Appendix A
PowerPoint slides
Plaintiffs’ slides
Plaintiffs’ slide 1
Plaintiffs’ slide 2
Plaintiffs’ slide 3
Plaintiffs’ slide 4
Plaintiffs’ slide 5
Plaintiffs’ slide 6
Defendant’s slides
Defendant’s slide 1
Defendant’s slide 2
Defendant’s slide 3
Defendant’s slide 4
Defendant’s slide 5
Defendant’s slide 6
Appendix B
Excerpts from opening statements
These excerpts from the parties’ opening statements contain the verbal descriptions of the evidence to which the slides dealing with statistical evidence correspond. (Due to space constraints we are not providing excerpts corresponding to the non-statistical slides.) Slide numbers correspond to the numbering of the slides in Appendix A.
Plaintiffs’ statement
Slide 1:
Here are a few pertinent numbers for you to consider: Even though, in 1988, 58% of the available Metropolitan workforce was African-American, only 18% of Tri State Railroad’s employees were African-American. Close to 15 years later, in 2001, even though it remains true that almost 60% of the workforce in general is African-American, the percentage of African-Americans in Tri State Railroad’s workforce has increased to a mere 20%. That’s just a 2% increase in almost 20 years and still a gross underrepresentation of African-Americans relative to the local workforce.
Slide 2:
Dr. Zellner [the plaintiffs’ expert] will show you that throughout the period 1988-2001, African-Americans were one-third less likely than white employees to be promoted. That’s the figure for the entire company, taken as a whole. And that’s a statistically significant difference, which means that it’s very unlikely to be due to chance or any factor other than the employees’ race.
Slide 3:
Dr. Zellner will testify that he ran regression analyses of the number of times employees were disciplined from 1990 to 1994 and from 1995 to 2001. Over both periods of time, African-Americans were disciplined more than their White counterparts and more severely. In fact, Black workers were more likely than Whites to be dismissed for any given infraction. The evidence will show: (1) On a company-wide basis, African-American employees were much more likely than Whites to be disciplined for the same infractions. (2) On a position by position basis, African-American employees were significantly more likely than Whites to be disciplined in no fewer than 20 of the 48 job titles represented by the plaintiff class.
Slide 4:
[This excerpt follows immediately after the one illustrated by Plaintiffs’ slide 3] (3) And on a company-wide basis, African-American employees were disciplined more severely than Whites, controlling for the severity of the infraction. . . . Dr. Evans [the defendant’s expert] will say that there is an alternative explanation [to racial discrimination], that African-Americans are disciplined more often than whites because African-American employees commit more of the infractions for which discipline is properly meted out. We will show that with respect to conductors’ failure to remit ticket money, an important infraction and one affecting the position in which the fifth-most African-Americans are employed, that is simply not true. African-American and White conductors fail to remit at statistically identical rates, so that the disparity in punishment – African-Americans are disciplined more often – can only be explained in terms of racial discrimination.
Defendant’s statement:
Slide 2:
The plaintiffs in this case will try to convince you that [the defendant’s] efforts [to evaluate prospective employees fairly for hiring and current employees for promotion] are insufficient because, they say, African-Americans were underrepresented at Tri State Railroad and especially in higher level positions, and that the picture hasn’t changed much in the nearly 20 years since Tri State took over from Conrail. . . . But [this] is not true. For instance, plaintiffs’ own evidence shows that between 1988 and 2001, the percentage of African-Americans in Tri State’s entire workforce has increased by 27%, not 2%, and that the percentage of African-Americans in management-level positions has increased, not by 2%, not by 4%, but by 72%.
Slide 3:
We will also show you that plaintiffs’ purported statistical evidence paints a false and wholly misleading picture. . . . For instance, Dr. Zellner [the plaintiffs’ expert] claims that African-Americans were promoted less often than Whites. How does he arrive at this conclusion? He does it by simply lumping together all of the promotion decisions throughout the company, throughout all departments and locations, over an extended period of time, “controlling” in some fashion for some simple variables like department, age, and tenure, and observing that there is an overall disparity in the rates at which African-Americans and Whites are promoted.
But you can’t do that. You can’t group together all sorts of employees who perform entirely different tasks and who, therefore, understandably would be expected to have different promotional experiences. . . . Dr. Evans [the defendant’s expert] will show you that by grouping the data together, one can actually create the false impression – a misleading impression – that there is a statistical difference when in fact there is none. Once we look at employees who have more common experiences and skill levels – that is, on a job by job or department by department basis – an entirely different picture emerges.
The most obvious example is when we look separately at the unionized employees, who comprise 90% of the workforce. Our reports will show, and Dr. Zellner will concede, that if he were to run his so-called regression analysis just on these union-represented employees, there is absolutely no statistical evidence of a difference in promotion rates between African-Americans and Whites. Nine out of 1,481 African-American unionized employees were promoted, or 0.6%; 47 out of 5,724 White unionized employees were promoted, or 0.8% – no significant difference. And if we look at those jobs where employees do get promoted more often, that is, within management, again we see no evidence of any significant disparity in promotion rates: for African-Americans, 16 out of 161management-level employees were promoted, or 9%; 106 out of 989 White management-level employees were promoted, or 11.8% – again, no significant difference. Only by aggregating promotion data across union positions, where neither Whites nor African-Americans get promoted very often, and managerial positions, where both African-Americans and Whites are regularly promoted (at 10 times the rate of unionized positions), does plaintiffs’ expert create an illusion of disparity in promotion rates. But that’s all it is, an illusion.
Slide 4:
Dr. Evans will also show you that Dr. Zellner’s analyses of Tri State Railroad’s disciplinary decisions are flawed, for much the same reasons. He will explain that once again Dr. Zellner combines together all departments and all positions to arrive at his conclusion that African-American employees are disciplined at a greater rate than White employees. By doing this, he again conceals from you important facts that demonstrate that there is in fact no company-wide pattern and practice of discrimination against African-Americans in discipline.
For example, in 25 of the defendant’s 37 departments, there was no disciplinary action at all during this period. None. It is nonsensical to claim “company-wide discrimination” when the incidence of discipline itself is so limited to certain departments. Furthermore, the evidence will show, and there will be no disputing, that of the 48 positions in the remaining 12 departments represented by the plaintiff class, in the majority of these positions, 28 out of 48, the difference in discipline was not statistically significant. What does that mean in plain English? It means that the differences do not show discrimination. But even further, for those remaining 20 positions, if we take into account length of service, it turns out that the so-called statistically significant disparity disappears.
Slide 5:
[This excerpt follows immediately after the one illustrated by Defendant’s slide 4] Even more important than Dr. Zellner’s failure to properly group the employees in his formulas is his failure to consider who is deserving of discipline. In other words, he condemns the discipline without looking at the misconduct. You will learn, for example, that there is objective, unbiased evidence collected on Tri State’s database showing who among Tri State’s employees were committing several of the types of infractions – attendance violations, drug abuse, insubordination – infractions that comprise more than half of the disciplines imposed. In other words, these data tell us who is deserving of discipline. It will show that the discipline experiences of African-Americans and Whites are in fact consistent with the rates at which the underlying infractions are committed. That is, African-American employees get disciplined more often because they commit more of these kinds of infractions.
Appendix C
Dependent measures of recall/understanding
Thinking back to the opening statements, please indicate whether each of the following items is TRUE or FALSE by CIRCLING the word true or false INSIDE THE PARENTHESES at the end of each item:
(a) The percentage of all African-American employees at the defendant company increased by 2% from 1988 to 2001 (True, False)
(b) The percentage of African-American employees at the defendant company increased by 72% from 1988 to 2001 (True, False)
(c) Considering only promotions of union employees to the management-level, African-American and White employees are promoted at the same rate (True, False)
(d) Considering only promotions within management-level (i.e., non-union) employees, African-American and White employees are promoted at the same rate (True, False)
(e) Across the defendant company as a whole, African-American employees are one-third less likely to be promoted than White employees (True, False)
(f) African-American conductors are disciplined three times more severely than White conductors for failing to remit fares (True, False)
(g) In 25 of the defendant’s 37 departments, there was no disciplinary action at all during the relevant period (True, False)
(h) Controlling for the underlying rates of infractions, Africa-American employees are disciplined more often than White employees for attendance problems, insubordination, and failing drug tests (True, False)
(i) There is no difference in the rates of disciplinary action against African-American and White employees in most of the defendant’s departments (True, False)
Note. In Study 1 the recall items were (c), (d), (e), (h) and (i). Items (a), (b), (f), and (g) were added while (i) was dropped in Studies 2A and 2B.
Appendix D
Judge’s instruction (Studies 2A and 2B)
According to the law, you should find the defendant, Tri State Railroad, liable (legally responsible) to the plaintiffs, its African-American employees, for racial discrimination only if the plaintiffs persuade you that: (a) Tri State Railroad has engaged in a policy, pattern, or practice of racial discrimination against the plaintiffs, or that (b) Tri State Railroad’s policies and practices have resulted in a statistically significant difference between how its African-American employees and its White employees have been treated.
The plaintiffs’ burden of proof is to persuade you that statement (a) and/or statement (b) above are more likely than not true (what is sometimes called “by a preponderance of the evidence”). So, you should find the defendant liable for racial discrimination if you believe that, more likely than not, either or both of those statements are true.
Your decision must be based solely on the facts and arguments you have heard. You should not be swayed or influenced by any sympathy or prejudice for or against either of the parties.