holding that the question in discrimination cases "is simply whether the evidence would permit a reasonable fact-finder to conclude that the plaintiff's race, ethnicity, sex, religion, or other proscribed factor caused the discharge or other adverse employment action"Summary of this case from Smith v. City of Janesville
Michael L. Resis, Chicago, IL, Jill A. Cheskes, for Defendant–Appellee. Nancy A. Temple, Chicago, IL, for Plaintiff–Appellant.
Michael L. Resis, Chicago, IL, Jill A. Cheskes, for Defendant–Appellee.
Nancy A. Temple, Chicago, IL, for Plaintiff–Appellant.
Before Posner, Easterbrook, and Hamilton, Circuit Judges.
Easterbrook, Circuit Judge.
Henry Ortiz worked as a freight broker for Werner Enterprises, Inc., for seven years until his discharge in 2012. Werner says that it fired Ortiz for falsifying business records. Ortiz says that Werner fired him because of his Mexican ethnicity, and he sued Werner under 42 U.S.C. § 1981 and the Illinois Human Rights Act, 775 ILCS 5/1–101 to 5/10–104. He also claimed that Werner subjected him to a hostile work environment in violation of the state law.
Werner is a shipping company that offers freight brokerage as one of its services; customers pay a fee and Werner finds transportation for their loads. Werner tracks its customers' loads through a proprietary records system. Loads that have not been matched with a carrier—a vehicle hired to transport the load—appear in the system, and for most of Ortiz's tenure at Werner any broker could locate a carrier for any load. After securing carriers, brokers update the system to confirm the transactions, the carrier's prices, and who gets credit for the deal. Werner's profit is the difference between what its customer pays and what it pays the carrier. Some loads produce a loss when the carrier charges more than the customer's payment. Werner compensates its brokers primarily through a base salary, but they can earn a commission of up to 4% for months in which they generate more than a specified profit. Ortiz usually received monthly commissions of around $1,000.
Permitting all brokers to book any load led to problems, as brokers cherry-picked the nation's most profitable loads. In 2012 Werner addressed this by assigning each broker a specific region. During Ortiz's final months with Werner the regional broker bore responsibility for matching every load in the region to a carrier whether the load would generate a profit or a loss. But the system did not automatically assign loads—brokers still needed to search for loads in their region and update the records.
Werner assigned Ortiz the West region, apparently not a lucrative assignment during late spring when high demand for transporting produce drives carriers' rates up. (The parties don't explain why Werner failed to change its fees to match predictable cycles in the business.) During the first week of June 2012 Michael Krikava, the branch's assistant manager, booked or directed another broker to book six unprofitable loads in Ortiz's name. There was some variation in how Ortiz's name wound up on each load: Krikava booked several loads in his own name and swapped in Ortiz's name after the loads had been picked up, but others were booked early in the morning when Ortiz may have been out of the office. Krikava did not notify Ortiz of any of these losing transactions. By the time Ortiz realized that he had been assigned responsibility for the loads, it was too late for him to search for a cheaper carrier. Ortiz later changed the records to reflect that he had not booked the loads or to show lower rates that he thought the carriers had agreed to charge. He then left for a planned vacation. On his return Werner fired him. The parties agree on this much, but here their stories diverge. We recount Ortiz's version, because we must view the evidence in the light most favorable to the party opposing a motion for summary judgment.
Ortiz contends that brokers and managers always notified one another when booking an unprofitable load in someone else's name, and that it was atypical for someone else to book so many unprofitable loads without so much as a courtesy email or phone call. He adds that Krikava assigned him unusually high losses. Ortiz repeatedly questioned Krikava about the unprofitable loads but he refused to answer until saying: “Why won't you just quit already?”
Ortiz maintains that he updated three records to show that he had spoken to the carriers, which had agreed to accept a lower rate because they had not picked up the goods on time. Ortiz adds that he often tried to negotiate a lower rate after a late delivery and that carriers would sometimes oblige. Ortiz says that this represents standard practice in the industry; other brokers offered similar testimony.
Ortiz justifies the removal of his name from three other loads with evidence that the branch's managers permitted brokers to delete their names from unprofitable loads. This practice allowed Werner to satisfy the needs of all of its customers without saddling brokers with undue losses, thus protecting each broker's commission. Removing a broker's name from a loss also moved the loss off the branch's books, increasing the branch's profits and the managers' bonuses, which are tied to branch performance. Other Werner brokers testified that this was called “spacing”, as in sending the record of the unprofitable load into cyberspace. (The evidence does not reflect what division of Werner ultimately bore the loss or how Werner reconciled its books.) Because spacing was common practice, Ortiz says, he thought it permissible for him to remove his name from the money-losing loads.
On his first morning back from vacation Ortiz met with Kip Lass, the branch manager. Lass told Ortiz that he had been fired for falsifying records. Ortiz argued that he had not falsified records but had corrected them by providing the rate that the tardy carriers had agreed to charge, and he offered to call the carriers to provide proof. Lass showed no interest in verifying Ortiz's allegations, noted “What's done is done,” and discharged Ortiz without further investigation.
Ortiz also tells us that Krikava and Lass subjected him to a barrage of ethnic slurs. They frequently used epithets such as “beaner,” “taco eater,” “fucking beaner,” “taco,” “bean eater,” “dumb Mexican,” “stupid Puerto Rican,” “dumb Puerto Rican,” “fucking Puerto Rican,” “Puerto Rican,” and “dumb Jew” to refer to Ortiz. Ortiz contends that he encountered such slurs throughout his tenure at Werner and that they increased in frequency and intensity in the months leading up to his discharge.
Werner asserts that the office did not regularly notify brokers of loads recorded under their names, and it contends that the losses assigned to Ortiz were not unusually high. It insists that its brokers do not have the authority to remove their names from unprofitable loads without a manager's permission. It also denies that it was customary to cut a carrier's rate for a late delivery—even if the carrier had consented to the lower rate. According to Werner, Ortiz improperly edited these three records as well. Finally, Werner denies that Lass and Krikava ever made derogatory remarks to Ortiz. But on all of these issues we must look at the evidence from Ortiz's perspective.
The district court granted summary judgment to Werner. 2015 WL 3961240, 2015 U.S. Dist. LEXIS 82952 (N.D. Ill. June 25, 2015). It dismissed Ortiz's claim of a hostile work environment for failure to exhaust his administrative remedies, a ruling Ortiz does not contest on appeal. The judge looked at the evidence through the “direct” and “indirect” methods that courts often discuss in employment-discrimination cases. See, e.g., Andrews v. CBOCS West, Inc. , 743 F.3d 230, 234 (7th Cir. 2014) ; Zaderaka v. Illinois Human Rights Commission , 131 Ill.2d 172, 178–79, 137 Ill.Dec. 31, 545 N.E.2d 684 (1989). Admissions of culpability and smoking-gun evidence were assigned to the “direct” method (the judge found no such evidence), while suspicious circumstances that might allow an inference of discrimination were assigned to the “indirect” method. The court did not try to aggregate the possibilities to find an overall likelihood of discrimination.
The district court treated each method as having its own elements and rules, even though we have held that they are just means to consider whether one fact (here, ethnicity) caused another (here, discharge) and therefore are not “elements” of any claim. See, e.g., Achor v. Riverside Golf Club , 117 F.3d 339, 341 (7th Cir. 1997). Pursuing two “methods” instead of a unified inquiry, the district court elaborated by saying that Ortiz could prevail only by coming up with “evidence that creates ‘a convincing mosaic of discrimination’ ”. 2015 WL 3961240 at *4, 2015 U.S. Dist. LEXIS 82952 at *12. It concluded that Ortiz had failed to present a “convincing mosaic” under the direct method because Lass's racial slurs did not have anything to do with Ortiz's discharge. The court also ruled that there was not enough of a “mosaic” under the indirect method because, by removing his name from the records and changing the rates, he fell short of Werner's expectations.
The district court's effort to shoehorn all evidence into two “methods,” and its insistence that either method be implemented by looking for a “convincing mosaic,” detracted attention from the sole question that matters: Whether a reasonable juror could conclude that Ortiz would have kept his job if he had a different ethnicity, and everything else had remained the same. See, e.g., Achor ; Troupe v. May Department Stores Co. , 20 F.3d 734 (7th Cir. 1994). The use of disparate methods and the search for elusive mosaics has complicated and sidetracked employment-discrimination litigation for many years. During the last decade, every member of this court has disapproved both the multiple methods and the search for mosaics. See, e.g., Sylvester v. SOS Children's Villages Illinois, Inc., 453 F.3d 900 (7th Cir. 2006) ; Coleman v. Donahoe , 667 F.3d 835, 863 (7th Cir. 2012) (concurring opinion joined by entire panel); Good v. University of Chicago Medical Center , 673 F.3d 670, 680 (7th Cir. 2012) ; Harper v. C.R. England, Inc. , 687 F.3d 297, 314 (7th Cir. 2012) ; Hitchcock v. Angel Corps, Inc. , 718 F.3d 733, 737 (7th Cir. 2013) ; Perez v. Thorntons, Inc. , 731 F.3d 699, 703 (7th Cir. 2013) ; Chaib v. Indiana , 744 F.3d 974, 981 (7th Cir. 2014) ; Bass v. Joliet Public School District No. 86 , 746 F.3d 835, 840 (7th Cir. 2014) ; Hutt v. AbbVie Products LLC , 757 F.3d 687, 691 (7th Cir. 2014) ; Simpson v. Beaver Dam Community Hospitals, Inc. , 780 F.3d 784, 789–90 (7th Cir. 2015) ; Castro v. DeVry University, Inc. , 786 F.3d 559, 564 (7th Cir. 2015) (collecting cases). The time has come to jettison these diversions and refocus analysis on the substantive legal issue.
We start with the “convincing mosaic.” This phrase originated in Troupe and was designed as a metaphor to illustrate why courts should not try to differentiate between direct and indirect evidence. Troupe explained that all evidence is inferential and cannot be sorted into boxes. All evidence should be considered together to understand the pattern it reveals. But instead of persuading district judges and litigants to merge “direct” and “indirect” methods into a unified approach, Troupe was understood by many as adding a new “test” that had to be satisfied. How a “mosaic” could be a “test” rather than a mental picture went unaddressed—as did the question how such a legal test could be rooted in the statutes that govern employment-discrimination cases.
We tried to set things straight in Sylvester, 453 F.3d at 903–04, reminding lawyers and judges that Troupe used “mosaic” as a metaphor that was designed to displace the unhelpful direct and indirect methods rather than add to them. The reminder did not work, as the district court's opinion in this case shows. Despite follow-up reminders that “convincing mosaic” is not a legal test of any kind, see Muhammad v. Caterpillar, Inc. , 767 F.3d 694, 699–700 (7th Cir. 2014) ; Fleishman v. Continental Casualty Co. , 698 F.3d 598, 603 (7th Cir. 2012) ; Davis v. Time Warner Cable of Southeastern Wisconsin, L.P. , 651 F.3d 664, 672 (7th Cir. 2011) ; it has continued to be misused as one.
To make matters worse, this court has itself occasionally treated “convincing mosaic” as a legal requirement, even while cautioning in other opinions that it must not be so understood. See, e.g., Hatcher v. Board of Trustees of Southern Illinois University , No. 15–1599, 829 F.3d 531, 767, 2016 WL 3770555 (7th Cir. July 14, 2016);Chaib , 744 F.3d at 981 ; Cloe v. Indianapolis , 712 F.3d 1171, 1180 (7th Cir. 2013) ; Smith v. Bray , 681 F.3d 888, 901 (7th Cir. 2012) ; Good , 673 F.3d at 674 ; Silverman v. Board of Education of Chicago , 637 F.3d 729, 734 (7th Cir. 2011) ; Phelan v. Cook County , 463 F.3d 773, 779 (7th Cir. 2006) ; Koszola v. Board of Education of Chicago , 385 F.3d 1104, 1109 (7th Cir. 2004) ; Rhodes v. Illinois Department of Transportation , 359 F.3d 498, 504 (7th Cir. 2004) ; Cerutti v. BASF Corp. , 349 F.3d 1055, 1061 (7th Cir. 2003) ; Robin v. Espo Engineering Corp. , 200 F.3d 1081, 1088–89 (7th Cir. 2000). Instead of simplifying analysis, the “mosaic” metaphor has produced a form of legal kudzu.
Today we reiterate that “convincing mosaic” is not a legal test. We overrule the opinions in the previous paragraph to the extent that they rely on “convincing mosaic” as a governing legal standard. We do not hold that any of those cases was wrongly decided; our concern is only with the treatment of “convincing mosaic” as if it were a legal requirement. From now on, any decision of a district court that treats this phrase as a legal requirement in an employment-discrimination case is subject to summary reversal, so that the district court can evaluate the evidence under the correct standard.
That legal standard, to repeat what we wrote in Achor and many later cases, is simply whether the evidence would permit a reasonable factfinder to conclude that the plaintiff's race, ethnicity, sex, religion, or other proscribed factor caused the discharge or other adverse employment action. Evidence must be considered as a whole, rather than asking whether any particular piece of evidence proves the case by itself—or whether just the “direct” evidence does so, or the “indirect” evidence. Evidence is evidence. Relevant evidence must be considered and irrelevant evidence disregarded, but no evidence should be treated differently from other evidence because it can be labeled “direct” or “indirect.”
Some opinions have observed that the methods largely rely on the same pieces of evidence. See Harper , 687 F.3d at 314 (“[T]he line between circumstantial evidence under the direct method and indirect evidence of discrimination or retaliation under the burden-shifting approach has been blurred by the gradual integration of these methodologies.”); Hutt , 757 F.3d at 691. Why have two tests if they consider the same information and answer the same question?
True, some cases permit easy inferences, such as the fabled employer who admits to firing an employee because of race. But even then factfinders must ask themselves what the admission means. Few discrimination cases are so straightforward—indeed they are often factually complex and require sifting through ambiguous pieces of evidence. And the direct-and-indirect framework does nothing to simplify the analysis. Instead it complicates matters by forcing parties to consider the same evidence in multiple ways (and sometimes to disregard evidence that does not seem to fit one method rather than the other).
Accordingly, we hold that district courts must stop separating “direct” from “indirect” evidence and proceeding as if they were subject to different legal standards. Once again, this court must accept its share of the responsibility—because even as some panels were disparaging the “direct” and “indirect” approaches, other panels were articulating them as governing legal standards. We need to bring harmony to circuit law, and the way to do that is to overrule Andrews , 743 F.3d 230 ; Silverman , 637 F.3d 729 ; Hemsworth v. Quotesmith.com , Inc. , 476 F.3d 487 (7th Cir. 2007) ; Rhodes , 359 F.3d at 498 ; Haywood v. Lucent Technologies, Inc. , 323 F.3d 524 (7th Cir. 2003) ; Oest v. Illinois Department of Corrections , 240 F.3d 605 (7th Cir. 2001) ; Radue v. Kimberly–Clark Corp. , 219 F.3d 612 (7th Cir. 2000) ; Chiaramonte v. Fashion Bed Group, Inc. , 129 F.3d 391 (7th Cir. 1997) ; Mechnig v. Sears, Roebuck & Co. , 864 F.2d 1359 (7th Cir. 1988) ; and La Montagne v. American Convenience Products, Inc. , 750 F.2d 1405 (7th Cir. 1984), to the extent that these opinions insist on the use of the direct- and-indirect framework. (As with the convincing-mosaic cases, we are not saying that any of these decisions produced a victory for the wrong litigant; our concern is with the legal standards they employed rather than with their outcomes.)
One point of clarification may be helpful. The burden-shifting framework created by McDonnell Douglas Corp. v. Green , 411 U.S. 792, 93 S.Ct. 1817, 36 L.Ed.2d 668 (1973), sometimes is referred to as an “indirect” means of proving employment discrimination. Today's decision does not concern McDonnell Douglas or any other burden-shifting framework, no matter what it is called as a shorthand. We are instead concerned about the proposition that evidence must be sorted into different piles, labeled “direct” and “indirect,” that are evaluated differently. Instead, all evidence belongs in a single pile and must be evaluated as a whole. That conclusion is consistent with McDonnell Douglas and its successors.
With the rat's nest of surplus “tests” removed from the law of the circuit, we can turn back to Ortiz's claim and his supporting evidence. Stripped of the layers of tests, our analysis is straightforward.
Ortiz relies on his own testimony plus depositions and declarations from other brokers, managers, and former employees. Some Werner employees testified that they always provided notice when booking a loss in another broker's name. Some went further and said they expected someone to ask their permission before recording an unprofitable transaction in their names. Several brokers testified that they (and other brokers) sometimes removed their names from unprofitable loads. Werner could not identify an express policy that forbids the practice. A former Werner broker with 20 years' experience submitted a declaration that while at Werner he had sometimes asked tardy carriers to lower their rates and that such action reflected industry practice. Finally, several current and former employees recounted that Lass and Krikava had directed ethnic slurs at Ortiz.
A reasonable juror could infer that Lass and Krikava didn't much like Hispanics (or Jews, though Ortiz is not Jewish) and tried to pin heavy losses on Ortiz to force him out the door. A juror also might infer that, because of Ortiz's ethnicity, Werner's managers fired him for using techniques that were tolerated when practiced by other brokers. In the end a jury might not credit Ortiz's evidence and could accept Werner's explanations. But given the conflict on material issues, a trial is necessary.
Ortiz also argues that the court erred by excluding as hearsay his affidavit that described what Lass said at the meeting where he was fired. This likely was not hearsay. See Fed. R. Evid. 801(d)(2)(D). But it matters little. Since we are remanding for trial Ortiz is free to call the attendees of the meeting to testify, and they can share their recollections subject to the rules of evidence.
Werner argues that it would have discharged Ortiz, regardless of the falsified records, based on information that surfaced after the discharge. In short, Werner discovered that Ortiz was using his work email to send and receive sexually explicit messages and photos. Werner asserts that this violates its policy on Internet use. This information might have justified a discharge, but the parties agree that Werner cannot use this alternate ground to support action that Werner took before the email problem came to light. So Ortiz's inappropriate emails may reduce his damages, but they cannot absolve Werner. See McKennon v. Nashville Banner Publishing Co. , 513 U.S. 352, 115 S.Ct. 879, 130 L.Ed.2d 852 (1995). The parties give conflicting accounts of whether Werner enforced its policy. But since they agree that the after-acquired evidence cannot retroactively justify Ortiz's discharge, we leave it to the jury to determine whether it affects Ortiz's damages, if any.
Because this opinion overrules two lines of cases in this circuit, it was circulated before release to all judges in active service. See Circuit Rule 40(e). None of the judges favored a hearing en banc.
The judgment is reversed, and the case is remanded for trial.