William Shea v. John Kerry

USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 1 of 43

United States Court of Appeals FOR THE DISTRICT OF COLUMBIA CIRCUIT

Argued January 20, 2015 Decided August 7, 2015

No. 13-5153

WILLIAM E. SHEA, APPELLANT

v.

JOHN F. KERRY, SECRETARY OF STATE, IN HIS OFFICIAL CAPACITY, APPELLEE

Appeal from the United States District Court for the District of Columbia (No. 1:02-cv-00577)

Joshua P. Thompson argued the cause for appellant. With him on the briefs were Meriem L. Hubbard and Ralph W. Kasarda.

Darrell C. Valdez, Assistant U.S. Attorney, argued the cause for appellee. With him on the brief were Ronald C. Machen Jr., U.S. Attorney, and R. Craig Lawrence, Assistant U.S. Attorney.

Before: ROGERS and SRINIVASAN, Circuit Judges, and WILLIAMS, Senior Circuit Judge. USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 2 of 43

2 Opinion for the Court filed by Circuit Judge SRINIVASAN.

Concurring opinion filed by Senior Circuit Judge WILLIAMS.

SRINIVASAN, Circuit Judge: From 1990 to 1992, the State Department had in place a hiring plan aimed to increase racial diversity among the officer corps in the United States Foreign Service. William Shea, a white Foreign Service Officer, brings suit alleging that the hiring plan violated Title VII. Although Shea challenges a plan that ceased to exist over twenty years ago, he joined the Foreign Service during the two years the plan was in effect. He alleges that, because of the plan, he entered the Foreign Service at a lower level than would have been the case had he been a minority applicant.

The district court viewed Shea’s claim to be controlled by the Supreme Court’s decisions in Johnson v. Transportation Agency, Santa Clara County, California, 480 U.S. 616 (1987), and United Steelworkers of America, AFL-CIO-CLC v. Weber, 443 U.S. 193 (1979). Those decisions upheld employers’ affirmative action plans against Title VII challenges. The district court, following Johnson and Weber, granted summary judgment in favor of the State Department. We agree with the district court and affirm its judgment.

I.

A.

The United States Foreign Service, a branch of the United States Department of State, works through its Foreign Service Officers to “advocate American foreign policy, protect American citizens, and promote American interests USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 3 of 43

3 throughout the world.” Taylor v. Rice, 451 F.3d 898, 900 (D.C. Cir. 2006). Foreign Service Officers “perform traditional diplomatic responsibilities, including trade promotion, political and economic reporting, and consular services and protection.” Id.

In 1990, Shea applied for an entry-level Foreign Service Officer position. At the time, the Foreign Service career ladder consisted of six pay grades, ranging from FS-06 (entry level) to FS-01 (upper level), with the Senior Foreign Service (SFS) a step above FS-01. The Department generally filled vacancies at more senior ranks through internal promotions rather than external hires. Applicants from outside the agency thus ordinarily entered the Officer corps only at the junior levels (FS-04, -05 and -06 levels). In May 1992, Shea joined the Foreign Service at the FS-05 level.

B.

In the years preceding Shea’s application to the Foreign Service, the State Department faced significant scrutiny about the lack of diversity of the Foreign Service Officer corps. In 1985, Congress perceived an underrepresentation of minorities among Foreign Service Officers. Congress therefore enacted legislation directing the Department to “develop . . . a plan designed to increase significantly the number of members of minority groups . . . in the Foreign Service,” with a “particular emphasis on achieving significant increases in the numbers of minority group members . . . in the mid-levels of the Foreign Service,” the FS-02 and -03 levels. Foreign Relations Authorization Act, Fiscal Years 1986 and 1987, Pub. L. No. 99-93, § 152(a), (b), 99 Stat. 405, 428 (1985). USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 4 of 43

4 Two years later, Congress remained unsatisfied. Concluding that the State Department “ha[d] not been successful in [its] efforts . . . to recruit and retain members of minority groups,” Congress instructed the Department to “substantially increase [its] efforts” to ensure that the “Foreign Service becomes truly representative of the American people throughout all levels of the Foreign Service.” Foreign Relations Authorization Act, Fiscal Years 1988 and 1989, Pub. L. No. 100-204, § 183(a), (a)(1), (b)(1), 101 Stat. 1331, 1364 (1987). Congress specifically directed the Department to “ensure that those [efforts] effectively address the need to promote increased numbers of qualified . . . members of minority groups into the senior levels of the Foreign Service.” Id. § 183(b)(2).

Congress did not stand alone in raising concerns about the diversity of the Foreign Service Officer corps. In 1989, the General Accounting Office (now known as the Governmental Accountability Office) released a report entitled “State Department: Minorities and Women Are Underrepresented in the Foreign Service.” The 1989 GAO Report evaluated the Department’s existing efforts, finding that, while “[p]rogress ha[d] been mixed” in increasing diversity,

[m]inorities . . . were still substantially underrepresented when compared with civilian labor force data that the EEOC ha[d] issued to measure federal agencies. . . .

In mid-level ranks of the officer corps, minority male representation ha[d] increased, but minority and white women ha[d] made less progress. In State’s Senior Foreign Service USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 5 of 43

5 positions, underrepresentation of minorities and white women [wa]s still pervasive.

U.S. Gen. Accounting Office, State Department: Minorities and Women Are Underrepresented in the Foreign Service 15 (1989) (1989 GAO Report).

The 1989 GAO Report compared the Department’s 1987 minority workforce with the racial breakdown of the American population possessing the skills required for Foreign Service employment. That comparison indicated that the Department generally fell short of “full representation”— the level at which a minority group would make up the same proportion of the workforce as its proportion of the American population possessing the relevant skills—at mid- and senior- level Foreign Service Officer positions, as follows: for women of each defined minority group at the SFS, FS-01, -02, and -03 levels; black, Native American and native Alaskan men at the SFS level; Hispanic men at the SFS and FS-01 levels; and Asian and Pacific Islander men at the SFS, FS-01, -02, and -03 levels.

The Civil Service Subcommittee of the House Committee on Post Office and Civil Service convened hearings focusing on the 1989 GAO Report’s findings and on the results of two other studies—the Bremer Study Group Report (commissioned by the Secretary of State on his own initiative) and the Thomas Commission Report (mandated by Congress as part of the 1988-1989 Foreign Relations Authorization Act). Representative Gerry Sikorski, the Subcommittee’s Chairman, interpreted those two studies to “disclose[] major problems of discrimination against . . . minorities in the Foreign Service.” Underrepresentation of Women and Minorities in the Foreign Service: Hearing Before the Subcomm. on the Civil Serv. of the H. Comm. on Post Office USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 6 of 43

6 & Civil Serv., 101st Cong. 3 (1989) (1989 Subcomm. Hearing). Those studies, he concluded, revealed that “management of the U.S. Foreign Service [was] seriously flawed.” The Department of State in the 21st Century: Joint Hearing Before the Subcomm. on Int’l Ops. of the H. Comm. on Foreign Affairs & the Subcomm. on the Civil Serv. of the H. Comm. on Post Office & Civil Serv., 101st Cong. 6 (1989) (1989 Joint Hearing).

As of 1989, minorities remained underrepresented in Foreign Service Officer roles. Id. And that was after years of concerns voiced by Congress and repeated warnings from the Equal Employment Opportunity Commission “that the State Department ha[d] not had an effective . . . plan or program for overcoming the underrepresentation [of minorities] in the Foreign Service.” U.S. Gen. Accounting Office, Testimony: Underrepresentation of Minorities and Women in the Foreign Service, Statement of Joseph Kelley, Director of Security and International Relations Issues, National Security and International Affairs Division, Before the Subcommittee on Civil Service, Committee on Post Office and Civil Service, United States House of Representatives 1 (1989). The Department undertook various measures in response, including creating a special hiring path for minorities into the Foreign Service’s mid- and upper-level ranks—the affirmative action plan in issue here.

C.

At the time of Shea’s entry into the Foreign Service, the State Department operated two distinct programs that enabled applicants to bypass the Department’s usual preference for internal promotions and allowed the direct hiring of outside applicants into mid- and upper-level (FS-01, -02 and -03) positions. One program, the Career Candidate Program USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 7 of 43

7 (CCP), was race-neutral. The other program, the 1990-92 Affirmative Action Plan (1990-92 Plan), targeted minority applicants.

Under the CCP, the Department accepted certain applications from outside candidates for FS-01, -02, and -03 positions. But the Department, in accordance with its general preference for filling vacancies through internal promotions, could hire an otherwise viable outside applicant through the CCP only if the Department issued a “certificate of need” attesting that no internal candidates could fill that vacancy. The Department would then consider the outside applicant consistent with its typical hiring procedures. In the absence of a certificate of need, no outside candidate could receive an offer of employment through the CCP.

Under the race-conscious 1990-92 Plan, the Department provided a special path for minorities seeking direct placement as outside hires into the FS-01, -02, and -03 ranks. The 1990-92 Plan gave one—and only one—advantage to minority applicants: an automatic waiver of the CCP’s certificate-of-need requirement for “American Indians, Alaska Native[s], Asians and Pacific Islanders, Blacks, and Hispanics.” U.S. Dep’t of State, Foreign Service Mid-Level Hiring Program Highlights 1 (1989). Apart from the certificate-of-need waiver at the threshold stage, the 1990-92 Plan granted no benefits to minorities in the course of the hiring process. That process was rigorous: The “vast majority” of minority candidates applying through the 1990- 92 Plan “were eliminated from competition at the preliminary review stage.” Id. USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 8 of 43

8 D.

In 2001, Shea filed an administrative grievance with the State Department. Among other claims, he argued that he started at a lower pay grade by virtue of the 1990-92 Plan’s preferential treatment of minority applicants, infringing his rights under Title VII as well as the equal protection component of the Due Process Clause of the Fifth Amendment. The Foreign Service Grievance Board dismissed his complaint for lack of jurisdiction, and Shea then filed suit in the United States District Court for the District of Columbia.

Shea’s case initially traveled back and forth between the district court and this court on the question of whether his Title VII and equal protection claims had been timely filed. (As to the remaining claims, Shea did not appeal their dismissal.) See Shea v. Kerry, 961 F. Supp. 2d 17, 22-25 (D.D.C. 2013). Ultimately, after Congress enacted the Lilly Ledbetter Fair Pay Act of 2009, Pub. L. No. 111-2, 123 Stat. 5, the district court found that Shea’s Title VII claims were timely under the Ledbetter Act but that his equal protection claims were untimely. See id. at 24, 29 & n.3.

Proceeding to the merits, the district court granted summary judgment to the State Department. Id. at 55. The court first determined that the Supreme Court’s Title VII affirmative action decisions in Weber, 443 U.S. 193, and Johnson, 480 U.S. 616, controlled the analysis. Those decisions, the district court explained, called for application of the three-step burden-shifting framework articulated by the Court in McDonnell Douglas Corp. v. Green, 411 U.S. 792 (1973). Shea, 961 F. Supp. 2d at 27-29. USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 9 of 43

9 At the first step, the district court concluded that Shea had established a prima facie case of discrimination in violation of Title VII. Id. at 31-33. Turning to the second step, the court found that the Department had proffered evidence that, if accepted as true, permitted the conclusion that the Department had acted pursuant to a lawful affirmative action plan. Id. at 33-44. Finally, at the third step, the district court considered whether Shea had shown that the affirmative action plan was, in fact, unlawful. The court rejected Shea’s proffer of lay statistical evidence to that end, and thus concluded that he had failed to raise any genuine issue concerning the validity of the Department’s affirmative action plan. The court therefore granted summary judgment in favor of the Department. Id. at 55.

II.

We review de novo the district court’s grant of summary judgment on Shea’s Title VII claim. Holcomb v. Powell, 433 F.3d 889, 895 (D.C. Cir. 2006). Before addressing the merits of that claim, we first assure ourselves of Shea’s standing to bring it. Although the Department raises no challenge to his standing, “it is well established that the court has an independent obligation to assure that standing exists, regardless of whether it is challenged by any of the parties.” Summers v. Earth Island Inst., 555 U.S. 488, 499 (2009).

To demonstrate his standing, Shea must show, inter alia, that he suffered an injury in fact that is both “concrete and particularized” and “actual or imminent, not conjectural or hypothetical.” Lujan v. Defenders of Wildlife, 504 U.S. 555, 560 (1992) (quotation marks omitted). In the context of an employment discrimination claim, a plaintiff may claim an injury in fact from the purported denial of the ability to compete on an equal footing against other candidates for a USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 10 of 43

10 job. See Ne. Fla. Chapter of the Associated Gen. Contractors of Am. v. City of Jacksonville, Fla., 508 U.S. 656, 666 (1993); Cerrato v. S.F. Cmty. Coll. Dist., 26 F.3d 968, 976 (9th Cir. 1994). Because the injury lies in the denial of an equal opportunity to compete, not the denial of the job itself, we do not inquire into the plaintiff’s qualifications (or lack thereof) when assessing standing. See Regents of Univ. of Cal. v. Bakke, 438 U.S. 265, 280-81 & n.14 (1978).

Shea alleges that the 1990-92 Plan denied him the opportunity to compete on an equal basis by extending a preference to minority candidates that was unavailable to him: the ability to gain consideration for entry to a mid-level position without any certificate of need. Shea could have sought direct mid-level placement through the race-neutral CCP program, however. He did not do so, instead applying only for an entry-level FS-05 position. There is thus a question whether Shea suffered an actual or imminent injury as a result of the 1990-92 Plan, or whether his injury was merely hypothetical.

The Supreme Court’s decision in Gratz v. Bollinger, 539 U.S. 244 (2003), found the existence of standing in parallel circumstances. In Gratz, one of the plaintiffs, Patrick Hamacher, sought to challenge the University of Michigan’s consideration of race in its undergraduate transfer admissions. At the time of the suit, Hamacher had yet to apply to transfer to Michigan. Indeed, the Court’s opinion indicated that he would not do so as long as Michigan’s race-conscious admissions program remained in place: Hamacher instead declared that he “intend[ed] to transfer to the University of Michigan when [it] cease[d] the use of race as an admissions preference.” Id. at 261 (emphasis added). The Supreme Court sua sponte questioned Hamacher’s standing to bring his challenge, ultimately concluding that he had shown an injury USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 11 of 43

11 in fact. The Court reasoned that, because of Hamacher’s stated intent to transfer should Michigan change its policy, he had established standing. Id. at 261-62.

Gratz controls our inquiry. Like Hamacher, Shea alleges that he possessed an intent to apply to the position in question, i.e., a mid-level position. Pl.’s Decl. in Supp. of Pl.’s Surreply at 3-4 (filed Dec. 14, 2012). If the mid-levels had been open to him for equal consideration on a race-neutral footing, he would have applied to the mid-levels instead of the entry-level. Thus, like Hamacher, Shea stood “able and ready to apply [to the mid-levels] should the [State Department] cease to use race” as a factor in mid-level hiring. Id. (quotation marks omitted). By choosing not to apply because the Department was considering race during the time of his application process, Shea did exactly what Hamacher alleged he would do: refuse to apply through the race- conscious program unless and until that program’s use of race-conscious preferences ceased. As a result, Shea, like Hamacher, has standing to challenge the Department’s affirmative action plan notwithstanding his failure to apply for a mid-level position through the CCP program.

III.

Title VII prohibits an employer from “discriminat[ing] against any individual with respect to his compensation, terms, conditions, or privileges of employment, because of” inter alia, “such individual’s race.” 42 U.S.C. § 2000e- 2(a)(1). The statute protects both minorities and non- minorities—the latter against “reverse discrimination.” See Mastro v. Potomac Elec. Power Co., 447 F.3d 843, 851 (D.C. Cir. 2006). Here, Shea alleges that the State Department’s 1990-92 Plan constituted impermissible reverse discrimination in violation of Title VII. USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 12 of 43

12

A.

At the outset, we consider the governing framework for resolving Shea’s reverse-discrimination claim. For nearly thirty years, we have examined Title VII challenges to affirmative action programs under the standards set forth by the Supreme Court in United Steelworkers of America, AFL- CIO-CLC v. Weber, 443 U.S. 193 (1979), and Johnson v. Transportation Agency, Santa Clara County, California, 480 U.S. 616 (1987). Shea argues that those standards have been displaced by the Supreme Court’s decision in Ricci v. DeStefano, 557 U.S. 557 (2009), such that Johnson and Weber no longer guide the analysis of reverse-discrimination claims under Title VII. We are unpersuaded.

1.

In Weber, the Supreme Court for the first time considered a Title VII challenge to an employer’s affirmative action plan. As of 1974, Kaiser Aluminum & Chemical Corp. had an “almost exclusively white craftwork force[],” with black employees making up only 1.83% of the company’s skilled craftworkers at its Gramercy, Louisiana, plant, even though the workforce in the area surrounding that plant was roughly 39% black. Weber, 443 U.S. at 198-99. As part of a collective-bargaining agreement, Kaiser promised to implement “an affirmative action plan designed to eliminate [that] conspicuous racial imbalance[].” Id. at 198. The company established job-training programs to teach both black and white employees the necessary skills for promotion to craftworker positions. Id. at 198-99. Selection of trainees for the program would be made on the basis of seniority, but “with the proviso that at least 50% of the new trainees were to be black until the percentage of black skilled craftworkers in USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 13 of 43

13 the Gramercy plant approximated the percentage of blacks in the local labor force.” Id. at 199. A white unskilled production worker from the plant sued, arguing that Title VII prohibited all race-conscious employer actions. Id. at 199, 201.

The Supreme Court disagreed and upheld Kaiser’s affirmative action plan. The Court declined to “define in detail the line of demarcation between permissible and impermissible affirmative action plans,” but concluded that Kaiser’s plan fell “on the permissible side of the line.” Id. at 208. The trainee-selection plan, the Court approvingly noted, aimed to “break down old patterns of racial segregation and hierarchy” and “open employment opportunities for [black workers] in occupations which have been traditionally closed to them.” Id. (quotation marks omitted). The Court set out the considerations that caused it to uphold the company’s plan as follows:

[T]he plan does not unnecessarily trammel the interests of the white employees. The plan does not require the discharge of white workers and their replacement with new black hirees. Nor does the plan create an absolute bar to the advancement of white employees; half of those trained in the program will be white. Moreover, the plan is a temporary measure; it is not intended to maintain racial balance, but simply to eliminate a manifest racial imbalance. Preferential selection of craft trainees at the Gramercy plant will end as soon as the percentage of black skilled craftworkers in the Gramercy plant approximates the percentage of blacks in the local labor force. USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 14 of 43

14 Id. at 208-09 (citation omitted). For those reasons, the plan fell “within the area of discretion left by Title VII to the private sector voluntarily to adopt affirmative action plans designed to eliminate conspicuous racial imbalance in traditionally segregated job categories.” Id. at 209.

2.

Nine years later, in Johnson, the Court again rejected a Title VII challenge to an employer’s affirmative action program. The case arose from the efforts of Santa Clara County, California, to increase diversity in portions of its workforce. The County sought to address a striking gender imbalance in certain positions: Women constituted 36.4% of the labor market in the area, but “none of [the County’s] 238 Skilled Craft Worker positions was held by a woman.” Johnson, 480 U.S. at 621. The County implemented a voluntary affirmative action plan with a stated “long-term goal” to “attain a work force whose composition reflected the proportion of minorities and women in the area labor force.” Id. at 621-22. The County’s plan “authorized the consideration of ethnicity or sex as a factor when evaluating qualified candidates for jobs in which members of such groups were poorly represented,” but it did not set aside a specific number of hiring slots for women or racial minorities. Id. at 622.

In upholding the County’s plan, the Court determined that the analysis should follow the three-step burden-shifting framework set forth in McDonnell Douglas Corp. v. Green, 411 U.S. 792 (1973). Under that framework:

Once a plaintiff establishes a prima facie case that race or sex has been taken into account in an employer’s employment decision [step one], USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 15 of 43

15 the burden shifts to the employer to articulate a nondiscriminatory rationale for its decision [step two]. The existence of an affirmative action plan provides such a rationale. If such a plan is articulated as the basis for the employer’s decision, the burden shifts to the plaintiff to prove that the employer’s justification is pretextual and the plan is invalid [step three].

Johnson, 480 U.S. at 626. Application of that framework, the Johnson Court emphasized, “does not mean . . . that reliance on an affirmative action plan is to be treated as an affirmative defense requiring the employer to carry the burden of proving the validity of the plan. The burden of proving its invalidity remains on the plaintiff.” Id. at 627.

The Johnson Court explained that it would “be guided by [its] decision in Weber.” Id. In Weber, the Court noted, it had blessed an affirmative action plan that (i) sought to “eliminate manifest racial imbalances in traditionally segregated job categories”; and (ii) did not “unnecessarily trammel the interests of white employees.” Id. at 628-30. The Court found the requisite “manifest imbalance” to exist in Johnson in light of the complete absence of women in the positions in question. Id. at 636. The Court further determined that the County’s plan did not “unnecessarily trammel[] the rights of male employees” based on a number of factors (without ascribing weight or rank to any single one). Id. at 637-40. In particular, the plan imposed “goals,” not “quotas.” Id. at 638. The plan worked such that “[n]o persons [were] automatically excluded from consideration; all [were] able to have their qualifications weighed against those of other applicants,” with gender considered only as a “plus.” Id. The plan did not abrogate any “absolute entitlement” of USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 16 of 43

16 male employees, as it operated only in the context of promotions, the denial of which would “unsettle[] no legitimate, firmly rooted expectation[s].” Id. And the plan was temporary, in that it “was intended to attain a balanced work force, not to maintain one.” Id. at 639.

For nearly three decades, Johnson has guided courts— including ours—in the analysis of Title VII claims alleging unlawful reverse discrimination. See, e.g., Hammon v. Barry (Hammon II), 826 F.2d 73 (D.C. Cir. 1987); see also Petitti v. New England Tel. & Tel. Co., 909 F.2d 28 (1st Cir. 1990); Taxman v. Bd. of Educ., 91 F.3d 1547 (3d Cir. 1996) (en banc); Smith v. Va. Commonwealth Univ., 84 F.3d 672 (4th Cir. 1996) (en banc); Edwards v. City of Houston, 37 F.3d 1097 (5th Cir. 1994); Janowiak v. Corporate City of S. Bend, 836 F.2d 1034 (7th Cir. 1984); Tharp v. Iowa Dep’t of Corr., 68 F.3d 223 (8th Cir. 1995); Doe v. Kamehameha Sch./Bernice Pauahi Bishop Estate, 470 F.3d 827 (9th Cir. 2006) (en banc); Cunico v. Pueblo Sch. Dist. No. 60, 917 F.2d 431 (10th Cir. 1990); In re Birmingham Reverse Discrimination Emp’t Litig., 20 F.3d 1525 (11th Cir. 1994).

3.

In 2009, the Supreme Court decided Ricci v. DeStefano, 557 U.S. 557. In Ricci, the Court considered the City of New Haven’s actions in the aftermath of the City’s administration of a firefighter promotional examination. The results of the exam showed a statistical racial disparity: White candidates had outperformed minority candidates. Id. at 562. Some firefighters threatened to bring a discrimination lawsuit if the City relied on the test in making promotions. Id. The City responded by throwing out the test results. A group of white and Hispanic firefighters sued the City under Title VII, USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 17 of 43

17 claiming that, by discarding the test, the City had engaged in unlawful reverse discrimination against them. Id. at 562-63.

The Supreme Court ruled in the firefighters’ favor. The Court understood that the City’s “objective” in discarding the tests was to “avoid[] disparate-impact liability” under Title VII. Id. at 579. But the Court concluded that, by rejecting the results of the promotional test “because of the statistical disparity based on race,” the City had engaged in “express, race-based decisionmaking.” Id. The Court held that “race- based action like the City’s in this case is impermissible under Title VII unless the employer can demonstrate a strong basis in evidence that, had it not taken the action, it would have been liable under” Title VII’s disparate-impact prohibition. Id. at 563; see id. at 585. That is, the Court held that the City could not invalidate the test results based on the race of the highest scorers for the asserted purpose of avoiding a disparate-impact lawsuit, unless the City had a strong basis in evidence to believe that it would be found liable in such a suit. The City could not meet that burden. Id. at 592.

Shea argues that Ricci upends Johnson and Weber such that those earlier decisions no longer guide our analysis here. Under Johnson and Weber, we would first assess the sufficiency of Shea’s prima facie case, then turn to the State Department’s proffer of a valid affirmative action plan, and finally examine Shea’s efforts to demonstrate the invalidity of that plan. See Johnson, 480 U.S. at 626. Throughout, Shea would retain the burden of proving the invalidity of the Department’s 1990-92 Plan. Id. at 627. Ricci changed all of this, Shea submits: After Ricci, Shea argues, we must jettison Johnson and Weber’s framework and instead ask whether the State Department can show “a strong basis in evidence that, had it not [instituted an affirmative action plan], it would have been liable” for discrimination under Title VII. Ricci, 557 USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 18 of 43

18 U.S. at 563 (emphasis added). And, if the Department proves unable to put forth the requisite “strong basis in evidence” in support of that showing, Shea contends, the Department would be liable under Title VII for impermissible reverse discrimination.

The Department initially argues that Shea forfeited any argument based on Ricci by failing to present that argument to the district court. We disagree. Although forfeiture principles apply to new arguments raised for the first time on appeal, see Potter v. District of Columbia, 558 F.3d 542, 547 (D.C. Cir. 2009), Shea’s argument has been consistent throughout the litigation: The Department’s 1990-92 Plan impermissibly discriminated against him in violation of Title VII. On appeal, Shea enjoys a measure of latitude to elaborate on his theory in service of the same argument. His reliance on Ricci for the first time on appeal lies within that latitude. Moreover, although Shea did not press a Ricci-based argument before the district court, the district court invoked Ricci on its own, observing that “nothing in Ricci directly overturns or modifies Johnson, at least as it applies to this case.” Shea, 961 F. Supp. 2d at 54 n.17. Shea is permitted to respond on appeal by explaining why he thinks Ricci governs this case. See Citizens United v. Fed. Election Comm’n, 558 U.S. 310, 330 (2010).

Shea’s argument based on Ricci fails on the merits, however. Johnson and Weber are directly applicable to this case. They set out the framework for “evaluating the compliance of an affirmative action plan with Title VII’s prohibition on discrimination,” Johnson, 480 U.S. at 640, the precise question in issue here. Those decisions unquestionably would control our analysis unless a subsequent decision dictates otherwise. Ricci is not such a decision. In reaching that conclusion, we draw guidance from USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 19 of 43

19 the Supreme Court’s admonition against concluding that its “more recent cases have, by implication, overruled an earlier precedent.” Agostini v. Felton, 521 U.S. 203, 237 (1997). Rather, if “a precedent of [the Supreme] Court has direct application in a case, yet appears to rest on reasons rejected in some other line of decisions, the Court of Appeals should follow the case which directly controls, leaving to [the] Court the prerogative of overruling its own decisions.” Id.

Here, Johnson and Weber have “direct application,” and we have no occasion or cause to conclude that Ricci, “by implication,” overruled those decisions. Id. Indeed, Ricci does not mention or even cite—much less discuss—Johnson and Weber. That is understandable, as Ricci, by its own description, addressed a particular situation not in issue here. Cf. Ricci, 557 U.S. at 626 (Ginsburg, J., dissenting) (“[Ricci] does not involve affirmative action.”). In Ricci, the Court’s “analysis beg[an] with this premise: The City’s actions would violate the disparate-treatment prohibition of Title VII absent some valid defense.” Id. at 579. The inquiry prescribed by Johnson and Weber, by contrast, pertains to assessing whether there is a violation of Title VII’s disparate-treatment prohibition in the first place, the same question we address here.

The specific question addressed in Ricci was whether, even though the City’s action in discarding the test results was assumed to violate Title VII’s disparate-treatment prohibition, that action could be justified based on a particular objective asserted by the City: avoiding liability in a Title VII disparate- impact lawsuit. The Court expressly framed its holding by reference to actions taken for that particular purpose:

We hold only that, under Title VII, before an employer can engage in intentional USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 20 of 43

20 discrimination for the asserted purpose of avoiding or remedying an unintentional disparate impact, the employer must have a strong basis in evidence to believe it will be subject to disparate-impact liability if it fails to take the race-conscious, discriminatory action.

Id; see id. at 580 (“We consider, therefore, whether the purpose to avoid disparate-impact liability excuses what otherwise would be prohibited disparate-treatment discrimination.”) (emphasis added).

The employers in Johnson and Weber did not modify the outcomes of personnel processes for the asserted purpose of avoiding disparate-impact liability under Title VII. Nor did the State Department here. The Department, like the employers in Johnson and Weber, instead acted to “expand[] job opportunities for minorities and women,” Johnson, 480 U.S. at 622, and to “eliminate traditional patterns of racial segregation,” Weber, 443 U.S. at 201; see id. at 209 & n.9. Ricci does not purport to reach the Department’s actions in pursuit of those purposes. Weber and Johnson therefore still control. The only other court of appeals of which we are aware to have addressed the interaction between Ricci and the Johnson-Weber framework reached the same conclusion. See United States v. Brennan, 650 F.3d 65, 102-04 (2d Cir. 2011).

IV.

Under the framework established by Johnson and Weber, we ask first if Shea establishes a prima facie case of discrimination. Second, we examine whether the State Department can articulate a nondiscriminatory reason—in this case, a valid affirmative action plan—for its actions. Finally, we assess whether Shea carries his burden to prove that the USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 21 of 43

21 Department’s plan is invalid. The district court found that Shea and the Department made the requisite showings at the first and second steps, respectively. The court then found Shea to falter at the third step and therefore granted summary judgment in favor of the Department. We agree at each step.

A.

We first address whether Shea has made out a prima facie case of reverse discrimination in violation of Title VII. At the outset, we note that neither party has addressed the potential implications of our decision in Brady v. Office of Sergeant at Arms, 520 F.3d 490 (D.C. Cir. 2008), for Johnson’s direction to assess whether the “plaintiff establishes a prima facie case,” Johnson, 480 U.S. at 626. Brady explained that, when “an employee has suffered an adverse employment action and an employer has asserted a legitimate, non-discriminatory reason for the decision, the district court need not—and should not—decide whether the plaintiff actually made out a prima facie case.” 520 F.3d at 494. Rather, “the district court must resolve one central question: Has the employee produced sufficient evidence for a reasonable jury to find that the employer’s asserted non-discriminatory reason was not the actual reason and that the employer intentionally discriminated against the employee on the basis of” a prohibited characteristic? Id. We have since invoked Brady in the context of a reverse-discrimination claim. See Ginger v. District of Columbia, 527 F.3d 1340, 1344 (D.C. Cir. 2008).

We have not, however, specifically applied Brady in the context of a reverse-discrimination suit challenging the validity of an employer’s affirmative action plan under Title VII. In that domain, Johnson has long set forth the governing approach. Because no party on appeal argues that Brady USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 22 of 43

22 should alter that framework, and because the existence of a prima facie case is readily resolved in this case in Shea’s favor, we leave for another day the resolution of the interaction between Brady and Johnson. We therefore proceed on the assumption that Johnson’s framework— including its call for examining the establishment of a prima facie case—is controlling for our purposes.

Here, the State Department contests Shea’s establishment of a prima facie case in only one respect. As part of the showing necessary to make out a prima facie case of discrimination (or reverse discrimination) in violation of Title VII, a plaintiff must establish that he has been subjected to an adverse employment action. George v. Leavitt, 407 F.3d 405, 412 (D.C. Cir. 2005). The Department argues on appeal, for the first time in this case’s long history, that Shea suffered no adverse employment action from his hiring at an entry-level (rather than mid-level) position because he never applied for direct mid-level placement, either through the 1990-92 Plan or through the race-neutral CCP. We do not reach the merits of that argument because the Department forfeited it by failing to raise it until this late stage.

Although “we may affirm a judgment on any ground that the record supports and that the opposing party had a fair opportunity to address,” Jones v. Bernanke, 557 F.3d 670, 676 (D.C. Cir. 2009) (internal citation and quotation marks omitted), an argument “never made below is waived on appeal,” id. (citing Marymount Hosp., Inc. v. Shalala, 19 F.3d 658 (D.C. Cir. 1994)). The Department at no point in the previous fourteen years of litigating this case contended that Shea’s failure to apply for a mid-level position could affect his establishment of a prima facie case. It has instead fought Shea’s prima facie showing on other grounds. “[A]bsent exceptional circumstances not present here, it is not our USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 23 of 43

23 practice to entertain issues first raised on appeal.” Marymount Hosp., 19 F.3d at 663 (quoting Roosevelt v. E.I. Du Pont de Nemours & Co., 958 F.2d 416, 419 & n.5 (D.C. Cir. 1992)) (quotation marks omitted). We adhere to that practice today. Because the Department has forfeited any argument that Shea suffered no adverse employment action, and because the Department otherwise does not challenge his establishment of a prima facie case on appeal, we agree with the district court that Shea has made that showing.

B.

At the second step of Johnson’s framework, the Department must “articulate a nondiscriminatory rationale for its decision.” Johnson, 480 U.S. at 626. Johnson observes that “[t]he existence of an affirmative action plan provides such a rationale.” Id. We do not understand Johnson to mean, however, that an employer establishes a legitimate, nondiscriminatory reason for its decision merely by showing that it acted pursuant to an affirmative action plan. See Hill v. Ross, 183 F.3d 586, 590 (7th Cir. 1999).

Rather, the Johnson framework maps onto McDonnell Douglas’s three steps. Johnson, 480 U.S. at 626-27. To satisfy its burden of production at the second McDonnell Douglas step, the State Department must “introduce evidence which, taken as true, would permit the conclusion that there was a nondiscriminatory reason for” its actions. St. Mary’s Honor Ctr. v. Hicks, 509 U.S. 502, 509 (1993) (emphasis omitted). And while a valid affirmative action plan is considered nondiscriminatory, see Parker v. Balt. & Ohio R.R. Co., 652 F.2d 1012, 1017 n.9 (D.C. Cir. 1981), an invalid affirmative action plan is discriminatory, see Taxman, 91 F.3d at 1567. As a result, the Department needs to produce “evidence which, taken as true, would permit the conclusion” USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 24 of 43

24 that it acted for a “nondiscriminatory reason,” i.e., pursuant to a valid affirmative action plan. Hicks, 509 U.S. at 509 (emphasis omitted); see Shea, 961 F. Supp. 2d at 33-34.

We have explained that, under Johnson and Weber, a valid affirmative action plan should satisfy two general conditions. First, a valid plan rests on an adequate factual predicate justifying its adoption, such as a “manifest imbalance” in a “traditionally segregated job categor[y].” Johnson, 480 U.S. at 631; see Hammon II, 826 F.2d at 74-75. Second, a valid plan refrains from “unnecessarily trammel[ing] the rights of [white] employees.” Johnson, 480 U.S. at 637-38; see Hammon II, 826 F.2d at 81. We take up those considerations in order.

1.

The district court concluded that the Department adequately grounded its 1990-92 Plan in evidence of a manifest imbalance in a traditionally segregated job category. See Shea, 961 F. Supp. 2d at 34-39. We agree.

a.

Ascertaining the existence of a “manifest imbalance” is a “fact-specific task” in a “sensitive and delicate area.” Hammon II, 826 F.2d at 75. One method that may be used to demonstrate such an imbalance—and the one relied on by the State Department here—entails a showing of statistical disparities between the racial makeup of the employer’s workforce and that of a “comparator population.” If the positions in question “require no special expertise,” the comparator population would be “‘the area labor market or general population.’” Id. (citation omitted) (quoting Johnson, 480 U.S. at 632). But for “jobs that require special training,” USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 25 of 43

25 the ‘“comparison should be with those in the labor force who possess the relevant qualifications.’” Id. (citation omitted) (quoting Johnson, 480 U.S. at 632).

When the Department adopted the 1990-92 Plan, the agency had before it two analyses comparing its own workforce with the labor pool possessing the relevant qualifications: (i) the 1989 GAO Report, and (ii) a formal analysis conducted by the Department itself when promulgating the 1990-92 Plan. The Department points to those two statistical studies as its principal evidence of a manifest imbalance between minority representation in the Foreign Service and the comparator population.

Shea contends that the Department cannot rely on either of those studies. He argues that the 1990-92 Plan amounted only to a continuation of a preexisting affirmative action plan in place from 1987-89, and that the 1990-92 Plan thus was actually adopted in 1987. Shea submits that any data on which the Department purports to justify any affirmative action plan must have been in its possession when it promulgated the plan—which, by Shea’s account, would have been in 1987, before either the 1989 GAO Report or the 1990- 92 Plan’s analysis. Consequently, Shea argues, the State Department is foreclosed from invoking either study as a justification for its actions.

We assume arguendo the correctness of Shea’s premise that the Department cannot justify its race-conscious actions by reference to post hoc data collection. Even so, Shea errs in contending that the Department cannot rely on the 1989 GAO Report or the findings contained in the 1990-92 Plan to justify the Plan. The district court concluded that the Department’s 1987-89 affirmative action efforts and the 1990-92 Plan in fact were two different plans. See Shea, 961 F. Supp. 2d at USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 26 of 43

26 30. We would tend to agree. But even if otherwise, the 1990- 92 Plan at the very least amounted to a review and overhaul of the Department’s affirmative action efforts. See, e.g., U.S. Dep’t of State, Multi-Year Plan, FY 1990-92 at 51 (rev. version Apr. 30, 1991) (1990-92 Plan Document) (assembling new diversity statistics for purposes of the 1990-92 Plan); id. at 61 (noting that the 1987-89 plan “has been refined . . . to define better the type of candidate to be recruited” (emphasis added)). “When a program that has been reauthorized is challenged, all evidence available to the [decisionmaker] prior to reauthorization must be considered in assessing” the program’s legality. Rothe Dev. Corp. v. U.S. Dep’t of Def., 262 F.3d 1306, 1328 (Fed. Cir. 2001). Accordingly, even if the 1990-92 Plan amounted to a reauthorization of the Department’s 1987-89 affirmative action efforts, and even if the Department cannot justify its actions based on post hoc data, the 1989 GAO Report and the findings contained in the 1990-92 Plan are a proper evidentiary proffer.

The version of the 1990-92 Plan in our record contains the Department’s employment data from 1989 and 1990. As the 1989 data represent the data in the State Department’s possession both at the time it promulgated the 1990-92 Plan and at the time Shea applied to the Foreign Service, we use that data (though we note that the minor differences between the 1989 and 1990 data would have no impact on our conclusions today). The 1990-92 Plan’s findings showed improvement in the Foreign Service’s diversity from the time of the 1989 GAO Report. The combined FS-02 and -03 levels, for instance, showed underrepresentation only for Native Americans and Alaskans. See 1990-92 Plan Document at 46a; 47a. For other minority populations at the combined FS-02 and -03 positions, there were no imbalances, manifest or otherwise. As a result, Shea contends, the Department cannot justify the 1990-92 Plan by claiming that USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 27 of 43

27 it addressed manifest imbalances for all minority groups at those levels.

The Department initially asserts that the 1990-92 Plan established “goals” only for groups specifically shown in the data to be underrepresented. We take this to mean, for example, that, at the combined FS-02 and -03 levels, “goals” would have been set only for Native Americans and Alaskans, and not for other minority populations. If the “goals” operated such that only members of the underrepresented minority groups received favorable treatment in the application process relative to Shea, the Department’s argument would have force. But the Department provides no information about how the “goals” would have worked in practice. And we find no description in the record. All that we can glean from the record is that all minority applicants received the main benefit available under the 1990-92 Plan— waiver of the certificate-of-need requirement for entry into the FS-01, -02, and -03 levels.

The Department’s defense of the 1990-92 Plan stands on stronger footing, however, with regard to more senior-level positions. Looking up the ranks from the FS-02 and -03 levels, the Department identified a more across-the-board manifest imbalance. The Department first points to the FS-01 level. According to the 1990-92 Plan data, all minority groups were underrepresented at the FS-01 level at the time of the plan’s promulgation. To achieve full representation, the number of black Officers at that level would have needed to increase by 62%, Hispanics by 14%, Native Americans and Alaskans by 256%, and Asians and Pacific Islanders by 47%. See 1990-92 Plan Document at 46a; 47a.

The Department also points to the ranks of the SFS. The 1990-92 Plan, so far as we can tell, contains no specific data USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 28 of 43

28 on diversity in the SFS. Accordingly, we look to the SFS findings from the 1989 GAO Report. Those findings show underrepresentation of all minority groups at the SFS level. And the imbalances are manifest: To achieve full representation, the number of black Officers in the SFS would have needed to increase by 154%, Hispanics by 163%, Asians and Pacific Islanders by 700%, and, for Native Americans and Alaskans, by an undefined percentage (because the Foreign Service had no SFS Officer of Native American or Alaskan origin). See 1989 GAO Report at 17.

b.

Johnson speaks in terms not just of any manifest imbalance, but of a manifest imbalance in a traditionally segregated job category. 480 U.S. at 631. As the Court explained, the “requirement that the manifest imbalance relate to a traditionally segregated job category provides assurance” that “race will be taken into account in a manner consistent with Title VII’s purpose of eliminating the effects of employment discrimination.” Id. at 632 (quotation marks omitted). That approach guards against licensing an employer to seek proportional representation purely for its own sake. The Department must make a showing that, if taken as true, would permit the conclusion that the manifest imbalance resulted from a “predicate of discrimination” rather than from benign forces. Hammon II, 826 F.2d at 74-75, 80-81. We find that the Department has done so.

First, the substantial imbalances at the SFS level themselves indicate that discriminatory practices may well have been afoot. While a significant disparity is not itself dispositive, “ranks [that are] overwhelming[ly] white” are “a powerful present-day demonstration of a prior regime of discrimination.” Hammon v. Barry (Hammon I), 813 F.2d USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 29 of 43

29 412, 427 (D.C. Cir. 1987) (referring to Weber). Here, the disparity between white and non-white SFS Officers qualifies as overwhelming. Of the 655 serving SFS Officers counted by the 1989 GAO Report, 631 were white. See 1989 GAO Report at 17. “[F]ine tuning of the[se] statistics could not have obscured the glaring absence of minority” officers. Int’l Bhd. of Teamsters v. United States, 431 U.S. 324, 342 n.23 (1977).

Second, testimony before Congress concerning the 1989 GAO Report, the Bremer Study Group Report, and the Thomas Commission Report provided Congress with evidence of pervasive historical discrimination in the Foreign Service tracing as far back as the 1960s. For example, one witness testified that, when he “entered on duty in State in late 1965 . . . [he] had [his] first experience with discrimination, aside from what [he] had experienced while stationed in the South,” and further noted that “[m]inorities have been underrepresented purposefully” within the Department. 1989 Subcomm. Hearing at 33, 39. Another witness, focusing on gender discrimination at the Department, explained that, while “[i]nstances of blatant sexism and discrimination have declined, . . . some do still take place,” and “a simple glance at the statistics contained in the [1989 GAO Report] . . . will confirm the continued existence of a problem.” Id. at 42. The former EEOC Commissioner offered his assessment that “[t]he State Department wants to hire what I call the mythical American, the 5’10,” 160 pound WASP man in perfect physical and mental health.” J.A. 342. And the Department’s Deputy Assistant for Equal Opportunity and Civil Rights provided testimony that, around the world, he had “encountered complaints of discrimination from [State’s] employees and criticisms from foreigners for that same discrimination as exhibited by our predominantly white male diplomatic corps.” Id. at 13. USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 30 of 43

30

To the House Subcommittee on the Civil Service, such testimony likely came as no surprise. While the record before us does not contain the underlying materials, Representative Sikorski, the Subcommittee Chairman, stated his belief that those reports confirmed a State Department inadequately concerned with diversity. Previous investigations and hearings by his subcommittee, he stated, “documented serious instances of discriminatory treatment by the Foreign Service of women, minorities, and people with handicaps.” 1989 Joint Hearing at 10. His testimony included the revelation that “more than 240 Equal Employment Opportunity (EEO) cases” had been filed and indications that previous efforts to diversify the Foreign Service championed by the Secretary of State were “largely ignored by the Department.” Id. Moreover, he noted, the Department had “been repeatedly cited by the [EEOC] for submitting deficient [diversity] reports.” Id. The 1989 GAO Report noted that, despite repeated criticism including suggestions of bias, the State Department never “conducted analyses of possible impediments to equal employment opportunity.” 1989 GAO Report at 4.

This case is therefore a far cry from our decisions in Hammon I and II, in which we determined that the District of Columbia had failed to demonstrate the predicate of discrimination necessary to justify an affirmative action program for its hiring of firefighters. The challenged plan purportedly addressed the District’s history of discriminatory hiring against black applicants. But during the relevant historical period, blacks made up an average of 41.8% of the firefighters hired each year, Hammon I, 813 F.2d at 427, and, at the time of the challenge, 37% of the firefighting workforce overall, Hammon II, 826 F.2d at 77. The proper comparator pool was 29.3% black. Hammon I, 813 F.2d at 428. In light USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 31 of 43

31 of those figures, the District “steadfastly and persuasively protested its innocence of any discriminatory activity,” and we agreed. Id. at 427.

Here, by contrast, evidence identified by the Department would permit the conclusion that there had been a past practice of discrimination with continuing effects through the early 1990s. We therefore agree with the district court that the Department made an adequate evidentiary proffer that the 1990-92 Plan “served to remedy the lingering effects of State’s past discrimination.” Shea, 961 F. Supp. 2d at 39.

2.

Having shown the necessary factual predicate for the 1990-92 Plan in the form of a manifest imbalance in a traditionally segregated job category, the Department faces one additional requirement: The plan must not have unnecessarily trammeled the rights of white applicants. Johnson, 480 U.S. at 637-38; Hammon II, 826 F.2d at 81. We, like the district court, conclude that the Department has made an adequate showing in this regard.

a.

There is “no precise formula for determining whether an affirmative action plan unnecessarily trammels the rights of non-beneficiaries.” In re Birmingham Reverse Discrimination Emp’t Litig., 20 F.3d at 1541. Rather, a number of considerations inform the inquiry. See Johnson, 480 U.S. at 637-40; Weber, 443 U.S. at 208-09; Hammon II, 826 F.2d at 81. Those considerations weigh in favor of the 1990-92 Plan’s validity. USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 32 of 43

32 First, the type of affirmative action plan matters. Affirmative action in hiring generally poses less of a concern than affirmative action in layoffs. See Johnson, 480 U.S. at 638. Hiring decisions upset settled expectations to a lesser degree (because an applicant has no absolute entitlement to a job), and they affect a more diffuse group (all potential applicants) than do layoffs, which target specific employees. See United States v. Paradise, 480 U.S. 149, 183 (1987) (plurality opinion) (“Denial of a future employment opportunity . . . is not as intrusive as loss of an existing job.”) (quotation marks omitted); cf. Firefighters Local Union No. 1784 v. Stotts, 467 U.S. 561, 578-79 (1984). Here, the 1990- 92 Plan awarded benefits to minority candidates only in the hiring process, and even then, only at the very initial stage.

Second, the degree of benefit, or “plus,” bestowed by the affirmative action plan can make a difference. Affirmative action resulting in the hiring only of qualified candidates more easily survives scrutiny than affirmative action resulting in the hiring of unqualified beneficiaries. See Johnson, 480 U.S. at 637-38. In this case, the Department’s 1990-92 Plan provided for hiring only of qualified candidates: Minority applicants considered through the 1990-92 Plan underwent the same rigorous application path as did white candidates considered through the race-neutral CCP, with the only difference coming in the form of the certificate-of-need waiver at the threshold.

Third, the goals of the affirmative action plan affect the inquiry. A plan that seeks to achieve full representation for the particular purpose of remedying past discrimination will generally be shorter in duration than one that pursues proportional diversity for its own sake. When a plan pursues only the former goal, it presumably would cease to operate once full representation is achieved. And the shorter the time USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 33 of 43

33 period for which a plan is in operation, the less it could be said adversely to affect non-beneficiaries. In Weber, for instance, the Court approvingly observed that the plan it upheld was “not intended to maintain racial balance, but simply to eliminate a manifest racial imbalance.” 443 U.S. at 208; see Johnson, 480 U.S. at 639-40. Here, the 1990-92 Plan sought to attain more proportional representation, not to maintain it in perpetuity. Indeed, the 1990-92 Plan ceased to operate in 1993 and has not been replaced. Shea, 961 F. Supp. 2d at 41-42.

Fourth, the extent to which the challenged plan limits opportunities for advancement by non-beneficiaries is a relevant consideration. In both Johnson and Weber, the Court observed that the plan in question created no “absolute bar” to the advancement of non-beneficiaries. Johnson, 480 U.S. at 637-38; Weber, 443 U.S. at 208. Here, Shea makes no argument that the 1990-92 Plan engendered any “absolute bar” to the advancement of non-minorities in the Foreign Service ranks. Non-minority candidates from outside the agency could apply directly to the mid-level ranks through the race-neutral CCP, and internal white candidates could—and did—gain promotion to mid-level positions from the Foreign Service entry-level ranks.

b.

Our court has understood the need to avoid “unnecessarily” trammeling the rights of non-minority candidates to indicate that a challenged affirmative action plan generally must be “tailored to fit the violation” sought to be addressed. Hammon II, 826 F.2d at 74; see id. at 81. Here, the 1990-92 Plan granted a certificate-of-need waiver to candidates applying to the FS-01, -02, and -03 levels. The Department’s identified manifest imbalances, however, USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 34 of 43

34 occurred at only the more senior levels. Why bestow benefits at the FS-02 and -03 levels if the manifest imbalances sought to be addressed existed only at more senior positions? To do so, Shea contends, means that the 1990-92 Plan was so over- inclusive as to unnecessarily trammel the rights of white applicants at the FS-02 and -03 levels.

The Department submits that there is a sound explanation for targeting the FS-02 and -03 levels to address an imbalance at more senior levels. The 1990-92 Plan satisfies the tailoring requirement, the Department explains, because the FS-02 and -03 levels serve as the training grounds for learning the skills necessary to perform at the SFS and FS-01 levels. We agree.

The plan upheld in Weber is instructive. The employer in Weber aimed to remedy the manifest imbalance in its ranks of skilled workers: a mere 1.83% of its skilled workers were black, while the labor force in the surrounding area was 39% black. 443 U.S. at 198-99. To address the identified imbalance in its skilled workforce, however, the employer could not simply hire laborers lacking the requisite skills. Rather, it needed to hire laborers after they had acquired those skills. The employer established a training program to tackle that problem, stipulating that 50% of all employees entering the training program would be black until the percentage of black skilled workers in its workforce approximated the percentage in the local labor force. See id. at 199.

Weber thus provides an example of an affirmative action plan going beyond strictly proportional representation in a training program: 50% of the spots would go to the company’s black workers, even though black persons made up only 39% of the area labor force. Employees who had completed Weber’s training program could then proceed to the rank of skilled worker, where the manifest imbalance USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 35 of 43

35 existed. The need to create an adequate pipeline of trained workers meant that the program was sufficiently tailored to target the “manifest imbalance” among skilled workers.

The State Department’s 1990-92 Plan worked similarly. In order to attain full representation at the SFS and FS-01 levels, the Department maintains, it had to go beyond strictly proportional minority representation at the FS-02 and -03 levels. It could then choose from qualified minority candidates at those levels to staff its SFS and FS-01 ranks.

With regard to the SFS, the Department’s hiring regulations in place at the time of the 1990-92 Plan demonstrate that the Department valued a certain set of skills in its SFS Officers and believed that the best way for SFS candidates to gain those experiences was through service in the mid-level Foreign Service ranks. The regulations provided that career SFS Officers “normally shall be appointed as the result of promotion of Mid-Level career officers,” and generally limited the SFS to a maximum of five percent external hires at any given time. Appointment of Members of the Foreign Service, 48 Fed. Reg. 38,606, 38,607 (Aug. 25, 1983). Additionally, career SFS applicants generally had to have completed at least five years of service in a position “of responsibility . . . equivalent to that of a Mid- Level Foreign Service officer (classes FS-1 through FS-3),” with “duties and responsibilities . . . similar to or closely related to that of a Foreign Service officer in terms of knowledge, skills, abilities, and overseas work experience.” Id. The difficulties encountered by those directly promoted to a mid- or high-level position in the Foreign Service, which usually included a “prolonged adjustment period” and experiencing “a competitive disadvantage,” further suggest that percolating through the ranks was, generally, a sounder career path. USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 36 of 43

36

The conclusion from the then-existing regulations is straightforward: The Department believed that the best training for the role of a SFS Officer was experience as an FS- 01, -02, or -03 Foreign Service Officer. The Department similarly valued skills gleaned from experience at the FS-02 and -03 ranks for the position of an FS-01, with outside hires into the FS-01 ranks serving as the small exception to the Department’s general internal promotion ladder. See J.A. 343, 516-17. Shea has introduced no evidence contradicting that understanding.

In view of the Department’s assessment that the most qualified candidates for the SFS and FS-01 ranks would come from its own mid-levels, the Department understandably saw a need to go further than strictly proportional representation in its mid-levels. That was necessary, the Department reasonably concluded, in order to have a sufficient reservoir of talented minority candidates from which to hire in order to achieve diversity in its SFS and FS-01 ranks. Otherwise, assuming that promotion rates were the same across races from the mid-levels to the SFS and FS-01 levels, the Department would need to await a great deal of turnover in the overwhelmingly white SFS and FS-01 ranks before the substantial imbalances at those levels would be rectified.

Congressional testimony on the 1989 GAO Report reveals that very concern. Joseph Kelley of the General Accounting Office, in response to questioning about when “the State Department [would] become representative of the American people,” told Congress that “[i]t is going to take a long time,” and noted that the EEOC had been pushing the Department “to have a program to move people around and to have upper-level promotions, but it ha[d]n’t worked out that well.” 1989 Subcomm. Hearing at 29-30. The Department USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 37 of 43

37 required a method by which to augment the flow of minority candidates to the SFS and FS-01 levels. As Representative Sikorski observed, “if the numbers [only] get[] better in . . . entry level and hiring,” then “there is no upward progress. There is no flow in the right direction. We are talking centuries.” Id. at 29-30.

It is no answer to claim that the Department could simply promote minorities to the SFS and FS-01 levels at higher rates than their non-minority peers. That itself would have been a race-conscious action requiring justification. That option, at any rate, appears to have been non-viable. Testimony before Congress indicated that promotions of minorities to high-level positions were already happening “too fast,” such that the Department began “to get a backlash” that promotions were “not [of] qualified . . . minorities” and that those promoted were “not really ready to make this jump.” Id. at 47. Title VII does not require the Department to promote unqualified candidates to execute the important mission of our diplomatic corps. For those reasons, the 1990-92 Plan’s emphasis on hiring at mid-level positions was adequately tailored to address manifest imbalances at the senior levels.

At its root, finally, the unnecessary trammeling inquiry amounts to an exercise in balancing a plan’s attempts to remedy past discrimination against the plan’s adverse impact on the rights of non-minorities. In this case, the latter impact was unquestionably limited. The 1989 GAO Report indicates that the State Department had 655 SFS Officers, 836 FS-01 Officers, and 2,032 FS-02 or -03 Officers. 1989 GAO Report at 17. Against that backdrop, the Department informs us that only sixteen minority candidates were hired into the mid- levels through the 1990-92 Plan over the three calendar years of its operation. With such a modest effect on the hiring process, the 1990-92 Plan was necessarily limited in the USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 38 of 43

38 extent to which it could “trammel” Shea’s rights, “unnecessarily” or otherwise.

c.

The tailoring inquiry, according to our decisions, also takes into account whether the employer considered race- neutral alternatives. See Hammon II, 826 F.2d at 81. While the program we considered in Hammon failed to pass muster because “reasonable alternatives were not seriously discussed,” Hammon I, 813 F.2d at 430, the district court in this case found the Department’s evidence to show that it turned to the 1990-92 Plan’s race-conscious measures only after race-neutral efforts failed to bear fruit. Shea, 961 F. Supp. 2d at 40-41. We agree.

The record documents a number of previous attempts to correct the identified imbalances without resort to explicit racial preferences, particularly through recruiting and outreach. From 1964 on, the Department targeted historically black institutions as part of its “diplomat in residence” program, through which it assigned a senior-level Foreign Service Officer to research, writing, and teaching duties at a university in an effort to generate interest in the Foreign Service among students. 1989 GAO Report at 24. From 1980 on, the Department made a concentrated recruiting push to stimulate an increase in minority applicants, including by “provid[ing] information packages to colleges . . . and ask[ing] college coordinators to encourage minorities . . . to take the annual written [Foreign Service] examination.” Id. at 22-23. The Department’s recruiters made special efforts to visit colleges and universities with large minority enrollments. Id. at 23. Ultimately, however, the Department concluded that its “recruiting efforts [did] not increase[] the number of minorities taking the FS examination for officer positions.” USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 39 of 43

39 Id. And in 1986, the Secretary of State implemented a recommendation from black Foreign Service Officers aimed at elevating minority written exam pass rates by increasing minority enrollment in university courses relevant to the exam. That initiative, too, apparently proved unsatisfactory. See id. at 25.

The Department also instituted “sensitivity training” between “senior management” and “senior minorities” to address the gap, with little success. J.A. 369-70. Moreover, it considered implementing an entirely race-neutral mid-level entry program, but rejected that option as unlikely to be effective—an understandable conclusion in light of the inadequacy of State’s earlier reliance on “the promotion of entry level FS officers to eliminate underrepresentation at more senior levels.” J.A. 301, 543. The 1990-92 Plan thus hardly constituted the Department’s maiden effort to solve its persistent diversity problem, and Shea points to no other race- neutral alternatives that should have been considered.

The Department, in short, has introduced evidence that the 1990-92 Plan worked to target manifest imbalances in senior-level positions in the Foreign Service Officer corps, and that those imbalances resulted from past discrimination. It has also introduced evidence that the Plan refrained from unnecessarily trammeling the rights of non-minority candidates. We therefore conclude that the Department satisfies its burden to introduce evidence that, if taken as true, demonstrates the 1990-92 Plan’s validity under Johnson and Weber.

V.

Having concluded that the Department met its burden of production at the second step of the Johnson-McDonnell USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 40 of 43

40 Douglas framework, we ask at the final step whether Shea has proven that the Department’s “justification is pretextual and the plan is invalid.” Johnson, 480 U.S. at 626. In the district court, Shea introduced his own lay statistical evidence in an attempt to show that the Department’s identified manifest imbalances did not exist. See Shea, 961 F. Supp. 2d at 45-53. The district court rejected every piece of statistical evidence proffered by Shea as inadmissible. See id. Shea does not appeal those findings, and he raises no other claims of the 1990-92 Plan’s invalidity for purposes of Johnson’s third step. He therefore necessarily fails to carry his burden at that step, warranting the entry of summary judgment in favor of the Department.

* * * * *

For the foregoing reasons, we affirm the district court’s grant of summary judgment. So ordered. USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 41 of 43

WILLIAMS, Senior Circuit Judge, concurring: I join the court’s opinion painstakingly applying the key Supreme Court cases, Johnson v. Transp. Agency, Santa Clara Cnty., 480 U.S. 616 (1987), and United Steelworkers v. Weber, 443 U.S. 193 (1979). I write separately to note that this area of the law continues to be rather amorphous and to call attention to a statistical problem disclosed by the record but not raised by the plaintiff on appeal.

Nearly three decades ago Judge Silberman observed that he was “uncertain as to the meaning of ‘manifest imbalance.’” Hammon v. Barry, 826 F.2d 73, 81 (D.C. Cir. 1987) (Silberman, J., concurring in denial of rehearing). I fully share that uncertainty, and would add that I have the same reaction to all of the key terms prescribed by the Supreme Court for assessing affirmative action plans under Title VII: whether there has been “manifest imbalance” in a “traditionally segregated job category,” and whether the plan “unnecessarily trammels the rights” of the persons disfavored. Court Op., 24. It may be that the Supreme Court selected these terms to assure that, without saying it in so many words, an employer can use race and gender for hiring or promoting minorities or women to the extent appropriate to assure that there is no “underrepresentation”—i.e., to amend any non- trivial deviation from proportionality to some more or less plausible applicant pool (at least so long as the employer can muster vague, generalized and/or hearsay assertions of past discrimination). This is not a self-evident interpretation of Title VII’s directive that employers are not “to discriminate against any individual . . . because of such individual’s race [or] sex.”

The effect is especially striking here: Shea neither challenged the district court’s ruling that his analysis of the State Department’s calculations was inadmissible, Court Op. 40, nor its ruling that the affirmative action plan’s repeated USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 42 of 43

2

declarations of “manifest imbalance” were sufficient without expert provision of statistical support. See Shea v. Kerry, 961 F. Supp. 2d 17, 51-52 (D.D.C. 2013). The figures underpinning State’s plan consist mainly of numerical comparisons of various subgroups of Foreign Service employees (“Administrative,” “Professional,” “Clerical,” etc.) with a selected comparison group based on the “National Civilian Labor Force” data for various types of workers, e.g., “Public Administration Administrators and Officials.” Joint Appendix (“J.A.”) 209, 216. The description of the study in the record, J.A. 209, does not state what statistical test or standard of statistical significance the authors used, or indeed whether they used any statistical method at all. Certainly they do not suggest that they made an adjustment in the standard for statistical significance to account for the multiplicity of subgroups, as would be necessary if we assume that State was seeking to identify only “imbalances” not attributable to random chance. “When interpreting . . . a table which summarizes results from a number of comparisons, one must bear in mind that when the number of comparisons is large [State’s report included hundreds], the probability may be substantial that at least one disparity with a P-value less than .05 will occur because of pure chance.” David C. Baldus & James W. L. Cole, Statistical Proof of Discrimination § 9.03 (Supp. 1987); see also id. at n.24a (“It is a mathematical fact that where 17 independent comparisons are to be tested, the probability of finding one or more to be statistically significant at the .05 level is .58, or almost 6 chances in 10.”).

Further impairing the value of the analysis is that many of the subsets are so small as to indicate a complete lack of intelligible criteria for State’s assertions of “manifest imbalance,” a term the report often uses but never explains. The report contains charts that split the workforce three ways (by occupational subgroup, ethnicity, and gender), and in one case it announces that it “reveals” a “manifest imbalance” of USCA Case #13-5153 Document #1566749 Filed: 08/07/2015 Page 43 of 43

3

American Indian females (who represent 0.2% of the labor force comparison data) in the Finance Officer division, which employs only 125 people. J.A. 224-25. It seems improbable that any statistical test or standard of significance could yield evidence of a non-random “imbalance” for so small a subgroup. To the extent the report is suggesting that some purported “imbalances” could be amended by the hiring of a single employee of the right ethnicity and gender in the occupational unit in question, that response would, in turn, presumably create “imbalance” in another direction—thus appearing to undermine whatever criteria may have been used to define “manifest imbalance.” See, e.g., J.A. 218-19; 224- 25. I recognize that Johnson is quite specific in stating that the proof of imbalance needed as a prerequisite for race- and gender-based affirmative action preferences is less than what is needed to establish a prima facie case of a Title VII violation, 480 U.S. at 632-33, but an employer performing this exercise should at least be able to state its criteria for “manifest imbalance.”

The State Department in this respect sounds rather like the defendant university in Hill v. Ross, 183 F.3d 586, 591 (7th Cir. 1999): “What the University appears to have in mind is a world in which the absence of discrimination means that every department would exactly mirror the population from which its members are hired. But that is statistical nonsense.” In Hill, Judge Easterbrook went on to explain in detail what made the university’s theories nonsensical. Without close attention, Johnson’s seeming license to pursue proportionality in a workforce can dissolve into a license to pursue proportionality in almost any subset of the workforce.