such distinctions is required. Cf. Vernonia, 515 U.S. at 685 (O’Connor, J., dissenting) (noting that “[t]he evidence of a drug-related sports injury problem . . . was considerably weaker” than the evidence indicating drug-related disruption in the classroom).

Finally, this Section reflects “a fundamental norm of administrative procedure [that] requires an agency to treat like cases alike.” Westar Energy, Inc. v. Fed. Energy Reg. Comm’n., 473 F.3d 1239, 1241 (D.C. Cir. 2007). In a variety of administrative contexts, courts have invalidated agency action when the agency applied its standards in arbitrary or discriminatory ways. See, e.g., Green Cnty. Mobilephone v. Fed. Commc’ns Comm’n, 765 F.2d 235, 237 (D.C. Cir. 1985) (“We reverse the Commission not because the strict rule it applied is inherently invalid, but rather because the Commission has invoked the rule inconsistently. We find that the Commissioner has not treated similar cases similarly”); Distrigas of Mass. Corp. v. Fed. Power Comm’n, 517 F.2d 761, 765 (1st Cir. 1975) (“[An administrative agency] has a duty to define and apply its policies in a minimally responsible and evenhanded way.”);Crestline Mem’l Hosp. Ass’n v. Nat’l Lab. Rels. Bd., 668 F.2d 243, 245 (6th Cir. 1982) (“The NLRB cannot ‘treat similar situations in dissimilar ways’”); Contractors Transport Corp. v. United States, 537 F.2d 1160, 1162 (4th Cir. 1976) (“‘[p]atently inconsistent application of agency standards to similar situations lacks rationality’ and is prohibited under the APA’s arbitrary and capricious standard”). See Christopher Slobogin, Policing as Administration, 165 U. Pa. L. Rev. 91, 144-146 (2016).

Although courts have long recognized the importance of the nondiscrimination principle in various settings, they have not always abided by that principle in evaluating suspicionless search-and-seizure programs. The DNA-testing and drug-testing examples described in Comment a are illustrative. In Maryland v. King, there was no evidence to suggest that DNA testing of arrestees was any more likely to solve cold crimes than DNA testing of the general population. As Justice Scalia pointed out in his dissent, because Maryland already collected DNA from individuals upon conviction, the additional DNA samples obtained through the arrestee collection program would all come from those who ended up being acquitted of crime. 569 U.S. 435, 481-482 (2013) (Scalia, J., dissenting). A similar comment can be made about the gratuitous search of a parolee walking down the street with a woman and child, which took place in Samson v. California, 547 U.S. 843 (2006). See id., at 862 (Stevens, J., dissenting) (“the notion that a parolee legitimately expects only so much privacy as a prisoner is utterly without foundation”). Likewise, the drug-testing policy approved in Board of Education v. Earls focused on students involved in extracurricular activities despite the absence of proof (and contrary to the likely probability) that those groups were more heavily involved in drug use than the general student body. 536 U.S. 822, 853 (Ginsburg, J., dissenting) (noting that “[n]ationwide, students who participate in extracurricular activities are significantly less likely to develop substance abuse problems than their less-involved peers.”). And the suspicionless bus-boarding drug interdiction program at issue in Bostick v. Florida, 501 U.S. 429 (1991), was similarly suspect. See United States v. Flowers, 912 F.2d 707, 710 (4th Cir. 1990) (sweep of 100 busses netted seven arrests). Absent some justification for singling out those particular groups, those programs would not be consistent with the nondiscrimination principle. And even if a reason exists, all suspicionless policing activities must comport with the constraints on discretion set out in § 5.05.

Table of Contents