JAMES HENDERSON, Plaintiff-Appellant, v. DAVID J. SHULKIN, Secretary, Department of Veterans Affairs, Defendant-Appellee.
James Henderson appeals the entry of summary judgment against him in this employment-discrimination suit asserting that the Department of Veterans Affairs failed to promote him because of his race, in violation of Title VII of the Civil Rights Act of 1964, 42 U.S.C. § 2000e-2. Because a reasonable jury could find that the agency's explanations were a pretext for racial discrimination, and there is other circumstantial evidence of discriminatory intent, we vacate the district court's judgment and remand for further proceedings.
Henderson, a 59-year-old African-American male and a former Marine, began working in 1986 as a patrol officer in the police department at Edward Hines Jr. Veterans Administration Hospital; through the years he rose through the ranks to become a captain. For personal reasons, Henderson later resigned from his captain position; at the time he brought this lawsuit, he was a detective, and he remains in that position at the present time. In 2013 he applied for a promotion to the position of criminal investigator. Gary Marsh, the police chief and the final decision-maker, did not select him, choosing a white officer instead. Henderson contends that he was not selected because of his race. (He has not pursued his age discrimination and retaliation claims on appeal. See FED. R. APP. P. 28(a)(8)(A).)
Henderson says that the discrimination began in 2012 under the previous police chief, who had stripped Henderson of his badge, identification, and authority to carry a weapon under the belief that he had a medical bar; however, Henderson had told him, accurately, that his doctor had cleared him to work and that he was current on his training. Henderson filed a charge of discrimination based on this event.
A new police chief, Gary Marsh, transferred to Hines in March 2013 and became Henderson's supervisor. Marsh also would not return Henderson's badge, identification, and weapon. At the time Marsh transferred, the hospital's police force had a vacancy for a criminal investigator. Later that year Marsh decided to fill it, reasoning that other employees, including the eventual selectee, Cary Kolbe, were picking up the slack. But David Scott, an African-American police lieutenant, attested that Marsh allowed Kolbe alone to act as criminal investigator while the position was vacant. Scott asked Marsh to rotate the position so that he and others could perform the duties of the job; Marsh replied by saying, “that's probably the right thing to do,” but that “he wasn't going to.”
Marsh worked with the human-resources department to create a vacancy announcement for public posting. Initially, Marsh wanted to consider only Hines employees. HR later expanded the desired applicant pool to include outside, nonfederal employees who were veterans with service-related disabilities of at least 30% or nonveterans with disabilities. The position was to be filled at the GS-9 or GS-11 level of the government's general schedule of salaries. Marsh, however, said that he would not consider any GS-9 certified applicants if there were viable GS-11 candidates.
After the deadline for receiving applications closed, Brian Cross, an HR specialist, prepared four lists of candidates: two lists for the external applicants eligible for hire at the GS-9 and GS-11 levels and two for the internal applicants at those same GS levels. In HR jargon the lists are called “certificates,” and they are valid for six months after a position is publically posted.
Henderson and the eventual selectee, Cary Kolbe, were listed on the eligibility certificates at both the GS-9 and GS-11 levels—meaning that they each qualified to be hired at either level. But although Cross had listed Kolbe's name on the GS-11 certificate, he was not eligible for a promotion to a GS-11 salaried position because he lacked the required “time in grade”—that is, a year of employment at the VA with at least a GS-9 salary. See 5 C.F.R. § 300.603(a), .604(b) (2017). Kolbe's resume erroneously, or falsely, listed his highest level of employment as GS-9 when really he had never surpassed the GS-8 level. Kolbe could not explain how the line “Highest Grade: GS-9” came to exist on his resume. Cross says he did not rely on Kolbe's resume when he erroneously certified Kolbe for GS-11 eligibility, but Cross could not explain how he made that determination.
Marsh worked with HR to create a process for reviewing applications. He created two panels to select finalists for the criminal-investigator position from the pool that Cross had certified as eligible. The first panel, which met on April 21, 2014, would review the candidates' resumes, assign each candidate a score, and select the applicants that would receive interviews. The second panel would conduct the interviews. Marsh would make the final decision.
Ostensibly, the applicants would remain anonymous to the first panel, but one reviewer admitted to being able to identify Henderson and Kolbe's respective resumes. She had also written Kolbe a character reference when he was going to be disciplined and admitted to socializing with him outside of work. Another reviewer said that he “may” have recognized Kolbe's resume. That reviewer was Marsh's “subject matter expert” because he was the only one on the panel with police experience, but the reviewer testified that he was not aware that he had that role, and the entire reviewing panel said they did not ask him any subject-matter questions during scoring.
The resume-review panel did not invite Henderson to interview, the agency says, because there was a natural break in scores between the third and fourth candidates, and Henderson's resume ranked seventh out of fifteen. The resume-review panel scored Henderson lower primarily because on his resume, he described his prior job duties as “helping” to perform tasks. None of the panel members took notes or wrote comments on their score sheets, even though the score sheet asked the reviewers to detail the candidate's strengths and weaknesses. Kolbe and two external candidates were the only ones invited to interview, and Kolbe received the highest score after the interview. Marsh selected Kolbe for the position on May 13, 2014.
But before Marsh selected Kolbe, the certificates of eligibility and the vacancy announcement had expired—on May 7, 2014. Under federal hiring rules, after the expiration date the position could not be filled without announcing the vacancy again and starting the competitive selection process anew. Rather than announce the vacancy again, on May 13, 2014, Marsh hired Kolbe through a Veterans Recruitment Appointment, 5 C.F.R. § 307.101–.105, which is an exception to competitive hiring for (as relevant here) disabled veterans. VRA hires must still meet both the job qualifications and salary requirements, but they do not go through the typical federal hiring process. After Kolbe was hired, Cross (the HR specialist) decided that Kolbe was qualified for a GS-11 salary based on his extensive experience as an investigator in the military (a substitute for the requirement of a certain period of time in grade). Henderson, however, disputes whether Kolbe was qualified for a GS-11 position, because he had only been a GS-8 before. And the record is silent regarding whether Kolbe's prior outside experience already had been taken into account in setting his initial salary when he was hired by the agency.
But it is undisputed that Kolbe, a 52-year-old Caucasian male, had disciplinary problems. In July, Kolbe received a letter of admonishment for “disrespectful conduct toward another police officer,” after he cursed at a fellow officer; “conduct unbecoming of a police officer,” for stating “I'll kill that old man” in reference to the same officer; and “failure to follow VA policies in the workplace,” based on that same incident. During that altercation he had become “volatile and loud and boisterous and had to be held back and pushed out of the area.”
Later, in September 2012, the agency released a full report of an unrelated internal investigation that began in May, detailing Kolbe's other improper conduct and recommending discipline. The report stated: “Interviews with many officers indicated that Lieutenant (Lt.) Kolbe is quick to be rude and angry towards his peers, subordinate staff, employees, and patients.” “The perception of facility staff towards Lt. Kolbe is they do not like him,” the report continued, and “officers indicated they try to avoid having anything to do with him.” The report recommended that Kolbe take anger-management classes, that progressive discipline be implemented, and that Kolbe be reassigned from a particular building so that he could be closely monitored by department leadership. Another accusation about Kolbe—that he brought a personal firearm to work—surfaced during the investigation, but the report did not make a finding on this point. Henderson did not have any record of discipline and was also a veteran, who, the agency's counsel admitted at oral argument, had also been eligible for a Veterans Recruitment Appointment that would have rendered his score in the competitive process obsolete.
It is disputed whether Marsh, who started about six months after the report on Kolbe issued, had heard any complaints about Kolbe or knew of his disciplinary record before he selected him for promotion. He says he did not know about Kolbe specifically, but that he knew there were a lot of disciplinary problems in general before he arrived at Hines. Marsh says that he told his subordinates they had a clean slate with him. But Henderson presents affidavits from other employees who attest that they complained to Marsh about Kolbe before he was selected. (One of the affiants also says he told Marsh he witnessed Kolbe drunk while at work.) And an employee who claims that Kolbe sexually harassed her for years also said that she told Marsh. Marsh agrees that she told him, but says that he investigated and determined the claims to be unfounded. During the sexual-harassment investigation, Marsh did not take Kolbe's badge or gun as he had done with an African-American officer accused of the same conduct.
And although Marsh said that all officers had a clean slate with him, he proposed that an African-American officer be removed from duty for conduct that occurred before Marsh became the police chief. Other employees attest that Marsh stated that he wanted to keep things “Black and White” when one officer asked for a specific African-American officer to join his shift, that a nickname for the facility was “The Plantation,” and that no African Americans maintain a supervisory position within the police force.
Based on these facts, the VA moved for summary judgment, arguing that there were no genuine issues of material fact because Marsh promoted Kolbe not out of animus for African Americans but because Kolbe came out on top of the merit-selection process. The district court concluded that there were no disputed material facts requiring a trial and that the VA was entitled to summary judgment under either the McDonnell-Douglas burden-shifting framework or the reasonable-factfinder standard set forth in Ortiz v. Werner Enters., Inc., 834 F.3d 760 (7th Cir. 2016). The district court reasoned that Marsh chose Kolbe solely because of the competitive selection process, that there was no evidence that Henderson was more qualified than Kolbe, and that the appointment of Kolbe under a noncompetitive hiring method was simply a bureaucratic nicety.
On appeal Henderson argues that he adduced sufficient evidence of discrimination and that the district court improperly viewed the facts and drew inferences in favor of the VA, the moving party. Henderson contends that material issues of fact existed as to whether Kolbe was more qualified and whether the agency's nondiscriminatory reason for hiring Kolbe was genuine.
The applicable standard at summary judgment is whether the evidence would permit a reasonable factfinder to conclude that racial discrimination caused the adverse employment action—here, the failure to promote. Ortiz v. Werner Enters., Inc., 834 F.3d 760, 765 (7th Cir. 2016). Although this court no longer divides evidence into the formal categories of “direct” and “indirect,” a plaintiff may still use the McDonnell-Douglas burden-shifting test to make the required showing. McKinney v. Office of Sheriff of Whitley Cty., 866 F.3d 803, 807 (7th Cir. 2017); Ferrill v. Oak Creek–Franklin Joint Sch. Dist., 860 F.3d 494, 499 (7th Cir. 2017).
In a failure-to-promote case, to establish a prima facie case a plaintiff must offer evidence that (1) he was a member of a protected class, (2) he was qualified for the position sought, (3) he was rejected, and (4) the employer promoted someone outside of the protected class who was “not better qualified” for the position, Riley v. Elkhart Cmty. Schs., 829 F.3d 886, 892 (7th Cir. 2016), or who “had similar or lesser qualifications,” Whitfield v. Int'l Truck & Engine Corp., 755 F.3d 438, 444 (7th Cir. 2014).
Henderson easily establishes that he is African American, that he met the minimum requirements for the criminal-investigator position, including the time in grade to be hired at the GS-11 level, and that the agency promoted a white applicant instead. And with respect to the last prong, Henderson provided sufficient evidence that Kolbe was “not better qualified” than he to survive summary judgment.
First, it is undisputed that, despite what he wrote on his resume, Kolbe did not spend any time at the GS-9 level, which was required before he could be certified as eligible for a GS-11 level position under competitive hiring. And Marsh wanted a GS-11 candidate. Marsh even told one African-American applicant that he did not receive the job because he was not a GS-11. Therefore, Kolbe's resume never should have been submitted to the resume-review panel at all. Henderson, but not Kolbe, was qualified, according to Marsh's stated preference. The agency has not pointed to any evidence that Kolbe would have survived the panel-selection process—or even entered it—if the agency had properly documented his eligibility.
The agency now admits that Kolbe was improperly classified as a GS-11, although an HR specialist had been tasked with ensuring that only qualified applicants were interviewed. After Kolbe was hired in May 2014, and Henderson filed an administrative charge, an EEOC investigator informed the agency that the EEOC could rule against it if it did not show that Kolbe was a qualified candidate before he was hired. The agency took months to reply to the EEOC's request for evidence of Kolbe's time spent at the GS-9 grade, until it finally admitted that there was not “any evidence” of Kolbe's GS-9 federal employment. (Ultimately, the EEOC found for the agency because Kolbe was retroactively deemed qualified based on relevant prior experience, not time in grade.) Henderson, however, was correctly put on the GS-11 certificate from the beginning.
More importantly, Kolbe's disciplinary record would allow a reasonable jury to conclude that he was no more qualified than Henderson. See Jordan v. City of Gary, Ind., 396 F.3d 825, 834 (7th Cir. 2005) (concluding that candidate failed to establish that she had “similar qualifications” to hired candidate, when she had an extensive disciplinary record). There is no evidence that Henderson ever had disciplinary problems in his 30 years with the VA, whereas there is undisputed evidence that Kolbe threatened fellow officers and even threatened to kill an officer, and was accused of appearing drunk on the job and sexually harassing a female coworker for years. Moreover there is evidence from which a jury could infer that Marsh, the selecting official, knew about Kolbe's disciplinary record and promoted him anyway. Although Marsh denies specific knowledge of the agency's formal admonition to Kolbe and its report of the investigation into his record (apparently, he was willing to promote someone without a glance at the candidate's personnel file), affidavits show that employees brought some of their problems with Kolbe to Marsh's attention. Further, Marsh disciplined another employee for conduct that occurred before Marsh began his tenure, which suggests that he was, in fact, aware of disciplinary records, calling into question his claim of ignorance about Kolbe. For these reasons, Henderson had adequate evidence to create a genuine issue of material fact as to whether Kolbe was a better qualified candidate.
Reaching the contrary conclusion—that because of the panel-selection process Kolbe was definitively “better qualified”—is to ignore the evidence in the record that the process was flawed—because, for example, a few candidates did not remain anonymous, and because Kolbe was not even supposed to be included in the candidate pool. To that point, the unexplained (yet very specific, and crucially important) error on Kolbe's resume tainted the process; the resume-review panel worked off of inaccurate information. As for the panelists, they were chosen by March; to the extent that he had a preferred candidate, he had the opportunity to set Kolbe up favorably (for example, he could have selected reviewers unlikely to know about Kolbe's abysmal disciplinary history). Furthermore, Marsh controlled who acquired the most relevant experience, which gave Kolbe an advantage from the start. The record also calls into question the rigor of the selection process; the reviewers took no notes on their score sheets and, many times when questioned during their depositions could give only vague, unenlightening answers about why they scored individuals a certain way. If even the reviewers cannot confidently explain why Kolbe was advanced, it is not appropriate for us to conclude that there was a truly independent, merit-based, selection process that shows Kolbe was “better qualified” as a matter of law.
Because a jury reasonably could find that Kolbe was not “more qualified” than Henderson, Henderson's prima facie case is made. On appeal, the agency maintains that Marsh hired Kolbe for the legitimate, nondiscriminatory reason that he was the candidate who scored the highest based on the resume review and his interview. So the burden shifts back to Henderson to establish that the alleged reason for hiring Kolbe is pretext—or, for purposes of the Ortiz inquiry, whether a reasonable factfinder could infer from the evidence that discrimination occurred. McKinney, 866 F.3d at 807. Pretext “ ‘involves more than just faulty reasoning or mistaken judgment on the part of the employer; it is [a] lie, specifically a phony reason for some action.’ ” Tibbs v. Admin. Office of the Ill. Cts., 860 F.3d 502, 506 (7th Cir. 2017) (quoting Burton v. Bd. of Regents of the Univ. of Wis. Sys., 851 F.3d 690, 698 (7th Cir. 2017)) (alteration in original).
Henderson says the agency's explanation was pretextual because it offered shifting explanations for why Marsh chose Kolbe, because Marsh could not have relied upon the voided panel-selection process, because Marsh gave Kolbe a noncompetitive appointment instead of posting the vacancy again, and because Marsh has a history of discriminatory conduct.
“One can reasonably infer pretext from an employer's shifting or inconsistent explanations for the challenged employment decision.” Applebaum v. Milwaukee Metro. Sewerage Dist., 340 F.3d 573, 579 (7th Cir. 2003). And there is something to Henderson's argument that the VA has been inconsistent in its explanations for why Kolbe was hired. During the EEOC investigation of Henderson's charge in June 2015, the agency told the EEOC investigator that Kolbe had been granted an interview as a “service connected veteran rated at 30% or more.” We know this is not true; Kolbe was granted an interview because he was in the top three after the resume-review process. Veterans' preference did not enter the discussion until after Marsh selected Kolbe. Perhaps because of this falsehood, the agency changed its story. The EEOC's written decision denying the charge cited another reason (presumably proffered by the agency): that Kolbe was promoted because he was a finalist in a panel-selection process. During discovery in this suit, in February 2016, Marsh testified at his deposition that he hired Kolbe as the “highest scored individual.” And that was the position the agency took at summary judgment. Thus the agency initially said that Kolbe was interviewed because of veterans' preference (note that Henderson also had veterans' preference points) and then changed the story, so there is reason to question the sincerity of the reason provided now.
Henderson's next arguments—that Marsh could not have relied upon the panel-selection process because it was void and that Marsh gave Kolbe a noncompetitive appointment instead of posting the vacancy again—also have force. In particular, Kolbe's and the agency's inability to explain why Kolbe was wrongly certified at the GS-11 level, which Marsh preferred, allows an inference that the process was fixed in favor of the white candidate Marsh wanted to hire. That would not be so if there were a plausible explanation for the error, but Kolbe could not explain why his own resume falsely said he had attained a GS-9 level salary. Did someone else prepare it? Was it a typographical error? Kolbe tells us nothing, and so a jury could certainly discredit his contention that he had “no idea” how the falsehood appeared on his resume.
Henderson also contends that the panel-selection process is a phony explanation because, although the VA now says that the only reason for selecting Kolbe was his high score, the advertised position, and all the applicants' certificates of eligibility, expired before anyone was promoted. The HR representative testified that the vacancy announcement was “pulled back” and there was “not a valid certificate,” i.e., that there was no longer a pool of applicants certified as eligible for hiring into the posted, but expired, notice of vacancy. So Marsh arguably was not permitted to choose among applicants to the expired posting; Kolbe and everyone else who were no longer certified as eligible would have had to reapply before a competitive selection could be made. And therefore an internal candidate like Kolbe would have had to meet the time-in-grade requirement. See 5 C.F.R. § 300.603(a) (2017). As Henderson argues, this makes the scores from the resumes and interviews suspect as a reason to hire Kolbe (particularly because Kolbe should not have gone through the process). And this is not a mere technicality, as the VA argues. Not only was Kolbe's rating invalid; so too was Henderson's—which deals a substantial blow to the VA's loud refrain that Henderson was rated too low to have been prejudiced by the irregularities.
Marsh's purported reliance on the panel-selection process is further undermined by the invocation of a noncompetitive appointment procedure after Marsh selected Kolbe. Although the agency says that it was only an administrative matter, it does not explain any further; it gives no reason why it chose to convert the position to an excepted appointment five days after Kolbe was selected. A jury could reasonably infer that it was a means of legitimizing Kolbe's hire because, as Henderson argues, the VA “knew it could not promote Kolbe via merit promotion because Kolbe was not qualified ․ and because the certifications for the Criminal Investigator position had expired and were no longer effective.”
A jury might also discredit the explanation that Marsh selected Kolbe because of disabled veterans' preference. The most obvious reason is that, if Marsh preferred to make a VRA hire, there was no reason the panel-selection process ever should have been invented and carried out. As long as the VRA candidate met minimum qualifications, he did not need to compete with other candidates—that is the whole point. Using a VRA reveals the panel-selection process as a waste of time and, perhaps, as Henderson argues, a mere fig leaf to disguise Marsh's preference for a white candidate. After all, Henderson was available for a VRA, too.1 And again, the Act gives preference to the enumerated classes of veterans who meet the position's qualifications; at the time of his hire, Kolbe did not. That the VA retroactively found a way to qualify him for a higher salary level once Henderson complained to the EEOC could be viewed by a jury as a cover-up.
Finally, Henderson has presented other circumstantial evidence from which animus to black employees could be inferred. Circumstantial evidence raising an inference of discriminatory intent can include “behavior toward or comments directed at other employees in the protected group.” Burks v. Union Pac. R. Co., 793 F.3d 694, 700 (7th Cir. 2015); see also Hutt v. AbbVie Prod. LLC, 757 F.3d 687, 691 (7th Cir. 2014) (same). There is evidence that Marsh made statements that he wanted “to keep it Black and White,” that he disciplined African-American employees more harshly than Caucasian employees, that employees call Hines “[t]he [p]lantation,” that no black members of the police force have supervisory authority, and that Marsh allowed only Kolbe to gain experience as the acting criminal investigator, despite the specific request of an African-American officer to be given a chance. The VA protests that this evidence is inadmissible, because it is based on the opinions and beliefs of Henderson's affiants. But the facts set forth above are within the personal knowledge of the Hines employees who attested to them, so the VA has not established that the testimony would be inadmissible at trial. See FED. R. EVID. 602; United States v. Joy, 192 F.3d 761, 767 (7th Cir. 1999).
The circumstantial evidence of Marsh's animus and the discriminatory environment of his police force, combined with the shifting explanations for Kolbe's selection, and the evidence that Marsh lied about his reason for choosing Kolbe, permit a reasonable juror to conclude that Marsh wanted to hire a white candidate even if he was “not better qualified” than a black candidate like Henderson who had a much cleaner record and met Marsh's preferred salary level.
And to address Henderson's argument that the district court improperly viewed the facts and drew inferences in favor of the agency, we agree that the district court appeared to credit the VA in a few instances. First, the court dismissed all “evidence regarding other employees and other matters not connected to Henderson” as an “attempt to paint Marsh as someone who generally discriminated against African-Americans.” As previously noted, a jury could infer discriminatory intent from that evidence, which is not all inadmissible from what we can discern. Another inference in the VA's favor was that “[t]here is no evidence in the record that would indicate Henderson was any more qualified for the position than Kolbe.” The law does not require Henderson to be superior, but there was evidence in the record that Henderson, not Kolbe, had the required time in grade to qualify for a GS-11 level position. Further, the evidence of Henderson's clean record compared to Kolbe's sullied one could support a reasonable inference that Henderson was an equally or more qualified candidate.
Another fact that the district judge viewed in the VA's favor was that “the undisputed facts indicate that an extensive selection process was used that included two separate panels in order to arrive at a selection.” This overlooks evidence tilting the other way: that the panel-selection process should not have gone forward after the vacancy announcement and the candidates' certificates had expired; that at times, the VA represented that it was Kolbe's status as a disabled veteran, not his relative merit, that caused the agency to select him through a noncompetitive process; that Kolbe should never have been part of the panel-selection procedure; and that the process did not function as intended. And it fails to consider that Marsh skewed the selection process prior to the panel-selection procedure, by allowing only Kolbe to gain experience as the acting criminal investigator and thereby inflating his credentials and preventing Henderson from demonstrating his abilities in the position prior to the panel selection. That alone renders the panel selection process tainted. Indeed, whether a neutral, “extensive selection process” resulted in Kolbe's promotion is the factual nut of this case, and the facts are not undisputed.
Finally, the VA emphasizes that Henderson would not have been hired no matter what because he finished seventh in the resume-review process. But there is no way to know what would have occurred if Kolbe had not been in the pool in error, if the position had been posted again (perhaps all Henderson needed to do was delete all references to “help” from his resume), if persons other than Kolbe had been allowed to gain experience as the acting criminal investigator, or if the process had been anonymous as intended. Further, the agency's argument implies that even if the fix was in, the very existence of the process immunizes the agency from liability. But an employer's failure to follow its own internal procedures is evidence of discriminatory intent. Rudin v. Lincoln Land Cmty Coll., 420 F.3d 712, 723 (7th Cir. 2005). It should not matter that Henderson's resume put him in the middle of the pack unless a jury were to conclude that the panel-selection process is the real reason Kolbe was selected and that the process was not tainted by Marsh. Because that fact is in dispute, summary judgment was not appropriate.
We VACATE the district court's judgment and REMAND for further proceedings.
1. The Veteran's Preference Act only applies to initial hires into the federal government, not to promotions. See Kerner v. Dep't of the Interior, 778 F.3d 1336, 1338 (Fed. Cir. 2015). The VRA, at issue here, covers promotions.