Opinion
CIVIL ACTION NO. 05-11832-GAO
09-28-2012
RONNIE JONES, ET AL., Plaintiffs, v. CITY OF BOSTON, ET AL., Defendants.
OPINION AND ORDER
O'TOOLE, D.J.
This case arises from the use by the Boston Police Department ("BPD") of hair tests to detect possible drug use among its officers. The plaintiffs here all failed the hair test and consequently suffered adverse employment actions. They have brought a number of claims against the BPD and its Commissioner. The parties have filed cross-motions, the defendants for summary judgment, the plaintiffs for partial summary judgment. I resolve the motions as follows.
A. Preemption (Counts I, II, and VI-IX)
The defendants have withdrawn so much of their motion for summary judgment as seeks dismissal of the plaintiffs' state law claims as preempted by the Labor Management Relations Act.
B. Disparate Impact Discrimination (Count I)
The plaintiffs contend that the BPD's use of a hair test to detect evidence of illegal drug use by its officers has a disparate impact on African-American officers and therefore constitutes an unlawful employment practice in violation of 42 U.S.C. § 2000e-2 and Mass. Gen. Laws ch. 151B. The plaintiffs bear an initial burden of making a prima facie showing that the test in fact "causes a disparate impact on the basis of race." § 2000e-2(k)(1)(A)(i). Here the plaintiffs rely on statistical evidence from a proposed expert witness to establish the prima facie showing. The defendants argue that the plaintiffs' statistical evidence is insufficient by itself to establish a prima facie case of disparate impact.
"The legal standards for establishing a violation of Mass. Gen. Laws ch. 151B, § 4 are the same as those for establishing a Title VII claim." DeRosa v. Mass. Bay Commuter Rail Co., 694 F. Supp. 2d 87, 92 n.2 (D. Mass. 2010).
The term "disparate impact" is not a statutorily defined term. See id. § 2000e, passim. The concept itself emerged from the recognition that employment practices that were "fair in form" might nonetheless be "discriminatory in operation." Griggs v. Duke Power Co., 401 U.S. 424, 431 (1971). "[T]he necessary premise of the disparate impact approach is that some employment practices, adopted without a deliberately discriminatory motive, may in operation be functionally equivalent to intentional discrimination." Watson v. Fort Worth Bank & Trust, 487 U.S. 977, 987 (1988).
Statistical evidence can help identify the existence and extent of a disparity. For the plaintiffs to meet the burden of establishing a prima facie case, the statistical disparities they are able to demonstrate "must be sufficiently substantial that they raise an inference" that the challenged employment practice causes a disparate impact on an identified racial group. Id. at 995. "The Supreme Court has said that no single test controls in measuring disparate impact." Langlois v. Abington Hous. Auth., 207 F.3d 43, 50 (1st Cir. 2000) (citing Watson v. Fort Worth Bank and Trust, 487 U.S. 977, 995-96 n.3 (1988) (plurality) ("[W]e believe that such a case-by-case approach properly reflects our recognition that statistics 'come in infinite variety and . . . their usefulness depends on all of the surrounding facts and circumstances.'") (citations omitted)). Rather, "[t]he utility of statistical evidence 'depends on all of the surrounding facts and circumstances.'" EEOC v. Steamship Clerks Union, 48 F.3d 594, 604-605 (1st Cir. 1995) (quoting Int'l Bhd. of Teamsters v. United States, 431 U.S. 324, 340 (1977)).
Although the Supreme Court's statement is preceded by the phrase, "At least at this stage of the law's development," the First Circuit and the District of Massachusetts, in more recent cases, have adopted the Court's holistic case-by-case approach.
As guidance, the Equal Employment Opportunity Commission has suggested what is known as the "Four-Fifths Rule":
"A selection rate for any race, sex, or ethnic group which is less than four-fifths (4/5) (or eighty percent) of the rate for the group with the highest rate will generally be regarded by the Federal enforcement agencies as evidence of adverse impact, while a greater than four-fifths rate will generally not be regarded by Federal enforcement agencies as evidence of adverse impact." 29 C.F.R. § 1607.4(D).
The First Circuit has "approved use of the four-fifths rule as a pertinent benchmark in the employment context." Id. (citing Boston Police Superior Officers Fed'n v. City of Boston, 147 F.3d 13, 21 (1st Cir. 1998)). This benchmark implicitly recognizes that in the real world some differences may inevitably arise and counsels that adverse impact should be found only when the difference is substantial. The Four-Fifths Rule is not literally a legal rule but a "rule of thumb."
In this case, over the range of eight years in which the hair test was required, it is not disputed that white officers passed the test at rates of 99% to 100%, and African Americans passed the test at rates between 97% and 99%. Thus, the passing rate for African Americans was at least 97% of the passing rate for whites. Under the Four-Fifths Rule, the question is not even close; the EEOC would not regard the results to amount to adverse impact. Cf. Smith v. Xerox Co., 196 F.3d 358, 369 (2d Cir. 1999) (overruled on other grounds) ("In some workforces, a disparate impact might well be actionable if older workers were retained at 88.79% of the rate for younger workers, but not if the comparison were 96.69%, especially considering the finding of statistical significance.").
The plaintiffs rely on evidence that differences in the rates at which African Americans failed, rather than passed the hair test, were statistically significant to the extent of between two to four standard deviations. Focusing on failure rates, rather than passing rates, eludes the point of the Four-Fifths Rule, where the EEOC addressed "selection" rates rather than "exclusion" rates. More importantly, the plaintiffs try to give statistical significance a significance it does not have. Statistical significance assesses the question of causation. That is, it assesses the likelihood that an outcome is the result of an identified factor rather than the product of randomness. It is not a measure of impact. By contrast, the Four-Fifths Rule is intended as a measure of impact. To accept the plaintiffs' argument based on statistical significance in failure rates alone, the Four-Fifths Rule must be wholly ignored, because if it is considered, a finding of actionable disparate impact is impossible.
The plaintiffs put misplaced reliance primarily on two cases, Castaneda v. Partida, 430 U.S. 482 (1977), and Ricci v. DeStefano, 557 U.S. 557 (2009). In Castaneda, the Supreme Court noted in a footnote that "as a general rule for such large samples . . . if the difference between the expected value and the observed number is greater than two or three standard deviations, then the hypothesis that the jury drawing was random would be suspect to a social scientist." Id. at 496 n.17. The passage does not mean what the plaintiffs try to make of it. The Court was merely acknowledging the general convention used by social scientists. In its holding the Court then affirmed the finding of a prima facie case of discrimination based on evidence of twelve standard deviations, not two or three.
870 people total over an eleven-year period, whereas in this case, the defendants administered over 15,000 tests over eight years.
The plaintiffs cite Ricci as stating that a prima facie case is "essentially a threshold showing of a significant statistical disparity and nothing more." Ricci, 557 U.S. at 587. Ironically, the Supreme Court never mentioned statistical significance anywhere in its opinion but instead pointed to the four-fifths rule to show that where the impact ratio was approximately 50%, the plaintiffs undisputedly established a prima facie case. See id. at 586-87. Thus, in discussing "statistical disparity" (emphasis added), id. at 586, the Court was addressing the degree of disparate impact, rather than the technical term "statistical significance." In sum, the plaintiffs have no authority for the proposition that a showing of two or more standard deviations in failure rates alone is sufficient to establish a prima facie case.
This is especially true here, where the sample size is so large, because statistical significance is highly sensitive to sample size. See RAMONA L. PAETZOLD & STEVEN L. WILLBORN, THE STATISTICS OF DISCRIMINATION: USING STATISTICAL EVIDENCE IN DISCRIMINATION CASES § 4:13 (2011) ("When the sample size is large enough, extremely small disparities can appear statistically significant. Thus if an employer has engaged in a large enough number of hires, disparities between actual and expected numbers of women or other minorities hired can appear statistically significant even when quite small in absolute terms."). Here, the sample size is quite large, ranging from approximately 1600 to 2000 tests each year.
C f. Apsley v. Boeing Co., 722 F. Supp. 2d 1218, 1239 (D. Kan. 2010) ("In certain instances, a large number of standard deviations simply will not be enough. This case presents one of those instances. It is important to recognize that here the Court is looking at thousands of hiring decisions, not hundreds. Due to this large sample, de minimus disparities may become statistically significant.") (emphasis added)).
In addition, the standard deviations here are relatively modest - plaintiffs' expert's calculations reflect 0.33 to 3.99 standard deviations. Various courts have found that a prima facie case had not been established where there were comparable standard deviations. See, e.g., Waisome v. Port Auth. of N.Y. and N.J., 948 F.2d 1370 (2d Cir. 1991) (2.68 standard deviations); Apsley v. Boeing Co., 722 F. Supp. 2d 1218 (D. Kan. 2010) (4 to greater than 5 standard deviations).
These calculations compare African-American officers to white officers, as well as minority officers to non-minority officers. (Reply Report of Sharon Kelly, Ph.D. Tables R-1, R-2 (dkt. no. 177-114.))
For these reasons, the defendants are correct that the plaintiffs have not satisfied the first step in making out a prima facie case. Accepting the plaintiffs' statistical evidence as reliable, it is nonetheless insufficient in itself to show the necessary disparate impact to support the claim. And its insufficiency is dramatically highlighted by the rest of the pertinent evidence. No reasonable jury could find that the plaintiffs have established a prima facie case of disparate impact.
C. Discrimination on the Basis of Race (Count II)
The plaintiffs also bring a claim of discrimination under 42 U.S.C. § 1981. This claim, however, runs counter to Jett v. Dallas Independent School District, which held that that § 1983 provided the exclusive federal remedy against state actors for a violation of rights declared in § 1981. 491 U.S. 701, 731-33 (1989). That is, § 1981 was merely one law among many capable of giving rise to a § 1983 claim. Contrary to the plaintiffs' suggestions, the Civil Rights Act of 1991 did not supersede Jett. Indeed, the legislative history nowhere purported to supersede Jett but targeted matters entirely different. See H.R. Rep. No. 40(I) (1991), reprinted in 1991 U.S.C.C.A.N. 549; 137 Cong. Rec. S15472-01 (daily ed. Oct. 30, 1991). Furthermore, § 1981 lacks the language regarding causation that allows an individual under § 1983 to bring an action against a municipality. See Monell v. Dep't of Soc. Servs., 436 U.S. 658, 692 (1978); Philippeaux v. N. Cent. Bronx Hosp., 871 F. Supp. 640, 655 (S.D.N.Y. 1994). The language, precedent, and history of § 1981 thus all indicate that the plaintiffs cannot bring here their independent cause of action under that statute.
As for the plaintiffs' state-law claim, their claim under Chapter 151B bars their claim under Chapter 93, section 103. See Agin v. Fed. White Cement, Inc., 632 N.E.2d 1197, 1199 (Mass. 1994) ("[S]hould a judge decide that G.L. c. 151B is or was available to the plaintiff, the plaintiff would have no viable c. 93, § 103, claim under the teaching of Charland.").
D. 42 U.S.C. § 1983: Equal Protection (Count III)
"The central purpose of the Equal Protection Clause of the Fourteenth Amendment is the prevention of official conduct discriminating on the basis of race." Washington v. Davis, 426 U.S. 229, 239 (1976). Accordingly, "[p]roof of racially discriminatory intent or purpose is required to show a violation of the Equal Protection Clause." Vill. of Arlington Heights v. Metro. Hous. Dev. Corp., 429 U.S. 252, 265 (1977). In some circumstances, a facially neutral law "may affect a greater proportion of one race than of another," and "an invidious discriminatory purpose may often be inferred from the totality of the relevant facts, including the fact, if it is true, that the law bears more heavily on one race than another." Davis, 426 U.S. at 242.
In the present case, the raw number of African Americans who actually failed the hair test was exceedingly small. It is especially significant, as discussed above, that the overall pass rate of African-American officers was between 97% and 99%. The pass rate for African Americans was at least 97% of the pass rate for white officers. No discriminatory intent is fairly inferable from these small variances. The plaintiffs have presented no additional evidence that would permit a reasonable jury to find an intent to discriminate in the present case. The question is not a close one. See Frazier v. Garrison, I.S.D., 980 F.2d 1514, 1526 (5th Cir. 1993) ("We have no basis for concluding that an exam with an overall pass rate for African-Americans of 95.58% . . . is evidence of intent to discriminate.").
E. 42 U.S.C. § 1983: Due Process (Count IV)
The plaintiffs argue that there are disputed issues of material fact as to whether they have established a valid procedural due process claim here.
In procedural due process claims, post-deprivation remedies alone can provide sufficient procedural remedies to satisfy the requirements of due process. See Cronin v. Town of Amesbury, 81 F.3d 257, 260 (1st Cir. 1996). The First Circuit has held that the Civil Service Law, Mass. Gen. Laws ch. 31, §§ 41-44, can provide sufficient due process by providing, among other things, the opportunity to file a grievance, to be given a hearing by a member of the Civil Service Commission or a disinterested person, and to have the possibility of reinstatement without loss of compensation. Id. "Although extraordinarily long delays may render a post-deprivation remedy inadequate," id., courts have found such remedies adequate even after years of delay, Decker v. Hillsborough Cnty. Attorney's Office, 845 F.2d 17, 22 (1st Cir. 1988). Adequate post-deprivation remedies were (and are) available to the plaintiffs here.
Besides the standard Civil Service protections, the defendants' own testing procedures had protections in addition. For example, positive drug tests had to be reviewed by a medical review officer before the test would be forwarded to the office of Internal Investigation. Furthermore, the plaintiffs were granted disciplinary hearings where they could present evidence to challenge the results of their drug tests and even had the opportunity to introduce evidence of other drug tests.
It was, of course, no violation of due process to insist that the plaintiffs establish a valid chain of custody to have their independent drug tests admitted and considered as evidence in their hearings. See Fed. R. Evid. 901(a).
In sum, the officers here were the beneficiaries of significant protections of their due process rights both pre- and post-deprivation. In light of these layers of protections, one cannot reasonably conclude that the plaintiffs were deprived of due process here.
F. 42 U.S.C. § 1983: Failure to Train and Supervise (Count V)
The plaintiffs have brought a claim for failure to supervise and train, claiming that the defendants inadequately supervised the administration of the hair tests. This theory, like the other theory in Count V that the defendants intentionally discriminated against the plaintiffs, seems at odds with their primary theory that the test, even when competently administered, had a disparate impact on African-American test subjects. In any event, in the absence of a showing of intentional discrimination, as discussed above, there can be no supervisory liability. See Bd. of Cnty. Comm'rs of Bryan Cnty., Okla. v. Brown, 520 U.S. 397, 407 (1997). Moreover, the record lacks any evidentiary basis for concluding that testing sloppiness that could have been corrected by better supervision and training would have had a greater impact on one ethnic group than another. The plaintiffs' § 1983 claim for failure to train and supervise fails.
G. Disability Discrimination (Count VI)
The plaintiffs bring claims for disability discrimination under state and federal laws protecting disabled individuals against discrimination. The relevant statutes, however, specifically exclude current drug use as a disability or handicap entitled to statutory protection. See 42 U.S.C. § 12114(a); Mass. Gen. Laws ch. 151B, §1(17); see also Mammone v. President & Fellows of Harvard Coll., 847 N.E.2d 276, 291-92 (Mass. 2006). That is precisely what the hair test indicates: whether someone used drugs recently enough that it is still in that person's hair. The plaintiffs were not targeted for testing on the basis of a perception of disability such as drug addiction. They not "targeted" for testing any more than any other officer. Similarly, they were not targeted for discipline because of a perception that they were disabled by addiction, but because they had flunked the test. The plaintiffs are not entitled to relief on this Count.
H. Massachusetts Civil Rights Act (Count VII)
The First Circuit has held that "under Massachusetts law a municipality cannot be sued under the [Massachusetts Civil Rights Act]." Kelley v. LaForce, 288 F.3d 1, 11 n.9 (1st Cir. 2002) (citing Howcroft v. City of Peabody, 747 N.E.2d 729, 744 (Mass. App. Ct. 2001) (concluding that a municipality is not a "person" under the MCRA)). Since that time, the Supreme Judicial Court has cast doubt on the issue. See Am. Lithuanian Naturalization Club, Athol, Mass., Inc. v. Bd. of Health of Athol, 844 N.E.2d 231, 243 (Mass. 2006) ("It is not resolved whether the Civil Rights Act applies to municipalities.").
Even if the defendants could be sued under the Massachusetts Civil Rights Act, the Supreme Judicial Court has held that "universal drug testing program[s]" do not constitute coercion for purposes of the Massachusetts Civil Rights Act where such programs involve "indiscriminate, impartially administered testing." Webster v. Motorola, Inc., 637 N.E.2d 203, 206 (Mass. 1994) (citing Bally v. Northeastern Univ., 532 N.E.2d 49, 52-53 (Mass. 1989)). The hair test was likewise universally and indiscriminately administered. Therefore, the plaintiffs' claim for relief under the Massachusetts Civil Rights Act must fail.
I. Declaratory Judgment (Count VIII) & Permanent Injunction (Count IX)
Declaratory judgment and permanent injunction are inappropriate, as the plaintiffs are not entitled to relief on any Count.
J. Conclusion
Accordingly, the defendants' motion for summary judgment (dkt. no. 154) is GRANTED as to all Counts. The plaintiffs' motion (dkt. no. 160) for partial summary judgment is DENIED. The defendant's motion (dkt. no. 181) to strike is DENIED. Judgment shall enter in the defendants' favor under all Counts of the Amended Complaint (dkt. no. 17).
It is SO ORDERED.
George A. O'Toole, Jr.
United States District Court