Haydar v. Amazon Corporate, LLC

Opinion

UNITED STATES DISTRICT COURT EASTERN DISTRICT OF MICHIGAN SOUTHERN DIVISION

ABDULLAH HAYDAR,

Plaintiff, Case No. 2:16-cv-13662 Honorable Laurie J. Michelson v. Magistrate Judge Stephanie Dawkins Davis

AMAZON CORPORATE, LLC, GARRET GAW, PETER FARICY, and JOEL MOSBY,

Defendants.

OPINION AND ORDER GRANTING IN PART DEFENDANT’S MOTION TO EXCLUDE REPORTS AND TESTIMONY OF DONNA BLANCERO [114] In 2015, Abdullah Haydar was terminated from his position as a senior manager at Amazon Corporate, LLC. The reasons for Haydar’s termination are disputed. According to Haydar, one significant reason was because he is of Syrian descent and is Muslim. Haydar alleges that Peter Faricy, effectively the head of Amazon’s “Marketplace” division at the time, “made references to you people, to you people need to learn how to treat your wives better, those types of comments repeatedly, . . . in reference to me uniquely differently than other people. He did not [even] make such comments about [non-Syrian, non-Muslim] people . . . who went through ugly divorces and whatnot.” (R. 68, PageID.6327.) Although Haydar was two management levels below Faricy, Haydar believes that Faricy’s view of him affected how his direct supervisors (and others under Faricy) viewed him. For its part, Amazon maintains that Haydar’s national origin and religion had nothing to do with his firing. From the company’s perspective, Haydar was terminated because he did not comply with several of the company’s 14 leadership principles. In a lengthy opinion, this Court explained that the evidence is not one-sided, and so a jury must decide if Haydar’s or Amazon’s view of the facts is correct. See generally Haydar v. Amazon Corp., LLC, No. 2:16-CV-13662, 2018 WL 4282777 (E.D. Mich. Sept. 7, 2018) (denying summary judgment on Haydar’s claims of national origin and religious discrimination). At the upcoming trial, Haydar hopes that the opinions of Donna Marie Blancero, Ph.D.,

will help the jury see things from his perspective. Among other things, Blancero intends to testify about bias against Muslims, unconscious bias, the difference between diversity and inclusion in the workplace, proper procedures for investigating an employee’s complaint, and the pitfalls of using “360 feedback” to evaluate an employee. (See ECF No. 114, PageID.11538–11546.) Amazon does not think the jury should hear from Blancero. To be more specific, Amazon asks this Court to take up its gatekeeper role under Federal Rule of Evidence 702 and preclude Blancero’s testimony and expert reports. Amazon argues Blancero is not qualified to give opinions on some issues, that her opinions are based on insufficient data, and that her opinions are based on a flawed methodology. (See ECF No. 114, PageID.11515.)

In large part, the Court agrees with Amazon. I. While this opinion is largely dedicated to specifics, the Court begins with two general observations that form a backdrop. The first is that nothing in this opinion should be construed as a critique of Blancero’s credentials. To the contrary, Blancero is quite accomplished. She holds a doctorate from Cornell, with a major emphasis in human-resource management and minor emphases in organizational behavior and labor relations. (ECF No. 114, PageID.11553.) Blancero is an adjunct associate professor in management at Bentley University and also serves as the interim dean of business and the interim dean of the McCallum Graduate School. (See id.) One of the courses Blancero teaches at Bentley is about managing diversity in the workplace. (ECF No. 114, PageID.11538.) She has “published over 35 papers/chapters, and [is] a national speaker on issues of diversity.” (Id.) Blancero’s credentials are not on trial here. But those credentials have led the Court to a second general observation: the two reports

Blancero prepared in this case are not the paradigm of what Rule 702 requires. In several places in her reports, Blancero makes broad generalizations based on limited data. It also appears that some of her conclusions are based on portions of the record favoring Haydar even though other, less-favorable portions of the record could have altered the conclusions. And at some points, Blancero does not bring her knowledge of the field to bear, instead relying on articles outside her expertise to make a point. Also, the overall organization of Blancero’s reports is difficult to follow. It is difficult for the Court to discern whether she is offering any opinions or simply intends to provide the jury with general principles. All of this contributes to the general concern the Court has in allowing Blancero to testify to everything in her reports.

II. Along with those two general observations, the law completes the backdrop for analyzing Blancero’s specific findings. Under Federal Rule of Evidence 702, this Court may permit opinion testimony only if the opinion witness is “qualified as an expert.” And even if Blancero is so qualified, any opinion she tells the jury must help the jury decide the disputed issues in the case, must be “based on sufficient facts or data,” must be “the product of reliable principles and methods,” and must be the result of reliably applying “the principles and methods to the facts of the case.” Fed. R. Evid. 702. Experts are permitted to educate a jury about general principles without applying those principles to the facts of the case. See Fed. R. Evid. 702 advisory committee’s note to 2000 amendment. Rule 702 grants this Court considerable discretion to allow or exclude opinion testimony. See Conwood Co. v. U.S. Tobacco Co., 290 F.3d 768, 781 (6th Cir. 2002) (“[W]e will reverse a district court only where we are left with a definite and firm conviction that it committed a clear error of judgment.”). III.

Now to the specifics. The Court will walk through Blancero’s January 2018 report1 section- by-section, explaining why some parts of the report clear Rule 702’s bar to admissibility and why others do not. (The Court takes the “Introduction” section last, as Blancero testified that it was effectively a summary of her report. (ECF No. 114, PageID.11756.)) A. In the “Diversity and Inclusion” section of her report, Blancero defines stereotyping and confirmation bias. (Report at 2.) In this Court’s view, the concept of stereotypes is within the knowledge of the typical juror. So there is no reason for Blancero to offer testimony on this topic. See United States v. Rios, 830

F.3d 403, 413 (6th Cir. 2016) (“A district court may commit manifest error by admitting expert testimony where . . . the . . . testimony is not beyond the ken of the average juror.” (internal quotation marks and alterations omitted)). As for confirmation bias, Blancero states that “confirmation bias is when one searches for data to support one’s preexisting beliefs.” (Report at 2.) She further states, “These type[s] of

1 The Court uses Blancero’s January 2018 report instead of her November 2017 report because the 2018 version is more comprehensive. Although this District’s citation practice is to use the “PageID” found in the CM/ECF header, because the parties are likely to refer to the January 2018 report during trial, the Court will use the report’s native page numbering in this opinion. The report can be found at ECF No. 114, PageID.11538–11546. Additionally, a highlighted copy of the report is attached to this opinion (see Appendix); assuming that an adequate foundation is laid at trial, Blancero may recount the highlighted portions to the jury. cognitive biases appear[] to be occurring in this case. Based on the facts of this case, Mr. Haydar was stereotyped in a negative manner and his supervisors searched for subjective data to support a poor evaluation rather than depending on the very clear objective data.” (Id.) “This,” Blancero continues, “is a classic example of confirmation bias.” (Id.) Blancero’s application of the theory of confirmation bias to this case is not based on an

adequate factual foundation. As the parties are aware, the Court has reviewed a large portion of the discovery in this case, including the depositions of Haydar’s supervisors, Garret Gaw and Joel Mosby. The Court recalls nothing—and Haydar has not pointed to anything—indicating that Gaw or Mosby (as opposed to Faricy) held any stereotypes about Muslims or Syrians, let alone that they searched for data on Haydar that fit their stereotypes of Muslims or Syrians. As such, Blancero’s vague reference to “the facts of this case” does not support her statement that Haydar’s “supervisors searched for subjective data to support a poor evaluation.” (Report at 2.) And to the extent that she intended to limit this to Faricy, the report in no way makes that clear. Further, Blancero faults Haydar’s supervisors for not relying on “very clear objective data,” but then she

cites a line from an email that appears to be Gaw’s subjective assessment. (Id.) Moreover—and this is something of a repeat theme in the report—Blancero ignores the fact that under Amazon’s rating system, objective performance, growth potential, and leadership were each assessed separately. See Haydar, 2018 WL 4282777, at *1–2 (describing how three separate metrics contribute to an overall rating). So Gaw and Mosby were not permitted to rely solely on “very clear objective data” in assessing Haydar’s leadership. Thus, Blancero will not be permitted to testify to the findings in the “Diversity and Inclusion” of her report. B. In the next section of her report, titled “Treatment of Mr. Haydar and Subsequent Investigation,” Blancero indicates that Amazon’s human resources department did not conduct a proper investigation into Haydar’s complaint of mistreatment by his supervisors. (Report at 3.) The Court hesitates to permit this testimony. It is problematic that Blancero critiques Anne

DeCleene’s HR investigation into Haydar’s internal complaints but did not even read her rather- lengthy and rather-detailed report. (See ECF No. 114, PageID.11655, 11767.) But, perhaps, Blancero does not take issue with how DeCleene conducted her investigation generally, but how DeCleene (or others in HR) investigated national-origin or religious bias specifically. (See Report at 3 (“When the investigation took place, . . . it appears that she did not investigate based on bias related to . . . nationality or religion.”); see also ECF No. 114, PageID.11653, 11752–11753.) To the extent that Blancero’s expertise allows her to reliably say that when an employee mentions “bias,” standard human-resources protocol is to investigate bias against the employee’s protected characteristics, Blancero may testify as such. (As this Court previously explained, the record is

mixed on whether Haydar told human resources about national-origin or religious bias or used the term “bias” in a more amorphous manner. See Haydar, 2018 WL 4282777, at *9.) The Court will determine at trial whether Haydar lays the proper foundation for Blancero to testify about how human-resources departments should investigate a claim of “bias.” Also in this section of her report, Blancero indicates that Amazon’s investigation did not follow industry standards for investigations into employee complaints (e.g., “procedural fairness rules”). The Court will permit Blancero to offer this background information to the jury. It appears that Blancero has sufficient education and experience to testify about proper internal investigations. (See ECF No. 114, PageID.11684; ECF No. 114, PageID.11763–11767 (“Q. [Y]ou have a Ph.D. in these kinds of [internal] investigations. Is that correct? A. Well, my Ph.D. is in human resources but my dissertation was on investigations, yes, and procedural fairness of investigations.”).) But the Court will not permit Blancero to opine on whether DeCleene’s investigation met industry standards (again, Blancero did not read DeCleene’s report (ECF No. 114, PageID.11655, 11767)). Instead, from the general standards Blancero provides, the jury will

evaluate for itself the quality of Amazon’s investigation into Haydar’s complaint. See Fed. R. Evid. 702 advisory committee’s note to 2000 amendment (“[I]t might also be important in some cases for an expert to educate the factfinder about general principles, without ever attempting to apply these principles to the specific facts of the case.”). In short, assuming proper foundation, Blancero may testify to the highlighted portions of the section titled “Treatment of Mr. Haydar and Subsequent Investigation.” (See Appendix.) C. In the “Culture of Lack of Inclusion” section of her report, Blancero introduces the concept of unconscious (i.e., implicit) bias.

To better understand this section of Blancero’s report, some background is helpful. The term “unconscious bias” captures the concept that a person holds a bias (positive or negative) toward a group of people but is unaware that he holds that bias. See (Report at 5); United States v. Robinson, 872 F.3d 760, 785 (6th Cir. 2017) (Donald, J., concurring in part and dissenting in part) (“Perhaps the most disturbing aspect of implicit bias is that it operates outside of a person’s conscious intent. Such biases often conflict with one’s consciously-held, egalitarian values, and indeed are more predictive of our conduct than are those explicitly-held values.”). Research indicates that everyone has some unconscious biases (see Report at 4), in part because “implicit biases allow individuals to efficiently categorize their experiences, and these categories allow people to easily understand and interact with their world,” Robinson, 872 F.3d at 785. With that background, Blancero’s opinion is more readily understood. Her report states, “In Mr. Haydar’s case, it is quite reasonable to come to the conclusion that there is implicit bias given the climate and tone of Mr. Faricy. Based on this I believe we can say that that bias

manifested itself in the poor treatment towards Mr. Haydar and, perhaps more importantly, his subsequent termination.” (Report at 3.) Effectively, then, Blancero opines that Faricy and perhaps others who evaluated Haydar were, unknown to them, negatively biased against Muslims or Syrians and that those biases contributed to Haydar’s negative feedback, performance reviews, and termination. Although the logic of this section of the report is somewhat difficult to follow, it appears that Blancero reaches her conclusion that unconscious bias played a role in Haydar’s treatment as follows. Blancero begins with the premise that “there is implicit bias against Muslims today.” (Report at 3.) Then, relying on a paper by researchers at the University of Maryland, Blancero

states that an organization’s climate for racial bias influences the organization’s ratings of job applicants. (Id.) She says that she can “extrapolate” from this study “a similar climate for religious bias.” (Id.) Blancero then attempts to establish that there was a “climate for discrimination” or bias against Haydar. (Id.) In support of this premise, Blancero says that Faricy perpetuated a culture of “white males” and that the climate was not inclusive of those who did not fit the white-male profile. (Report at 4.) Blancero then faults Faricy for, on the one hand, testifying that he went to a training on unconscious bias, but, on the other hand, testifying in a way that, in Blancero’s opinion, revealed “no such awareness of his own biases.” (Id.) Specifically, the report characterizes Faricy’s testimony as suggesting that he or Amazon was “colorblind.” (Id.) The problem with this, says Blancero, is that it misunderstands that biases can be unconscious. (Id.) Blancero maintains that the colorblind concept also reduces engagement by minorities and increases minorities’ perception that their white colleagues are biased. (Id.) One broken link in this chain of reasoning is the premise that Faricy perpetuated a culture of “white males” and that the climate was not inclusive of those who did not fit the white-male

profile. It appears that Blancero’s opinion about the culture that Faricy perpetuated—in an organization with thousands of employees—is based on her review of Faricy’s treatment of Haydar standing alone (e.g., Faricy’s “you people” comments) and Faricy’s treatment of Haydar relative to Faricy’s treatment of Stefan Haney, a white man at one management level higher than Haydar’s. (See Report at 3 (stating that Faricy mocked Haydar about how he treated his wife); id. at 6–7 (quoting Faricy’s comments about Haney); id. at 8 (comparing Haney’s and Haydar’s behavior and asserting disparate treatment); ECF No. 114, PageID.11757.) But, in this Court’s view, how Faricy treated two people is an insufficient factual basis to infer that Faricy perpetuated a white- male culture. Again, thousands of Amazonians—of various races, nationalities, and religions—

were under Faricy’s branch of the Amazon organization chart. Yet Blancero never went onsite to observe the ranks under Faricy. Blancero never interviewed any of Faricy’s direct reports. Blancero never read DeCleene’s interviews of some of Faricy’s direct reports. And Blancero never compared, in aggregate, the evaluations of whites and non-whites (otherwise similarly situated) who worked under Faricy. So Blancero lacks the factual basis to opine that Faricy perpetuated a culture favoring white men and a climate that was not inclusive of non-white men. And even setting aside the specific reasoning set out in the report, the Court is not convinced that the concept of unconscious bias would help the jury decide the key issue of intentional discrimination in this case. The jury will hear about Faricy’s “you people” comments. And the jury will be able to compare how Faricy (and others) viewed the leadership deficiencies of non-Muslims and non-Syrians (e.g., Haney) and how they viewed Haydar’s leadership deficiencies. From that evidence, a jury is fully capable of inferring that Faricy or others were (or were not) biased against Muslims or Syrians. With case-specific evidence, the jury does not need to base a finding of negative biases on the general concept that all people hold some type of

unconscious bias. See E.E.O.C. v. Wal-Mart Stores, Inc., No. 6:01-CV-339-KKC, 2010 WL 583681, at *4 (E.D. Ky. Feb. 16, 2010) (“The burden . . . is on the plaintiff to prove that intentional discrimination occurred at this particular distribution center, not just that gender stereotyping or intentional discrimination is prevalent in the world.”). Worse, introducing the concept of unconscious bias based on inadequate data and analysis runs the risk of unfairly tilting the scales in Haydar’s favor. See Fed. R. Civ. P. 403. Aside from comments about and treatment of Haydar (and, again, a jury can draw its own inferences from that evidence), Blancero has little evidence that those who provided feedback or evaluated Haydar held unconscious, negative biases toward Muslims or Syrians. She does not have the results of any

implicit-association test, she did not observe the workplace, and she did not conduct any type of statistical analysis suggesting hidden bias. So apart from Haydar’s situation, Blancero has little to infer that those who assessed Haydar’s leadership held a negative, unconscious bias toward Syrians or Muslims. And Blancero has even less to infer that Haydar’s evaluators acted on any unconscious bias that they might have had. See Jones v. Nat’l Council of Young Men’s Christian Ass’ns of the U.S., 34 F. Supp. 3d 896, 900 (N.D. Ill. 2014) (finding plaintiffs had not established a “logical connection” between the mere existence of unconscious bias and “the results of employment decisions made by supervisors and managers who are armed with abundant data and are personally invested in the results of the process”); Childers v. Trustees of the Univ. of Pennsylvania, No. 14- 2439, 2016 WL 1086669, at *5–6 (E.D. Pa. Mar. 21, 2016). So there is weak foundation for Blancero to infer that unconscious bias played a significant role in Haydar’s termination. (See ECF No. 114, PageID.11678–11679 (testifying that it was “likely” but not “reasonably certain” that negative, unconscious biases affected Haydar’s evaluations).) Yet if the jury were to hear that every person holds some type of unconscious bias, it may well leap to concluding that Haydar’s

negative feedback was tainted by an unconscious bias against Muslims or Syrians. Given the weak foundation, that would be undue advantage. In short, the Court will not permit Blancero to recount to the jury the general concept of unconscious bias or the role it might have played in Haydar’s assessment, treatment, or termination. In the “Culture of Lack of Inclusion” section of her report, Blancero also provides general information about the difference between diversity and inclusion. (Report at 4.) According to her, diversity looks at numbers (e.g., a certain percentage of a corporation’s employees are Muslim) whereas inclusion looks at whether an individual is “made to feel part of the organization.” (Report

at 4.) Relatedly, she states that while almost all organizations promote diversity, ultimately it comes down to how specific leaders in the organization treat specific individuals. (Id.) The Court will allow Blancero to provide this general background information to the jury. Blancero’s distinction appears to be solidly within her expertise. (See Report at 4.) Indeed, it is based in part on her co-authored article, “Hired for diversity, rewarded for conformity.” (Report at 4.) So even if the typical juror is familiar with the difference between diversity and inclusion, the reliability of Blancero’s testimony ensures that the only prejudice is a bit of trial time, not misinformation. And because Amazon will undoubtedly stress to the jury that it has a very diverse workforce and that it values diversity, the testimony is relevant. Finally, in this section of her report, Blancero provides general information about discrimination against Muslims. (Report at 3, 4.) Although Blancero has expertise in certain concepts about diversity and inclusion that cut across groups, she admitted that she has never conducted any research specific to Muslims. (ECF No. 114, PageID.11600–11601.) It thus appears that Blancero’s specific knowledge about discrimination against Muslims comes from the articles

she cites in her report. But because Blancero does not have expertise in this area, her reliance on these articles is not much different from a lay person finding articles that discuss discrimination against Muslims. And again, Blancero did no analysis of how Amazon, or even its Marketplace division, treats Muslims beyond Haydar. Moreover, it appears that the purpose of establishing the existence of discrimination against Muslims is to make more plausible that Haydar was the victim of unconscious bias. Yet, as explained, the Court will not be permitting testimony on unconscious bias. For all these reasons, the Court will not permit Blancero to tell the jury that after the 9/11 attacks, the EEOC saw a 250 percent increase in discrimination against Muslims or that “Muslims may not be fully accepted in contemporary American society.” (Report at 3, 4.)

In short, assuming that an adequate foundation is laid at trial, Blancero may testify to the highlighted portions of the “Culture of Lack of Inclusion” section of her report. (See Appendix.) D. The last section of Blancero’s report is titled “Summary of Major Issues.” In this section, Blancero makes five points. The Court takes them in order. 1. Blancero first opines that the “subjective opinions” of various Amazonians about Haydar’s leadership were not “actually independent.” (Report at 5.) She points out that Faricy, the head of Marketplace, had expressed negative views about Haydar. (Id.) Blancero then says, “If someone in a leadership position speaks poorly about my peer it is likely that I’m going to echo those concerns. It could be groupthink or it could be political astuteness. Why would you voice a different opinion than your boss?” (Id.) But in answering this question—which seems to have any number of responses—the only support Blancero cites is a published article that appears to be a critique of “360 feedback”—not

an article on groupthink or political astuteness. And Blancero’s follow-the-leader concept appears to be grounded in common sense, i.e., it is likely within the knowledge of a typical juror. That said, if Blancero can in good faith testify at trial that she has special knowledge or expertise in the follow-the-leader concept (see ECF No. 114, PageID.11758), she may educate the jury as to the underlying psychology, e.g., groupthink or political astuteness. The Court will not permit extended testimony on this topic or testimony from Blancero that indicates that Gaw, Mosby, or others had a follow-the-leader mentality. Further, if Blancero testifies to the psychology underlying the follow-the-leader concept, Amazon would be free to show that while that mentality exists generally, it was not the reason that Amazonians under Faricy gave similar feedback about

Haydar’s leadership. 2. Next, Blancero says “most troubling is the ‘disagree and then commit’ leadership principle.” (Report at 6.) Amazon has 14 leadership principles; one is “have backbone; disagree and commit.” Amazon explains, “Leaders are obligated to respectfully challenge decisions when they disagree, even when doing so is uncomfortable or exhausting. Leaders have conviction and are tenacious. They do not compromise for the sake of social cohesion. Once a decision is determined, they commit wholly.” (ECF No. 68, PageID.8226.) Blancero thinks that this principle was applied unfairly to Haydar because he was retaliated against when he challenged his negative performance reviews. (Report at 6.) The Court will not permit Blancero to opine on whether Amazon fairly applied the “have backbone; disagree and commit” principle to Haydar. It is within a jury’s ability to decide whether Haydar complied with the principle (or, more importantly, whether Haydar’s superiors were

justified in finding that Haydar did not comply with the principle). Indeed, Blancero identifies nothing specific about her education or training that gives her special insight into Haydar’s compliance with the have-backbone-disagree-and-commit principle. And to the extent that her special insight is based on her knowledge of unconscious bias or a claim that Faricy promulgated a white-male culture, the Court has already described problems with those grounds for drawing an opinion. 3. “Third,” says Blancero, “the treatment of Mr. Haney compared to the treatment of Mr. Haydar exemplifies how the culture was such that a white male’s actual or perceived weaknesses

can be not merely overlooked, but can result in a promotion while an Arab Muslim man has a hyper focused review of subjective data—in the face of strong success with objective results.” (Report at 6–7.) The Court will not permit Blancero to testify in this manner. The Court has already explained (and explains further at the end of this opinion) that Blancero lacks an adequate factual basis or reliable methodology for inferring that the “culture” under Faricy was biased against anyone who was not a white male. And the Court has explained that the jury is competent to draw proper inferences from any disparate treatment between Haydar and Haney. The Court also has explained that strong “objective results” are not necessarily inconsistent with negative leadership scores as the two are independent axes in Amazon’s evaluation matrix. (Indeed, the matrix contemplates high performance and low leadership scores. (See R. 68, PageID.8227.)) To all of this, the Court adds that Blancero makes no attempt to explain how she reconciled distinctions between Haney’s and Haydar’s assessments, including that Haney’s performance reviews were not completed by Haydar’s supervisors. (Indeed, it is not even apparent that she reviewed Haney’s

performance reviews.) 4. Blancero next highlights that two people in Amazon’s HR department (Kaitlin McVey and Michael Beary) were not on the same page as to whether Haydar should be terminated. (Report at 7.) There is no reason for Blancero to testify about this inconsistency. The jury will hear evidence about McVey’s and Beary’s positions on Haydar’s employment and is more than competent to identify any inconsistencies. 5.

As her fifth “Major Issue[],” Blancero opines that “the overall focus on subjective data is just dangerous.” (Report at 7.) She cites a second article about the problems with “360 feedback” and opines that the tool is best used to develop—not assess—an employee. (Id.) (As the term “360 feedback” suggests, it includes feedback from an employee’s supervisors, peers, and reports.) While the Court has concerns about whether Blancero selected articles that support her view of 360 feedback, the Court will permit Blancero to testify about the weaknesses of 360 feedback. The topic appears to be within Blancero’s education and research background: she holds a doctorate in human-resource management and has taught a course on employee selection and appraisal. (ECF No. 114, PageID.11565.) E. Lastly, the Court addresses what appears to be Blancero’s bottom line. Blancero says that the “Introduction” section of her report was essentially an abstract of her entire report. (ECF No. 114, PageID.11756.) In the introduction, she states, “In my professional opinion, these subjective and biased comments [about Haydar] have resulted in a culture that allowed for discriminatory

actions to occur.” (Report at 2.) And in the final portion of the “Summary of Major Issues” section of her report, Blancero states, “It is my opinion, based on my analysis above that Mr. Haydar was not treated fairly with regard to his performance evaluations, overall treatment, retaliation from bringing up these issues, and subsequent termination.” (Report at 8.) The opinion in the introduction is neither based on sufficient data nor a reliable methodology. It appears that Blancero is extrapolating from a few Amazonians’ treatment of Haydar that Amazon’s “culture” allowed for discriminatory actions to occur. In this Court’s view, Blancero inappropriately expands a single data point—the treatment of Haydar—into a sweeping generalization about Amazon. And even if Blancero meant the culture of those under Faricy,

thousands were under Faricy’s branch of Amazon’s organization chart. So it would remain that Blancero is opining on the culture among thousands of Amazonians from her review of a single incident. As for the opinion found at the end of Blancero’s report, it appears to be an opinion on the ultimate issue in this case: whether Haydar’s reviews, feedback, and treatment were “fair[]” or whether they were the partial product of bias toward Muslims or Syrians. (See Report at 8.) In his response to Amazon’s motion to exclude Blancero’s opinions from trial, Haydar has agreed that Blancero will not opine “on the ultimate issue of whether intentional discrimination occurred against Plaintiff.” (ECF No. 116, PageID.11948.) The Court believes this concession includes Blancero’s opinion that “Haydar was not treated fairly with regard to his performance evaluations, overall treatment, retaliation from bringing up these issues, and subsequent termination.” (Report at 8.) IV. For the reasons given, the Court GRANTS IN PART and DENIES IN PART Amazon’s

motion to exclude the opinions, reports, and testimony of Blancero. Assuming proper foundation, Blancero may testify in accordance with the portions of her report highlighted in the appendix to this opinion. SO ORDERED.

Dated: October 10, 2019

s/Laurie J. Michelson LAURIE J. MICHELSON UNITED STATES DISTRICT JUDGE Report on Abdullah Haydar Prepared by Donna Maria Blancero, PhD January 5, 2018

My name Donna Maria Blancero and I have been researching issues of diversity and inclusion for 20 years. My PhD is in Human Resources from Cornell University and my dissertation was on procedural justice and fairness in complaints. I am currently Associate Dean of Business at Bentley University in Waltham, Massachusetts. Included in my work at Bentley is teaching a course in managing diversity in the workplace. As well, I am heavily involved in training the faculty and staff in diversity and inclusion. I have published over 35 papers/chapters, and am a national speaker on issues of diversity. Additionally, I work with several business organizations on inclusion issues. My curriculum vitae is attached. My report is based on my reading of the materials provided to me on Mr. Abdullah Haydar and my own expertise in this area. Specifically, these materials included: 1. Deposition of Abdullah Haydar, June 28, 2017 2. Deposition of Garret Gaw, July 17, 2017 3. Deposition of Derek Oehler, July 21, 2017 4. Deposition of Shelly Cerio, July 24, 2017 5. Deposition of Peter Faricy, July 28, 2017 6, Deposition of Stefan Haney, November 7, 2017 7. Deposition of Kaitlin McVey, November 14, 2017 8. Deposition of Anne Decleene, November 14, 2017 9. Deposition of Mr. Berry, November 17, 2017 10. Selected documents from defendant’s document preparation 11. Review of the relevant academic and practitioner literature BLANCERO/HAYDAR

Introduction Issues of inclusion and discrimination are never simple and typically include several layers of complexity. My job here was to review the materials provided and, based on literature in the field as well as my own expertise, provide my professional opinion. That said, I find several issues in this particular case to be amazingly straightforward and obvious. There are evaluations of Mr. Haydar that are objective — based on goals and hard data. These cannot be contested and these evaluations are positive ones. Clearly, Mr. Haydar was effective in his job. We also find subjective data that are inconsistent and change within a very short time period. These subjective data include positive attributions regarding him and certainly some areas of improvement. In my professional opinion, these subjective and biased comments have resulted in a culture that allowed for discriminatory actions to occur. The data provided are inconsistent with Amazon’s statement of non-discrimination; moreover, the data support a culture that would allow for discrimination to occur. Diversity and Inclusion Stereotyping is when we take entire groups and assign characteristics to everyone based on their membership of that group or that social identity. Additionally, confirmation bias is when one searches for data to support one’s preexisting beliefs. This happens when individuals gather — or remember — data or information that supports their behavior; they interpret events in a biased manner. These type of cognitive biases appear to be occurring in this case. Based on the facts of this case, Mr. Haydar was stereotyped in a negative manner and his supervisors searched for subjective data to support a poor evaluation rather than depending on the very clear objective data. This is a classic example of confirmation bias. As just one example, we see an email from Garret Gaw to Mr. Haydar saying, in part, “... Since you moved to the office, you've been phenomenal Detroit team member and Marketplace leader. You lead by example ...” Mr. Gaw later becomes Mr. Haydar’s immediate supervisor. Why was this not considered in his overall evaluation?

BLANCERO/HAYDAR

Treatment of Mr, Haydar and Subsequent Investigation Mr. Haydar has stated that he was treated differently — poorly — based on his religion and/or nationality. This was done in private and public, for example he was mocked about how he treated his wife by Mr. Faricy, yet Mr. Faricy didn't even know his wife. (When the investigation took place, under Shelly Cerio’s direction, it appears that she did not investigate based on bias (elated to national nationality Or religion, Clearly, that was the reason for the treatment, what else would it have been? If we look at standard human resources procedure for investigating internal ‘complaints they need to follow specific processes.' The literature on complaint systems strongly recommends adherence to procedural fairness rules. For example, the data need to be accurate, unbiased, ethical, and consistent, among others.’ None of these seem to be followed. Culture of lack of inclusion It's well-established that there is implicit bias against Muslims today. After September 11, 2001, the Equal Employment Opportunity Commission (EEOC) saw a 250% increase in discrimination against Muslims.* A study conducted by researchers at the University of Maryland look at climate for racial bias, and we can extrapolate from this study, a similar climate for religious bias. In this study, they found that organizational climate influenced applicant ratings’. In Mr. Haydar’s case, it is quite reasonable to come to the conclusion that there is implicit bias given the climate and the tone of Mr. Faricy. Based on this I believe we can say that that bias manifested itself in the poor treatment towards Mr. Haydar and, perhaps more importantly, his subsequent termination. It seems clear that there was a climate of discrimination or at least bias against Mr. Haydar. What is striking to me was that not only was Mr. Haydar a sought-after mentor from many people but that he was actually invited to speak at an internal Amazon conference. It begs the question — why did you let allow someone to speak at such a conference when that person is not doing the job and did not embody leadership principles? There are so many examples of inconsistencies

a BLANCERO/HAYDAR

here that show excellence and leadership success — measured objectively — and then on the other hand, incredible bias and inappropriate treatment.

Blacks, Muslims, Latinos, etc. in the organization. That's quite objective, we can state that with certainty the same way that we can state Mr. Haydar was effective in his job. HS/objective, We

and if they are made'to feel part of the organization) For me it's quite clear that there was not inclusion for Mr. Haydar, based on his identity as a Muslim and/or his national origin. His objective success as well as the feedback that he asked for (including HR telling him that he was doing a fine job) and then he is managed out? (While almost all organizations promote issues people. It is the actions of the specific leaders involved and how they are treating individuals, It appears that the culture, perpetuated by Mr. Faricy, was one of white males and it was a poor climate for inclusion for those who did not fit that specific profile. Given the size of Amazon, I would expect more attention to be paid to such issues. Perhaps one of the things that is most unbelievable is that both Ms. Cerio and Mr. Faricy state that no one has been discriminated against at Amazon. This appears to be next to impossible, in particular given the statistics provided earlier. A study in Personnel Psychology, a leading journal in the human resource area, states the challenges encountered by Muslim individuals in selection contexts and highlights the need to understand and reduce discrimination toward religious minorities. The authors state that “despite claims to recognize and respect a wide range of religious affiliations, Muslims may not be fully accepted in contemporary American society,”' This is an unfortunate reality. I found it revealing that Mr. Faricy states that he is aware of the concept of unconscious bias and acknowledges that all human beings hold such biases®. He even states that he went to training ne ee BLANCERO/HAY DAR

offered by Amazon and mentions the research done at Harvard in this area. Yet, his responses to questions on unconscious bias illustrate that he has no such awareness of his own biases by suggesting that everyone is treated the same and that he has never intentionally been biased against someone. Unconscious bias, by definition, is that one is unaware or ‘unconscious’ of such biases. More concerning to me and more illustrative of the culture that seems to be allowed is the comment regarding everyone being treated the same. His responses about different cultures, stereotypes, and how everyone is treated the same is not merely showing a lack of understanding of unconscious bias but suggests he is ‘colorblind.’ As Emerson states in her article, “downplaying demographic differences reduces the engagement of underrepresented employees and increases their perceptions of bias from their white colleagues.”? The comments from such a high ranking executive suggests a lack of awareness of what it takes to have a positive diversity climate and, in my opinion, allows for a diversity climate that would be blind to these issues. Once again, I do not see any data that supports a positive or aware diversity climate or culture.

Summary of Major Issues First, Mr. Haydar’s objective performance reviews are positive. For example, there is agreement that he hired and retained an effective team. His subjective reviews are mixed. There are several issues here. How can Mr. Haydar exhibit behaviors that are characterized as ineffective and yet still succeed in his work? Are the subjective opinions of others actually independent? I am questioning that perceived independence; you have a strong and powerful leader in Mr. Faricy who voices a negative opinion about Mr. Haydar in front of others that can certainly influence those individuals. If someone in a leadership position speaks poorly about my peer it is likely that I’m going to echo those concerns. It could be groupthink or it could be political astuteness, (Why Would You Voice a different opinion than your boss? | am not suggesting that they were told that they had to agree with Mr. Faricy, but in my opinion they would be inclined to do so if they wanted a successful career under him, (ml@reView article of Various performance review methodologies, Aggarwal & Thakur (2013)'° report that a weakness with 360 reviews is that Pr ee ee Se Se ee ly BLANCERO/HAYDAR

they are “Prone to political and social games played by people.” Additionally, they argue that Such reviews are “Sensitive to organization and national culture” Both of these statements apply in this case. There was some political awareness of the ‘boss’s’ view, and, in my professional opinion, the organizational culture that allowed for such behavior. An example of this in in Philipsen’s 360 comments!'. First, he provides input about objective outcomes from Haydar, specifically about hiring “ahead of the ramp plan” and bringing in a “great set of people.” Then he has this long list of subjective criticism. Gaw then, in his deposition, appears to rely on this when he is discussing the ratio of Haydar’s feedback with Haydar having a smaller ratio of ‘good to bad’ than he would have wanted to see. In my opinion, it is taking poor data and using them a second time to draw conclusions that are subjective. Second, most troubling is the ‘disagree and then commit’ leadership principle. Based on the facts that I read Mr. Haydar was exemplary in voicing his concern about his review. A leader should not just ‘fall in line’ but should voice concern and provide data to support the concern. It appears to be quite arbitrary to decide when and how one stops voicing his concern. Ethically, one should expect a leader to disagree — especially with such powerful facts (i.e., objective review) — to stand up for himself. That is exactly the behavior exhibited by Mr. Haydar. Retaliation is an important human resource principle yet it appears that he retaliated against for his complaint. Mr. Gaw and Mr. Oehler say that Mr. Haydar won’t give up on fighting his reviews and that this is a problem, which speaks directly to their intent to retaliate against him. Examples of this include Faricy’s email of June 9, 2013!* where he advises Haydar to “keep your head down and shut up.” So, we have a tangible example of inconsistency here — Haydar working to the leadership principle of disagreeing and then being told to shut up. Further, we have Faricy then directing a subordinate (Ramiah) to exclude Haydar from meetings. '° Third, the treatment of Mr. Haney compared to the treatment of Mr. Haydar exemplifies how the culture was such that a white male’s actual or perceived weaknesses can be not merely overlooked, but can result in a promotion while an Arab Muslim man has a hyper focused review

EE BLANCERO/HAYDAR

of subjective data — in the face of strong success with objective results. The difference in the treatment of these two men is striking. Note the feedback that Faricy gives to Haney": Stefan was the very first I hired to the Marketplace team in January of 2009, This was for great reasons. Stefan has unique skill set to effectively lead both tech and product teams to excellence. Since 2009 you've been our positive face of the seller business within Amazon, my go-to buyer experience expert and go take-the- hill leader, I will miss the sound of your booming voice and contribution to the Marketplace team but am equally excited to see your impact within NA sales team.: The ultimate compliment is to be respected by your peers and are well respected. You made a big difference in 2014. Thanks for your hard work, leadership and loyalty.

Fourth, there are inconsistent statements surrounding the discussions of Mr. Haydar by Ms. McVey and Mr. Berry. On one hand, Mr. Berry says that they had an open mind, while McVey suggests that he should be terminated. In fact, the suggestion for termination is before Mr. Haydar is provided a performance plan. Incredibly, Berry did not know that McVey even sent an email to Gaw making this suggestion and, moreover, was something he would not expect.'> Fifth, the overall focus on subjective data is just dangerous. The literature on best practices for 360 reviews recognizes the potential weaknesses in such a system. These subjective data points are best utilized for development, not for evaluation. As stated in the Harvard Business Review in an article entitled: The Fatal Flaw with 360 surveys.'° In this article, performance review expert Marcus Buckingham articulates two statements that I find most relevant to the case in hand. When discussing the flaws of such data he states that “It's the data itself. The data generated from a 360 survey is bad. It’s always bad.” As well, he states (bolded is original author’s markings) “The bottom line is that, when it comes to rating my behavior, you are not objective. You are, in statistical parlance, unreliable. You give us bad data.” Of course, such subjective evaluations can be quite useful as additional information for developmental purposes, but not for performance reviews and/or grounds for termination. BLANCERO/HAYDAR

It is my opinion that, based on my analysis above that Mr. Haydar was not treated fairly with regard to his performance evaluations, overall treatment, retaliation from bringing up these issues, and subsequent termination. Additionally, the investigation did not follow standard processes for complaints. If one looks closely at the facts, you will inconsistent treatment, and inconsistent reports on what was being done (performance plan vs. termination plan). Mr. Haydar did quite well objectively, was treated differently than another employee (Mr. Haney) with regard to outcome (fired vs. promoted), and was retaliated for strongly questioning his review. Then we have Ms. Cerio and Mr. Faricy stating that no one has ever been discriminated against at Amazon. Let that sink in fora moment. It is indeed a ludicrous statement. With hundreds of thousands of people employed at Amazon, regardless of what official policies are in place, processes are implemented by so many people that, statistically, one would expect some issues of discrimination to occur. Finally, it is clear that the culture was unaccepting of people who were not of the ‘dominant’ culture, in other words, the dominant culture of those who identify as White (non Arab/Muslim),. This connects to my earlier discussion of diversity and inclusion. Based on my review of the materials provided, the culture was such that those who did not identify as such were not given the latitude of behaviors that White employees were provided. This is a classic example of a lack of inclusion — it becomes an issue of ‘you can work here, but you need to fall in line and act a certain way.’ So, in this case, behaviors exhibited by Mr. Haney are okay — and lead to a promotion, while for Mr. Haydar these behaviors result in termination. In my view, this illustrates a lack of inclusion and an inclusion climate that is not accepting. Amazon has good policies that should allow for a lack of discrimination yet, in practice, these do not appear to have been followed in this case. In my opinion, if you review these issues, you will see that the culture that prospered here is one that allowed discrimination to occur and possibly flourish.

a BLANCERO/HAYDAR

1Blancero, D.M, DelCampo, R. G., & Marron, G. F, (2010). Just tell me! Making alternative dispute resolution systems fair. Industrial Relations: A Journal of Economy and Society, 49(4), 524-543. 2 Colquitt, J. A., & Rodell, J. B. (2015), Measuring justice and fairness. Oxford handbook of Justice in the workplace, 187-202. 3 https://onlabor.ore/workplace-discrimination-against-muslims/ Retrieved 11/2/17. 4 Ziegert, J. C., & Hanges, P. J, (2005). Employment discrimination: the role of implicit attitudes, motivation, and a climate for racial bias. Journal of Applied Psychology, 90(3), 553. 5 Social identities include race/ethnicity, gender, age, religion, sexual orientation, socioeconomic status, nationality, among others. 6 Blancero, D. M., DelCampo, R. G., Marron, G. F. (2007). Hired for Diversity yet Rewarded for Conformity: Hispanics in Corporate America. The Business Journal of Hispanic Research, 1(1), 12-25, 7 King, E. B., & Ahmad, A. S. (2010). An experimental field study of interpersonal discrimination toward Muslim job applicants. Personnel Psychology, 63(4), 881-906, *Faricy deposition, July 28, 2017. ° Emerson, J. (2017) Colorblind Diversity Efforts Don’t Work Harvard Business Review https://hbr.org/2017/09/colorblind-diversity-efforts-dont-work#comment-section 0 Aggarwal, A., & Thakur, G, S. M. (2013). Techniques of performance appraisal-a review. International Journal of Engineering and Advanced Technology (IJEAT), 2(3), 2249-8958. 1 Bates 0473 22 Bates 2201 3 Bates 642 44 Haney deposition 45 Berry deposition 6 Buckingham, M. "The fatal flaw with 360 surveys." Harvard Business Review (2011).

era BLANCERO/HAYDAR