Plaintiffs' Reply to Defendants' Memorandum in Opposition to Plaintiffs' Motion in Limine

    IN THE UNITED STATES DISTRICT COURT 

FOR THE EASTERN DISTRICT OF NORTH CAROLINA 

EASTERN DIVISION 

Civil Action No. 4:96-CV-104-BO(3) 

MARTIN CROMARTIE, et al., 

Plaintiffs, 

V. 

JAMES B. HUNT, in his official capacity 

N
a
r
?
 

N
r
?
 

N
m
 

N
u
 

N
m
 

N
a
 

N
w
 

N
w
 

N
a
 

N
a
 

N
u
 

N
o
 

N
o
 

N
m
 

N
a
”
 

  

as Governor of the State of North Carolina, PLAINTIFF'S REPLY TO 

et al., DEFENDANTS’ MEMORANDUM 

State Defendants, IN OPPOSITION TO PLAINTIFFS’ 

MOTION IN LIMINE 

and 

ALFRED SMALLWOOD, et al., 

Defendant-Intervenors. 

INTRODUCTION 

Advancing a novel view of Daubert, the defendants would have this Court believe that 

“there is no need for a gatekeeper in this case,” Defendants’ Memorandum at 3, and “when there 

is no jury, however, there is no need for special protection against allegedly unreliable evidence, 

and the admissibility of Dr. Peterson’s testimony should be determined in the traditional fashion, 

after presentation of the expert’s qualifications and voir dire.” Id. at 3. However, Daubert itself 

applies Rule 702 of the Federal Rules of Evidence to triers of fact, not just juries. See Daubert v. 

Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 591, 113 S.Ct. 2786, 2796, 125 L.Ed.2d 469, 

492 (1993). Recognizing this fact, a number of federal courts have excluded testimony under 

 



  

Daubert in connection with bench trials.’ 

Defendants also point out that “vigorous cross-examination and presentation of contrary 

evidence are the ‘traditional and appropriate means of attacking shaky but admissible evidence.’ 

Defendants’ Memorandum at 5 (quoting Daubert, 509 U.S. at 596). Plaintiffs agree that Dr. 

Peterson’s evidence is “shaky” at best, and believe that this will be readily apparent to experienced 

judges. But it is also inadmissible under F.R.E. 702, since it is neither reliable nor relevant as 

these terms are interpreted by Daubert. 

I. DR. PETERSON’S SEGMENT ANALYSIS IS NOT RELIABLE. 

In their Response, defendants criticize the plaintiffs’ thorough discussion of the factors or 

“signposts” of Daubert on the ground that “plaintiffs fail to recognize that the so-called Daubert 

factors are intended to be no more than ‘useful signposts, not dispositive hurdles that a party must 

overcome in order to have expert testimony admitted.” Defendants’ Memorandum at 17 

(quoting Heller v. Shaw Industries, Inc. 167 F.3d 146, 152 (3™ Cir. 1999). Then, defendants 

  

'See, e.g. Bradley v. Brown, 42 F.3d 434 (7" Cir. 1994) (upholding the district court’s 

exclusion during a bench trial of the testimony of two physicians on the subject of Multiple 

Chemical Sensitivity disorder); In re: Unisys Savings Plan Litigation, 173 F.3d 145, 157 (3" Cir. 

1999) (explaining that “we would be hard pressed to require a District Court judge sitting in a 

non-jury case who credibly and with reason found that he could not believe a witness to 

nevertheless hear the witness’s direct examination, cross-examination, and rebuttal examination in 

an extended trial when he knew that he would only reject it as unbelievable”); Jones v. United 

States, 933 F.Supp. 894 (N.D.Cal. 1996) (excluding plaintiffs’ scientific evidence regarding 

interaction between antibiotics and oral contraceptives in wrongful birth action brought under 

Federal Tort Claims Act); Smith v. Rasmussen, 57 F.Supp.2d 736 (N.D. Iowa 1999) (excluding 

state defendants’ witness in a suit seeking Medicaid benefits for a sex reassignment surgery), In re 

Dow Corning Corporation, 237 B.R. 364 (Bankr. E.D. Michigan 1999) (excluding a CPA’s 

proposed testimony as to sufficiency of litigation fund given lack of evidence that accountant’s 

testimony was based on reliable data and methodology). 

2Plaintiffs in fact quoted this very language, See Plaintiffs’ Memorandum at 16, and 

addressed this issue in Section IIB of their memorandum under the heading: “Dr. Peterson’s 

2 

 



J? << 

mischaracterize the plaintiffs’ “three or four primary concerns with Dr. Peterson’s analysis which 

are then repeated again and again as support for the absence of almost every factor.” Defendants’ 

Memorandum at 7. 

A. Dr. Peterson’s testimony is unreliable because it does not support his purported conclusion. 

While defendants pretend that the plaintiffs’ objection to Dr. Peterson’s testimony is that 

he uses statistics, See Defendants’ Memorandum at 7, the plaintiffs’ actual objection is that the 

methodology used by Dr. Peterson in his statistical analysis can not reliably answer the questions 

he poses. Daubert requires that the gatekeeper look behind the facade at the methodology used 

and assess its reliability.> That is what plaintiffs are requesting the Court to do. 

Plaintiffs have consistently contended that Dr. Peterson’s testimony is unreliable because, 

instead of counting people, Dr. Peterson counted precincts. Instead of looking at precincts as a 

political scientist would in this context, as a group of uneven building blocks, each with its own 

unique population, shape, and characteristics, Dr. Peterson treated them as equal units. Because 

Dr. Peterson did not weight his segments but only assigned one binary comparison to each of 

them, his analysis is a meaningless mathematical exercise that has nothing to do with the reality of 

the Twelfth District. This is plaintiffs’ chief underlying concern with Dr. Peterson’s analysis. This 

  

‘segment analysis’ does not support what it purports to show, violating the essence of Daubert.” 

Id. at 22. If the Daubert factors have any meaning, surely the fact that all of the various factors 
or “signposts” point to inadmissibility must have significance to this Court. 

>This Court under Daubert has the responsibility of determining whether Dr. Peterson’s 
testimony is based on scientific validity, i.e. whether or not the principle supports what it purports 

to show. See Daubert, 509 U.S. at 590 n.9, 113 S.Ct. at 2795 n.9, 125 L.Ed.2d at 481 n.9. 

3  



defect alone makes everything Dr. Peterson concludes unreliable and therefore inadmissible. * 

Although the plaintiffs time and time again repeated this criticism in their memorandum, 

defendants have given no answer to it. 

This is also why it is significant that Dr. Peterson cannot identify precincts or segments 

and their particular characteristics on a map. The issue is not whether the precinct and segments 

have been accurately entered into Dr. Peterson’s database,’ the issue is how Dr. Peterson’s 

analysis treats them as binary bits of data, weighted equally when in fact each one of them is 

different from all the others. 

The second recurring concern of the plaintiffs has been accurately identified by the 

defendants: the anomalies in the data. Dr. Peterson and the defendants claim that these anomalies 

are simply irrelevant because this is the data that the state used to create the 1997 Twelfth 

District. See Defendants Memorandum at 12-13. However, the evidence shows that the General 

Assembly did not use this data in the way that Dr. Peterson used it.” Also, this new position is 

  

“Any step that renders the analysis unreliable, renders the expert’s testimony 
inadmissible” Curtis v. M&S Petroleum, Inc. 174 F.3d 661, 670 (5" Cir. 1999) (quoting In re 
Paoli Railroad Yard PCB Litigation, 35 F.3d 717, 745 (3" Cir. 1994). Plaintiffs have pointed 
out numerous steps that render Dr. Peterson’s analysis unreliable and inadmissible. 

>The plaintiffs are not, however, willing to concede that this is the case since a preliminary 
examination of several of the maps used by Dr. Peterson seems to show that boundary segments 
around the 12" District were not entered with complete accuracy. 

®This differs from Dr. Peterson’s deposition testimony where he said that if inaccurate data 
were included in his analysis the proper recourse would be to substitute the correct information. 
See Peterson Deposition at 104. 

"For example, in his analysis Dr. Peterson compared the racial percentages of the (a) total 
population, (b) voting age population, and (c) registered voters, with the percentage of voters 
registered as Democrats and with the percentage of votes for the Democratic candidate in each of 
three statewide elections. On these twelve comparisons he based his conclusions. On the other 

4  



  

inconsistent with Dr. Peterson’s deposition testimony.® Furthermore, as to the Davie County 

data, it is highly unlikely that the General Assembly ever considered the alternative of including 

any of Davie County in the 1997 Plan District 12.° 

Plaintiffs’ third recurring concern is with Dr. Peterson’s treatment of the “convergent” 

segments. These segments-about 80% of the total segments- were arbitrarily and improperly 

excluded from any further consideration when exploring the differences between the possible race 

and party hypotheses. This undercuts the defendants’ contention that Dr. Peterson performed a 

“straightforward and scientifically objective analysis to the racial and political data available in this 

case.” Defendants’ Memorandum at 7.'° 

  

hand, Gerry Cohen, the redistricting plan’s chief draftsman, has testified that “I looked almost 

exclusively at total population and election returns.” See Cohen Deposition at 34. 

*Dr. Peterson testified that he did not know whether the legislature relied on inaccurate 

data in the data set he used. See, e.g. Peterson Deposition at 103. Perhaps Dr. Peterson was 

given the assurance he reiterates in § 3(c) of his Third Affidavit after the deposition. Since that 

information could not have guided his decision about the data to use, this is nothing more than a 
post hoc rationalization. 

’Furthermore, Dr. Peterson’s segment analysis gives the Davie County precinct/township 

segments the same weight as the segments involving racial splits in Charlotte, Winston-Salem, and 

Greensboro, even though 44% of the black population of the Twelfth District is in Mecklenburg, 

27.% is in Guilford, and 16.7% is in Forsyth. In contrast, the bridge counties each have less than 
5% of the total black population of the Twelfth District. 

'% Defendants have taken issue with plaintiffs’ characterization of Dr. Peterson’s use of his 
“convergent” segments as arbitrarily discarding 80% of the data. See Defendants’ Memorandum 
at 10 n.6. However, after determining that various segments were in fact convergent and 

counting the respective numbers of segments for which party and race converged, Dr. Peterson 
did ignore the differences in race and party in the convergent segments. Thus, he arbitrarily 
discarded approximately 80% of his data and thereby disregarded many of the precincts whose 

inclusion in the Twelfth District displayed the General Assembly’s predominately racial motive. 
Plaintiffs submit that any reliable and relevant analysis should consider the magnitude of 
differences in the respective race and party percentages of Dr. Peterson’s “convergent segments.” 

5 

 



  

Last, but not least, plaintiffs submit that any methodology that purports to show whether 

or not race predominated in the drawing of the Twelfth District must take into account not only 

the boundary but also the populous core precincts. Surely, any anlaysis which concludes that race 

was not the predominate factor is unreliable if it treats as immaterial whether the inside precincts 

of the Twelfth District not touching its boundary are 100% white or 100% black! See Peterson 

Deposition at 70-71. Dr. Peterson’s analysis is flawed because it does not consider what 

alternative land bridges were available to the General Assembly in linking together the racial cores 

of Forsyth and Guilford with the racial core of Mecklenburg County while keeping the District 

population around 550,000. Thus, it doesn’t consider the overarching decisions that led to the 

formation of the Twelfth District, including which cities and counties to split along racial lines and 

how to run the rural land bridges between the urban population clusters. 

B. Dr. Peterson’ ment analysis is found lacking under all relevant factors or sign 

For some unexplained reason, the defendants have characterized the comments in 

plaintiffs’ memorandum about the tested/testable factor in Daubert as one of “three or four 

primary concerns with Dr. Peterson’s analysis which are then repeated again and again as support 

for the absence of almost every factor.” Defendants’ Memorandum at 7. Plaintiffs see no basis 

for this characterization. Rather, the tested/testable factor was just one of eight that the plaintiffs 

discussed-of which the defendants only attempt to refute four. See Plaintiffs’ Memorandum at 

  

For example, if an inner precinct had 73% African Americans and 79% Democrats, and an outside 

precinct had 5% African Americans and 78% Democrats, this would be a “convergent segment” 

according to Dr. Peterson’s definition, and would not have been used to explore any potential 

difference between party and race in the General Assembly’s motive. To be reliable, Dr. 

Peterson’s analysis should have taken into account the differences between all the segments along 

the border, in addition to weighting the segments and using correct data. 

6 

 



  

The defendants contend that Dr. Peterson’s analysis is testable because it poses three 

testable questions and goes “about answering these questions objectively by applying standard 

statistical methods using the available racial and party preference data.” Defendants’ 

Memorandum at 16. 

However, even if the three questions that Dr. Peterson poses were “hypotheses capable of 

empirical testing,” Daubert, 509 U.S. at 593, the methodology of Dr. Peterson is completely 

inadequate to the task. Moreover, for testability plaintiffs believe that the bottom line issue is not 

whether Dr. Peterson’s data and methodology can be reproduced for the 12" District of the 1997 

plan, but whether Dr. Peterson’s “segment analysis” has been or even can be reproduced and 

applied to other congressional districts previously found to be drawn with race as a predominate 

factor. Dr. Peterson’s “segment analysis” has not been tested on any other district; and for 

districts that split precincts frequently, it can not be tested." 

As for peer review, plaintiffs were quite surprised to learn that their counsel and experts 

have been designated by the defendants as Dr. Peterson’s peers for the purposes of reviewing his 

“segment analysis.” See Defendants’ Memorandum at 17. Plaintiffs doubt that this is the peer 

review that the Daubert Court had in mind." 

With respect to the unacceptably high known or potential rate of error of Dr. Peterson’s 

  

See Peterson Deposition at 16. 

This lack of testing Dr. Peterson’s method on previously adjudicated unconstitutional 

districts also affects other Daubert factors, such as the rate of error and lack of standards. 

3paintiffs believe that the Supreme Court intended a favorable evaluation from a review 

associated with acceptance of research in peer-reviewed professional journals. If the defendants’ 

interpretation of peer review is accepted, then whenever any expert testimony is reviewed and 

challenged by the opposing party, that would constitute peer review. | 

7 

 



  

“segment analysis,” the defendants would have this Court dismiss plaintiffs’ argument that “the 

‘variable results’ from the multiple analyses of divergent segments ‘show the unreliability of the 

analysis’ as “patently absurd.” To the contrary, “Dr. Peterson repeatedly has stated that each of 

his several analyses conveys the same message- that no matter how you measure, there is almost 

the same level of support for the race hypothesis and for the political hypothesis, although on 

balance the data tip slightly in favor of the political hypothesis.” Defendants Memorandum at 16. 

Furthermore, “what plaintiffs fail to comprehend is that the results are in fact highly consistent 

and, hence, reliable.” 1d. 

Plaintiffs have contended that the flaws in the data change the outcomes of two out of the 

three analyses of Dr. Peterson. This seriously undercuts the claim that “no matter how you 

measure, there is almost the same level of support for the race hypothesis and for the political 

hypothesis, although on balance the data tip slightly in favor of the political hypotheses.” Id. 

Furthermore, the defendants wish to have it both ways on statistical significance. On the 

one hand, when asked in his deposition whether “the degree of difference in the explanatory data 

(was) sufficient to permit you to say that one is a better fit than the other,” Dr. Peterson replied: 

“Yes.” Peterson Deposition at 13. On the other hand, when confronted with an analysis of his 

data that would correct it and tip the balance in favor of the racial hypothesis, the defendants reply 

that the proper statistical conclusion to be drawn from Dr. Peterson’s analysis is “that there is 

very little difference between the two alternative theories.” Defendants’ Memorandum at 14. 

Also, while defendants maintain that it is appropriate for Dr. Peterson to conclude that a six to 

 



  

four majority would tip the balance statistically in favor of the political hypothesis," they respond 

that a balance of four to three in favor of the racial hypothesis is “not statistically significant.” 

Defendants’ Memorandum at 13.7 

Finally, the defendants have resubmitted Dr. Peterson’s resume and pointed out his long 

history of testifying in federal courts. However, as the Supreme Court has made clear, a person 

who is qualified as an expert in one area of expertise may be precluded from offering opinions 

beyond that expertise or which are not founded on a reliable methodology. See Kumho Tire Co. 

v. Carmichael, U.S. 119 S.Ct. 1167, 143 L.Ed.2d 238 (1999) (the proffered expert 

qualified as an expert in mechanical engineering, but his methodology in analyzing a particular tire 

failure was not reliable). 

II. DR. PETERSON’S TESTIMONY IS NOT RELEVANT. 

In their discussion concerning the relevance prong of Daubert, the defendants simply 

ignore the plaintiffs’ argument and assert that “it is absurd to argue that this testimony is not 

relevant to the issue at hand.” Defendants’ Memorandum at 18. “Absurd” as plaintiffs’ argument 

may or may not be, the defendants have not refuted it. Defendants assume that Dr. Peterson’s 

  

14 «The statistical support for the Political Hypothesis is at least as strong as that for the 

Race Hypothesis, and indeed, slightly stronger.” First Peterson Affidavit at 20 (commenting on 

the six to four breakdown of the unequivocally divergent segments). 

I5Dr. Peterson also testifies to this effect in § 23 of his Third Affidavit. However, in his 

deposition he refused to comment whether differences that were less than .01 were statistically 

significant or trivial. “When I use the term “significant” it is within the context of a particular 

probability model. And I can’t think of a probability model that would be appropriate to employ 

here.” Peterson Deposition at 56. At any rate, plaintiffs’ arguments concerning the reliability of 

Dr. Peterson’s testimony do not hinge on a change in the outcomes when rerunning Dr. 

Peterson’s analyses using corrected data. Rather, they believe that these corrections merely 

highlight one more wrong step in Dr. Peterson’s stairway of errors. 

9 

 



  

testimony is relevant simply because its conclusions speak to a matter at issue in this case. 

However, as plaintiffs have demonstrated, the fit required by the Daubert relevance prong 

requires much more, and Dr. Peterson’s testimony fails in this regard. See Plaintiffs’ 

Memorandum at 24-28. 

Nor are plaintiffs inconsistent in arguing that there is no fit or relevance to Dr. Peterson’s 

testimony because they previously presented evidence of inclusion or exclusion of voters along the 

boundary precincts of the Twelfth District. Contrary to defendants’ suggestion otherwise, 

plaintiffs have never contended that the boundary precincts are irrelevant. Plaintiffs contend that 

Dr. Peterson’s testimony is irrelevant under Daubert because it tells the Court nothing meaningful 

about the boundary precincts. Dr. Peterson’s analysis tells this Court nothing about how many 

people with what characteristics were included or excluded along the boundary of the Twelfth 

District. That, in essence, is why it is irrelevant. 

Respectfully submitted, 

This 17 day of November, 1999 A | /| : { 
ia Ke { 

CA W {J 
\ 

Robinson O. Everett 

Everett & Everett 

N.C. State Bar No.: 1385 

Attorney for the Plaintiffs 

P.O. Box 586 

Durham, NC 27702 

Telephone: (919)-682-5691 

  

10 

 



  

Williams, Boger, Grady, Davis & Tuttle, P.A. 

io Py SEE 11 es 
; 

by: 4 2dr 
Martin B. McGee 

State Bar No.: 22198 

Attorneys for the Plaintiffs 

P.O. Box 810 

Concord, NC 28026-0810 

Telephone: (704)-782-1173 

  

  

Douglas E. Markham 

Texas State Bar No. 12986975 

Attorney for the Plaintiffs 

333 Clay Suite 4510 

Post Office Box 130923 

Houston, TX 77219-0923 

Telephone: (713) 655-8700 

Facsimile: (713) 655-8701 

Robert Popper 

Attorney For Plaintiffs 

Law Office of Neil Brickman 

630 3" Ave. 21* Floor 
New York, NY 10017 

Telephone: (212) 986-6840 

Seth Neyhart 
Attorney For Plaintiffs 

N7983 Town Hall Road 

Eldorado, WI 54932 

Telephone: (920) 872-2643 

11 

 



  

CERTIFICATE OF SERVICE 

I certify that I have this day served the foregoing Plaintiffs’ Reply to Defendants’ Memorandum in 
Opposition to Plaintiff’s Motion in Limine by hand delivery to the following addresses: 

Ms. Tiare B. Smiley, Esq. 

Special Deputy Attorney General 

North Carolina Department of Justice 

114 W. Edenton St., Rm 337 

P.O. Box 629 

Raleigh, NC 27602 

Phone # (919) 716-6900 

Mr. Adam Stein 

Ferguson, Stein, Wallas, Adkins, Gresham, Sumter, P.A. 
312 W. Franklin St. 

Chapel Hill, NC 27516 

Phone # (919) 933-5300 

In addition, I have served the above mentioned document by U.S. mail to the following 
address: 

Mr. Todd A. Cox 

NAACP Legal Defense & Educational Fund, Inc. 

1444 Eye Street, NW 10" Floor 
Washington, DC 20005 

This the 17" day of November, 1999 

BR 
Robinson O. Everett 

Attorney for the Plaintiffs 

  

12