Waller v. Butkovich (593 F. Supp. 942, 1984)

MEMORANDUM AND ORDER

MERHIGE, District Judge.

Plaintiffs have applied to the Court for permission to appeal, pursuant to 28 U.S.C. § 1292(b), this Court’s order of July 11, 1984. That order denied plaintiffs’ motion to stay trial proceedings in this case. Plaintiffs based their motion to stay on the grounds that the jury plan now in effect in the Middle District of North Carolina substantially underrepresents black people in the district.

In addition to plaintiffs’ materials, two amicus briefs have been submitted in support of the plaintiffs’ application. A number of organizations and individuals have joined in each brief. Pending before the Court are motions by these groups for permission to file the two already-submitted amicus briefs. The Court will direct the filing of the briefs in support of the application.

The Court has carefully considered plaintiffs’ application, their supporting memorandum, and the briefs of amici. It has also reviewed plaintiffs’ memorandum in support of their prior motion to stay the trial proceedings. In addition, the Court has considered more recent data about the proportion in the various stages of the jury selection process. Having thoroughly considered the relevant facts, statutes, and case law, the Court must deny plaintiffs’ application for the reasons stated below.

Background

Plaintiffs filed a motion to stay the trial proceedings (hereinafter “Motion to Stay”), pursuant to 28 U.S.C. § 1867(c), on July 2, 1984. They complained in their motion that the jury selection plan for the Middle District of North Carolina impermissibly underrepresents black people; thus, they sought the modification of the current jury selection plan to correct for the alleged underrepresentation prior to the selection of a petit jury in this case. The legal basis for plaintiffs’ claims was primarily two sections of the Jury Selection and Service Act, codified at 28 U.S.C. §§ 1861, 1982. Although plaintiffs also alluded to the due process clause of the Fifth Amendment and the trial-by-jury clause of the Seventh Amendment in their motion, see Motion to Stay, ¶ 9, their memorandum addressed itself exclusively to their Jury Selection and Service Act claim.

Plaintiffs’ memorandum presented various statistics to describe the extent to which black people are underrepresented in the Middle District’s jury selection process. Their statistics were based on a study conducted by Dr. John Ruoff, whose research specialties include advanced work in demography. His study relied on three sources: the U.S. Census (for the racial composition of the voting-aged population of the Middle District); a 1980 report on registration statistics prepared by the Executive Secretary-Director of the North Carolina State Board of Elections (for the racial composition of voter registration lists in the Middle District); and the Court’s JS-12 report dated December 31, 1981 (for the racial composition of potential jurors).

Plaintiffs have presented statistics about the proportion of non-whites in various phases of the juror selection process in order to compare those proportions with the proportion of non-whites in the population who are age-eligible to vote. A brief description of the relevant phases of the Middle District’s plan helps to place these statistics in context. This description is based on the Middle District’s “Plan for Random Selection of Jurors,” approved by the Reviewing Panel on January 29, 1981, amended February 14, 1983 (the “Jury Plan”). See APPENDIX “A.” At least once every four years, the District’s Clerk compiles a “master jury wheel” from voter registration lists maintained by each of the counties in the Middle District. See Jury Plan at (VI)(A)(3). Those voter lists are the official records of persons registered to vote in the most recent national general election. See id. at (V)(B). Because the four-year periods commenced in January, 1981, the most recent national general elections will ordinarily be presidential elections. The master jury wheel must include at least one-half of one percent of all voters on the lists. The Clerk selects names from the voter lists for the master jury wheel according to a method designed to ensure randomness. See id. at (VI)(B).

From the master jury wheel, the Clerk generates a “qualified jury wheel” periodically, when directed by the Court. See Jury Plan at (VI)(C)(1). The Clerk first mails out questionnaires to names selected at random from the master jury wheel. See id. at (VI)(C)(1). The Clerk then reviews the returned questionnaires, culling out people who are unqualified, exempted, or excused. See id. at (VII)(A). The remaining names constitute the qualified jury wheel. The qualified jury wheel currently in use in the Middle District was generated at least as recently as June, 1983. See Memorandum, from J.P. Creekmore (Clerk), to Chief Judge Hiram H. Ward (August 17, 1983) (APPENDIX “B”).

According to Dr. Ruoff’s study, 19.87% of the population that is age-eligible to vote (hereinafter the “age-eligible population”) of the Middle District is “non-white.” The proportion of the registered voters in the Middle District that was non-white as of April 8, 1980 was 15.88%. In addition, 14.88% of the sample of registered voters selected on July 6, 1981 from the master jury wheel to whom questionnaires were sent was non-white. Finally, 13.60% of the qualified jury wheel, as of December 31, 1981, was non-white.

Plaintiffs provided other statistics from the Ruoff study as well, concerning absolute disparities, comparative disparities, and standard deviations. Table A, below, shows some of the most significant statistics that plaintiffs use in their argument for impermissible underrepresentation.

TABLE A:

Plaintiffs’ statistics concerning disparities in the Middle District of North Carolina between the percentage of non-whites who are age-eligible to vote and the percentage of non-whites at various stages of the jury selection process (1981).

Level of Analysis

People who are age-eligible to vote:

People who are registered voters:

People who were mailed jury questionnaires:

People who were on qualified jury wheel:

Total Number of People in Level of Analysis

1,347,097

856,126

3,341

2,250

% of NonWhites in Level of Analysis

19.87%

15.88%

14.88%

13.60%

Absolute Disparity

-3.99%

-4.99%

-6.27%

Comparative Disparity

-20.07%

-25.13%

-31.55%

Standard Deviations

92.49

7.23

7.45

The plaintiffs also briefly discussed, in their Memorandum Supporting the Motion to Stay, two facts relating to discrimination against blacks in North Carolina in the voter registration process. First, plaintiffs referred to findings by a three-judge district court in Gingles v. Edmisten, 590 F.Supp. 345 (E.D.N.C.1984), appeal pending. The three-judge court in Gingles set aside portions of North Carolina’s legislative reapportionment on the basis of Section 2 of the Voting Rights Act of 1965, 42 U.S.C. § 1973. The Court also made express findings of past discrimination and its continuing effects in the form of relatively depressed levels of black voter registration, with respect to two of the twenty-four counties in the Middle District. Second, plaintiffs note that seven other counties in the Middle District are designated for coverage under Section 5 of the Voting Rights Act of 1965, 42 U.S.C. § 1973c, “because of their use of discriminatory literary tests for voting and their continuing low levels of voter registration or voter turnout.” Memorandum in Support of Motion to Stay, at 9. The plaintiffs do not appear to claim that intentional discrimination against blacks in voter registration continues. In any event, plaintiffs did not focus on supporting a claim of intentional discrimination, past or present; and little if any of their legal argument is based on such claims.

In a two-page order, this Court, on July 11, 1984, denied plaintiffs’ Motion to Stay. On August 7, plaintiffs filed the instant application for permission to appeal the order.

In order to better evaluate plaintiffs’ assertions of underrepresentation, the Court also considered more recent data on the disparities involved. The Clerk’s Office of the Middle District of North Carolina furnished data on the proportion of blacks in various stages of the jury selection process based on the Middle District’s qualified jury wheel, which was assembled between February and June of 1983. The Clerk’s Office also furnished data, based on the 1980 census, on the proportion of voting-aged blacks in the total population of the Middle District. This data is more accurate than that tendered by plaintiffs, in two important ways. First, the information concerning jury selection is more recent by two years. Second, all the information is more relevant than plaintiffs’ data to plaintiffs’ claim of underrepresentation of blacks, because its data is directly about blacks (by contrast, plaintiffs’ proferred data is about “non-whites” and is simply a surrogate for blacks).

This most recent data is assembled in Table B, which follows. The data reveals disparities somewhat smaller than those in Dr. Ruoff’s study on which the plaintiffs rely.

TABLE B:

Clerk’s Office statistics concerning disparities in the Middle District of North Carolina between the percentage of blacks who are age-eligible to vote and the percentage of blacks at various stages of the jury selection process (August 1983).

Level of Analysis

People who are age-eligible to vote:

People who are registered voters:

People who were mailed jury questionnaires:

People who are on Qualified Jury Wheel:

Total Number of People in Level of Analysis

(not provided)

(not provided)

5,013

3,079

% of Blacks in Level of Analysis

19.06%

15.50%

15.36%

14.62%

Absolute Disparity

-3.56%

-3.70%

-4.44%

Comparative Disparity

-18.68%

-19.41%

-23.29%

Standard Deviations

(cannot be computed with data provided).

6.67

6.29

Discussion.

I. Appeal Under 28 U.S.C. § 1292(b).

To grant permission to appeal under 28 U.S.C. § 1292(b), the Court must:

... be of the opinion that such order involves a controlling question of law as to which there is substantial ground for difference of opinion and that an immediate appeal from the order may materially advance the ultimate termination of the litigation____

The Court assumes for the purposes of decision that the question of the representativeness of the Middle District’s jury plan is a controlling one. Nevertheless, the Court must also find that there is substantial ground for difference of opinion on the question, and that an immediate appeal from the Court’s challenged order may materially advance the ultimate termination of the suit. After a careful review of the case law, the Court, for the reasons which follow, strongly believes that there is no substantial ground for difference of opinion on the question. Having so concluded, granting permission to appeal would be a dereliction of the Court’s responsibility. Further, this lawsuit’s just and already long delayed resolution will very likely be significantly advanced by denying permission to appeal at this stage. The parties and several non-party government agencies are currently intensely involved in pre-trial discovery. Appealing the jury selection plan now would interrupt their discovery proceedings and distract the parties and their counsel from the issues of privilege and relevance. They have spent much time becoming familiar with the law and the facts relating to these issues here. The pace of discovery and the fact-finding process would suffer if permission to appeal now were granted. Given the Court’s conviction that the jury selection plan is clearly not unlawful, an appeal on the issue now would undermine the imminent resolution of this suit both by consuming time in the appeal itself and by interrupting the intense discovery currently underway.

II. Violation of the Jury Selection and Service Act

A. Overview of the Jury Selection and Service Act.

The Jury Selection and Service Act (hereinafter “the Act”) states, in pertinent part, that:

It is the policy of the United States that all litigants in Federal courts entitled to trial by jury shall have the right to grand and petit juries selected at random from a fair cross section of the community in the district or division wherein the court convenes.

28 U.S.C. § 1861. The Act further states that “no citizen shall be excluded” from juror service in the federal district courts “on account of race, color, religion, sex, national origin, or economic status.” 28 U.S.C. § 1862. The Act also provides that the names of prospective jurors “shall be selected from the voter registration lists or the lists of actual voters of the political subdivisions within the district or division.” 28 U.S.C. § 1863(b)(2). However, it also requires that a district’s or division’s jury selection plan prescribe some other source or sources of names in addition to voter lists where necessary to foster the policy and protect the rights secured by sections 1861 and 1862 of this title. Id. These rights include the one at issue here: the right to a petit jury selected from a “fair cross section of the community.”

The Act gives civil litigants a right of enforcement. 28 U.S.C. § 1867(c). Through that subsection, any party acting in a timely fashion may move to stay proceedings on the ground of “substantial failure to comply with the provisions of this title in selecting the petit jury.” Id. Thus, plaintiffs here contend that the Middle District’s plan substantially fails to provide a fair cross section of the community from which jurors can be drawn because of the underrepresentation of blacks on voter registration lists, and that the plan therefore also substantially fails to prevent the exclusion of blacks from jury service on account of their race.

B. Discriminatory Intent.

Plaintiffs correctly maintained, in their Memorandum in Support of Motion to Stay (hereinafter “plaintiffs’ Memorandum”), that intentional discrimination in the jury selection process is not required in order to establish a claim based on the Act.

As many courts have held, the Act’s fair-cross-section requirement establishes the same standards, in terms of securing federal juries from a “fair cross section of the community,” as the Sixth Amendment establishes for criminal defendants. See, e.g., United States v. Test, 550 F.2d 577, 584-85 (10th Cir.1976) (en banc) (discussing the basis, in both the case law and the legislative history of the Act, for that assumption); United States v. Blair, 493 F.Supp. 398, 405-06 (D.Md.1980), aff'd, 665 F.2d 500 (4th Cir.1981). And the Supreme Court has held, in the context of the Sixth Amendment fair-cross-section requirement, that discriminatory intent is unnecessary to establish a violation. Duren v. Missouri, 439 U.S. 357, 368 n. 26, 99 S.Ct. 664, 670 n. 26, 58 L.Ed.2d 579 (1979). The Supreme Court in Duren distinguished equal protection cases, where discriminatory intent must be shown, from Sixth Amendment fair- cross-section cases. Under the Sixth Amendment, “systematic disproportion itself demonstrates an infringement of the defendant’s interest in a jury chosen from a fair community cross section.” Id. Given the equivalence of the fair cross section requirement of the Sixth Amendment and the Act, it follows that discriminatory intent need not be shown in order to establish a violation of the Act. At least one court reached this conclusion even before Duren. See United States v. McDaniels, 370 F.Supp. 298, 301 (E.D.La.1973), aff'd sub nom United States v. Goff, 509 F.2d 825 (5th Cir.), cert. denied, 423 U.S. 857, 96 S.Ct. 109, 46 L.Ed.2d 83 (1975).

C. Elements of a Jury Selection and Service Act Claim.

The Supreme Court in Duren clearly set out the three elements of a prima facie case for violation of the fair-cross section requirement of the Sixth Amendment. Those elements are:

(1) that the group alleged to be excluded is a “distinctive” group in the community; (2) that the representation of this group in venires from which juries are selected is not fair and reasonable in relation to the number of such persons in the community; and (3) that this under-representation is due to systematic exclusion of the group in the jury-selection process.

439 U.S. at 364, 99 S.Ct. at 668. As previously discussed, the Act creates the same standard for what constitutes a fair cross section of the community as the Sixth Amendment. Thus, the Duren test for violations of the fair cross section standard is appropriate for claims based on violations of the Act’s fair cross section requirement.

1. Distinctive Group in the Community.

A fair-cross-section violation requires that the underrepresentation involve a “ ‘distinctive’ group in the community,” according to Duren. 439 U.S. at 364, 99 S.Ct. at 668. The Court of Appeals for the Tenth Circuit has formulated the following test to determine whether a group is “distinctive”:

(1) the presence of some quality or attribute which defines and limits the group; (2) a cohesiveness of ‘attitudes or ideas or experience’ which distinguishes the group from the general social milieu; and (3) a “community of interest” which may not be represented by other segments of society.

U.S. v. Test, supra, 550 F.2d at 591. Blacks in the Middle District of North Carolina easily satisfy these standards. Their color, their common history of de facto and de jure segregation, and their continuing efforts to overcome the effects of that discrimination, need only be mentioned. In addition, at least one other court in this circuit has assumed that blacks are a “distinctive group” for purposes of fair-cross-section analysis. See U.S. v. Blair, supra, 493 F.Supp. at 407.

2. Fair and Reasonable Representation.

The second element of a prima facie case for violation of the fair-cross-section is that the “representation of [the distinctive group] in venires from which juries are selected is not fair and reasonable in relation to the size of the group in the community.” Duren, 439 U.S. at 364, 99 S.Ct. at 668. Thus, the next question in this case is to determine whether the representation of blacks on the voter registration lists and qualified jury wheel is fair and reasonable in relation to their number in the community.

Plaintiffs have submitted a variety of statistics, many of which were introduced in the “Background” section of this memorandum, supra, to support their claim that the Middle District’s plan does not fairly and reasonably represent black people. Plaintiffs’ statistics are not, however, for blacks. Rather, they are for “non-whites.” While “non-whites” may not be a distinctive group for purposes of underrepresentation of claims brought under the Sixth Amendment and the Act, the Court considers plaintiffs’ statistical representations here about non-whites as a surrogate for statistical representations about blacks. According to the affidavit of plaintiffs’ statistician, “Indians and Others represent only 4% of the Non-White population of the Middle District.” Affidavit of John C. Ruoff (hereinafter “Ruoff Affidavit”), at 2, note. The court accepts Dr. Ruoff’s conclusion that “references to non-whites are essentially references to Black people.” Id. Assuming that blacks constitute the same percentage of the other categories about which plaintiffs have submitted statistics (that is, “registered voters,” “people to whom questionnaire samples were mailed,” and “qualified jury wheel”), then the computations of absolute disparity, comparative disparity, and standard deviation should be accurate. For purposes of discussion, the Court makes this assumption and treats all of plaintiffs’ statistics about non-whites as reasonably accurate statistics about blacks.

a. Review of Disparities in Other Cases.

To place the plaintiffs’ claim of substantial underrepresentation in context, the Court has reviewed the statistics in a large number of cases, including cases found through the Court’s research as well as cases cited by plaintiffs and amici, that involve claims (based on the Sixth Amendment, the Jury Selection Act or both) of substantial underrepresentation. A striking fact about this review is that not a single court, state or federal, has concluded that a substantial underrepresentation existed where statistics such as those in this case have been mustered. Indeed, those cases include several where the disparities were greater in all respects than the disparities alleged here.

In United States v. Goff 509 F.2d 825, 826 (5th Cir.), cert. denied, 423 U.S. 857, 96 S.Ct. 109, 46 L.Ed.2d 83 (1975), the jury selection plan for the Eastern District of Louisiana was upheld against a challenge based on the Act. The District relied on voter registration lists as its source of jurors. The challengers asserted that blacks were substantially underrepresented on the lists. Blacks constituted 26.33% of the voting age population in the District, but only 21.06% of the registered voters there. Accordingly, the absolute disparity was 5.27%, and the comparative disparity was -20.02%. The Fifth Circuit in Goff nevertheless affirmed the district court’s conclusion that the voter registration lists did not illegally underrepresent blacks.

In U.S. v. Facchiano, 500 F.Supp. 896, (S.D.Fla.1980), the jury selection plan for the Southern District of Florida (Fort Lauderdale Division) was challenged. The plan relied exclusively on voter registration lists as the source of jurors. Blacks constituted 11.3% of the Division’s population, but only 3.7% of the sample from the master wheel. (The case report does not indicate the percentage of blacks among registered voters.) 500 F.Supp. at 898. The resulting absolute disparity was 7.6%, and the comparative disparity was -67.3%. Yet, the Court held that such underrepresentation was not so substantial as to be unfair, for Sixth Amendment purposes. Id. at 899. The court did go on to suggest, however, that disparities of this magnitude might in some circumstances be enough to trigger the duty, under the Act, of supplementing voter lists as juror sources. Id. at 903.

In United States v. Test, supra, 550 F.2d 577, the Colorado jury selection plan was challenged. As in Goff and Facchiano, the challenge included a claim, based on the Act, that distinctive groups (“Chicanos” in Test, as well as blacks) were unfairly underrepresented in voter registration lists. 550 F.2d at 581-82. Those lists were the “primary” source of jurors. Id. at 582. In the Grand Junction Division, 8.89% of the population was Chicano, while only 4.81% of those who returned questionnaires were Chicanos. (The case report does not provide the percentage of Chicanos on the voter registration lists.) Id. at 583. The resulting absolute disparity was 4.08%. The comparative disparity was -46%. Id. at 589 (chart). The Court of appeals for the Tenth Circuit, en banc, rejected the argument that these figures established any substantial departure from a fair cross section of the community. Id. at 589.

Table C, below, compares the statistics in Goff, Facchiano and Test with the statistics in this case.

TABLE C:

Comparison of disparities in Waller with disparities in other fair-cross-section cases where no substantial underrepresentation was found.

Case Waller:

(plaintiffs’ statistics)

Goff:

Facchiano:

Test:

Distinctive Group in Population (%)

19.87

26.33

11.3

8.89

Distinctive Group Registered to Voters (%)

15.88

13.60»

21.06

3.7 b

4.08 c

Absolute Disparity (%)

-3.99

-6.27

-5.27

-7.6

-4.81

Comparative Disparity (%)

-20.07

-31.55

-20.02

-67.3

-46

a This percentage expresses the proportion of the distinctive group in the qualified juror wheel, rather than the proportion registered to vote.

b This percentage expresses the proportion of the distinctive group in the sample of the master wheel, rather than the proportion registered to vote.

cThis percentage expresses the proportion of the distinctive group returning questionnaires, rather than the proportion registered to vote.

As Table C shows, the disparities involved in this case are on the same order as the disparities in the three cases discussed above. Indeed, even plaintiffs’ statistical measures of the disparities between nonwhites in the population and non-whites on the voter registration lists show less disparity in this case than in any of the three other cases, except that the comparative disparity here is a shade higher than that in Goff. Even upon considering plaintiffs’ level of analysis that has the greatest disparity here (comparing the percentage of non-whites in the voting-age population with the percentage of non-whites on the qualified jury wheel), the resulting absolute and comparative disparities are within the range set by these cases.

To further place plaintiffs’ statistical evidence in context, the Court has considered cases where the fair- cross-section requirement in the juror pool has been held violated. In Duren v. Missouri, 439 U.S. at 362, 99 S.Ct. at 667, the Supreme Court found the fair-cross-section requirement of the Sixth Amendment violated where 26.7% of persons summoned — and only 14.5% of persons on the postsummons weekly venires— were women, even though women constituted 54% of the adult population. The absolute disparity is 27.3%, and the comparative disparity is -51%, computed at the level of persons summoned. (Of course, these disparities are even greater when computed for the percentage of women on the postsummons weekly venires).

In a more recent federal case, the Court of Appeals for the First Circuit found that where a group constituted 38.4% of the population but only 13.5% of the grand juries and 10.8% of the petit jury venires, the fair- cross-section requirement was also violated. LaRoche v. Perrin, 718 F.2d 500, 502-04 (1st Cir.1983). Thus, the case presented an absolute disparity of 24.9% and a comparative disparity of 64.8% (computed at the level of grand jury venires).

Finally, in People v. Harris, 36 Cal.3d 36, 201 Cal.Rptr. 782, 679 P.2d 433 (Cal.1984), the California Supreme Court found the fair- cross-section requirement of both the United States and California Constitutions violated by the jury selection procedure of Los Angeles County. The County’s procedure used only voter registration lists as the juror source. The black population of the County was 12.6% and the Hispanic population was 27.6% according to the 1980 census. Id. 201 Cal.Rptr. at 787, 679 P.2d at 438. But of the potential jurors appearing at the courthouses over a three-and-one-half month period of 1979, only 5.5% were black and only 3.4% were Hispanic. Id. This meant, for blacks, an absolute disparity of 7.1% and a comparative disparity of 56.3%; for Hispanics, the absolute disparity was 24.2% and the comparative disparity was 87.7%.

This review of the disparities in cases where the fair-cross-section requirement has been violated suggests that the disparities presented here are on a different order altogether. Table D, below, summarizes the statistics to facilitate their cornparison.

TABLE D:

Comparison of disparities in Waller with disparities in cases where substantial underrepresentation has been found.

Case Waller:

(plaintiffs’ statistics)

Duren:

LaRoche:

Harris:

Distinctive Group in Population (%)

19.87

54

38.4

27.6 a

12.6 b

Distinctive Group on Voter Registration Lists (%)

15.88

13.60 c

26.7 d

13.5 »

3.4 f

5.5 f

Absolute Disparity (%)

-3.99

-6.27

-27.3

-24.9

-24.2

-7.1

Comparative Disparity (%)

-20.07

-31.55

-51

-64.3

-56.3

-87.7

a Hispanics

b Black

c Proportion of distinctive group on qualified jury wheel

d Proportion of distinctive group summoned

c Proportion of distinctive group on grand juries

f Proportion of distinctive group appearing at courthouse as potential jurors.

A quick review of Table D shows that in each of the three cases — including Harris, which the Black Caucus Brief calls to the Court’s attention — the comparative disparities involved were nearly double those involved here; and in some instances more. Further, the absolute disparities were almost always triple those involved here, except for the blacks in Harris. Plaintiffs have not provided, and the Court has been unable tb find, any cases where disparities such as those presented to the Court here have been held to constitute “substantial underrepresentation,” much less a violation of the fair-cross-section requirement.

Case

Waller (plaintiffs' statistics)

Stephens

Quadra:

—Non-white Ethnic

Minorities:

—Women

Sanford

Group in Population(%)

38.8 52.4 25.2 19

Group in Selection Process(%)

15.88b 13.60* 23.95* 23.60* 11.7r 7.248

Absolute Disparity(%)

-3.99

-6.33

-14.85

-28.90

-13.3

-11.86

Comparative Disparity(%)

-20.07

-31.55

-38.27

-54.96

-53.2

-63

a This was the lowest estimated proportion of blacks on the voting lists, not in the population. The Sanford court focuses on the disparity between the proportion of blacks on voting lists and the proportion of jury venires, not the disparity between blacks in the population at large and jury venires.

b Percentage of blacks on voter registration lists.

c Percentage of blacks on qualified jury wheel.

d Average percentage of blacks in six venires in year in question.

e Average percentages of these groups, nominees to grand jury panels, over six-year period.

f Average proportion of blacks on jury venires over four-year period.

g Highest proportion of blacks on jury venire lists over a several-year period.

b Black

c Proportion of distinctive group on qualified jury wheel

d Proportion of distinctive group summoned

e Proportion of distinctive group on grand juries

f Proportion of distinctive group appearing at courthouse as potential jurors.

b. Standards for Determining Under-representation.

(i) Absolute disparity. Some courts have focused solely on the absolute disparity involved, insisting on an absolute disparity of at least 10% to establish a prima facie case of substantial underrepresentation. See, e.g., United States v. Butler, 611 F.2d 1066, 1070 (5th Cir.) cert. denied sub nom. Fazio v. U.S., 449 U.S. 830, 101 S.Ct. 97, 66 L.Ed.2d 35 (1980); United States v. Maskeny, 609 F.2d 183, 190 (5th Cir.), cert. denied, 447 U.S. 921, 100 S.Ct. 3010, 65 L.Ed.2d 1112 (1980); U.S. v. Test, supra, 550 F.2d at 587. The standard of 10% derives from Swain v. Alabama, 380 U.S. 202, 85 S.Ct. 824, 13 L.Ed.2d 759 (1965). Swain was an equal protection challenge based on the alleged abridgement of blacks’ rights to participate in the justice system as jurors. The petitioners had to prove purposeful discrimination against blacks in their participation as jurors. The evidence included a showing that, while the population of black males over 21 was about 26%, only 10% to 15% of the grand and petit jury panels over time had been black. The Supreme Court held that “we cannot say that purposeful discrimination based on race alone is satisfactorily proved by showing that an identifiable group in a community is underrepresented by as much as 10%.” Id. at 208-09, 85 S.Ct. at 829-30. Thus, in Swain the issue was whether an absolute disparity of 10% established purposeful discrimination — not whether, as here, an unfair cross section was used. A fair-cross-section claim does not require a showing of discriminatory intent. Several courts have acknowledged this distinction, but have applied the 10% absolute-disparity standard anyway. See, e.g., United States v. Maskeny, supra, 609 F.2d at 190. The panel in Maskeny defended its use of the 10% absolute disparity figure from Swain as its benchmark, by referring to Duren: the panel explained that in Duren, which established that discriminatory intent is unnecessary in a fair cross section claim, the Supreme Court did not indicate that the disparity needed to show substantial under-representation for Sixth Amendment purposes was any different from that needed to show discriminatory intent for equal protection claims, even though the Court discussed both equal protection and Sixth Amendment violations. 609 F.2d at 190. This line of reasoning does not persuade this Court, however. Whether a fair cross section exists is entirely different from whether intentional discrimination occurred. Further, adopting the 10% absolute disparity rule would have the effect of legitimizing the total or near exclusion from the jury pool of groups constituting 10% or less of the voter-eligible population. This Court does not adopt the 10% absolute-disparity rule in concluding that no substantial underrepresentation of blacks has been shown in the Middle District’s plan.

(ii) Standard Deviation Analysis (Statistical Significance). Plaintiffs and one set of amici insist that the Court of Appeals for the Fourth Circuit has established a standard deviation analysis (also known as a statistical significance test) for determining whether underrepresentation in jur- or sources is substantial. They rely on Moultrie v. Martin, 690 F.2d 1078 (4th Cir.1982). The case involved a habeas corpus challenge to a state-imposed prison sentence. The petition was based on an alleged equal protection violation, in that blacks were underrepresented on the grand jury that indicted the petitioner. The petitioner’s evidence consisted entirely of statistics comparing the proportion of blacks in the county’s population with the proportion of blacks actually on county grand juries. 690 F.2d at 1081-82. The comparisons showed “underrepresentations between 8% and 41% over the 1971-77 period, with an average of 22%.” Id. at 1082. The Court refused to consider these percentages as evidence of disparity, however. That refusal was premised on a careful consideration of statistical principles governing hypothesis-testing. One important principle that the Court focused on is that “the precision and dependability of statistics is directly related to the ■ size of the sample beign evaluated.” Id. at 1083. Applying a standard deviation analysis, the Moultrie court rejected the statistical showing as any indication of underrepresentation. The samples in which the alleged underrepresentation appeared were so small that they could easily reflect random deviations in every year except 1971, according to commonly applied statistical standards. The appropriate statistical standard, the Moultrie court indicated, is 2 or 3 standard deviations. Id. at 1084.

Plaintiffs and one of the amici have focused on the “2 or 3 standard deviations” as the critical figure for the Fourth Circuit in underrepresentation claims. They point out that standard deviations here range from 92.49 (blacks on voting lists compared to blacks in age-eligible population) to 7.45 (qualified black jurors compared to blacks in age-eligible population). Because these numbers are well above the 2-3 standard deviations discussed in Moultrie, they conclude that they have made out a prima facie case of substantial underrepresentation. But this is not the lesson of Moultrie. Although the court in Moultrie did not pursue its statistical analysis of the alleged underrepresentation beyond the stage of considering the standard deviations involved, comparing the standard deviations is only the starting point of a proper analysis of a claim of substantial under-representation. A high standard deviation for an “underrepresentative” distribution of a given sample indicates only that the chances of such an underrepresentative distribution, in a sample of that size, being randomly selected are very small. It is not a useful indicator of the magnitude of the underrepresentation, however.

As one law review article has pointed out, standard deviation varies considerably with sample size. See Kairys, Kadane, and Lehoczky, Jury Representativeness: A Mandate for Multiple Source Lists, 65 Cal.L.Rev. 776, 794, 794 n. 101 & 102 (1977). Even small underrepresentations can be statistically significant if the sample size is large enough. This phenomenon can be dramatically illustrated by considering what the standard deviation would be in this case for the non-whites among the registered voters (compared to non-whites in the age-eligible population) if 19% of the registered voters were non-white. Recall that 19.87% of the age-eligible population is contended to be non-white. Because the “sample” — all registered voters in the Middle District — is so large (856,126), a standard deviation of 20.17 would exist. This is well above the “two or three standard deviations” that plaintiffs claim establishes a prima facie case of substantial underrepresentativeness. It would appear obvious, however, that if 19.0% of a group is on voter registration list where 19.87% of the group is in the age-eligible population, no unfair underrepresentation exists by anyone’s standards.

Furthermore, even with sample sizes much smaller than the voter registration lists, the standard deviation test still produces unsatisfactory results. For example, the qualified jury wheel here in plaintiffs’ sample included 2,250 people. Even if 16.93% of the people on the Middle District’s qualified jury wheel were non-white, the standard deviation would be 3.5, (computed with respect to 19.87% of the age-eligible non-whites in the Middle District). Using the “two or three standard deviations” that plaintiffs insist is the Fourth Circuit’s measure of “substantial underrepresentation,” such a situation would be “substantially underrepresentative,” even though the absolute disparity is only -2.9% and the comparative disparity is only -14.80%.

It is inconceivable that the Fourth Circuit, in Moultrie, established a test for substantial underrepresentation in jury source lists that is based solely on statistical significance (standard deviation), given that statistical significance is so easily established where samples are of the size of voter registration lists and qualified jury wheels. A more realistic reading of Moultrie is that statistical significance is the first step in analyzing any claims of substantial underrepresentation in juries or juror source lists. Unless statistical significance is established, it is unnecessary for a court to consider percentage disparities in the source lists or jury venire. This is because until statistical significance is shown, one cannot be reasonably certain that any such disparities are not due to random variations. Requiring statistical significance as a threshhold matter, undoubtedly, is especially useful in cases such as Moultrie, where a claim of impermissible underrepresentation is based on disparities in very small samples.

This reading of the proper use of standard deviation analysis is supported by the en banc decision of the Court of Appeals for the Tenth Circuit in U.S. v. Test, supra, 550 F.2d 577. The court there held that:

the mathematical conclusion that the disparity between ... two figures is “statistiealy significant” does not ... require an a priori finding that these deviations are “legally significant.”

Id. at 584. The court went on to examine the actual percentages of underrepresentation of blacks and Chicanos in jury source pools and concluded that the underrepresentations were not impermissible, even though they were “statistically significant.”

In sum, this Court rejects plaintiffs’ argument that the Fourth Circuit, in Moultrie, established a standard deviations test for determining whether underrepresentation demonstrated by disparities in the voter registration lists and jury wheels are in fact substantial.

(iii) Totality of Circumstances. Some cases involve disparities that are so stark, in terms of both absolute and comparative disparities, that substantial underrepresentation is established by simply eye-balling the disparities. Duren was such a case. Women constituted 54% of the community’s population, but only 26.7% of the persons summoned and only 14.5% of the post-summons weekly venires were women. 439 U.S. at 362, 99 S.Ct. at 667. Thus the Supreme Court did not face the problem of having to articulate a test of substantial underrepresentation. In other cases it may be just as easy to conclude without analysis that no substantial underrepresentation exists. For example, if a group constitutes 20% of the population but 19% of the jury pool, few if any would argue that the group is substantially underrepresented in the jury pool. There is a large area of middle ground, however, where the disparities are real and yet their substantiality vel non cannot be determined instinctively.

A statistical guideline would be easy to apply and would quickly resolve the fair cross section question. Congress would be the institution that one would expect to fix such a universal statistical guideline. But Congress chose not to specify any such facile statistical test. Indeed, the legislative history shows that Congress committed the question of determining what underrepresentations are “substantial” to the courts:

The voting list need not perfectly mirror the percentage structures of the community. But any substantial deviations must be corrected by use of supplemental sources. Your committee would leave the definition of substantial to the process of judicial decisions.

S.Rep. No. 891, 90th Cong., 1st Sess. at 17 (1967), reprinted in U.S. v. Test, supra, 550 F.2d at 584. Accordingly, this Court considers a variety of factors in deciding whether the disparities shown here, which are almost entirely derived from disparities in voter registration, are “substantial.”

First, the Court considers comparative disparity. The Court notes that some courts have rejected any consideration of comparative disparity because where small groups are involved, very small absolute disparities produce large comparative disparities. See United States v. Jenkins, 496 F.2d 57, 65 (2d Cir.1974), cert. denied, 420 U.S. 925, 95 S.Ct. 1119, 43 L.Ed.2d 394 (1975). For example, if 5.45% of the population is black but 3.3% of the registered voters are black, a large comparative disparity of -39.45% exists. No such small population is involved here, however.

Plaintiffs show that while 19.87% of the age-eligible population is non-white, only 15.88% of the registered voters in the District are non-white; the resulting comparative disparity is -20.07%. Further, they show that 13.60% of the qualified jury wheel in 1981 was non-white, resulting in comparative disparity of -31.55%. These are not negligible underrepresentations. It means that the likelihood that a non-white who lives in the Middle District will serve as a jury is at least 20.07% less than if the voter registration list fully represented non-whites. One other district judge characterized this underrepresentation as about evenly balanced between “substantial” and “insubstantial,” although he ultimately decided that blacks were not underrepresented. See United States v. McDaniels, supra, 370 F.Supp. at 304.

Furthermore, several commentators have suggested that comparative disparities of this order should be the demarcation of substantial underrepresentation. See Kairys, Kadane, & Lehoczsky, Jury Representativeness, 65 Cal.L.Rev. at 799 (15% comparative disparity); J. Van Dyke, Jury Selection Procedures, 98 (1977); (20% comparative disparity) see also Judge Gewin’s Appendix to Foster v. Sparks, 506 F.2d 805, 818 (5th Cir.1975), (20% comparative disparity). Their policy recommendations have not been unequivocally adopted by any courts of which this Court is aware, however. The closest thing to adoption of this standard appears in U.S. v. Facchiano, supra, 500 F.Supp. at 903 (20% comparative disparity is a “helpful guideline”). Given that the 20.07% comparative disparity between non-whites on voter registration lists and non-whites in the age-eligible population that plaintiffs present here is on the borderline of “substantial underrepresentation” suggested by a number of commentators, the Court has considered other evidence carefully before concluding that substantial underrepresentation does not exist.

The Court notes that plaintiffs’ statistics exaggerate the comparative disparities actually involved here. More recent data, directly about blacks rather than about “non-whités,” has been provided by the Clerk’s Office of the Middle District, as discussed in the “Background” section supra. The comparative disparity between registered black voters and age-eligible blacks in the population is only -18.68%, as compared to -20.07% in plaintiffs’ older computations for non-whites. Further, the comparative disparity between blacks in the qualified jury wheel and age-eligible blacks in the population is -23.29%, as compared to 31.55% in plaintiffs’ older computations for non-whites. These more recent statistics about the black composition of the qualified jury wheel suggest that the comparative disparities at issue here are somewhat smaller than plaintiffs assert.

In any case, comparative disparity is not the sole factor to consider in determining whether the disparities here constitute substantial underrepresentation. A second factor is trends in voter registration, especially where, as here, most of the disparity stems from underrepresentation of blacks on voter registration lists. At least one other court has considered voter registration trends in its determination of whether substantial underrepresentation exists. See U.S. v. McDaniels, supra, 370 F.Supp. at 303, 304. Voter registration trends are certainly not dispositive of the question of substantial underrepresentation. But they do help in determining the issue in cases such as this, where the disparity on its face appears to be near the borderline between substantial and insubstantial. In such borderline cases, if voter registration is increasing among the underrepresented group relative to other groups, and if underrepresentation in the voter registration lists is the main source of underrepresentation in the jury selection process, it makes little sense to order the District to supplement its source list with other lists to correct the disparity.

Although the evidence here on voter registration trends is limited, a case cited by plaintiffs and previously referred to by the Court, does contain some relevant information. In Gingles v. Edminsten, supra, at 361, the court recited data on voter registration trends in a number of North Carolina counties. In the state as a whole, as well as in each of the fifteen counties for which information was provided, the proportion of voting-aged blacks registered to vote increased relative to voting-aged whites between October 1978 and October 1982. Two of those counties, Forsyth and Durham, are in the Middle District. The information about these two counties and the state as a whole is reproduced in Table E, below.

TABLE E

Voter Registration Trends in the State of North Carolina and Two Counties in the Middle District

10/78 10/82 White(%) Black(%) White(%) Black(%)

State 61.7 43.7 66.7 52.7

Forsyth 65.8 58.7 69.4 64.1

Durham 63.0 39.4 66.0 52.9

The trend of increasing voter registration among blacks and diminishing disparities between black and white voter registration here contrasts sharply with the trend in People v. Harris, supra, 36 Cal.3d 36, 201 Cal.Rptr. 782, 679 P.2d 433, cited by amici in the Black Caucus brief. In Harris, the Supreme Court of California found that a jury selection plan relying solely on voter registration lists violated the fair-cross-section standard of the Sixth Amendment, (based on comparative disparities ranging from -56.3% (blacks) to -87.7% (Hispanics)). Besides these dramatic disparities — which are, incidentally, much larger than the disparities in this case — the court noted a nationwide trend of declining voter registration among blacks and Hispanics from 1972 to 1978, Id. at 790, n. 6, not rising voter registration among blacks as exists here.

A third factor bearing on the Court’s conclusion that the underrepresentation of blacks here is not substantial is the weight of the case law. As the Court indicated in its review of cases at the beginning of this discussion of underrepresentation, not a single court has held that disparities such as those here — including even the highest comparative disparity involved here (-31.55%, relating to the proportion of nonwhites on the qualified jury wheel in 1981), which exaggerates the disparity actually at issue in this case, as explained supra — constitute substantial underrepresentation. Rather, courts have found disparities that are greater in all respects to be insubstantial. On those occasions where courts did not find underrepresentation to be substantial, the disparities were of a different magnitude altogether than those here. The Court cannot simply ignore the accumulation of judicial experience as to what constitutes “substantial underrepresentation.”

Finally, the Court considers the extent to which the District’s jury selection plan was reviewed before its adoption. The statute requires that judges of both the District and the Fourth Circuit review the plan before it is implemented. See 28 U.S.C. § 1863(a). While this too is not dispositive, it does suggest that a court should at least have “some sense of conviction” before declaring that a district’s jury selection plan has substantially failed to comply with the fair-cross-section requirement. See U.S. v. McDaniels, supra, 370 F.Supp. at 304. This the Court does not have.

3. Systematic Exclusion

Concluding that there is no substantial underrepresentation here, the Court need not consider the third element of the Duren test for fair-cross-section claims. That element is whether the substantial underrepresentation is due to the “systematic exclusion” of blacks in the jury selection process.

An appropriate order will issue.

Appendix A

UNITED STATES COURT OF APPEALS For the Fourth Circuit JUDICIAL COUNCIL

In the Matter of the Review of the Amendment to the Jury Selection Plan submitted by the United States District Court for the Middle District of North Carolina

Order No. 8

ORDER

The Jury Selection Plan of the United States District Court for the Middle District of North Carolina is hereby amended by the district court order dated the 14th day of February 1983 and shown in a copy of the order attached hereto, and it is so ordered.

For the Council

s/ Harrison L. Winter Chief Judge

IN THE UNITED STATES DISTRICT COURT

FOR THE MIDDLE DISTRICT OF NORTH CAROLINA

PLAN FOR THE RANDOM SELECTION OF JURORS

Pursuant to 28 U.S.C. § 1863(a),

IT IS ORDERED that the first sentence of Section VII.C. (“Service of Summonses”) of this court’s Plan for the Random Selection of Jurors is modified and amended to read as follows:

“Each person drawn for jury service may be served personally, or by registered, certified, or first-class mail addressed to such person at his usual residence or business address.”

This the 14th day of February, 1983.

s/Hirman H. Ward Chief Judge

United States District Court

BEFORE THE REVIEWING PANEL OF THE FOURTH CIRCUIT

In the Matter of the Review of the Jury Selection Plan Submitted by the United States District Court for the Middle District of North Carolina

ORDER

This Reviewing Panel, consisting of the undersigned members of the Judicial Council of the Circuit and the undersigned Chief Judge of the United States District Court for the Middle District of North Carolina, having examined the Jury Selection Plan attached hereto and made a part of this Order, and having ascertained that it complies on all particulars with the Jury Selection and Service Act of 1968, as amended, and the Jury System Improvements Act of 1978, 28 U.S.C. § 1861 et seq., now therefore,

IT IS ORDERED that the said Plan is hereby approved by this Reviewing Panel as of the 29th day of January, 1981.

s/ Clement L. Haynsworth, Jr. Chief Judge, Fourth Circuit

s/ Harrison L. Winter United States Circuit Judge

s/ John D. Butzner, Jr. United States Circuit Judge

s/ Donald Russell United States Circuit Judge

s/ H. Emory Widener United States Circuit Judge

s/ K.K. Hall United States Circuit Judge

s/ James Dickson Phillips United States Circuit Judge

s/ Francis D. Murnaghan, Jr. United States Circuit Judge

s/ James M. Sprouse United States Circuit Judge

s/ Sam J. Ervin, III United States Circuit Judge

s/ Eugene A. Gordon Chief Judge, Middle District of North Carolina

THE UNITED STATES DISTRICT COURT

FOR THE MIDDLE DISTRICT OF NORTH CAROLINA

PLAN FOR THE RANDOM SELECTION OF JURORS

I.

PREAMBLE

Pursuant to the Jury Selection and Service Act of 1968, as amended, and the Jury System Improvements Act of 1978, 28 U.S.C. Section 1861 et seq., (the Act) the following Plan for the Random Selection of Jurors is adopted by this Court superseding the Plan now in effect, subject to approval of this Plan by a reviewing panel of members of the Fourth Judicial Circuit Council and to such rules and regulations as may be adopted from time to time by the Judicial Conference of the United States.

II.

DECLARATION OF POLICY

It is the policy of this Court that all litigants in this Court entitled to trial by jury shall have the right to grand and petit juries selected at random from a fair cross section of the community of the Middle District of North Carolina; and that all citizens shall have the opportunity to be considered for service on grand and petit juries, and shall have an obligation to serve as jurors when summoned for that purpose.

III.

DISCRIMINATION PROHIBITED

No citizen shall be excluded from service as a grand or petit juror on account of race, color, religion, sex, national origin, or economic status.

IV.

MANAGEMENT AND SUPERVISION OF THE JURY SELECTION PROCESS

The Clerk of Court shall manage the jury selection process under the supervision and control of the Honorable Hiram H. Ward, U.S. District Judge, Middle District of North Carolina.

V.

VOTER REGISTRATION LISTS AS SOURCE OF NAMES

The Court finds that the persons whose names appear on the voter registration lists in the twenty-four counties comprising this District (on and after October 1, 1981) represent a fair cross section of the community. The procedures prescribed in this PUn to be followed in selecting names Lorn voter registration lists are designed to ensure the random selection of a fair cross section of the persons residing in the community of this district, and to ensure that each county in the district is substantially proportionately represented in the Master Jury Wheel.

A. Random Selection from Voter Registration List.

The “random” selection of names from the voter registration lists shall be made by choosing a starting name by a purely random method; and selecting each subsequent name systematically at regular intervals throughout the source list(s) as hereinafter described. This procedure ensures that: (a) names chosen will represent all segments of the source list or file from which drawn, (b) the mathematical odds of individual names being picked are substantially equalized, and (c) the possibility of human discretion or choice affecting the selection of any individual’s name is eliminated.

B. The term “voter registration list(s)” as used herein shall mean the official records of persons registered to vote in the most recent national general election, established and maintained by county boards of elections in compliance with Chapter 163 of the General Statutes of North Carolina and any rules and regulations promulgated pursuant thereto by the North Carolina State Board of Elections.

C. The selection of names shall be made by the Clerk, or made under his direction or supervision. The Clerk is authorized to use personnel of county boards of elections and to use automated data processing equipment in counties having such equipment to make the name selections, provided names are selected at random in compliance with the Act and this Plan.

VI.

MASTER JURY WHEEL

A. General Requirements.

(1) The Clerk shall establish and maintain, or cause to be maintained, one Master Jury Wheel for this District, as defined in section 1869(g) of the Act. The Master Jury Wheel shall contain the names of those persons selected at random for prospective jury duty. The physical form of records on which names for the Master Jury Wheel are kept may include such automated data storage devices as punched cards, magnetic tapes, or magnetic disc files.

(2) The minimum number of names to be placed in the Master Jury Wheel shall be at least one-half of one per centum of the total number of persons on the voter registration lists in all the counties within the District. The Court may direct that additional names be placed in the Master Jury Wheel at any time. Any such additional names shall be selected at random from voter registration lists in compliance with the Act and this Plan.

(3) Commencing in January of 1981, and every fourth year, thereafter, the Master Jury Wheel shall be emptied and refilled in the manner set forth in this Plan.

(4) To ensure the random selection of a fair cross section of the persons residing within the District, whose names are to be placed in the Master Jury Wheel, the names appearing on the county-wide active voter registration lists shall be considered in alphabetical order of their appearance on the voter registration lists, or in numerical order of their appearance on the voter registration lists, if they are found to have been numbered in a random manner.

(5) If precinct lists of registered voters are used as a source of names, they shall be considered in numerical order (if the precincts are designated by number in a given county) or in alphabetical order (if the precincts are designated by name) and the total considered to be a county-wide list of registered voters.

B. Selection of Names.

(1) The Court shall determine the total number of names to be selected at random from voter registration lists and placed in the Master Jury Wheel. The Clerk shall select, or cause to be selected as provided herein, the number of names from each county voter registration list to be placed in the Master Jury Wheel, which will give each county substantially the proportionate representation in the Master Jury Wheel that its total number of registered voters bears to the total number of registered voters in the District. The selection may be made either manually or through the use of automated data processing equipment or a combination of both methods.

(2) To secure proportionate, or substantially proportionate, representation in the Master Jury Wheel, the total number of registered voters in the District, as reflected in the latest statistical report published by the North Carolina State Board of Elections, shall be divided into the total number of registered voters in each county within the District. The resulting percentages of the total shall determine the number of names from each county to be placed in the Master Jury Wheel for the District.

(3) When the number of names needed from the voter registration list of each county has been thus determined, a ratio number or quotient shall be determined and used to ensure a random, objective and fairly distributed selection of names from each county. The ratio number shall be determined by dividing the total number of registered voters in the District by the number fixed by the Court to be placed in the Master Jury Wheel. For example, if 10,000 names are to be placed in the Master Jury Wheel and there are 920,000 voters registered, the ratio number or quotient would be 920,000 -h 10,000 or 92. Therefore, every 92nd name on the voter registration list would be selected. If the ratio number selected, for example, is 92, there shall be placed 92 numbers in a container. The Clerk shall publicly draw one of those numbers. The number drawn shall be known as the starting number. If, for example, the starting number drawn is 15 and ratio number is 92, the first name to be placed in the Master Jury Wheel from each county shall be the 15th name appearing on a given county voter registration list; the second name shall be the 107th name; the third shall be the 199th name, and so on thro