1 1 UNITED STATES DISTRICT COURT EASTERN DISTRICT OF MICHIGAN 2 SOUTHERN DIVISION 3 4 BARBARA GRUTTER, For herself and all others 5 Similarly situated, 6 Plaintiff, 7 v. Civil Action No. 97-CV-75928 8 LEE BOLLINGER, JEFFREY LEHMAN, DENNIS SHIELDS, and REGENTS OF 9 THE UNIVERSITY OF MICHIGAN, 10 Defendants. _________________________________________/ 11 12 BENCH TRIAL - VOLUME 13 13 MONDAY, FEBRUARY 12TH, 2001 14 15 BEFORE THE HONORABLE BERNARD FRIEDMAN United States District Judge 16 Theodore Levin United States Courthouse 231 West Lafayette Boulevard, Room 238 17 Detroit, Michigan 18 - - - 19 Appearances: 20 Kirk O. Kolbo, Esq., 21 R. Lawrence Purdy, Esq., 22 On behalf of the Plaintiff, 23 John Payton, Esq., 24 Craig Goldblatt, Esq., Stuart Delery, Esq., 25 On behalf of the Defendants Bollinger, et al, 2 1 - - - 2 APPEARANCES (Continued): 3 4 George B. Washington, Esq. Miranda K. S. Massie, Esq. 5 On behalf of Intervening Defendants. 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Joan L. Morgan, Official Court Reporter 21 Proceedings recorded by mechanical stenography. Transcript produced by computer-aided transcription. 22 23 24 25 3 1 2 I N D E X 3 WITNESS: PAGE: 4 STEPHEN RAUDENBUSH 5 Direct Examination by Mr. Delery 6 Cross-Examination by Mr. Kolbo 24 6 FRANK WU 7 Direct Examination by Ms. Massie 36 8 Cross-Examination by Mr. Payton 88 Cross-Examination by Mr. Purdy 99 9 FAITH SMITH 10 Direct Examination by Ms. Massie 153 11 12 13 E X H I B I T S 14 RECEIVED 15 Trial Exhibits Nos. 226, 227 4 Trial Exhibit No. 175 41 16 Trial Exhibit No. 171 178 Trial Exhibit Nos. 202-210 179 17 18 19 20 21 22 23 24 25 GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 4 1 2 Detroit, Michigan 3 Detroit, Michigan 4 Monday, February 12th, 2001 5 9:40 a.m. 6 - - - 7 THE COURT: I'm sorry I'm late. I had an 8 Immigration Swearing In. I wasn't going to read the names, 9 but I ended up having to read all the names. Stephen did it 10 for me. He told me on the way back he now knows why on Ellis 11 Island people's names are changed after he had to read all the 12 names. I apologize. I'm sorry I'm late. 13 Okay, ready to roll? 14 MR. KOLBO: Your Honor, just briefly a couple of 15 things. We looked at things again this weekend, and decided 16 that in view of the trial -- in the interest of assuring that 17 we get finished here this week, we have withdrawn Gail Heriot 18 as a witness. I've informed counsel of that last evening so 19 they are aware of that as well. 20 THE COURT: Okay. 21 MR. KOLBO: The other thing is, your Honor, we have 22 taken care of the drawings of Professor Larntz's larger 23 drawing. I've shown that to counsel. I believe there are no 24 objections to it. At this time we would offer Exhibit 226 and 25 227. GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 5 1 THE COURT: Both of them are saying yes so we'll 2 accept them in that form? 3 MR. KOLBO: Yes, and I have two copies here. 4 THE COURT: Great. I appreciate it. 5 MR. KOLBO: And your Honor, we will -- I think you 6 indicated you would like someone to take custody of the actual 7 large drawing and we would be happy to do that. 8 THE COURT: That's great. Perfect. Okay. 9 MS. MASSIE: Good morning, Judge. Just a couple 10 scheduling things. In all the shifting that's happened around 11 today and tomorrow, we have two inefficiencies that I 12 apologize for. One is that our two witnesses for today won't 13 be getting in probably in time to testify until about 14 something like 11:00 or even 11:15. I apologize for that. 15 And the second thing is we don't have any witnesses who can 16 testify tomorrow afternoon. As it happened we filled up today 17 -- anyway I won't go into all the logistics -- 18 THE COURT: No problem. 19 MS. MASSIE: But we should without any question be 20 done with the testimony by Thursday. 21 THE COURT: You have no witnesses this morning? 22 MS. MASSIE: This morning we do, but we may have to 23 take a break after Professor Raudenbush. In other words, 24 before our people get here -- 25 THE COURT: Not a problem. And then tomorrow GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 6 1 everybody agrees we won't go at two, or any other time 2 tomorrow? 3 MS. MASSIE: Correct. 4 THE COURT: And Wednesday we'll be off. And then 5 we'll go Thursday, and we'll finish up with Closing Arguments 6 Friday? 7 MS. MASSIE: Right. 8 THE COURT: Good. 9 MS. MASSIE: Great. 10 THE COURT: It will work out perfect. 11 Are we going to recall -- 12 MR. DELERY: Yes, your Honor. It will be very -- 13 THE COURT: It can't too much. 14 MR. DELERY: It will be brief and expeditious. 15 The defendants call Professor Stephen Raudenbush. 16 THE COURT: I know you've been sworn once but we've 17 been re-swearing the witnesses. 18 Do you solemnly swear or affirm to tell the truth in 19 the matter now pending before this Court? 20 THE WITNESS: I do. 21 THE COURT: Please have a seat. 22 S T E P H E N R A U D E N B U S H , 23 being first duly sworn by the Court to tell the truth, was examined 24 and testified upon his oath as follows: 25 DIRECT EXAMINATION GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 7 1 BY MR. DELERY: 2 Q Good morning, Professor Raudenbush. 3 A Good morning. 4 Q You were here on Saturday for Dr. Larntz's rebuttal 5 testimony; is that right? 6 A Yes, I was. 7 Q I'd like to talk about the main points that he raised on 8 Saturday, just on Saturday. And I think I'm just going to take 9 them in the order that he raised them on Saturday. The first 10 thing he did was present a new chart that compared odd ratios 11 to probabilities for various baseline probabilities; do you 12 remember that? 13 A Yes, I do. 14 MR. DELERY: And, your Honor, this is the last page 15 of Exhibit 225. 16 BY MR. DELERY: 17 Q Does that chart as you remember it satisfy your 18 criticisms about the difficulty of interpreting global or 19 composite odds ratio? 20 A No, it doesn't. As I've testified and as that table 21 shows the practical meeting of an odds ratio depends on the 22 baseline probability, and as we know, the baseline 23 probabilities depend on the grades and test scores of the 24 students. What I've also found in my work is that the odds 25 ratios themselves depend substantially on the grades and test GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 8 1 scores of the students. So when both the odds ratios 2 themselves vary as a function of those things and the meaning 3 of the odds ratios varies as a function of grades and test 4 scores it becomes exceedingly difficult to make a practical 5 interpretation of the global or composite odds ratio. 6 Q The second topic or the second issue that Dr. Larntz 7 addressed on Saturday was the excluded data and he presented 8 charts, versions of the grids that he had highlighted in 9 yellow, the cells that included data in his composite odds 10 ratios calculations and the non highlighted cells or cells that 11 did not contribute information to that calculation; do you 12 recall that testimony? 13 A Yes, I do. 14 Q Now, in part of that discussion Dr. Larntz indicated that 15 he believed that the amount data, in other words, the number of 16 applicants in the non highlighted cells, the amount of data 17 that had been excluded was really irrelevant to interpreting 18 his results; do you recall that? 19 A Yes, I do. 20 Q Do you agree with that? 21 A I strongly disagree with that. In fact, I would say that 22 the amount of excluded data can be more informative than the 23 odds ratio itself in his analysis. 24 Q Why is that? 25 A Well, in order to explain why that's true, I'd like to GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 9 1 present just a simple example if the Court's permission. 2 Simple. 3 Q Would you like to draw it? 4 THE WITNESS: Could I? 5 THE COURT: Sure. 6 A In this example, we have a law school that bases its 7 admissions decisions on just two factors: Test scores and 8 race. So what I would like to do, I'll just draw a number line 9 that gives the possible values of test scores. They start at 10 one twenty and they go to one eighty. 11 Now, in this law school, starting at one eighty at 12 the top, everyone is admitted without regard to race until we 13 reach a certain point that I'm going to call a cut point. At 14 that point we find candidates who are sitting right on the 15 border of that cut point, some of whom are minority, and some 16 are majority, and the policy is: Admit all minorities, and 17 some majorities. Below that cut point, all candidates 18 regardless of race are rejected. So this is the reject area, 19 and this is the admit area. 20 Now, what I would like to do is contrast that law 21 school with a second law school that also uses just test 22 scores and race in making admissions decisions. I'll draw the 23 same number line. This law school has a very, very different 24 policy. Starting at the top it admits all minority 25 candidates. And it maintains that policy until we get down to GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 10 1 the very, very lowest possible test scores. So in this area, 2 admit all minorities, and admit some majorities. 3 Now, I think we would agree that in both law schools 4 race is taken into account in admissions. However, the extent 5 to which it is taken into account is tremendously different. 6 It is a very modest extent only at the very border that could 7 possibly have an effect, whereas in the second law school, law 8 school number two, it has a much -- it's taken into account a 9 great deal more because of what we see here, that all 10 candidates in this very wide range where minorities are 11 automatically admitted. 12 Now the question is if we apply Professor Larntz's 13 methodology can we discern the difference between these two 14 law schools? If we apply that methodology what will happen in 15 the case of law school number one is that all of the data 16 except at the border -- 17 THE COURT: But see some of your premises -- in the 18 first one, it says admit all minorities at that point. The 19 second one says all minorities at a different point; isn't 20 that what you're saying? 21 THE WITNESS: In this one here, your Honor? 22 THE COURT: Right -- no, that one first. All 23 minorities at that point. 24 THE WITNESS: Yes, just in that border. Just in 25 that little area -- GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 11 1 THE COURT: The second one says, admit all 2 minorities throughout the whole -- 3 THE WITNESS: Except unless the grade -- 4 THE COURT: The reason -- and I want to hear what 5 you have to say, but that's not what happened here. The 6 University of Michigan takes the position we don't admit all 7 of the -- all of them. 8 THE WITNESS: Right, I agree. 9 But what I want to show is here's a case where we 10 have all the relevant information. There's no missing 11 information. We know the test scores, we know the race. Can 12 this methodology reveal the extent to which race is taken into 13 account in this simple situation. If it can't, how can we 14 except that methodology to reveal the extent to which it's 15 taken into account in a much more complex situation -- 16 THE COURT: Explain it to me because -- 17 THE WITNESS: So what will happen is -- using 18 Professor Larntz's methodology, all of the candidates in this 19 -- where they are all rejected and all of the candidates here, 20 where all are admitted, will be discarded because they have 21 the same admissions decision regardless of race. They provide 22 in his words, no comparative information. The odds ratio will 23 be computed simply for this group where there is, in fact, a 24 difference, where some people are being admitted. And what he 25 would find would be an infinite odds ratio in this case. GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 12 1 Now, when we go down to the second law school, what 2 would happen is there is only a small fraction of the data 3 that would actually that would actually be discarded because 4 here is the small area where everyone is rejected regardless 5 of race. Admits and rejects are occurring in this area. This 6 would have -- this area, here, would have the "comparative 7 information" in his words. We would compute the odds ratio 8 and we would find an infinite odds ratio. 9 In both cases, we would find an infinite odds ratio. 10 What would actually be far more revealing about these cases is 11 the fraction of information discarded. The fact that 12 virtually all of the data is discarded in case number one 13 reflects the fact that this policy in case number one applies 14 to very few students. Race has no effect on these decisions. 15 It has no effect on these decisions, they're discarded. 16 THE COURT: How about -- see -- I understand what 17 you're saying, but his whole theory was to what extent is race 18 because the University of Michigan takes a position that it's 19 very little, you know, it's not a trump card in the words of 20 Mr. Payton several times. And so this we know it's not very 21 little. We know that race is a predominant factor in making 22 decisions. So his analysis is different than your analysis. 23 THE WITNESS: Well, in these two cases, simplified 24 as they are, we have all the knowledge we need. Race is taken 25 into account, I would say, very little in this first school. GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 13 1 Only if you're at the very border -- 2 THE COURT: We don't need a statistician to tell us 3 that because they've already told us that. One of your 4 premises is that race is taken into consideration as a major 5 factor. 6 THE WITNESS: Right. See, here's the way we 7 evaluate a methodology. We set up a simple situation where we 8 know the truth. We have all the relevant information, and we 9 evaluate the capacity of the methodology to answer -- to 10 discover the what we know. 11 THE COURT: I'll never be able to discuss it on your 12 level. I'm discussing it as a layperson who's never taken a 13 statistic class. But I do know that -- just by listening to 14 Dr. Larntz, and I'm not arguing his position, I'm not saying 15 it's good or bad, what I'm saying is the premise is different 16 because he is trying to figure out to what extent is race 17 taken into consideration. And in your analysis here we know 18 to what it's taken into consideration. So we're not comparing 19 apples to apples. 20 THE WITNESS: Well, what I'm showing you here, what 21 I'm trying to argue here is that his analysis cannot reveal, 22 the odds ratio cannot reveal the extent to which race is taken 23 into account because it can't discriminate between two cases. 24 In the first case it's taken into account very little. In the 25 second case, it's taken into account an enormous amount. And GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 14 1 in both cases the odds ratios are the same. And what is 2 actually far more relevant to the evaluation of these two 3 cases is the fraction of data that he would discard. 4 THE COURT: Okay. 5 BY MR. DELERY: 6 Q How does that, Professor Raudenbush, tie back to the 7 highlighted grids that Dr. Larntz had here on Saturday? 8 A Well, obviously the Michigan case is more complicated, 9 but in the highlighted grid a large fraction of the cases, 10 especially a large fraction of the minority cases, were 11 discarded. 12 Those are cases where the admissions decision is 13 being made without respect to race. The fact that so many 14 candidates, forty percent, are having decisions made that 15 don't take into account their race is very relevant to the 16 issues in this case. 17 Q Is the approach that you just took here in terms of 18 setting out two simplified hypothetical cases using that to 19 evaluate a methodology, is that a standard practice in your 20 field? 21 A That's the standard practice. To evaluate a methodology 22 we need to know whether it can recover the distinctions that 23 are important when we know what the truth is. If it can't do 24 it then, you can't discover the reality when we know the 25 reality and when we have all available information, we can't GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 15 1 expect that methodology to give us the answer about the extent 2 o which race is taken into account in a much more complexed 3 situation where we don't know the information. It's not going 4 to work better when we have less information. 5 Q Why don't you take the stand again, Professor Raudenbush. 6 And -- I guess just one final question before we move onto the 7 next topic, and that is, did Dr. Larntz's testimony on Saturday 8 about the excluded data and the highlighted charts, does that 9 change your opinions at all concerning the excluded data issues 10 that you've discussed -- 11 A No, actually, sitting and thinking about it, actually 12 during the testimony it made it even more clear to me how 13 important this excluded data actually is in looking at his 14 results. 15 Q Okay. Let's move now to the third issue from Saturday, 16 and it's related to the issues of assumptions. On Saturday, Dr. 17 Larntz said that he thought an analysis that used all of the 18 data, that didn't exclude any of the data, would have to be 19 based on a model with more assumptions than his was. Do you 20 think that's right? 21 A It depends strongly on what the broad methodological 22 goals of the analysis are. In my analysis where I was 23 assessing the causal impact of the policy of taking race into 24 account the issue of discarding data simply didn't arise. Just 25 referring back to this case, my methodology would attempt to GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 16 1 predict the probability of admission for everyone in law 2 school, number one, based strictly on test scores, and to see 3 how closely those predictions correspond to what actually 4 happened in practice, and we would find that they correspond 5 very closely, where as in the second case, they would not 6 correspond closely so I would argue that my methodology could 7 sharply distinguish between these two cases. The issue of 8 discarding data just doesn't come up when that's the goal. We 9 use standard methods, and we don't need to discard data. 10 Q Do you think that Dr. Larntz made fewer assumptions than 11 you do? 12 A Well, I wouldn't necessarily say that he made fewer or 13 more. We all have to make assumptions when we do statistical 14 analyzes. What we need to do is to be as explicit about what 15 those assumptions are, to whenever we cann to test the validity 16 of those assumptions and then to evaluate the possible impact 17 of those assumptions on our results. My criticism -- or one of 18 my criticisms, I should say of Professor Larntz's analysis is 19 that -- he makes a very strong assumption, namely, that the 20 same global odds ratio applies to everyone regardless of grades 21 and test scores. And I actually did a series of analyzes to 22 assess that, and I found that assumption not to be tenable. 23 And I believe, and I think he would disagree that the impact of 24 that could be important in this case. 25 Q On that last point I think Dr. Larntz did say on Saturday GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 17 1 that he thought that while the assumption might not technically 2 be true, the assumption of a uniform odds ratio might not be 3 technically be true, it wasn't important to evaluating his 4 results; do you disagree with that? 5 A Well, I disagree, and I think I could represent why I 6 disagree if we could just briefly refer to the chart that we 7 had on Saturday. 8 Q Dr. Larntz's chart? 9 A Dr. Larntz's chart, yes. 10 Q Okay. This is Exhibit 226. 11 A What this chart was really building up -- maybe I need to 12 get up because I can't see it -- 13 THE COURT: You may. 14 MR. DELERY: Can you see it, your Honor. 15 THE COURT: I'm fine. 16 MR. DELERY: Okay. Thanks. 17 A Is that the lower bounds of these constant variables for 18 the odds ratios are large numbers. And I think that was really 19 the crucial point. If that assumption is false, that is, if 20 the offs ratios vary as a function of where you with respect to 21 grades and test scores, then for some students and possibly 22 many students these lower boundaries should be much lower than 23 these numbers. And that in a practical sense is probably the 24 key potential problem. 25 THE COURT: When you say "much lower" could they GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 18 1 have used that in the figuring? 2 THE WITNESS: Well, I have -- in the analysis I did, 3 I did some inspection. Yes, there are certainly cases where 4 these are much, much lower than these. 5 THE COURT: When you say "much" give me a relative 6 figure. 7 THE WITNESS: Okay, in one analysis I did, the 8 average odds ratio was twelve. But in my analysis I also 9 estimated how much the odds ratio -- 10 THE COURT: You used a different methodology. 11 THE WITNESS: I used a different methodology. 12 THE COURT: Again, comparing apples to oranges. 13 THE WITNESS: It only makes one -- it only really 14 has one really important difference and that is it allows the 15 odds ratios to vary over the cells of the table just. It's 16 still a logistic regression. It introduces that possibility 17 that the odds ratio would not be invariant across the cells of 18 the table. 19 BY MR. DELERY: 20 Q And using that method -- I'm sorry, I'm sure that you 21 finished -- 22 A Oh, yeah, so while the odds ratio was twelve, the 23 standard deviation of the log odds was about one, and that 24 leads to odds ratios that are much smaller than twelve, even as 25 small as one and two. GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 19 1 THE WITNESS: Now, I didn't highlight that, your 2 Honor, because as you can tell I'm not really buying into the 3 idea of using the odds ratio to -- 4 THE COURT: I understand that. You made that very 5 clear. 6 THE WITNESS: I just to make sure. I really was 7 just doing it evaluate his assumptions. 8 BY MR. DELERY: 9 Q Let's go now, Dr. Raudenbush, to the last of the major 10 issues from Saturday and that was the stability of the odds 11 ratios across the years. On Saturday, Dr. Larntz said that he 12 thought that your concern about instability across the years 13 might be due to a computational error that you had made. Do you 14 remember that? 15 A I do remember that. 16 Q Was he right? 17 A Well, actually on Saturday, he solved a puzzle that has 18 been in my mind for some time. As he pointed out the standard 19 output for a logistic regression analysis includes regression 20 co-efficient which is in a log metric, a standard error, and 21 then what we call a Z test which is a test of significance. I 22 had been confused by his reports because his reports were a 23 little bit non standard. They had an odds ratio, and then 24 another column that was labeled "standard deviation." 25 Now, it certainly would be possible to compute for GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 20 1 each odds ratio a standard error in the metric of the odds 2 ratio. It would not be conventional and it would not be the 3 optimal way to proceed, but it could be done. And I wasn't 4 sure whether he had used that method or the more standard 5 method from the results of his report. So I asked you to 6 pursue this issue in the deposition of Professor Larntz. I 7 mean, this is a very different situation from how we usually 8 operate. In our profession we would simply ask each other 9 these questions, but in an adversarial thing, I have to try to 10 get you to find out the answer. So I read the deposition, and 11 my reading of the deposition suggested to me that he was 12 actually using the odds ratios and the standard errors in that 13 metric. And on Saturday, he made it very clear he was not 14 doing that. So I was actually under a false -- operating 15 under a false assumption as to what he was actually doing. 16 What I was able to do then, now that I could see how 17 he actually operated in the log odds metric, is to recompute 18 the stability, a test of stability using a very simple test. 19 I would be happy to show you how it's computed. It can be 20 done on a calculator. It would take about fifteen minutes, 21 but -- it wouldn't take fifteen minutes to explain it, but it 22 would take maybe fifteen minutes on a calculator. It's a very 23 simple procedure. I think it would be useful to just tell you 24 what the result is -- 25 Q Yes. GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 21 1 A It's a Chi square test of homogeneity very commonly used 2 in this area. And the Chi square value I computed was 22.9. 3 The probability associated with that was less than .001. 4 Q And that's the P value? 5 A That's the P value. And what that says is that there is 6 statistically very significant heterogeneity across the years 7 in the log odds ratios which is rather than the odds ratios 8 themselves. 9 Q From that what do you conclude about your earlier 10 criticism concerning the instability of the odds ratios? 11 A I remained concerned about the instability for the 12 following reason: When we look at the process of admissions, 13 it seems to be stable. When we look at the average 14 probabilities of admission, they're stable. The results are 15 bi-causal analysis, year-to-year, in the exhibits, in my own 16 testimony. If you review those you'll see the results are 17 stabling. We see here quite significant heterogeneity and we 18 also see very significant heterogeneity across the models that 19 are also in the exhibit of my testimony. And what worries me 20 is when we see heterogeneity in a model, when we don't see it 21 in more basic summaries of the data is that something is going 22 on in the methodology that is creating or perhaps identifying 23 the instability. 24 Q Okay. When you were here last month -- a few weeks ago 25 -- you talked about something called the standard error or the GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 22 1 difference between -- through the odds ratios is what I'm 2 talking about the instability -- and not a test of homogeneity 3 that you just mentioned. Why did you use the other one then 4 and this one then? 5 A Well, it's much easier to explain if I have two numbers 6 asked, are these numbers different than it is to explain the 7 test of homogeneity, and I thought that would be a useful way 8 to go. But the test of homogeneity is the better test because 9 it doesn't -- Professor Larntz made the point when you select a 10 big difference from a whole set of possible differences it you 11 have -- it's hard to get -- you have to be careful to get the 12 right P value. And so to simply test the homogeneity across 13 the six years is a more straightforward statistical way of 14 doing this although it's more difficult to explain. 15 Q And you reached the same bottom line conclusion both ways 16 is that correct? 17 A Yes, I did. 18 Q Okay. Well, Professor Raudenbush, I want to ask you just 19 one final question: We've heard this testimony back and forth 20 between you and Dr. Larntz, and my question is: Based on your 21 twenty-five years as an educational statistician and having now 22 heard the rebuttal testimony in your view is there a way to 23 make a sense of the areas of real dispute you and the 24 significance of those areas? 25 A Well, let me try to summarize the disagreements as I see GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 23 1 them. We both would agree that there is an association between 2 race and admissions controlling for test scores and grades. 3 And that finding would not be surprising given the stated 4 policy of the law school which does take race into account. We 5 would disagree about the extent to which that relationship 6 varies as a function of grades and test scores and we would 7 seemingly disagree on how important that variation is. We 8 would disagree about the instability of the results that we've 9 just described or the importance of that. Those in my view, 10 however, would be secondary disagreements in my view. The key 11 disagreement is the one we've discussed earlier. From a broad 12 methodological point of view I am convinced that one cannot 13 compute odds ratios that will reveal the extent to which race 14 is taken into account in admissions. And I did say that in my 15 testimony. I think in statistics we have to do what we can do 16 and limit what we do to what we really think we can do. And 17 what I did -- what I think we can do is give an assessment of 18 the impact of the policy on those who apply. What's far more 19 difficult to do from a statistical point of view than is to 20 understand the process, the causal process, the process is 21 cognitive process in the admissions office that generate that 22 impact. So we can access the impact, but to say what extent 23 it's taken into account requires what we know more about the 24 process than we actually know. And that's really the key 25 disagreement. GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 24 1 Q And in your view do you have any doubt as to what the 2 correct answer is? 3 A Well, I think I've made it clear that I am convinced that 4 we cannot use the statistical data and certainly not the odds 5 ratio to reveal in this case the extent to which race is taken 6 into account, but we can with -- with an amount of uncertainty 7 -- there is some uncertainty that I have tried to quantified, 8 we can access the causal impact of the policy. 9 MR. DELERY: I no further questions, your Honor. 10 THE COURT: Intervenors, any questions? 11 MS. MASSIE: No. 12 THE COURT: Plaintiff? 13 MR. KOLBO: Yes, sir. 14 CROSS-EXAMINATION 15 BY MR. KOLBO: 16 Q I just have a few questions. 17 I'm sure I'll struggle with some of the language 18 here, but I'll do my best. 19 Dr. Raudenbush, on the first subject matter with 20 respect to the drawing that you've got there, am I correct 21 first of all with respect to number 2, you would agree that 22 this is a case in which race is taken into an extensive -- to 23 a great degree? 24 A I would. 25 Q So it's something that one can ascertain from example GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 25 1 number 2 even though you had a calculated odd ratio infinity. 2 A Yes. 3 Q And the second -- or in the first example, it's your 4 conclusion that this is a case in which race is not taken into 5 account to a great extent. 6 A To a great extent, that's correct. 7 Q And your testimony as I understand it, one wouldn't be 8 able to discern the difference between those two scenarios 9 because in both cases you get a calculated odds ratio of 10 infinity; right? 11 A Well, we can discern the difference between those cases, 12 but not using the approach of discarding data and then 13 calculating an odds ratio. The reason, by the way, that we can 14 discern the difference in this case is because we have all the 15 available information. There are only two things that can go 16 into the admissions decision. 17 Q But as I understood your testimony, and correct me if I'm 18 wrong, one of the problems was looking at odd ratios in that 19 case is that in both cases you've got infinite calculated odd 20 ratios; right? 21 A It's not because they're infinite. Even if they were 22 just the same, let's say they were both twenty-three, the fact 23 that the odds ratio comes out the same in both cases, and could 24 easily come out the same, shows that the methodology can reveal 25 the difference between two cases, one in which there's a great GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 26 1 extent and one in which there is a lesser extent. 2 Q But one could also in that scenario, in addition to the 3 odds ratio analysis, one could also look and see that there is 4 a difference between these two cases; right, because one would 5 know that in most cases, under number one, either you have all 6 admits or all rejects, and the other a very small area where 7 you have admit all minorities and admit only some majority 8 students. 9 A That would be the crucial piece, would be the amount of 10 data that were discarded by Professor Larntz which are those 11 cases in which all are rejected or all are admitted. That is 12 actually a much more informative, I'll say, piece of evidence 13 about these two cases than the odds ratio. 14 Q And you can actually discern looking at one or two that 15 they're different in that respect; correct, in terms of the 16 amount of data that you call discarded? 17 A You would certainly -- yes, that would be a decisive -- 18 in this limited case, where we have all the information, that 19 by itself would tell us an enormous amount and to what extent. 20 Q And the data that we have in this case including the data 21 that Dr. Larntz submitted graphically we can see very clearly, 22 can't we, how many cases -- how many applications fall in this 23 category of admitting all students of all races? 24 A We can see -- the problem is that we don't know all of 25 the factors that are creating those decisions. If we assumed GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 27 1 -- if we're willing to assume that grades and test scores are 2 the only thing that count other than race, we could do the kind 3 of analysis that I have here. But I don't think any of us would 4 make that assumption. 5 Q But my question is: By looking at the data as Dr. Larntz 6 has presented it, and the data that we have here, we can 7 actually determine what percentage of files are those under 8 scenario number one in which all students, of all races, are 9 admitted; true? We can see what the size of that group is; 10 true? 11 A We can -- you mean conditioning or controlling for grades 12 and test scores? 13 Q Yes. 14 A Controlling for grades and test scores we can certainly 15 assess -- see what cells have all admits or all rejects, that's 16 right. 17 Q We could see how large of a representation that is across 18 all the grid; correct? 19 A Yes, we can. 20 Q And we could do the same analysis with respect to the 21 number of students who are rejected, whether all minorities or 22 all -- 23 A Controlling for what we know, again, just grades and test 24 scores, yes. 25 Q Okay. And would you agree that there is a very GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 28 1 substantial number of files in the University of Michigan case 2 in which there are comparative differences, that is, in which 3 there are some minority students admitted, and some majority 4 students admitted, less than all of those groups, in which we 5 can calculate the differences in probabilities? 6 A We can calculate the differences in probabilities -- we 7 can calculate the differences in probabilities I should add for 8 every, every person who applies in all test scores and grades. 9 We don't need to discard any different information in order to 10 calculate the two probabilities -- proportion, I should say. 11 Where we come in discarding data is when we compute 12 an odds ratio where -- which prohibits certain calculations 13 because of the division by zero. 14 Q But scenario number one, for example, you've got here, 15 you don't want to assume any cases, do you, in which there are 16 less than a hundred percent minority students admitted and less 17 than a hundred percent of majority students admitted; true? 18 A I'm sorry, could you please rephrase that? 19 Q Example number one, doesn't account for any situation in 20 which there is less than a hundred percent minority students 21 admitted, and less than a hundred percent majority students -- 22 A Right. In scenario number up there, everything to the 23 left of that little bubble, all are rejected regardless of 24 race. 25 Q Right, and you're assuming no students in which the GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 29 1 grouping is such as: fewer than one hundred percent minority 2 students admitted, and with comparable grades and test scores, 3 fewer than all majority students admitted. You made no 4 provision for that in example number one. 5 A For an occasion in which every minority student is 6 admitted and every majority student is rejected? 7 Q No, no, that's not what I -- 8 A Sorry. 9 Q I'm sorry if I wasn't clear. You have not provided in 10 your model number one or model number two, for that matter, for 11 situations in which less than one hundred percent -- 12 A Oh, I see -- 13 Q -- of minority -- 14 A Right -- excuse me -- 15 Q Can I just finish? You have not provided in your model 16 here for either one or two scenarios in which less than one 17 hundred percent minority students are admitted. And for 18 comparable majority students where less than a hundred percent 19 of majority students are admitted; true? 20 A That assumption follows though exactly from saying there 21 are only two factors. If there are only two factors that can 22 affect the admissions decision, then if it's not test scores, 23 it can only be race because that's the only other factor. Now, 24 I agree that's oversimplified. But under that oversimplified 25 assumption, what I'm saying about all minorities admitted in GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 30 1 the bubble would follow from the assumption. We can certainly 2 imagine a situation in which other things were done. Like, you 3 could flip coins within that bubble, or you could assign 4 different probabilities, but I was trying to create a very 5 simple example. 6 Q Well, your example just assumes just test scores and 7 race; right? 8 A Just -- actually, this example simply assumes test scores 9 and race. 10 Q Okay, test scores and race. 11 A Right. 12 Q But you're assuming no cases in which just looking at 13 test scores and race, no cases in which less than a hundred 14 percent minority students are admitted, and less than a hundred 15 percent majority students are admitted with comparable tests; 16 true? You're not even accounting for that in your example. 17 A If only -- right. That logically follows from the 18 assumption that those two things can count, yes. 19 Q And what the odds ratios would tell us here, whether 20 they're infinite or whether they're less than infinite in these 21 scenarios it would be telling us something about the groups of 22 students where there's comparative information; correct? 23 A I strongly disagree with the idea that there is a 24 principle of comparative information that tells which data we 25 can look at and which data we cannot look at. The principle of GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 31 1 comparative information that Professor Larntz described is a 2 principle only within the methodological framework that he 3 selected for analyzing these data. 4 Q But am I correct that the calculated odd ratios that you 5 have here in this case, infinite, those calculated odd ratios 6 relate only to students for which there is comparative 7 information, whether there are differences; correct? 8 A In -- once -- yes, once you have decided to use the odds 9 ratio, they can't be computed except in the bubbles of the two 10 cases, right. 11 Q That's all I have with respect to that. A couple of 12 questions on modeling and assumption. Do I understand that 13 your analysis with respect to what you did in this case, you 14 created -- you devised a model that modeled the entire grid; is 15 that right? 16 A Yeah, let me try to explain. I did -- sir, my primary 17 thing that I did was a causal analysis which I've described and 18 I think we all understand. And then as a secondary activity, I 19 did some analyses to check certain assumptions that Professor 20 Larntz was making. And in that second analysis I did do what 21 you said, yes. 22 Q You modeled the entire grid? 23 A That's right. 24 Q And did you have to make certain assumptions in modeling 25 the entire grid? GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 32 1 A Yes, I did -- 2 Q But if you were to do an analysis simply on a 3 cell-by-cell basis you wouldn't have to make any assumption 4 based on modeling the entire grid; correct? 5 A Oh, you have make -- absolutely, you do have to make 6 assumptions especially if you come up with a composite odd 7 ratio, you have to assume that the underlying two odds ratios 8 are the same for every cell. 9 Q Would you agree that the fewer parameters you have means 10 the more assumptions that will be in the model? 11 A The fewer parameters the more assumptions? 12 Q Yes. 13 A It might be true, but it's not necessarily true what you 14 say. The number of assumptions is not strictly a function of 15 how many parameters are in the model. 16 Q Just the last few questions on the last point that was 17 covered here and that was with respect to the -- your opinions 18 with respect to instability over years. You were here 19 yesterday when Dr. Larntz went through the calculations or the 20 difference in terms of standard deviations between the 1997 and 21 2000 year? 22 A Yes. 23 Q Would you agree that your testimony suggesting that the 24 standard deviation separating those two numbers -- would you 25 agree that it would not represent a number by merely eleven -- GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 33 1 A Yes, now that I understand exactly what he did, I agree 2 that's true. 3 Q You agree with calculation? 4 A Yes, the 3.7 was the more relevant number. 5 Q And your testimony as I understand it was you simply had 6 a misunderstanding based on your reading of Dr. Larntz's 7 deposition? 8 A Yes. 9 Q Were you produced -- did Mr. Delery produced to you as 10 I've produced to him, Dr. Larntz's computer output in this 11 case? 12 A Yes, I was. 13 Q Did you review it and consider that? 14 A I did. I actually looked very, very carefully through 15 that output. Professor Larntz does his own programming. I 16 received hundreds and hundreds of pages of output. The 17 regression output doesn't come in a standard format. What you 18 see in his output primarily are simply large blocks of numbers. 19 Now, Professor Larntz, I have no doubt knows exactly 20 how to translate those numbers into the relevant tables in his 21 report. I could not make the translation. I was not able to 22 make the translation. 23 Q So you didn't see any columns in there for his different 24 analyses where he had a column for one -- one colum of 25 regression co-efficient? GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 34 1 A There were regression co-efficients. There were several 2 -- there were many analyses and there were hundreds of them. 3 Q And did you see columns for represented values of 4 standard error? 5 A Yes. 6 Q And I am correct it was determined standard deviation -- 7 and I probably get this wrong, and correct me, but you divide 8 the regression co-efficient by the standard error; is that 9 right? 10 A That would give you what we call a Z test. It's not 11 normally labeled standard deviations. 12 Q Is it sometimes? 13 A I've never seen it labeled that way until this -- until I 14 saw this report. 15 Q Did you ever make any attempts to take a look at some of 16 the numbers that showed up on Dr. Larntz's computer output and 17 see if they corresponded to the standard deviation value as 18 reported in his report for different years? 19 A I did. 20 Q And you couldn't find any correlation? 21 A Well, as I've said, there were hundreds of pages of 22 output. There were many, many regressions. And he had well 23 over one hundred predicted variables in every equation. So 24 every single regression he did produced hundreds of regression 25 co-efficients and I -- maybe I didn't spend enough time or do a GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 35 1 more -- perhaps I could have found those if I spent an enormous 2 amount of time. But it was very difficult for me to reproduce 3 what was his reports from the output that he sent me. I didn't 4 make the effort. 5 MR. KOLBO: I have nothing further, your Honor. 6 THE COURT: Thank you. Anything else? 7 MR. DELERY: Nothing further, your Honor. 8 THE COURT: Thank you, Doctor. We appreciate it, 9 very much. Sorry to have taken up your weekend. 10 THE WITNESS: I'm glad it's over. Thank you. 11 THE COURT: Thank you, very much. 12 Ms. Massie, it's my understanding your next witness 13 will be here at eleven? 14 MS. MASSIE: If we could make it at 11:15. I'm 15 sorry about the delay, but I'm sure we will be able to get 16 through the two people today. 17 THE COURT: Sure. I understand. We'll do it. 18 Okay, we'll stand in recess in the case until 11:00 19 a.m. 20 (Court recessed, 10:25 a.m.) 21 (Court reconvened, 11:45 a.m.) 22 MR. DELERY: Your Honor, one housekeeping matter 23 while we're waiting. We would like to do the same thing with 24 Dr. Raudenbush's drawings that we with Dr. Larntz's -- 25 THE COURT: Yes. GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 36 1 MR. DELERY: So we'll mark it as Exhibit 228 for 2 identification now, and we'll translate it into a small copy 3 and move it in. 4 THE COURT: Perfect. 5 MR. DELERY: Okay. Thank you, your Honor. 6 MS. MASSIE: Hi, Judge. Our next witness is 7 Professor Frank Wu. 8 THE COURT: Okay, Professor Wu. 9 Good morning. 10 MR. WU: Good morning. 11 F R A N K W U , 12 being first duly sworn by the Court to tell the truth, was examined 13 and testified upon his oath as follows: 14 DIRECT EXAMINATION 15 BY MS. MASSIE: 16 Q Hi, Professor Wu. 17 A Good morning. 18 Q If you could spell your name for the record, please. 19 A Sure. Frank, F-r-a-n-k, Wu, W-u. 20 Q Tell us about the teaching you have done or are 21 contracted to do if you would, sir. 22 A Sure. I'm currently an associate professor of law at 23 Howard University in Washington, D. C. where I've taught since 24 1995. I also serve as the director of our clinical program, 25 and I supervise students practicing in the D. C. Superior GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 37 1 Court. I teach civil procedure, and I teach federal courts on 2 some of their subject. I have previous taught as a fellow at 3 Stanford University. I've taught civil rights at American 4 University, in one of its summer sessions. I would be a 5 scholar and residence teaching again on Asian American civil 6 rights at Eeps Spring College in about one month. 7 Q I understand you got a JD, a law degree from the 8 University of Michigan. 9 A That's right, class of 1991. 10 Q Professor Wu, how long have you been doing academic work 11 on questions relating to Asian Americans, civil rights, and 12 social policy? 13 A Probably for about fifteen years or so. 14 Q And how long have you been doing academic work on 15 questions related more specifically to Asian Americans and 16 affirmative action? 17 A At least ten years. 18 Q That work has involved extensive publications. 19 MS. MASSIE: I'd like to move Professor -- well, the 20 publications are in his CV which I know the Court has, at tab 21 175 in the exhibits. 22 BY MS. MASSIE: 23 Q You have a book that's about to be published that's a 24 comprehensive treatment of the question of Asian Americans and 25 civil rights as I understand it. GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 38 1 A That's right. My book entitled, "Yellow, Civil Rights 2 Beyond Black and White" is in the catalog basic books scheduled 3 for publication later this year. It's just about done. I just 4 have to revise a few more chapters. 5 Q You have numerous chapters in other books on Asian 6 American and public policy. 7 A That's right. 8 Q A number of law review articles, articles in the "Asian 9 American Policy Review. A law case book which I understand is 10 not published yet, but it's forthcoming. 11 A That's right. I have a co-authored case book concerning 12 Asian Americans and civil rights with a focus on the 13 Japanese-American Internment experience. It was written with 14 four other law professors, and we have a contract with Aston 15 Books, one of the leading case book publishers. That also 16 should be out later this year. 17 Q A chapter either just published or forthcoming in Asian 18 American politics? 19 A That's right. I have a chapter, a final chapter, in a 20 book that was just published by Stanford University Press and 21 Woodrow Wilson Center, edited by Historian Gordon Chan. It's 22 entitled, "Asian Americans and Politics." 23 Q And you've recently been appointed by the Washington, D. 24 C. City Council to serve on the city's Human Rights Commission 25 and on -- in part so that your expertise on that is of Asian GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 39 1 American civil rights and social policies will be made 2 available to the Commission as a whole, is that right? 3 A That's right. I'm the first and only Asian American to 4 serve on that Board. I was nominated by the mayor and 5 confirmed by the City Council. 6 Q You've testified before Federal Governmental entities on 7 questions involving Asian Americans and civil rights, and 8 specifically, Asian Americans and affirmative action including 9 the judiciary committee of the House of Representatives and the 10 United States Civil Rights Commission? 11 A That's right. I've testified before the House in 1995, 12 and the Civil Rights Commission in I believe 1998. 13 Q Admist Is a very long list of public appearances you've 14 hosted a PBS series entitled "Asian America" for some time? 15 A That's right. I've hosted about thirty episodes of that 16 series which is syndicated by PBS. 17 Q And have also done a number of invited academic 18 presentations at universities all across the country on a range 19 of subjects involving Asian Americans civil rights and social 20 policy? 21 A That's right. I spoke to the University of Texas, 22 University of Nebraska, Smith College, and a few others that I 23 can't recall. 24 Q Is it fair to say, Professor Wu, that you're one of the 25 foremost nationally recognized experts on issues of Asian GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 40 1 Americans civil rights and social policy? 2 A Well, at the risk of being immodest, I would say, yes, I 3 have written very widely, and probably have studied it more 4 extensively than just about anyone else who looks at these 5 issues. 6 THE COURT: You know, Frank Lloyd Wright -- did you 7 ever read the story about Frank Lloyd Wright, who testified 8 for the first time in his life, and they said -- the attorney 9 said something like, tell us about you in terms of your 10 ability to architecture and so forth. And he said, well, I'm 11 the best architect in the world. And then he looked over to 12 the Court and said, I'm under oath, your Honor. 13 THE WITNESS: Thank you, your Honor, I'll use that 14 next time. 15 THE COURT: There's a whole book of those kinds of 16 things that are really interesting. When we take a break 17 remind me to tell you about the police officer one which is my 18 favorite. 19 MS. MASSIE: I have a police officer one, too. 20 THE COURT: I guess we all could write a book on 21 police officers. 22 MS. MASSIE: With that, I'll offer Professor Wu as 23 an expert on Asian Americans civil rights and social policy 24 and also move into evidence his expert report which as I said 25 is tab 175. GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 41 1 THE COURT: Any objection by anyone? 2 MR. PURDY: No, your Honor. 3 THE COURT: He may testify as an expert. 4 (Trial Exhibit number 175 received into evidence.) 5 BY MS. MASSIE: 6 Q Professor Wu, partly in response to a question that the 7 Court asked of another witness who was very well qualified, but 8 somewhat less qualified in this particular area than you, I'd 9 like to start by asking you to tell us all who are Asian 10 Americans? 11 A Well, the term Asian American" is usually used to refer 12 to some ten million or more Americans who can trace their 13 ancestry to Asia, to any more than some two dozen countries in 14 Asia, or to a Pacific Island. It includes people of diversed 15 backgrounds. Some people, relative newcomers to the United 16 States, some people who may be third, fourth, or fifth 17 generation Californian. People of different ethnicities, 18 different faiths, different linguistic backgrounds, different 19 walks of life. But what they have in common is their Asian 20 heritage, Asian ancestry. And that makes them a minority in 21 the U. S. According to the 2000 census, it looks like Asian 22 Americans comprise approximately four percent of the nation's 23 population. 24 I think to get a real sense though of who Asian 25 Americans are it might be best for me to just tell you about a GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 42 1 few who I know, have known. 2 My father-in-law, for example, my late 3 father-in-law, who passed away recently at the age of 4 seventy-five, was born in Wakayama, in a small fishing village 5 in Japan in 1905. At the age of sixteen, Shinsuke Izumi, 6 that's S-h-i-n-s-u-k-e, I-z-u-m-i, persuaded his parents that 7 they should let him come to the United States to join his 8 older sister who already lived here and with her husband ran a 9 shoe store in Los Angeles. So he boarded a boat and after a 10 arduous of many months arrived in the United States where he 11 enrolled as a relative youngster at a business college in Los 12 Angeles. 13 While he was there he felt pressure to have an Anglo 14 name so he picked the name Edwin, following the dapper of King 15 Edward who was much in the news. So Eddie as he was then 16 known after graduating from college was unable to find work 17 that suited his qualifications because in Los Angeles, in 18 California, in the United States at that time, there just 19 weren't many opportunities for people of Asian descent. So he 20 opened a small supermarket, a fruit stand in Hollywood. And 21 that was his vocation until World War II broke out. 22 When World War II broke he, along with a hundred and 23 ten thousand other American citizens and residents of Japanese 24 were rounded up by their own government and were incarcerated 25 in internment camps because there is a fear that they would be GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 43 1 -- be aliens. That's what they were deemed to be, and that 2 they would be disloyal and proved to engage in acts of 3 sabotage or things of that sort. 4 Well, when he had the opportunity to leave 5 internment camps to work as an army translator he jumped at 6 that opportunity. And from his work as an army translator, 7 later went on to become topographer for the Defense 8 Intelligence Mapping Agency where he worked for the rest of 9 his life. He is an example of someone who is an Asian 10 American. 11 My own father who was born in Mainland China near 12 Shanghai who then moved to Taiwan as a youngster was offered 13 an opportunity to go to college in the U. S. And his older 14 brother had come to the U. S. before him, to Iowa. So my 15 father followed his footsteps. He had a scholarship. In the 16 late '50s he moved to snowy cornfields of Iowa City. And 17 there because there was segregation and because Chinese 18 students couldn't find housing some six or seven of them lived 19 in a one-bedroom apartment with no frig, no real appliances, 20 and in the winter they would keep their food outside to keep 21 it fresh, and there was plenty snow which he had never seen 22 before. 23 Well, my father who played on an intramural 24 basketball team while he was in college, a basketball team 25 called the Orientals who then went on to join Ford Motor GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 44 1 Company, and to start a family, here in this area, first 2 living in Dearborn, and who then help found the Detroit 3 Chinese American Engineers Association. And who would win, 4 and was very proud of winning his bridge tournament and his 5 tennis tournament. He is an example of an Asian American. 6 My friend Kaying Yang, that's K-a-y-i-n-g, Y-a-n-g, 7 an immigrant who arrived in this country as a child, who is of 8 Hmong, that's H-m-o-n-g, whose family had served the United 9 States military in it's covert operations in Southeast Asia 10 and who in recognition of their service were brought over in a 11 dramatic effort to bring them out after the fall of Saigon who 12 then moved to Denver where as she told me just a few months 13 ago as we traveling together where she encountered racial 14 discrimination, but didn't know what it was, whereas she and 15 her brother, some of her cousins and other friends went to 16 school and had kids, called them Chinc and Gook and say they 17 should go back to where they came from, that they weren't 18 welcomed, who was spat upon, who had her hair pulled, beaten 19 up on a daily basis. Who now having gone to college where she 20 studied Asian and American studies and learned that there were 21 other people like her, that her experience was not unique that 22 she was not alone. And that there was a name that we give to 23 this phenomenon and that it was not her fault. She now runs a 24 non profit, a leading non profit group in Washington, D. C. 25 called the Southeast Asian Resource Action Center. She's an GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 45 1 example of an Asian American. 2 It's easy though to think that all Asian Americans 3 are people who have just moved here, who are: Fresh off the 4 boat" as people sometimes say. But Asian Americans also 5 include people of second, third, fourth and fifth generation 6 whose ancestors have worked, for example, on the railroad in 7 the 1880s, whose ancestors were in California before it was a 8 state, and before it joined the Union, who were in the south, 9 in the U. S. as part of a fantastic scheme to import Chinese 10 laborers shortly after 1864, to compete with the recently 11 freed Black slaves. 12 There are people who can trance their ancestry back 13 more than a hundred years from this country, people such as 14 the Japanese American soldiers, the Nisei, N-i-s-e-i, meaning 15 the second generation soldiers of the Army 440 Second Unit 16 that fought in World War II, in the segregated armed forces, 17 the most highly decorated unit on a per man basis to have 18 served in the U. S. Army. A unit which lost eight hundred me 19 in rescuing the so-called Lost Battalion, that were behind 20 enemy lines. Those were native-born American citizens who by 21 birth, by birth right were part of this country, they along 22 with their families in locked up in internment camps. They 23 are also Asian Americans. 24 By brothers and I, born in the United States are 25 Asian Americans. The Hmong, who like Kaying Yang, who are GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 46 1 younger, were born in Wisconsin, in the Twin Cities or 2 elsewhere, across the American midwest, after they arrived, 3 some of whom I've met a year ago when I spoke at Wisconsin, at 4 its law school, they're also Asian Americans. They're native 5 born Asian Americans. 6 So Asian Americans include people of Korean descent, 7 of Indian descent, of Philippino descent, of Pakistani 8 descent. It includes people from Tonga, people from Guam. 9 And what we have in common is an experience, an experience in 10 the United States of being called Chinc, and Jap, and Gook, of 11 being told you should go back to where you came from, of being 12 asked where are you really from, as if we're going to go back 13 some place. Being asked how do you like it in our country, 14 when are you going home. Of being told, my, you speak English 15 so well, which I'm always attempted to rely, why, thank you, 16 and so do you. That's the experience that defines what it 17 means to be Asian American. It's a set of experiences that 18 there's no easy definition, and I think it depends on each 19 particular person, each particular community. But most 20 sociologists, census would agree that there is a distinct set 21 of communities that we can properly call Asian American. 22 Q Professor Wu, when did the term Asian American -- and 23 also I'll ask you in a minute about the term Asian Pacific 24 American -- when did that term come into being if I understand 25 is an aggregate term reflecting the many national peoples who GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 47 1 are included within the category? 2 A Well, those two terms are roughly synonomous. They first 3 came into use in the late 1960s. There was a time -- certainly 4 when my parents came where, when they were called Oriental, 5 which congers up images of exoticism, of the Far East notion of 6 otherness. And that's why that term is no longer usually used. 7 It's favored. And Asian American was invented. It was an 8 effort by people by to claim an identity for themselves, to 9 proclaim that this is who we are. We want to use this name 10 which shows and recognizes our roots as well as our American 11 status which says not only to ourselves, but to others that we 12 are here to stay as part of this country. 13 So Asian American first started to be used as a term 14 in the late 1960s. There were student movements in 15 California. At San Francisco State University, for example, 16 that followed the lead, like the Black Power Movement as they 17 looked at African-American students as they organized, and 18 protested, and were part of the multi-racial civil rights 19 movement. And they said, we, too, can do that. And so at San 20 Francisco State University in the late 1960s you saw people 21 suddenly stand up and proclaim that they were in favor of 22 Yellow power, and they invented the term Asian American. 23 Now as it happened, the census categories changed in 24 1970, as they often do, and the census began to offer the 25 option of checking off Asian American. So you see two GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 48 1 different strains here in this history that come together that 2 leaves this term Asian American. 3 The use of the term Asian Pacific American is meant 4 to embrace the many different islands from people also come, 5 that also share this common experience. 6 Q With that introduction I'd like to ask you to turn your 7 attention to questions about anti-Asian, anti-Asian Pacific 8 American racism. Can you tell us about the forms of 9 discrimination that persist against Asian Americans in the 10 United States? 11 A Sure. I would divided this into different types. First, 12 there is the straightforward racial prejudices and 13 discrimination and bias. The sort of thing that a consensus 14 now recognizes is wrong. And, second, a more a subtle form of 15 discrimination. A form of discrimination that may be in some 16 instances unconscious, or unintentional or even on its surface 17 not look like bias, but if you examine it a little closer, more 18 clearly reveals itself. Let me start though by talking about 19 the type. 20 Asian Americans face straightforward, plain old 21 racial discrimination. You see that, for example, in the 22 glass ceiling. If you look at the federal government's 1995 23 glass ceiling study, what you find is that in many categories 24 Asian American individuals who have the same qualifications as 25 their white peers, the same educational background, working in GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 49 1 the same types of jobs, earn less money. So when everything 2 is controlled for, what you find is there are racial 3 disparities, disparities that can be explained by nothing 4 other than racial backgrounds. So they face the glass ceiling 5 at the work place where even though they may have a Ph. D., 6 they simply can't make as much money. 7 There's a sociologist, Joyce Tang, who has studied 8 phenomenon and has taken a look at Asian Americans working in 9 technical fields. And she has found that many of the reasons 10 that people offer turn out to be false. Sometimes people say, 11 well, it's probably because all these Asians are coming from 12 someplace else, maybe they don't have good language skills. 13 Well, she did a study where controlled for nativity. She 14 looked at people born in the U. S. native-born Asian American, 15 and compared them with native-born Caucasian. And what she 16 found was not only did you still see these same disparities, 17 equally -- well-educated, qualified people, yet mysteriously, 18 at companies that presumably are not actively discriminating, 19 at companies that did hire these people, but they're just not 20 getting promoted, just not getting paid at the same rate. 21 In fact, in some instances, she found a real oddity, 22 that Asian Americans who are native born in many instances 23 make less than whites who are foreign born. So it has nothing 24 to do with whether or not you're born in the country. And it 25 doesn't have to do with language. GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 50 1 She found that what some people say about Asians not 2 being interested in management, that turned out to be false as 3 well. Sometimes you hear people, well, Asians are more 4 interested in the technical aspects of these things. They want 5 to be engineers. They don't want to managers. They don't want 6 to vice president. They don't want all that hassle stress. 7 They don't want to rise in the company. 8 Well, by using extensive surveys, that's simply 9 false. Asian Americans working in these technical fields are 10 just like their white peers. They do want to be in 11 management. They'd like to be in charge. They wouldn't mind 12 being the vice president, and they're just not offered those 13 opportunities. 14 Again, this happens not in every instance but often 15 enough that structurally, systematically a pattern emerges and 16 you see it again, and again, and again in rigorous, empirical 17 research. That's one example of glass ceiling. 18 Another example is if you look at housing. Housing 19 segregation for Asian Americans exist. It is not quite as bad 20 as a housing segregation for African-American, it is housing 21 segregation. Asian American of the same socio-economic status 22 as the whites who own houses in neighborhoods where they'd 23 like to buy can't buy into those neighborhoods quite often. 24 Asian Americans tend to live in segregated area. Again -- not 25 entirely, but you still find a persistent housing segregation. GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 51 1 You also see hate crimes. We've seen a rise in hate 2 crimes towards Asian Americans, ranging from the 1982 brutal 3 beating of Vincent Chin who was killed by two out-of-work auto 4 workers, who took a baseball bat, a Louisville slugger to his 5 head and just beat his head until he was senseless and in a 6 coma and died a few days later. Those out-of-work auto workers 7 who pursued Vincent Chin from a nightclub where they had all 8 been, blamed him. They had called him -- you'll have to 9 pardon my language, your Honor -- they called him you dirty, 10 fucking Jap as they were killing him. And that case I think 11 stands a symbol for many Asian Americans of the sort of 12 violence that can still occur to someone who is no different 13 than anyone else other than because of their racial 14 background. 15 Those two people who killed Vincent Chin, who 16 received probation, and a three thousand seven hundred dollar 17 -- three thousand seven hundred eighty-dollar fine said that 18 they blamed him because they were out-of-work auto workers, 19 and they thought it was because of people like him that they 20 were out of work. 21 So you see hate crimes. You see other instances. 22 We saw in the 1980s and 1990s a gang, a white gang in New 23 Jersey that attacked South Asian women. They called 24 themselves the "Dot Busters" in reference to the popular 25 movie, "Ghost Busters" and they would assault Asian women, GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 52 1 basically picked out at random, but selected because of their 2 ethnic background. 3 So you see hate crimes. You see these different 4 forms of racial discrimination that persist to this day. You 5 no longer -- and I hope we don't see anything like the 6 Internment again, we don't see the sort of discrimination 7 perpetrated by the government itself, but you see widespread 8 societal discrimination in instances which sometimes are 9 condoned, are condoned because people think, well, it's 10 different than other racial discrimination. These are 11 foreigners, they're not American, they don't have the same 12 rights. 13 Now, there is a second type of racial discrimination 14 though that Asian Americans face. And it's more subtle, but 15 in some ways every bit as dangerous. There is a myth called 16 the model minority myth. There's a stereotype of Asian 17 Americans. I think in order to understand how this stereotype 18 works, I first have to describe a stereotype. At first it 19 might strict people as quite a positive stereotype. It's the 20 stereotype of the Asian immigrant who comes here penniless 21 with nothing more than the shirt on his or her back, but who 22 by dent of hard work, confusion work, ethnic, good values, by 23 opening a small business, that they operate seven days a week, 24 twenty-four hours a days rises, that even though they speak 25 broken English, though they have a Ph. D. that they can't use, GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 53 1 and have to work running a drycleaners or a small store, they 2 nonetheless persevered and succeed. And then you see their 3 children become the whiz kids, the proteges who play the 4 violin at the age of five, and become valedictorian of the 5 high school, so when the top ten of the graduating class are 6 read off it's Chang, and Kim, and Betel. You see these images 7 of Asian Americans taking over college campuses, winning all 8 the scholarships, and going off to Harvard or Yale or Stanford 9 and breaking the curve in calculus or physics. Starting to 10 use science experiments at the age of thirteen. They're whiz 11 kids, geniuses. 12 You see this positive image of Asian Americans as 13 the so-called model minority. We see that in newspaper 14 articles. You see it in television programs. There was a 15 "New York Times" article that was entitled "Asian Going To the 16 Head of the Class" for example. In some of my published work, 17 I cite dozen of examples through the '80s and '90s of this 18 very positive glowing image of Asian Americans as super 19 successful. "Fortune Magazine" dubbed them the "super 20 minority." So it's this notion that somehow that Asian 21 Americans have triumph. They represent as another magazine 22 writer put it, "the triumph of the dream." 23 You might look at this and say what could possibly 24 be wrong with this, this is a wonderful celebration of 25 opportunity. It shows how well Asian Americans are doing. I GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 54 1 guess there are three things wrong with this, and I'd like to 2 go through them in order. 3 First, this is a stereotype, it's false. Second, if 4 that's not enough this causes the backflash for Asian American 5 who should be suspicious of any stereotype not matter how 6 positive because of what it can conceal. Third and finally, 7 it often is used as it was when it was first mentioned by the 8 "New York Times Sunday Magazine" an article by William 9 Peterson in 1966, to make an explicit comparison between Asian 10 Americans and African-Americans to say in effect, they made 11 it, why can't you. 12 Let me start with the first problem. The stereotype 13 is simply as a factual matter wrong. It is not an accurate 14 stereotype. Now, truth be told some Asian Americans have been 15 successful. They deserve praise. They deserve credit. I mean 16 to take nothing away from them. But if you take a look at 17 Asian Americans what you find is that Asian Immigration is 18 selective. Before 1965, before comprehensive changes to the 19 laws that were passed in 1965, there were tiny quotas for 20 Asian ethnic groups. For example, a total of no more than one 21 hundred and eighty-five individuals of Japanese descent could 22 come into the U. S. per year before 1965. So what you found 23 before 1965, people who came here tended to be well-to-do, or 24 had already gone to school and got a great deal of schooling 25 or both, well-to-do and people who had at least gone to GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 55 1 college if not had already gotten a master's or Ph. D., tended 2 to people who represent the cream of the crop of their home 3 land. This was a phenomenon we all know "brain trained." They 4 would come here and get Ph. D.s' and do well. 5 So when you look at Asian immigrants what you 6 realize is they're not representative, they're not 7 representative in Asian. They represent the luck few, the 8 ones who had the means to get out, or the talent to get out. 9 Now, that's important. It's important because it 10 means that the Asian Americans once they naturalized and 11 stayed here where their children have advantages. One of the 12 way you can predict how well educated a person will become is 13 to look at how well educated their parents were before them. 14 In fact, look at how well educated their father was. That's 15 one of the most robust social science factors that you can 16 look at if you want to get, if you just take any person, 17 you're going to ask what is the likelihood this person will 18 complete college, or this will get a master's, or a Ph. D. 19 One of the ways to figure out the answer to that is to ask, 20 well, what was the last year of schooling their father 21 completed. 22 Let me give you a concrete example. If you look at 23 South Asians, what you find is according to some studies as 24 many as two-thirds of them arrived in the United States with 25 better than a bachelor's, with at least a master's, a Ph. D. GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 56 1 or an M. D. So this creates a terrible skewing of our 2 picture. That means you're comparing Asian immigrants and 3 Asian Americans who are doctors when they arrive here against 4 a general U. S. population that haven't completed college, 5 that has just slightly on average done more than completed 6 high school. 7 So the first reason that this image is false is 8 because it wrongly suggests that all of the success is due 9 solely to opportunities in the United States. And, again, 10 don't get me wrong. This is a wonderful country. I am very 11 pleased that my parents came here, and that I was born here. 12 It does offer tremendous opportunity, and some people are able 13 to avail themselves of it, but it would be highly misleading 14 to suggest that Asian Americans by themselves as a racial 15 group represent in some way the triumph solely of the system 16 here. They represent instead a complicated table of factors 17 some of which have to do with who we open our doors to, and 18 who we welcome. That's one reason it's false. 19 Another reason it's false is because the most often 20 cited statistic that you hear is family income. You sometimes 21 hear as you did when the 2000 Census came out that Asian 22 Americans have attained parity. That average income for a 23 family of Asian Americans is equal to or greater than the 24 average income o, f whites. This is extraordinarily misleading 25 for several reasons. Let me detail some of them. GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 57 1 First of all, Asians are geographically 2 concentrated. A majority of Asian Americans live in 3 high-income, high-cost states. Hawaii, California and New 4 York constitute those three states, constitutes -- if not a 5 majority then certainly a plurality of the Asian American 6 population in the United States. Well, as it happens, those 7 three states also have people of all racial backgrounds a 8 higher than average family income. There are just not as many 9 Asian Americans in South Dakota or Alabama so when you look at 10 Asian American family income it's terribly inflated because of 11 this geographic skewing. 12 Asian American family income is also distorted by 13 the fact that on average, Asian Americans have larger families 14 with more wage earners. The typical Asian American family has 15 two wage earners. People of color tend to have families with 16 more wage earners. I mean, sometimes the Asian American 17 families with extended families living in one household, 18 everyone putting their income into a common pot. 19 Now, clearly, it doesn't make sense to compare a 20 household where yo have both adults working to make an income 21 of sixty thousand, let's say, against a household where you 22 have one wage earner making fifty-nine thousand, and then to 23 say that the two-earner household at sixty thousand is somehow 24 better off than the one wage earner household at fifty-nine 25 thousand. It may be true, but it's true only in the most GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 58 1 superficial and misleading sense. 2 Now, Asian Americans also tend to be much more 3 enterpreneural, tend to engage in small businesses endeavors 4 that are much higher risks. So what you find with Asian 5 Americans if you just look at the simple question of income, 6 Asian Americans have not obtained parity. They have obtained 7 parity only when you ignore the different factors. When you 8 look at individual Asian Americans as the 1995 Federal 9 Government Glass Ceiling Study did, what you find and get is 10 comparing individual Asian Americans controlling for education 11 level and occupational field, that Asian Americans make less 12 money on average than whites. It's unambiguous data so this is 13 just false in the sense that if you look at the condition of 14 Asian Americans, most Asian Americans are not the super 15 minority. 16 There are also significant ethnic differences, true 17 that Chinese Americans and Japanese Americans have incomes 18 that cluster toward the top if you do an ethnic breakdown. 19 But you also find that Southeast Asians, you find Phillipinos 20 and you find others clustered toward the bottom. Their 21 socio-economic status us much more similar to that of 22 African-Americans than it is of White Americans. So there are 23 tremendous ethnic variations as well. 24 So the stereotype is like most stereotypes, thin and 25 flimsy and just doesn't stand up to scrutiny. So no matter GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 59 1 what you think public policy should be if you simply ask 2 yourself where are Asian Americans, what is their status, this 3 notion that Asian Americans have made it and are well-to-do, 4 is incorrect. 5 Second, this image leads to backlash. Every part of 6 the positive stereotype is correlated to the connected 7 counter-part, and it gets flipped around very easily. Let me 8 offer a few examples. You sometimes here Asian Americans 9 described as hard-working. Well, hardworking very quickly 10 becomes unfair competition. You sometimes hear Asian 11 Americans described as good at math and science. I'm often 12 told, oh, could you fix my computer. You must be good with 13 computers. Yet, that quickly turns into they're nerdy and 14 geeky, and can't be lawyers, they can't be managers, they lack 15 of people's skills. You sometimes hear Asian Americans 16 praised for strong families, family values, a nuclear family 17 that stays together. 18 Yet that can be turned around. Asian Americans the 19 next can be criticized for being too clannish, too ethnic, too 20 insular, not mixing enough, self- segregating. 21 Let me give you concrete examples of when these 22 turnarounds occur. They tend to occur when there's some sort 23 of economic crisis. They did when the Chinese Exclusion Act 24 was passed in 1882, when the federal government for the first 25 time started to regulate the borders in a comprehensive GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 60 1 fashion. The first set of laws they passed were racial laws, 2 and they barred Chinese from coming to the United States, and 3 eventually a work steady created an Asiatic barred zone so 4 that people so that people of Asian descent could not come to 5 the United States. Those Asians who were in the U.S couldn't 6 naturalize because in order to naturalize you had to be a free 7 white person. And despite two Supreme Court challenges the 8 Ozawa case and the Thind case, in 1922, and 1923, in which a 9 Japanese person and a South Asian argued we are white, those 10 claims were turned down. 11 So what you found when the Chinese Exclusion Act was 12 first being proposed there was the working man's party in 13 California. It was an early labor movement led by Dennis 14 Currney. And he organized rallies in sandbox. And his 15 organizing cry was that Chinese must go. And the central claim 16 that the white laborers made because at that time San 17 Francisco was more than one third Asian, and San Francisco was 18 majority foreign born so that even the people who weren't 19 Asian, the people where of white ethnics, German, Italian, or 20 Polish, were foreign born and not native born. They 21 distinguished though the Asians and they said, the Asians were 22 competing unfairly, they work too hard. They said things such 23 as well, white workers can't just eat a bowl of rice a day. 24 The Chinese workers are inhuman. You're going to reduce to 25 their standard. If we have the Chinese, we can't have the GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 61 1 whites in California. California must be either White or 2 Yellow. Things like that. 3 And what you found was that the previous image, the 4 Asian immigrants as hard-working, that commended them for the 5 work of railroad where some thirteen thousand were hired and 6 organized into racial crews to compete against Irish laborers 7 to see who could lay more track in a day. That very same 8 notion that the Chinese laborers would work diligently, work 9 hard and not complain was turned against them, and then they 10 were said to compete unfairly. 11 You see the same thing with this notion of being too 12 nerdy or too good at math or science and not good with other 13 skills because that's the excuse most often offered when Asian 14 Americans ask well, why am I'm not being promoted to 15 management, why am I not being groomed, why I am not being 16 trained? The assumption is that they're only good at math and 17 science. 18 You hear stereotypes, you know, in the 19 1980s -- Brace Ellis, for example, published a novel, "Less 20 than Zero" where he referred to UCLA which was then becoming 21 predominantly Asian American, he said, UCLA, those words, 22 UCLA, stand for United Caucasian Lost Among Asians. You heard 23 people refer to MIT as Made in Tiawan. And these were 24 comments by white students who said, well we can't compete 25 these Asian students, they're just too good, they're just too GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 62 1 nerdy, and too geeky. 2 Gary Trudeau did a series of "Doonesbury Cartoons" 3 where he satirized this phenomena. We saw for example, the 4 white president, then president of Stanford University tell a 5 story about how when -- white students sign up for classes in 6 different technology fields and go to th class and they find 7 out that there are too many Asian American students, their 8 only choice is to drop the class because they figure they 9 won't be able to do any better than to get a C. 10 So you see example, after example, where every 11 positive trait is correlated exactly to the negative trait. So 12 that to be called hard-working as a double edge. 13 Another reason the model minority myth is dangerous 14 is because it is explicitly a comparison that's used not to 15 praise the Asian Americans at all but to insult African 16 Americans. 17 In 1966, a sociologist named William Peterson taught 18 at Berkeley wrote an article, "Success Story, Japanese 19 American Style." It was a popular article for the "New York 20 Times Sunday Magazine." He later followed it up with a book. 21 This same old article has been called the most influential 22 article ever written about Asian Americans. I think that's an 23 accurate of this single article. In it, Dr. Peterson was very 24 sympathetic. He talked about the Internment. He reviewed and 25 gave a summary of Japanese-Americans in the United States from GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 63 1 the early part of the century. He said they had done so well 2 that they had overcome every obstacle that race could put in 3 front of them. Then he said, I'm quoting here, "They examples 4 stands in contrast to what we might term problem minority." 5 And you didn't need the author there to nod and wink at you to 6 see who those problem minorities were because he then went on 7 to say that the only Japanese Americans who weren't successful 8 were juvenile delinquents who ran with as he put it, Negro and 9 Mexican gangs. So he very clearly set up this contrast 10 between Asian Americans as the successful minorities and 11 African Americans as the unsuccessful ones. 12 Q Is there anything about the content of stereotype against 13 Asian Americans that interacts specifically with the practice 14 of law? 15 A Absolutely. One of the popular stereotype of Asian 16 Americans is the sort of thing that I used to hear as a child 17 growing up. I still hear it now and then. A personal example, 18 you sometimes hear people say oh, Asians, you are all so 19 polite. I was once at a convention of the AAJA, The Asian 20 American Journalists. This was in 1987, in Los Angeles. There 21 was a guest speaker who had been brought to talk to us. He 22 opened up by saying, you know, I'm so pleased to be here, to 23 speak before all of you Asian Americans because you are all so 24 polite. And at that moment, everyone in the room hissed him. 25 To be polite, that seems like a compliment, oh, you're so GRUTTER - BOLLINGER, ET. AL. BENCH TRIAL -v- VOLUME 13 MONDAY, FEBRUARY 12TH, 2001 64 1 polite. But that's just someth