State v. Gibbs, No. Cr-93-89935 (Jun. 17, 1998) , 1998 Conn. Super. Ct. 6809 ( 1998 )

[EDITOR'S NOTE: This case is unpublished as indicated by the issuing court.]

MEMORANDUM OF DECISION RE: THIRD AMENDED CHALLENGE TO THE JURY ARRAY The defendant, David Gibbs, was charged by information with capital felony, in violation of General Statutes §53a-54b(8), and two counts of Murder, in violation of General Statutes § 53a-54a(a)¹. Pursuant to Practice Book § 842,² now Practice Book (1998 Rev.) § 42-4, the defendant filed a challenge to the jury array on March 15, 1996. The defendant subsequently filed a second amended challenge to the jury array dated December 17, 1997 and a third amended challenge dated April 27, 1998. The defendant claims that the jury array as constituted violates the jury summoning statutes, federal and CT Page 6810 state constitutional rights, and the right to heightened procedural protection including the right to reliable sentencing proceedings. The court scheduled an evidentiary hearing on the motion for February 3, 1998 and continued jury selection to February 18, 1998. The array hearing begain on February 3, 1998 and continued through February 9, 1998 and on February 25, 1998 through March 10, 1998. The evidence at the hearing included testimony of statistical and other experts and various exhibits.

The evidence also included testimony and exhibits from Statev. Rodriguez, Superior Court, judicial district of Hartford/New Britain at Hartford, Docket No. CR-94-456268, as stipulated by the parties. The stipulated materials included testimony from jury clerks, Richard Gayer, Jury Administrator, voter registrars; and experts.³ (Court's Ex. 2.) The parties also stipulated to certain facts concerning the procedure for selecting and summoning potential jurors for jury service. (Courts, EX. 3.)

I

(A)THE SELECTION PROCESS

(1)The Connecticut Jury Selection System⁴ In Connecticut, the jury administrator prepares a master list from a larger pool of names called the source list annually for each judicial district. The voter list and motor vehicle list comprise the source list in effect at the pertinent time.⁵ The master list is generated from town voter's registries and the state motor vehicle list of licensed drivers, minus the duplicates. (Court's Ex. 3, ¶¶ 1-7.)

Based on the population of each town in the judicial district, the jury administrator advises town jury committees of the number of prospective jurors needed from the town. By January 1, the jury committees randomly select the required number of jurors from the list of registered voters in that town. The jury committee list is then compared to the list of licensed drivers in the state provided by the Department of Motor Vehicles and duplicate names are removed from the jury committee list. (Court's Exhibit 3, ¶¶ 3-6.) Because the master list pulled from the source list based on town population, it is referred to as a "stratified" list.

The jury administrator randomly selects a list of names of CT Page 6811 licensed drivers per town and merges this list with the list of names on the town's jury committee list to from a combined list for the town. The combined lists from each town in the judicial district are merged to form the master list for the district. (Court's Ex. 3, ¶¶ 7-8.) From the master list, names are randomly selected for jury service at courthouses within the judicial district based on anticipated requirements. A summons is sent by first class mail to the selected individuals directing them to appear at a certain courthouse on a certain day. The summons includes a list of statutory exemptions so that a potential Juror can indicate a disqualification; those who do not claim an exemption must appear as summoned, although postponement for one year is available. Individuals who do not appear within a year of summons or postponement date are listed as delinquent. (Courts Ex. 3, ¶¶ 9-13.)

If a summons is returned by the Post Office as undeliverable but contains a forwarding address, the summons is sent to the new address. If there is no forwarding address, no action taken. (Defendant's Ex. S1.) It should be noted that if a person fills out a card giving a forwarding address (State's Ex. 21), the post office will forward first class mail free of charge for one year, and provide the sender with the forwarding address for the next six months. (Testimony of Joseph Gavin, February 27, 1998.) Thus, there is an eighteen month period during which the summons will either be forwarded to the new address by the Post Office or sent to the new address by the jury administrator.

2.Evaluations of the Connecticut System

The parties presented evidence with regard to jury selection systems. The state's expert, George Thomas Munsterman of the Center for Jury Studies at the National Center for State Courts, testified that Connecticut's one year cycle for source lists is typical and generally as good as it gets. The Connecticut system, according to Munsterman, is very good. The levels of undeliverables in the Hartford/New Britain judicial district are not high; 36-40% undeliverables to a particular ethnic group is high but is expected in urban jurisdictions. With regard to a hypothetical bias of 23% in favor of Hispanics on the master list receiving summonses, he stated that he did not favor such a result but that it could be a result of randomness. He noted that it would be very hard to increase the representation of Hispanics. CT Page 6812

Munsterman testified that first class mail is the only delivery method used because registered/certified mail is not more effective and may even decrease minority representation. Enforcement efforts are very rare across jurisdictions. Some states send out a second notice as does Connecticut but follow-up is rare and generally restricted to making an example of small groups of non-responders. Monitoring of source lists is rare but should be a regular practice. Munsterman stated that if a jury administrator were aware of an effect on a certain group, he should do something if there is a remedy, e. g., using the National Change of Address system to get better addresses.⁶ Other possible ways of increasing response include providing day care, increasing pay to jurors, and increasing safety around urban courthouses. Munsterman testified that no jurisdiction uses personal service of summonses by sheriffs to large metropolitan areas because no one wants to commit that level of law enforcement resources.

In the defendant's rebuttal case, the defense subpoenaed Magistrate Judge Joan Margolis, who supervises the jury plan for the United States District Court for the District of Connecticut. Judge Margolis testified that the wheel used under the federal jury plan is compiled every three years.⁷ Judge Margolis testified that the state system is better because the addresses on the master list become less stale over its one year period, versus the three year federal wheel. Judge Margolis agreed that the district court has more flexibility than state courts in taking steps to remedy underrepresentation, within the parameters of Congressional legislation.

The district court uses four methods to attempt to ensure representation on the jury panels. First, Judge Margolis monitors the panels to determine whether there are unusually low proportions of minorities on a panel without explanation.⁸ If so, she will ask the clerk to randomly select a supplemental panel. If the supplemental panel would increase representation of minorities, the supplemental and original panels are combined. If not, no action is taken.

Second, the district sends a second, follow-up questionnaire to those who have not responded to the initial questionnaire. Judge Margolis testified that although response to the second letter is small, it does increase minority representation.

Third, the district once sent a remedial mailing to certain CT Page 6813 areas where the percent of undeliverables, non-response, and disqualification was disproportionately high. These areas were sent more questionnaires to increase participation. This method was used only once and was not particularly helpful.

Fourth, the district recently began using the National Change of Address system to check addresses for undeliverable questionnaires. Judge Margolis testified that this effort was too new to determine whether it would be effective in increasing minority representation.

Like the state system, the federal jury plan in the district court disqualifies potential jurors if they do not speak English. These is no provision of interpreters for non-English speaking jurors nor is it planned for the future. Judge Margolis testified that there are four requirements which decrease Hispanic representation on the qualified wheel: potential jurors must (1) speak English; (2) be U.S. citizens; (3 respond to summonses; and (4) provide correct addresses.

(B)HARTFORD/NEW BRITAIN JUROR QUESTIONNAIRES

Beginning in March of 1996, the courthouses in the Hartford/New Britian judicial district were ordered by the court, Norko, J., to distribute a supplemental questionnaire to potential jurors who report for jury duty. The supplemental questionnaire was designed to elicit information regarding the race and ethnicity of the reporting jurors. (Defendant's Ex. Gl.) The supplemental questionnaires do not yield an exact picture of the race/ethnicity of reporting jurors. Not all reporting jurors filled out the questionnaire or answered all the questions. In addition, the presiding judge of the Manchester courthouse ordered the clerks to refuse to distribute the supplemental questionnaires. (Defendant's Exs. F, G1.) Supplemental questionnaires were distributed to potential jurors in at least one courthouse for only a limited period of time, ending in September 1996. (Defendant's Ex. G1.) These differences in distribution of questionnaires lead to distortion, i.e. the "courthouse effect," because not all courthouses distributed the questionnaires equally.

In the Hartford and New Britain courthouses, however, the supplemental questionnaires were distributed since March, 1996. Although incomplete, the questionnaire date was relied on by the defendant to verify the accuracy of other estimates made based on CT Page 6814 Judicial Information Systems (JIS) data. (Defendant's Ex. X20.)

The supplemental questionnaires were also used as the numerical basis for the defendant's statistical calculations attempting to show underrepresentation. (Defendant's Exs. X21, X22.)

(C)EVIDENCE OF UNDELIVERABLES AND NO-SHOWS

The defendant introduced numerous exhibits through its expert, David Pollard, Processor of Statistics and Mathematics at Yale University. Pollard examined JIS data and tabulated the results of the summonses sent in the Hartford/New Britain judicial district, i.e., returned with claimed disqualifications, returned as undeliverable ("undeliverables"),⁹ returned by jurors who did not report for jury duty ("no-shows")¹⁰ and returned by jurors who were either cancelled or who did report for service ("OK"). The data showed a higher percentage of undeliverables and no-shows for Hartford and New Britain as compared to other towns in the judicial district over the 1992/1993 to 1995/1996 court years.¹¹ (Defendant's Ex. X5.) The data also showed an increase in the number of undeliverable summonses as the court year progressed, which Pollard attributed to addresses becoming "stale" over time. (Defendant's Ex. X4.)

In order to estimate to whom the undeliverable summonses were sent (e.g. to Hispanics, non-Hispanics), Pollard used the geocoding method with Census tract estimates of ethnicity. A tract is a small, relatively homogenous subdivision of a town as determined by the Census Bureau. In geocoding, the summonses are targeted to the tracts by locating the street addresses on the tracts, using Census tables. A tract is estimated by the Census Bureau to consist of a certain percentage of Hispanics based on the 1990 Census. Pollard used this proportion of Hispanics in the tract to determine the ethnicity of those receiving the summonses in that tract. The data showed that many of the tracts which had high percentages of undeliverables were also tracts which contained high percentages of Hispanics. (Defendant's Ex. X18.)

The defendant's expert also used the surname matching method to determine percentage estimates of undeliverables to Hispanics. Surname matching¹² was applied to the JIS data that had been used in the geocoding method and to all the JIS records. (Defendant's Ex. X19.) Under either the surname method as applied to geocoded JIS records or the geocoding method, the data shows CT Page 6815 that of all the summonses for the court years 1992/1993 through 1995/1996, a higher percentage of undeliverables was sent to Hispanics as compared to non-Hispanics or to all. (Defendant's Ex. X19.) For example, in 1992/1993, approximately 29% of summonses sent to Hispanics were undeliverable, compared to 12% of all summonses sent. (Defendant's Ex. X19.) The gap between Hispanics and all, and Hispanics and non-Hispanics, remains relatively constant over the 1993/1994, 1994/1995 and 1995/1996 court years. (Defendant's Ex. X19.) Similarly, a higher percentage of summonses sent to Hispanics end up in the no-show category.¹³ (Defendant's Ex. X19.)¹⁴

With regard to undeliverables, Pollard stated that some undeliverables would undoubtedly go to people who would have been ineligible to serve as jurors. Pollard indicated that factors other than race/ethnicity, such as poverty, showed a similarity with undeliverables but noted that the similarity could not be said to be causal.¹⁵

Based on all the data, the defendant's expert concluded that undeliverables have a larger effect on Hispanics over time throughout the judicial district. After elimination for undeliverables, approximately 70% of Hispanics, versus approximately 87% of non-Hispanics or of all, remain available to move through the system. (Defendant's Ex. X20.) Pollard concluded that the staleness of addresses is the one certain cause of undeliverables. The no-shows also affect Hispanics to a greater extent (10-12% Hispanic versus 5% of all). (Defendant's Ex. X20.) Pollard also concluded that he believed that he would continue to see the same effects over time provided that the system did not change.

Dr. Stephan Michelson, the state's expert criticized Pollard's calculations for rate of loss of Hispanics due to various factors. According to Michelson, the defendant's expert measured the rate of loss of Hispanics from the jury selection system incorrectly because he compared the loss of Hispanics against all mailed summons, rather than accounting for the people lost at each stage of the process. (State's Exs. la, 1b.) The rates for undeliverables to Hispanics and non-Hispanics are the same in both Michelson's and Pollard's calculations for 1994-95, using surname coding only. According to Michelson, the rate of loss of Hispanics due to being out of the jurisdiction, lack of citizenship, language and other disqualifications is actually higher than in Pollard's calculation. For instance, Pollard CT Page 6816 determined that approximately 10% of Hispanics are lost due to the no-show category, as compared to 4% of non-Hispanics. Michelson, by contrast, calculated that 28% of Hispanics are no-shows, based on the number of summonses sent out minus people lost due to earlier disqualifications, and that 7.8% of non-Hispanics are no-shows. (State's Exs. 1a, lb.)

II.

(A)THE DEFENDANT'S CASE

The defendant asserts the following grounds for his challenge to the jury array: violation of the jury summoning statutes; federal and state constitutional rights to due process, equal protection, and a jury summoned from a fair cross section of the community; and the right to heightened procedural protection in a capital felony prosecution, including the right to reliable sentencing proceedings. (Defendant's Third Amended Challenge to Jury Array.) The defendant focuses on three factors relating to the representation of Hispanics: (1) English ability requirement; (2) undeliverables; and (3) no-shows.

With regard to the exclusion of non-English speakers, the defendant argues that it violates (1) the defendant's and jurors' state and federal constitutional rights to equal protection and (2) the state constitutional anti-discrimination right.

The defendant asserts that both non-deliverables and no-shows disproportionately affect Hispanics such that Hispanics are underrepresented on the jury array.¹⁶ The defendant claims that this violates (1) the state and federal constitutional rights to equal protection; (2) the federal sixth amendment right to a jury drawn from a fair cross section of the community and (3) the state constitutional right to due process.¹⁷

(B)DUE PROCESS AND THE SIXTH AMENDMENT

(1)Sixth Amendment Right

Although the United States Supreme Court has held that juries must be drawn from a fair cross section of the community, there is "no requirement that petit juries actually chosen must mirror the community and reflect the various distinctive groups in the population. Defendants are not entitled to a jury of any particular composition, . . . but the jury wheels, pools of CT Page 6817 names, any panels or venires from which juries are drawn must not systematically exclude distinctive groups in the community and thereby fail to be reasonably representative thereof." (Citations omitted; emphasis omitted.) State v. McDougal, 241 Conn. 502,516, 699 A.2d 872 (1997).

In order to put forth a prima facie case under the sixth amendment, the defendant must show: (1) that the group alleged to be excluded is a "distinctive" group in the community; (2) that the group's representation in the source 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 underrepresentation results from systematic exclusion of the group in the jury selection process. Duren v. Missouri,439 U.S. 357, 364, 99 S.Ct. 664, 58 L.Ed.2d 579 (1979). Once the defendant has shown underrepresentation of a distinctive group due to systematic exclusion, the burden shifts to the state to show that a significant state interest is manifestly and primarily advanced by those aspects of the jury selection process that result in the disproportionate exclusion of a distinctive group. Id., 368; State v. Webb, 238 Conn. 389, 450, 680 A.2d 147 (1996). Whether the defendant has met the three requirements under Duren will be analyzed below in conjunction with the test for an equal protection violation.

(2)Due Process Riqhts

"A fair trial in a fair tribunal is a basic requirement of due process." Peters v. Kiff, 407 U.S. 493, 501, 92 S.Ct. 2163,33 L.Ed.2d 83 (1972). "[A]ny attempt to stack a jury panel by intentionally including or excluding any members of a discernable class runs afoul of both due process and the right to a jury trial." (Internal quotation marks omitted.) State v. Webb, supra, 238 Conn. 448, citing State v. Nims, 180 Conn. 589,595-96, 430 A.2d 1306 (1980). A due process challenge to the grand jury is "analogous to a sixth amendment claim that a petit jury was not `impartial' because it was not drawn from fair cross section of the community. . . . Accordingly, in its analysis, the trial court was guided by Duren v. Missouri, supra, which enunciated a test for analyzing a fair cross section challenge." (Citations omitted.) State v. Castonquay, 194 Conn. 416, 420,481 A.2d 56 (1984). See also State v. Tillman, 220 Conn. 487, 495 n. 3, 600 A.2d 738 (1991), cert. denied,' 505 U.S. 1207,112 S.Ct. 3000, 120 L.Ed.2d 876 (1992) ("the United States Supreme Court has since construed Theil's fair cross-section requirement CT Page 6818 to be demanded by the sixth amendment and by the due process clause of the fourteenth amendment"); State v. McCarthy,197 Conn. 247, 248, 496 A.2d 513 (1985) (applying Duren test to fair cross section challenge of jury array based on due process);State v. Couture, 194 Conn. 530, 549-551, 482 A.2d 300 (1984) (applying Duren test to due process challenge to composition of grand jury).

Nevertheless, a jury selection process that does not violate the Duren standard may violate the constitution if unconstitutional criteria are used in selecting its members underState v. Nims, 180 Conn. 589, 430 A.2d 1306 (1980). State v.Tillman, 220 Conn. 487, 493, 600 A.2d 738 (1991). In light of the state's evidence regarding overrepresentation of Hispanics on the master list, the defendant has amended his challenge to assert a State v. Nims violation. The overrepresentation of Hispanics on the master list, according to the state, results from the sampling process of the selection system which is stratified by town. Because Hispanics tend to live in towns with populations that are not reflected on the source list and tend to be younger than non-Hispanics, adult Hispanics on the source list are more likely to be on the master list and thus to receive a summons.

The state's expert, Michelson, testisfied that rather than being underrepresented in the jury system, Hispanics are overrepresented on the master list. First, the use of population to allocate slots by town affects Hispanics because they are less likely to live in towns where the composition of the source list mirrors the total adult population. Second, the allocation by population also affects Hispanics because total population figures are used and Hispanics comprise a younger population than non-Hispanics.¹⁸ Michelson stated that he had derived different estimates of bias on the master list. One estimate was 23% in favor of Hispanics. (State's Ex. 33, p. 35.) Michelson also determined that there was a bias in favor of Hispanics based on the data from the voter lists of 17% and of 35% bias on the DMV list. (State's Ex. 1, p. 47) The defense's expert, Pollard, disagreed with Michelson's figure of 25% bias on the master list in favor of Hispanics. Pollard asserted that the figure was based on bad interpretations of regressions. Pollard also contended that Michelson's derivation of the source list figure by "reverse engineering" from the master list yields a nonsense figure.

Regardless of the exact measure of overrepresention on the CT Page 6819 master list, if any, the defendant has not shown a violation of due process in line with State v. Nims. Although a jury selection process may violate the constitution if unconstitutional criteria are used in selecting its members; State v. Tillman, supra,220 Conn. 493; the criterion of town population as used in the jury summoning statutes is not unconstitutional.

In State v. Nims, 180 Conn. 589, 430 A.2d 1306 (1980), the court found that the jury clerk's separation of juror cards by gender "was a conscious effort by the jury clerk to take a person's sex into account in the selection of the jury panel." Id., 593-94. The court concluded that "any attempt to stack a jury panel by intentionally including or excluding any members of a discernable class runs afoul of both due process and the right to a jury trial." Id., 595-96. "The end result . . . is irrelevant. It is the method of selection which offends the constitution." (Internal quotation marks omitted.) State v.Tillman, supra, 220 Conn. 487, 493.

"Under Nims, even if there is no underrepresentation of any group, it is possible for the jury to be selected according to unconstitutional criteria. The sex-based classification in question in Nims had, however, been previously held by the United States Supreme Court to constitute a distinctive group both in the specific context of jury selection . . . and as a suspect classification that must be subjected to heightened judicial scrutiny. . . ." (Citations omitted.) Id., 499-500. By contrast, town population has not been held to be a suspect or unconstitutional criterion. The deliberate separation of juror cards based on gender is far more constitutionally suspect than the stratification based on town population used here. Moreover, the town criterion by itself does not affect the proportion of Hispanics on the master list; rather, it is the way in which the Hispanic population is distributed across towns and age that might result in overrepresentation. This supports the determination that the criterion used here is not unconstitutional. Thus, any overrepresentation of Hispanics on the master list is not due to the use of unconstitutional criteria or the deliberate inclusion of a group and therefore does not constitute a violation of due process.

(C)EQUAL PROTECTION

The United States Supreme Court has "recognized it is a denial of the equal protection of the laws to try a defendant of CT Page 6820 a particular race or color under an indictment issue by a grand jury . . . from which all persons of his race or color have, solely because of that race or color, been excluded by the state . . . While the earlier cases involved absolute exclusion of an identifiable group, later cases established the principle that substantial underrepresentation of the group constitutes a constitutional violation as well, if it results from purposeful discrimination." Castaneda v. Partida, 430 U.S. 482, 493,97 S.Ct. 1272, 51 L.Ed.2d 498 (1977). The fourteenth amendment applies to the venire stage of jury selection as well as to grand juries. See State v. Ellis, 232 Conn. 691, 700 n. 12,657 A.2d 1099 (1995) (method for challenging grand jury array articulated in State v. Castonguay is equally applicable to petit jury selection process); State v. Couture, 218 Conn. 309, 315,589 A.2d 343, (1991) ("The fourteenth amendment has long been held to forbid unequal treatment at the venire stage of jury selection").

To state a prima facie case of violation of equal protection, the defendant must show: (1) cognizable class; (2) degree of underrepresentation of the identifiable group over a significant period of time; and (3) selection procedure susceptible of abuse or not racially neutral, supporting the presumption of discrimination raised by the statistical showing. The burden then shifts to the state to dispel the inference of intentional discrimination. Castaneda v. Partida, supra, 430 U.S. 494. The state can refute the existence of an equal protection violation by showing that race neutral selection criteria have produced the result. State v. Wright, 207 Conn. 276, 282, 542 A.2d 299 (1988).

Under Castaneda, the defendant must show that he is a member of the cognizable group being underrepresented. Castaneda v.Partida, supra, 430 U.S. 494. The defendant asserts, and the states has conceded, that the defendant, a black Jamaican, has standing to challenge the underrepresentation of Hispanics on the jury array. In Powers v. Ohio, the United States Supreme Court held that a white defendant may object to the prosecution's peremptory challenges of black venirepersons. Powers v. Ohio,499 U.S. 400, 111 S.Ct. 1344, 113 L.Ed.2d 411 (1991). In Campbellv. Louisiana, 118 Supreme Court, 1419 (1998), the court held that a white defendant has standing to raise an equal protection objection to discrimination against black persons in the selection of grand jurors. Id.¹⁹ Accordingly, the defendant here has standing to raise an equal protection claim based on the exclusion of Hispanics. Whether the defendant has shown the elements under the Castaneda test will be discussed below in CT Page 6821 conjunction with the Duren test.

(1)First Element: Distinctive Group/Cognizable Group

"With respect to the first requirement of Duren, we have noted on several occasions . . . that Hispanics constitute a `distinctive' group in the community for purposes of this fair cross section claim." State v. McCarthy, supra, 197 Conn. 250. Similarly, Hispanics are an identifiable group for the purpose of equal protection claims. See State v. Gonzalez, 206 Conn. 391,396, 538 A.2d 210 (1988).

The state argues that the defendant is basing his claim not on the exclusion of Hispanics, but on the exclusion of a particular subgroup of Hispanics. The state contends that Michelson showed that the defendant's evidence goes to underrepresentation of a "highly mobile subgroup among the Hispanic population" rather than of Hispanics as a whole. (State's Post-Hearing Memorandum of Law, p. 6.) The state relies on State v. Couture, supra, 218 Conn. 316, which held that "observant Jews" are not a cognizable group, unlike Jews as a whole and State v. Haskins, 188 Conn. 432, 440, 450 A.2d 828 (1982), which held that apartment-dwellers do not constitute a cognizable group. These cases are distinguishable because the defendant in each case was alleging that the subgroup was underrepresented; here, the defendant alleges that Hispanics as a whole were underrepresented and bases his complex data analysis on all Hispanics, not a subgroup thereof. The fact that some subsection of Hispanics may be driving the numbers of Hispanics overall does not transform the defendant's claim into a challenge based on a non-cognizable group.

The state has cited no case law which supports its position that where factors other than race or ethnicity appear to explain underrepresentation as does race or ethnicity, the court should assume that racial and ethnic factors drop out. Assuming Hispanics are underrepresented because they are mobile, the fact remains that Hispanics are missing from the jury array. The reasons for Hispanic underrepresentation go to systematic exclusion/racially suspect process, not to distinctiveness of group. Therefore, this court concludes that Hispanics as a whole are at issue in this challenge and that the defendant has alleged underrepresentation of a distinctive or cognizable group. The first prong of the Duren and Castaneda tests is reached. CT Page 6822

(2)Second Element: Underrepresentation

In order to establish that Hispanics are unconstitutionally excluded from the jury system, the defendant must show that Hispanics are significantly underrepresented. Under Castaneda, "the degree of under representation must be proved, by comparing the proportion of the group in the total population to the proportion called to serve as grand jurors, over a significant period of time. . . ." Internal quotation marks omitted.) Statev. Wright, supra, 207 Conn. 281. Under Duren, the defendant must show that "the representation of [Hispanics] in venires from which juries are selected is not fair and reasonable in relation to the number of such persons in the community." Duran v.Missouri, supra, 439 U.S. 364.

"The jury array need not mirror the sociological composition of the community . . . or be mathematically proportionate to the percentage of the group in the community. . . . Substantial underrepresentation is required to make out a prima facie case." (Citations omitted; internal quotation marks omitted.) State v.Castonguay, supra, 194 Conn. 425-26. The underrepresentation must be "constitutionally significant"; id., 425; or "substantial."Castaneda v. Partida, supra, 430 U.S. 494.

(A)The Relevant Population

In Duren, the court refers to the "actual percentage of women in the community," presumably the total population of females in the county. Duren v. Missouri, supra, 439 U.S. 365. In Taylor v.Louisiana, 419 U.S. 522, 531, 95 S.Ct. 692, 42 L.Ed.2d 690 (1975), the court utilized the stipulated figures of the population eligible for jury service. In Castaneda, the relevant population was the total population of the county. Castaneda v.Partida, supra, 430 U.S. 486. It is therefore unclear what the relevant population should be, according to United States Supreme Court precedent.²⁰

Under Connecticut precedent, the total population figure, rather than the eligible population, is utilized. See State v.Robinson, 227 Conn. 711, 719, 631 A.2d 288 (1993); State v.Couture, supra, 218 Conn. 316; State v. McCarthy, supra,197 Conn. 249; State v. Castonguay, supra, 194, Conn. 424. In State v.Castonguay, the court specifically addressed the relevant population. The court concluded that the trial court was correct in using the "percentage of Hispanics in the general population CT Page 6823 of Hartford County" rather than the eligible population. State v.Castonguay, supra, 194 Conn. 424. The court noted that "in Statev. Haskins, 188 Conn. 432, 440, 450 A.2d 828 (1982), weerroneously utilized figures representing the eligiblepopulation." (Emphasis added.) Id., 425 n. 8.8

Despite the above references to the county population, the Connecticut Supreme Court has held that the relevant community for purposes of selecting venire panels is the population of thejudicial district. In State v. Faust, 237 Conn. 454, 678 A.2d 910 (1996), the court examined the General Statutes and concluded that "`[t]hrough these statutes, the Connecticut legislature has defined the community for purposes of jury selection to be the judicial district instead of the county.'" Id., 467-68, quotingState v. Carolina, 40 Conn. App. 762, 769-70, 673 A.2d 562 (1996). Thus, if the court were to follow these cases, the relevant population would be the total population of the Hartford/New Britain judicial district.²¹

The Connecticut cases utilizing total population as the relevant measure, however, are inapplicable to the unique facts here. Unlike typical challenges to the jury selection system, the challenge here is not based on lack of representation in the initial pool of potential jurors. Rather, the defendant's challenge is based on factors which deter Hispanics from appearing as jurors after the summonses are sent to a randomly selected group of persons from the master list. The defendant argues that the English-language requirements, undeliverables and no-shows disproportionately affect Hispanics. That is, after summonses are mailed based on the master list, the fact of disqualification for lack of English comprehension, the inability to deliver summonses because of stale addresses, and the failure of Hispanics to report for jury duty are the grounds for arguing unconstitutional jury selection. Because the defendant's claim defines the particular population to be examined; United Statesv. Rioux, supra, 930 F. Sup. 1568; the population of Hispanics on the master list is the appropriate pool to which to compare the number of Hispanics on the jury arrays.

The defendant conceded the composition and constitutionality of the source list. The defendant challenges the composition of the master list to the extent that stale addresses on the master list contribute to the undeliverable problem. The defendant is also challenging the master list under State v. Nims foroverrepresenting Hispanics, see discussion, supra. As a result, CT Page 6824 the appropriate population to use as the relevant community is the number of Hispanics on the master list.²² By comparing the number of Hispanics who report for jury duty to the number of Hispanics on the master list, the significance of underrepresentation of Hispanics due to the English requirement, undeliverables and no-shows can be determined.²³

The defendant's expert, Pollard, stated that Hispanics constitute approximately 7% of the people on the master list.²⁴ The state's expert, Michelson, calculated that Hispanics make up approximately 6.73% of the master list. This number can be compared to the number of Hispanics who report for jury duty to assess whether the defendant has shown substantial underrepresentation of Hispanics in the jury selection system.

It is found that Hispanics are 8.32% of the total population according to the 1990 Census and 6.57% of the over-eighteen population. (State's Proposed Findings of fact; Defendant's Ex. X22.) According to the Census Estimates for 1996, Hispanics make up 10.02% of the total population and 7.77% of the over-twenty population of Hartford County. (Defendant's Ex. A.) Even if total population were the appropriate unit of comparison, the Census Estimates are not sufficiently reliable. The Bureau of the Census report itself warns that "we do not consider these data to be accurate for each individual cell" and that "[t]he estimates presented here were produced by a method which is still in a developmental stage. These figures should be used with caution." (Defendant's Ex. A.)

The defendant's assertion that the 1996 Census Estimates be employed as the measuring touchstone in this challenge is understandable but misplaced. The Census Estimates are projections based on samples, found inaccurate by the Bureau supplying them and are in violation of our statutes requiring decennial figures. §§ 51-219c; 51-220, Connecticut General Statutes.

The parties presented two measures of Hispanic jurors who are available to serve on jury venires. The defense's expert, Pollard, calculated that Hispanics constituted 4.21% of those jurors who filled out the supplemental questionnaires. (Defendant's Ex. X22.)²⁵ The state contends that this figure is unreliable and has put forth a figure of 3.8% Hispanic for the measure of qualified jurors. (State's Ex. 37, p. 9.) Although the court entertains some doubts regarding the veracity of the 4.21% CT Page 6825 figure given the difficulties surrounding the distribution of the supplemental questionnaires, the defendant's figure will nevertheless be used. Even assuming that the pool of qualified jurors is 4.21% Hispanic, this pool does not establish significant underepresentation.

In order to assess underrepresentation, the proper method of comparison of the percentage of Hispanics on the master list with the population of appearing jurors must first be determined. The results are then evaluated to determine whether the underrepresentation is constitutionally significant.

(B)Statistical Methods

"[T]he choice of a statistical method depends on the facts and circumstances of each case. An important consideration is whether the claim is an equal protection challenge, where the issue is whether the discrimination was purposeful or the result of chance, or whether the claim is a fair cross section challenge, where the issue is whether the underrepresentation is unfair. Another important factor is the size of the population at issue. Ultimately, however, the decision is not one of numbers but rather a subjective determination of whether the disparity is constitutionally significant." State v. Castonguay, supra,194 Conn. 426-27. There are essentially four statistical models that have been used to analyze jury array challenges: (1) absolute disparity; (2) comparative disparity; (3) substantial impact; and (4) statistical decision theory (SDT). Id., 427-30.

1.Absolute Disparity

Absolute disparity is also known as absolute difference. Absolute disparity measures the difference between the percentage of the group in the relevant population and the percentage in the pool being challenged as unconstitutional. Id., 427. Here, absolute disparity would equal the difference between the percentage of Hispanics on the master list and the percentage of Hispanics in the population of appearing jurors. Absolute disparity, however, is "inadequate when the percentage of persons in the group is small in relation to the entire population. The result obtained is a distortion of reality." Id., 428. Thus, absolute disparity is an inappropriate measure here because Hispanics comprise less than 8% of the master list population.

2.Comparative Disparity CT Page 6826

Comparative disparity or relative disparity measures "the diminished likelihood that members of an underrepresented group, when compared to the population as a whole, will be called for jury service." (Internal quotation marks omitted.) United Statesv. Rioux, supra, 97 F.3d 655. It is calculated by dividing the absolute disparity by the percentage of the group in the relevant population, here, the percentage of Hispanics on the master list. See State v. Castonquay, supra, 194 Conn. 429. Like absolute disparity, comparative disparity "tends to magnify the size of the disparity as the relevant group's percentage of the population decreases. . . . [I]t is ordinarily inappropriate where a very small proportion of the population is [made up of the relevant group]. A comparative characterization in such circumstances distorts reality." Id. As a result, comparative disparity is also an inappropriate measure of underrepresentation here.

3.Substantial Impact

Substantial impact, also known as absolute impact, focuses on "whether the underrepresentation substantially affects the composition of the [jury array]." Id., 430. The Connecticut Supreme Court has settled on "the `substantial impact' test as best suited to a fair cross section claim." State v. McCarthy, supra, 197 Conn. 251.

Under the substantial impact test, the first step is usually determining the number of the group in the venire based on the percentage of the group in the relevant population. The first step would be to determine how many Hispanic jurors would have appeared if the proportion of Hispanics who appeared was the same as the proportion of Hispanics on the master list. Comparing this figure to the number of Hispanics who actually did appear will yield the number of Hispanic appearing jurors needed in order to reach the percentage Hispanic of the master list.

As an example, assume that a distinctive group makes up 10% of the relevant population and 2% of the venire. The absolute disparity, or the difference between expected and actual representation, would be 8 percentage points. If the venire is made of 60 persons, an increase of 8% of 60 would be necessary to reach 10% representation. Thus, four to five people would have to be added to the sixty person venire.²⁶ United States v.Rioux, supra, 97 F.3d 655. CT Page 6827

Here, using 7% as the percent of Hispanics on the master list, the substantial impact is 2.79 or approximately three Hispanic people per a 100 person venire. First, calculate absolute disparity: (7% Hispanics on master list) minus (4.21% Hispanics who appear) = 2.79%. Assuming a 100 juror panel, two to three people would have to be added to the jury panel to reach the proportion of Hispanics on the master list.

4.Statistical Decision Theory (SDT)

"SDT has gradually become the favored method of the United States Supreme Court in analyzing equal protection challenges to grand jury arrays. . . . The import of this method is that it is a highly sophisticated and precise indicator of whether the disparity between the group's representation among grand jurors is the result of a neutral and random selection process." Statev. Castonguay, supra, 194 Conn. 427-28. SDT, therefore, "calculates probabilities and measures the likelihood that underrepresentation could have occurred by sheer chance. Under this method, if one can determine that it is statistically improbable that the jury pool resulted from random selection, then there is an imperfection in the system." United States v.Rioux, supra, 97 F.3d 655. Because SDT is based on random selection, "[i]t is illogical to apply [SDT] when assessing the constitutionality of a qualified wheel.²⁷ By definition, the qualified wheel is not the product of random selection; it entails reasoned disqualifications based on numerous factors. It is irrational to gauge the qualified wheel — an inherently non-random sample — by its potential for randomness." Id.

Similarly, the population of appearing jurors is non-random; disqualifications decrease the population of the pool of prospective jurors. As the defendant's expert has asserted, SDT is "superfluous" here. Indeed, the defendant conceded that SDT does not logically apply here, but felt bound by case law to assert that SDT is the appropriate measure. The state's expert agreed that SDT is inappropriate because it measures the outcome of the process, not the process itself.

The defendant has not calculated the SDT based on the difference between the master list percentage of Hispanics (7%) and the percentage of Hispanics appearing (4.21%). There are calculations using 4.21% Hispanic as compared to 7.77% Hispanic (1996 Census Estimates for over-twenty population) and 6.57% CT Page 6828 Hispanic (1990 Census Estimates for over-twenty population). Using 7.77% and 6.57%, SDT is less than or approximately 10 (Defendants Ex. X22.) Thus, the SDT using the 7% figure from the master list will be similarly minuscule. The likelihood of achieving the result of 4.21% Hispanics appearing and answering yes to the Hispanic question from a random sample pulled from a pool comprised of 7% Hispanics is extremely minute.

The impractical application of SDT to the issues at bar was best articulated by Dr. Pollard in his testimony of February 9, 1998, excerpts of which follow:

Dr. Pollard: "If you were totally determined to do this, to keep drawing until you got 4.21 percent, if you kept doing it over and over and over again, then the theory says eventually you get a number this small. But it would be one in every ten to the sixty-third. So, if you did one every day it would take ten to the sixty-third days. What's that? How many billions and billions of years into the future would it take until this happened? (Emphasis added).

Court: "I'm into the Paleolithic stages."

Dr. Pollard: "Well beyond the age of the universe. I would say." (Emphasis added).

This result is not probative of the significance of underrepresentation, however, because the prospective jurors who appear are not selected randomly; rather, disqualifications and other factors intervene to exclude some of the persons on the master list.

(C)Defendant's Proposed Calculation Because the defendant's expert believes that none of the four tests is measuring the correct thing, `Pollard put forth a test calculating the yield of Hispanic appearing jurors. That is, based on the master list population, he calculated the number of Hispanics and non-Hispanic who fell into the OK category. Assuming a 10,000 person master list, of which 7% is Hispanic, 9300 persons would be non-Hispanic and 700 persons would be Hispanic. Based the percentages of each group who end up in the OK category (see Defendant's Ex. X19), approximately 210 Hispanics and 4650 non-Hispanics fall into the OK category. Defendant's Exhibit X19 indicates that over the court years CT Page 6829 1992/1993 and 1993/1994, approximately 30% of Hispanics fell into the OK category, as compared to approximately 50% of non-Hispanics.²⁸

This yield analysis has no basis in case law. The defendant's argument that it should be adopted by the court under the constitution of Connecticut is without merit, see discussion infra. The comparative analysis of the yield assumes that Hispanics and non-Hispanics are similarly situated with regard to disqualifications and receipt of summonses. The evidence, put forth by the defendant, has shown that Hispanics are less likely to receive summonses and more likely to be subject to disqualification. The fact that fewer Hispanics receive summonses, make it through the disqualification screen, and report for jury duty does not support a finding of significant underrepresentation. The factors which lead the Hispanics to report for jury duty in fewer numbers are not controlled by or influenced by the system and cannot support a finding of substantial underrepresentation. As discussed infra, the undeliverable problem, no show problem and requirement of English ability are either non-systematic factors or legitimate qualifications. Although this determination blurs the line between underrepresentation and systematic exclusion, "the unfair representation and systematic exclusion prongs of the Duren tests are intertwined inextricably." United States v. Rioux, supra,930 F. Sup. 1568. Thus, the "comparative analysis of yield" calculation advocated by the defendant merely serves to highlight the legitimate differences between Hispanic and non-Hispanic participation — based on factors external to the system — and does not prove significant underrepresentation. The court therefore declines to adopt the defendant's method of measuring underrepresentation.

(3)Sufficiency of Underrepresentation

As discussed above, the Connecticut Supreme Court has adopted the substantial impact test to evaluate underrepresentation underDuren. The court has found insufficient underrepresentation where the substantial impact is one Hispanic for every other eighteen-member grand jury. State v. McCarthy, supra, 197 Conn. 252. The court has also rejected a claim of substantive underrepresentation where "slightly more than one additional Hispanic should have been included on every other [eighteen-member] grand jury." State v. Castonguay, supra,194 Conn. 430. In Castoguay, the court concluded that "[c]ommon sense dictates that the impact of this underrepresentation is not CT Page 6830 substantial."Id.

Putting the figures from Castonguay in terms of 100-person venire, 2.78 Hispanics should have been added.²⁹ Here, using 7% as the percentage of Hispanics or the master list, the absolute disparity is 2.79.³⁰ Thus, 2.79 Hispanics would have to be added to a 100-person venire. As the Connecticut Supreme Court has held, this underrepresentation is not constitutionally significant. See also United States v. Rioux, supra, 97 F.3d 658 (two or three additional minorities to pool of 125 is not significant); United States v. Purdy, 946 F. Sup. 1094, 1103 (D. Conn. 1996) (citing United States v. Biaggi, 909 F.2d 662, 678 (2nd Cir. 1990) figures of 3.6 and 4.7 absolute impact for venire of 100 persons as insufficient).³¹ Thus, the defendant has failed to show the second required element under Duren.

As noted above, the SDT calculation shows that the outcome found is extremely unlikely to occur due to sheer chance. The relevance of this finding is undermined by the fact that the selection process is by definition non-random. The SDT calculation, nevertheless, is the calculation applied by the courts to evaluate equal protection claims. See, e.g., State v.McCarthy, supra, 197 Conn. 250; State v. Castonguay, supra,194 Conn. 427; United States v. Rioux, supra, 97 F.3d 659. Given the acceptance of SDT, this court will not attempt to fashion a new calculation of underrepresentation despite its reservations about the efficacy of SDT. It is unnecessary to do so because the defendant has failed to meet the third prong of the Castaneda test, see infra. Thus, even though per arguendo defendant can meet the second prong using SDT, the defendant cannot establish a prima facie case.

(4)Third Element: Systematic Exclusion and ProcessSusceptible to Abuse

(A)Systematic Exclusion

In order to show a prima facie violation of the sixth amendment right to a jury drawn from a fair cross section of the community, the defendant must show that the distinctive group is significantly underrepresented due to systematic exclusion. Durenv. Missouri, supra, 439 U.S. 357. "There is systematic exclusion when the underrepresentation is due to the system of jury selection itself, rather than external sources." United States v.Rioux, 97 F.2d 648, 658 (2nd Cir. 1996). See also United StatesCT Page 6831v. Fields, 113 F.3d 313, 326 (2nd Cir. 1997) (affirming district court's opinion based on reasoning of Rioux).³²

Here, the defendant argues that "[j]ury official's knew or should have known of this serious and longstanding problem with undeliverable summonses and its disproportionately high impact on potential Hispanic venirepersons. The state has made no effort and has taken no action to remedy this undeliverable problem as it affects the jury array from which this defendant must select his petit jury." (Defendant's Third Amended Challenge to the Jury Array.) The defendant further asserts that "[t]he state has made no effort and has taken no effective action to remedy this `no-show' problem as it affects the jury array from which this defendant must select his petit jury. Additionally, state efforts to assure those summoned do actually appear for jury duty have been ineffectual, and have now been curtailed or discontinued.³³ (Defendant's Third Amended Challenge to the Jury Array.)

The problems of undeliverables and no-shows, however, are not due to systematic factors but to private sector effects. A summons may be undeliverable for a number of reasons including: failure to provide the Post Office with a forwarding address when moving; providing an inaccurate, incomplete or illegible address when registering to vote, obtaining a license or giving a forwarding address; or even human error with regard to delivery.

There is no evidence of any basis for undeliverables centered in the jury selection system itself. Failure to remedy private sector influences does not constitute systematic exclusion. Similarly, the failure of prospective jurors to appear at the courthouse at the designated time is clearly outside of the jury selection process.³⁴

Further, the defendant has put forth no evidence that remedial efforts such as personal service would increase the number of summonses delivered. Even if the number of undeliverables to Hispanics did drop, that would not guarantee that the recipients would then report for jury duty. There is similarly no evidence that strict enforcement of penalties in the form of a fine against no-shows would increase the numbers of Hispanic appearing jurors. To the contrary, the testimony of Federal Magistrate Margolis and Mr. Munsterman of the Center for Jury Studies at the National Center for State Courts indicates that enforcement procedures in general are ineffective.³⁵ Furthermore, efforts at deliberate inclusion of a racial or CT Page 6832 ethnic group might constitute a due process violation under Statev. Nims.

"In Rioux, the court held that under the systematic exclusion requirement the assessment of jury representativeness should take into account only `affirmative governmental action' and not `private sector influences.'. . . Thus, the court stated that `failure to take affirmative measures to rectify underrepresentation caused by demographic patterns and a high disqualification rate for Hispanics' does not constitutesystematic exclusion. . . . Because the Sixth Amendment does not impose an affirmative obligation on the courts to counteract [such private sector influences as voting patterns, demographic trends, and cultural differences], the failure to do so cannot constitute systematic exclusion." (Citations omitted; emphasis added.) United States v. Purdy, 946 F. Sup. 1094, 1103 (D. Conn. 1995), quoting United States v. Rioux, 930 F. Sup. 1558, 1572-78 (D. Conn. 1996), aff'd, 97 F.3d 648 (2nd Cir. 1996).

"The inability to serve juror [summonses] because they were returned as undeliverable is not due to the system itself, but to outside forces, such as demographic changes". United States v.Rioux, 97 F.3d 648, 658 (2nd Cir. 1996). "Moreover, the Sixth Amendment does not comprehend detailed procedural obligations, such as personal service of summonses. . . ." United States v.Rioux, supra, 630 F. Sup. 1558, 1578 (D. Conn. 1995). "Because the district does not have an obligation under the Sixth Amendment to affirmatively counteract `private sector influences,' neither the failure to mail follow-up questionnaires to persons who did not respond to [earlier mailings] nor the failure to follow . . . recommendations for increasing minority participation constitute `systematic exclusion.'" United Statesv. Purdy, supra, 946 F. Sup. 1104. It is noted that efforts to specifically target minority populations in the summoning process "might be self-defeating since it would undermine the randomness of selection — an attribute of jury selection that is keenly desired." United States v. Rioux, supra, 97 F.3d 659. Moreover, targeting Hispanics to intentionally increase their numbers on jury venires might constitute a State v. Nims violation; see discussion, supra.

The defendant contends that the Rioux court "at least hinted that the case may be different if it were established that there was a significant impact on Hispanics due to undeliverables." (Defendant's Brief, p. 34.) It is true that the defendant in CT Page 6833Rioux failed to show that the undeliverable questionnaires went to Blacks or Hispanics. Id., 658. The court did not, however, state that undeliverables would be a systematic defect if it were shown that the undeliverables went to a particular group. Rather, the court was merely emphasizing that even if the use of stale addresses from two-and-a-half-year-old voter lists were considered part of the system, the defendant had failed to establish any link between an undeliverable questionaire and a Black or Hispanic. Id. While the use of voter lists over two years old might approach a systematic flaw, there is no such allegation here.

The defendant accused the state of knowing that Hispanics are more likely to be undeliverable and not remedying the problem by other procedures. There was no contention that the state should update the master list in intervals of less than one year.

In all of the defendant's briefs and arguments not one remedy was advanced to alleviate the undeliverable summonses. Instead, the defendant proposed that this court initiate any needed remedies.

The defendant relied on United States v. Gometz, 730 F.2d 475 (7th Cir. 1984) and United States v. Carmichael, 685 F.2d 903 (4th Cir. 1982). In particular the following statement in UnitedStates v. Gometz, supra, 730 F.2d 478: "Still, a low rate of response to juror questionnaires could lead to the underrepresentation of a group that is entitled to be represented on the qualified jury wheel." Id. The court did not hold that a low response rate would lead to significant or unconstitutional underrepresentation. Furthermore, the court was discussing only the Jury Selection and Service Act of 1968, 28 U.S.C. § 1861 et seq. The court concluded that the Act did not require the district court clerk to take measures to correct a low response rate,³⁶ based in part on the facts that district court clerks lack the resources to issue summonses to persons who do not return forms and to follow up on summonses which are ignored; that overworked district judges should not be required to cite nonresponders for contempt; and that a person forced to serve on a jury will be an angry juror and therefore a bad juror. Id., 480, 482.

In United States v. Carmichael, supra, 685 F.2d 912, the court stated: "A low return rate of questionnaires without some evidence that such rate caused underrepresentation of some group, CT Page 6834 does not amount to a substantial violation of the [Jury Selection and Service] Act. Without other evidence, unfair cross section simply cannot be inferred." Id. The defendant asserts that this statement implies that a low rate of return that is shown to cause underrepresentation does constitute an unfair cross section claim.

While this may be true in the context of the Act, the Act is not at issue here. Furthermore, the defendant has not shown thatsubstantial underrepresentation has resulted from the undeliverables to Hispanics. While the defendant is to be commended for providing the court with a comprehensive analysis of the effect of undeliverables on Hispanic representation, it does not follow that such underrepresentation is constitutionally significant or that it stems from sytematic exclusion or a system susceptible to abuse. Moreover, the defendant has failed to show that remedial efforts such as personal delivery of summons would increase the number of Hispanics reporting for jury duty.

Magistrate Judge Margolis addressed the issue of delivery of summonses. She asserted that she had not suggested that marshals deliver questionnaires and that such personal delivery would not be an appropriate use of resources. She further noted that the Second Circuit categorically rejected the suggestion in Rioux. The state's expert on jury selection systems, Munsterman, testified that first class mail is the only delivery method used because registered/certified mail is not more effective and may even decrease minority representation.

Moreover, the state has introduced evidence indicating that the underrepresentation of Hispanics is not due to their ethnicity, but rather to factors which many Hispanics happen to have in common, to a greater extent than non-Hispanics at this time. The state's expert, Michelson, performed three studies of mobility using the Census data, Post Office data, and DMV files regarding renewal of licenses. Michelson found that there is a strong relationship between high percent Hispanic in a zip code and high numbers of rental units as well as high numbers of people who have moved in the past five years. (State's Ex. 34, p. 8.) Similarly, zip codes with high levels of Hispanics are more likely to have high numbers of change of address cards marked "Moved, left No Address" (MLNA) because postal carriers have filled out the cards upon determining that the party has moved but left no forwarding address. (States's Ex. 34, p. 13.) CT Page 6835

Michelson indicated that Hispanics are less likely to obtain Connecticut driver's licenses than non-Hispanic. Among those Hispanics who do have licenses, they are less likely to renew those licenses. According to Michelson, this is true because they are more likely to have moved. (State's Ex. 34 p. 19.)

Michelson concluded that the greater number of undeliverables to Hispanics is due to the greater mobility of the Hispanic population, not due to their "Hispanic-ness." Michelson used multi variate regressions to show that factors other than Hispanic cause the high numbers of undeliverables. (State's Exs. 32d 32f.) Pollard, however, contended that the conclusions Michelson draws from his regressions are suspect. Pollard discussed the assumptions inherent in regression models and the difficulty of interpretation. Pollard specifically pointed to the city of Hartford as an influencing factor that could affect the regression outcome. (Defendant's Exs. DD, FF.) Pollard also illustrated the difficulty of relying on regressions by pointing to state's exhibits 32a and 32g in which Michelson performed different multi variate regressions using the isolated factor; the sign flipped, meaning that the number of linguistically isolated households had a positive effect on undeliverables in one equation and a negative effect in the other.

Pollard also pointed out that when a multi variate regression is performed which shows that Hispanics are not related to the dependant variable, interpretation is especially suspect. This is true because the other variables taken together may describe the Hispanic variable and thus cancel out its effect. It is difficult to determine which variables cause the effect. Variables may also be a proxy for other factors. Pollard did not, however, attempt to identify which variables are causal and which act as Proxies.

Pollard also criticized Michelson's report on mobility as unconvincing. Pollard stressed that one cannot conclude that the studies show moving out of the judicial district, as opposed to movement within or movement in. The court finds that the evidence of several experts³⁷ as to the higher mobility of Hispanics is convincing, although the extent to which Hispanics are more mobile is not clear. Regardless of greater mobility however, it has been shown that Hispanics are less likely to leave forwarding addresses when they do move, an external factor which contributes to the undeliverable problem. (State's Ex. 34, p. 13.) Similarly, Hispanics are more likely to live in areas where there are high numbers of vacancies and of people who do CT Page 6836 not speak English. (State's Exs. 32d 32f.) These factors also contribute to the lower numbers of Hispanics who appear for jury duty. Thus, it appears that factors other than ethnicity are the causal connections to underrepresentation. While these factors may be related to ethnicity, the exclusion of Hispanics is not based on their ethnic status. Common sense and history indicate that as Hispanics, like other ethnic groups before them, acquire English as a primary language, own homes, and live in more stable neighborhoods, the representation of Hispanics will increase on juries.

This court concludes that the current underrepresentation of Hispanics is not a consequence of systematic exclusion or defects.

This conclusion is buttressed by the fact that the defendant has stipulated that the persons to whom summonses are sent are picked at random. Cf. United States v. Pion, 25 F.3d 18, 23, cert. denied, 115 S. Ct 326, 130 L.Ed.2d 286 (1st Cir 1994) ("since the names included in the Master Jury Wheel³⁸ are randomly drawn from the most inclusive data available, and random selection also determines to whom juror questionnaire are mailed, there can be no reasonable inferenoe that the jury — selection process itself systematically excludes Hispanics at any stage up to and including the distribution of juror questionnaires").

Thus, any disproportionate effect on Hispanics with regard to undeliverables and no-shows is due to non-systematic factors and cannot be considered systematic exclusion. Furthermore, the failure to remedy private sector effects does not constitute systematic exclusion. Accordingly, the defendant has failed to state a prima facie violation of the sixth amendment.

(5)Process Susceptible to Abuse

In order to show a prima facie violation of the equal protection clause, the defendant must show that the selection procedure is "susceptible to abuse or not racially neutral."State v. Castonguay, 194 Conn. 416, 421, 481, A.2d 56 (1984) "Although the equal protection test is similar to the cross section test, the critical difference is that in an equal protection claim the defendant must prove discriminatory purpose." Id. Here, the defendant argues that the selection procedure generates disproportionately high numbers of undeliverables, resulting in a system that is not racially CT Page 6837 neutral/susceptible to abuse. The defendant further argues that inaction upon the part of the state is sufficient to show that the system is susceptible to abuse.

In order to show that the system was susceptible to abuse, the defendant must show that it was capable of being applied in such a manner as practically to proscribe any group thought by the law's administrators to be undesirable; Smith v. Texas,311 U.S. 128, 130-31, 61 S.Ct. 164, 85 L.Ed. 84 (1940); or that it "provided a clear and easy opportunity for racial discrimination."Alexander v. Louisiana, 405 U.S. 625, 630,92 S.Ct. 1221, 31 L.Ed.2d 536 (1972).

The defendant has put forth no evidence that the jury officials acted in ways which deliberately excluded Hispanics through the use of undeliverables, e.g., by intentionally utilizing outdated addresses for Hispanics but new addresses for non-Hispanics. The defendant has not introduced any evidence indicating that the mailing of summonses was based on race or ethnicity. Rather, the parties have stipulated that the summonses are sent by first class mail to people whose names are "randomlyselected" from the master list. (Court's Ex. 3.) Nor has the defendant introduced any evidence indicating that the intervening events between a potential juror's receipt of summons and his failure to report for jury duty are in any way affected by an official response or action based on race or ethnicity. Disqualifications, for instance, have not been shown to be granted or denied on ethnic grounds. The random selection of persons to whom to send summonses and the lack of state participation in, or influence on, failure to appear indicates a lack of intentional discrimination.³⁹

In Castaneda v. Partida, supra, 430 U.S. 482, the court held that the "key man" system of jury selection as applied was not racially neutral and was susceptible to abuse. That selection system is easily distinguishable from the procedure at issue here. Under the key man system, jury commissioners choose prospective grand jurors from the community at large. Id., 484. The court determined that the system was not racially neutral with respect to Mexican-Americans and that the selection was "highly subjective."Id., 496-97. Similarly, in Alexander v.Louisiana, supra, 405 U.S. 630, a non-racially neutral selection procedure existed where questionnaires and cards from which jury commissioners picked grand jurors were clearly marked with the person's race. The court noted that the sets of questionnaires CT Page 6838 and information cards were "placed on a table, and the papers of 400 persons were selected, purportedly at random, and placed in a box from which the grand jury panels" were drawn. Id., 628.

By contrast, here the names of prospective jurors are chosen at random from the master list. This system of choosing persons to whom to send summonses is not susceptable to abuse the way the procedures in Castaneda or Alexander were. Similarly, the procedure here is racially neutral; ethnicity is not a factor in the random selection. Nor has the defendant shown that the system of mailing summonses is in any way manipulated to exclude Hispanics. Moreover, the defendant has not shown that the disproportionately high number of Hispanics who fail to appear for jury duty is in any way due to a racially-charged procedure or system open to abuse. The failure of Hispanics to report for jury duty has not been shown to be connected to the jury system's procedures, let alone linked to a procedure susceptible to abuse.

The defendant attempts to overcome his lack of evidence by arguing that intent is not required, at least under the state constitution. The defendant attempts to rely on Sheff v. O'Neill, 238 Conn. 1,(1998), in which the court held that the state's failure to remedy substantial inequalities in educational opportunities is unconstitutional. From this, the defendant tries to extrapolate that integration in the context of jury arrays is a goal which must be met, regardless of the reasons for underrepresentation of Hispanics.

The Sheff case and its reasoning are inapplicable for numerous reasons. First, the Sheff court interpreted the anti-discrimination clause of article first, § 20, to prohibit segregation in schools in the context of the state's affirmative duty to provide school children with an education under article eight, § 1.⁴⁰ There is no such parallel constitutional requirement regarding juries that provides an affirmative constitutional obligation beyond the right to an impartial jury. Furthermore, the children whose rights were at issue in Sheff have no choice but to attend school whereas Hispanics and others can get on source lists in higher numbers by registering to vote and obtaining drivers' licenses, can receive summonses after moving by filing a change of address form, and have the potential to learn English. Thus, the defendant's argument based on Sheff is without merit.⁴¹

The defendant has not shown that the jury selection system CT Page 6839 is susceptible to abuse or not racially neutral. Because the defendant has failed to show intentional discrimination, the third prong of the Castaneda test has not been met. Therefore, the defendant has failed to state a prima facie case of violation of the equal protection clause with regard to undeliverables and no-shows.

(D)ENGLISH REQUIREMENT

The defendant argues that the requirement that jurors understand English, General Statutes § 51-217 (a)(3),⁴² unconstitutionally results in lower representation of Hispanics on jury arrays.⁴³ The defendant argues that this violates the equal protection clause and the anti-discrimination clause of the constitution of Connecticut, article first, § 20, as amended by articles V and XXI of the amendments to the constitution of the state of Connecticut.⁴⁴ The defendant also argues that the provision of interpreters for the hearing impaired but not non-English speakers violates equal protection.

According to the defendant's expert, the language disqualification eliminates disproportionately more Hispanics (11-13% of Hispanics versus 2% of all). (Defendant's Ex. X20.) Assuming that Hispanics are significantly underrepresented on jury arrays due to the English requirement, the state has an overwhelming interest in ensuring that those who serve on juries comprehend the proceedings and can fulfill their role as fact-finders. The United States Supreme Court has recognized that the states can enact reasonable qualifications for service in the jury system. See Taylor v. Louisiana, 419 U.S. 522, 538,95 S.Ct. 692, 42 L.Ed.2d 690 (1975).

Even assuming that the defendant has established "a prima facie case of discriminatory purpose by virtue of showing substantial underrepresentation of a cognizable group [then the burden shifts to] the state to refute the existence of an equal protection violation by showing the use of racially neutral selection criteria." State v. Wright, 207 Conn. 276, 281-282,542 A.2d 299 (1988). The state's interest in a judicial system in which all jurors can understand the proceedings is a proper selection criteria.

"It has long been accepted that the Constitution does not forbid the States to prescribe relevant qualifications for their jurors." Carter v. Jury Commission, 396 U.S. 320, 332, 90 S.Ct. CT Page 6840 518, 24 L.Ed.2d 549 (1970). "The requirement that jurors speak English is unquestionably reasonable. . . ." United States v.Rioux, supra, 97 F.3d 659. "The courts that have considered the problem [of whether the sixth amendment allows the requirement that jurors be able to speak and understand English] have found that the requirement that conduct of judicial affairs be in English is both reasonable and important." (Internal quotation marks omitted.) Commonwealth v. Tolentino, 663 N.E.2d 846, 850 n. 8 (Mass. 1996). See United States v. Flores-Rivera,56 F.3d 319, 326 (lst Cir. 1995) (if English-only requirement is systematic exclusion, the overwhelming national interest justifies the requirement even where defendant argues that the requirement effectively excludes two-thirds of population);United States v. Benmuhar), 658 F.2d 14, 19-20 (lst Cir. 1981), cert. denied, 457 U.S. 117, 102 S.Ct. 2927, 73 L.Ed.2d 1328 (1982) (holding that English proficiency requirements does operate in systematic manner but that state interest is significant and precisely tailored to jury exclusion criteria). Thus, even if the defendant can show underrepresentation of Hispanics due to the systematic exclusion of non-English proficient people, the state interest is significant and sufficient to rebut a prima facie case based on the English requirement.

Additionally, the defendant has introduced no evidence that the state is using the English requirement to intentionally discriminate against Hispanics. There is no evidence that the English requirement has been abused in order to exclude Hispanics. In fact, the disqualification for lack of English proficiency is self-reported; as potential jurors receive summonses, they are given the opportunity to avoid jury service by indicating a disqualification such as lack of English ability. People of all races and ethnic backgrounds are excluded by the English requirement; the defendant has not shown that the English requirement was intended to be, or is used as, a proxy for race or ethnicity. Because the mechanism of disqualifying for lack of English proficiency is devoid of state action, it is not likely to be susceptible to abuse.⁴⁵ Moreover, the state has introduced evidence showing the high burden, both fiscal and administrative, of eliminating the English requirement for Hispanics. Therefore, assuming arguendo that the defendant has established a prima facie case under Castaneda v. Partida based on the English requirement, the state has submitted an effective rebuttal. CT Page 6841

William Aalvarado, the state's chief court interpreter conservatively estimated the annual cost of interpreters for Hispanic jurors at $2,513,986 and for all non-English speaking jurors at $7,500,000.

The above analysis is similarly applicable to the defendant's claim under the state anti-discrimination clause. This court finds that the ability to understand English is a necessary and appropriate requirement under both the state and federal constitutions. The defendant has not shown that the state constitution should be interpreted differently in this context.

With regard to the defendant's equal protection argument based on the provision of interpreters for deaf jurors, providing interpreters for a disability such as hearing impairment but not for non-English speakers does not violate equal protection. The legislature has made the public policy decision to provide, by statute, interpreters for hearing impaired jurors.⁴⁶ This decision was likely influenced by the insignificant numbers of jurors who use these interpreters and the nature of the disability. By contrast, here the provision of interpreters for non-English speaking jurors would involve larger numbers or people and constitute a significant financial burden to the state. Thus, the defendant's arguments based on the English language requirement are unavailing.

(E)DEFENDANT'S ADDITIONAL ARGUMENTS

1.Heightened Protection

The defendant also argues that the underrepresentation of Hispanics on the array violates "the right to heightened procedural protection including the right to reliable sentencing proceedings in this capital felony prosecution."(Defendant's Third Amended Challenge to Jury Array.) While it is true that there is a requirement of heightened reliability in capital punishment proceedings, this requirement is not violated by alleged underrepresentation where the defendant has failed to make a prima facie case under Duren or Castaneda.

"Recognizing the unique nature of the death penalty and the need for heightened reliability in death penalty deliberations . . . the United States Supreme Court has derived from the eighth amendment's proscription of cruel and unusual punishment two principal prerequisites to assure the propriety of CT Page 6842 the imposition of the death penalty in any particular case."State v. Ross, 230 Conn. 183, 230-31; 646 A.2d 1318 (1994), cert. denied, 513 U.S. 1165, 115 S.Ct. 1133, 130 L.Ed.2d 1095 (1995). First, sentencers must not be given unbridled discretion. Second, the defendant must be allowed to introduce any relevant mitigating evidence. Id., 231. Neither of these prerequisites imply a lesser burden with regard to challenging a jury array nor are they violated where the defendant has not shown significant underrepresentation of an identifiable class due to systematic or intentional exclusion.

In David v. Zant, 721 F.2d 1478, 1482 (11th Cir. 1983), the court stated that "[t]he importance of affording a defendant trial by a representative jury of his peers is magnified in capital cases, where juries are required to consider as amitigating factor, any aspect of a defendant's character or record and any of the circumstances of the offense that the defendant proffers as a basis for a sentence less than death." (Internal quotation marks omitted.) Id. Nevertheless, the court applied theDuren and Castaneda tests without alteration. Id., 1482-85.

Further, the Connecticut Supreme Court has rejected the contention that it should "adopt a `heightened reliability standard' when reviewing constitutional claims raised in a capital case." State v. Webb, supra, 238 Conn. 449. Although the courts have recognized a requirement for heightened reliability in capital cases, this requirement does not alter the standards for reviewing a challenge to the jury array.

2.Jury Summoning Statutes

The defendant asserts that the selection process violates the jury summoning statutes, General Statutes § 51-217 et seq. The defendant refers specifically to the English language requirement; General Statutes § 51-217 (a)(3); and the "mechanism for summoning jurors" as provided in General Statutes § 51-219a et seq. The defendant has not briefed this argument beyond the bare reference to the several statutory provisions. Nor has the defendant introduced evidence supporting a conclusion that the jury summoning statutes were violated. Rather, the defendant's case is based on the premise that the jury summoning statutes, as implemented, violate the defendant's constitutional rights. Accordingly, the defendant has failed to show a violation of the jury summoning statutes. CT Page 6843

3.State Constitutional Arguments

In his third amended challenge to the jury array, the defendant argues that, under the state constitution, the burden shifts to the state under the Duran and Castaneda tests after the defendant has established distinctive/cognizable group and underrepresentation. (Third Amended Challenge to the Jury Array, Parts II.B., III.B.) In his brief, the defendant expands his argument to assert that independent provisions of the state constitution also establish that the court should require a higher level of protection for the right to an impartial jury and, specifically, adopt the defendant's method of calculating underrepresentation. (Defendant's Brief, pp 23-25.)

The defendant, following State v. Geisler, 222 Conn. 672,610 A.2d 1225 (1992), attempts to enlist the six tools of analysis for determining whether our state constitution affords greater rights than under the federal constitution. These tools are (1) textual analysis; (2) holdings/dicta of our appellate courts; (3) federal precedent; (4) sister state precedent; (5) historical considerations; and (6) economic/sociological or public policy. State v. Linares, 232 Conn. 345, 379, 655 A.2d 737 (1995).

This court is not persuaded by the defendant's analysis of the six considerations. While the evidence suggests that the state constitution provides protection for an impartial jury and may provide greater protection for criminal defendants in some instances, there is no compelling indication that the state constitution provides greater protection than under the federal constitution in this context. This court will not create a state constitutional right to shift the burden under the applicable tests of unconstitutional underrepresentation or to require the defendant's proposed test of significant underrepresentation.⁴⁷

III CONCLUSION Logic, history, experience and common sense, as supported by the evidence, degree that undeliverable jury summonses, in the main, result from stale addresses; that the cause of stale addresses is tenant mobility; that tenant mobility is highest amongst the poor, the newly arrived to these shores, non-owners CT Page 6844 and the linguistically isolated. Nowhere is mobility a cachet particular to any distinctive or cognizable group. Ethnicity is irrelevant to the causes of mobility.

The evidence was incontrovertible that the Hispanic population in Connecticut is in a high state of flux. Dr. James Connelly, Bridgeport's superintendent of schools described the excessive traffic of Puerto Rican school children traveling back and forth from Connecticut to Puerto Rico and its impact upon the state's schools and economy. To minimize learning down time, both Connecticut and Puerto Rico agreed to an "Educational Passport" for these itinerant students. Connelly testified that in Bridgeport 30% of all students, grades 1 to 6, are new to their particular school.

A plethora of reasons were submitted to explain this migratory phenomena. Ranging from cultural preferences to economy of travel, what is clear is that all of the reasons are external forces and disconnected from any state action.

The costs of this trial will exceed $1,000,000. The trial was an adversarial contest between two state agencies. Most of the issues raised were previously litigated in a recent federal case. See Rioux, supra.

A concern for husbanding precious judicial resources dictates that a decision, in the future, to challenge the array anew be predicated on facts warranting judicial intervention.

In conclusion this court finds that the defendant has failed to state a prima facie case for a violation of the sixth amendment under Duren v. Missouri, 439 U.S. 357, 99 S.Ct. 664,58 L.Ed.2d 579 (1979) or for a violation of equal protection under Castaneda v. Partida, 430 U.S. 482, 97 S.Ct. 1272,51 L.Ed.2d 498 (1977). Even if the defendant had put forth a prima facie case, the evidence introduced by the state would clearly rebut the defendant's claims.

Additionally, the defendant failed to show a violation of due process pursuant to State v. Nims or under the constitution of Connecticut. Finally, the defendant's arguments with regard to heightened protection, jury summoning statutes, and independent state constitutional protection are without merit.

Accordingly, the defendant's third amended challenge to the CT Page 6845 jury array is denied.

SO ORDERED,

ARTHUR L. SPADA, JUDGE.