Commonwealth v. AponteAnnotate this Case
391 Mass. 494 (1984)
462 N.E.2d 284
COMMONWEALTH vs. JOSE APONTE (and companion cases).
Supreme Judicial Court of Massachusetts, Essex.
November 8, 1983.
March 19, 1984.
Present: HENNESSEY, C.J., WILKINS, LIACOS, ABRAMS, & O'CONNOR, JJ.
*495 Lila Heideman, Assistant District Attorney, for the Commonwealth.
Daniel E. Callahan (Lawrence J. McGuire, Albert S. Conlon & Paul Morton with him) for the defendants.
The Commonwealth has appealed from an order issued by a Superior Court judge dismissing indictments returned against fifteen Hispanic defendants by Essex County grand juries. Mass. R. Crim. P. 15 (b) (1), 378 Mass. 882 (1979). The judge granted the Commonwealth's motion to stay his order pending decision by an appellate court. We granted the Commonwealth's petition for direct appellate review.
Essex County grand juries returned indictments against the fifteen Hispanic defendants in February, March, and August, 1981. The attorneys representing these defendants filed a motion in the Superior Court in Essex County to dismiss the indictments on July 3, 1981. Mass. R. Crim. P. 13 (c), 378 Mass. 871 (1979). The defendants contended that the procedures used in selecting and empanelling grand juries for Essex County for the years 1976-1981 were not neutral and resulted in substantial underrepresentation of Hispanic persons. They claimed that these procedures violated the rights guaranteed them by the Fourteenth Amendment to the Constitution of the United States, by art. 12 of the Massachusetts Declaration of Rights, and by G.L.c. 234.
The judge conducted evidentiary hearings spanning eleven days from December, 1981, through June, 1982. Based on the testimony of the defendants and the definition provided by the 1980 census of population and housing, the judge, in an interim *496 order dated June 11, 1982, found that all the defendants were Hispanic. The Commonwealth does not contest this finding.
The Commonwealth contends, however, that the judge erroneously decided that the indictments returned against the defendants violated their rights under the equal protection clause of the Fourteenth Amendment to the United States Constitution and art. 12 of the Declaration of Rights of the Massachusetts Constitution. We conclude that we need not decide whether the evidence in this case supports the judge's findings of purposeful discrimination in violation of the equal protection clause under the "rule of exclusion" espoused by the United States Supreme Court. See Castaneda v. Partida, 430 U.S. 482, 494 & n. 13 (1977). We agree with the judge that the defendants demonstrated a prima facie case of systematic, albeit unintentional, discrimination against Hispanic persons in violation of art. 12 of the Declaration of Rights. The order of the judge dismissing the indictments against the defendants is therefore affirmed.
1. The factual background. A. Evidence of underrepresentation. The parties stipulated, based on the 1980 Federal census, to the number of persons of Hispanic origin in the total Essex County population, as well as to the percentage of these individuals in the county as a whole and in the various cities and towns. Of 465, 477 adult persons in Essex County, 9,225, or 1.98%, were of Spanish origin; 94.92% of the Spanish origin population was concentrated in ten cities and towns, with Lawrence and Lynn containing the largest percentages.
*497 The parties also agreed that between 1976 and 1981, 328 individuals were summoned for grand jury duty in Essex County. The record does not indicate how many of these persons actually served on a grand jury. The defendants mailed to these 328 individuals a questionnaire in which each person was asked to designate his or her ethnic identification. Based on the returned questionnaires and on personal investigation, the parties agreed that 321 of the 328 individuals summoned for the grand jury between 1976 and 1981 were non-Hispanic persons. Although the other seven persons could not be found, both parties agreed that the odds were very small indeed that they were of Spanish origin, based on their surnames and on interviews with their neighbors and other persons who knew them. The court-appointed expert testified that the probability of three of the seven remaining persons being of Hispanic origin was approximately three in 10,000.
During the evidentiary hearings, the defendants and the Commonwealth presented witnesses who were experts in statistical analysis. Dr. Thomas J. Marx, the defendants' expert, testified that a random selection of prospective grand jurors from Essex County would have produced, at minimum, three Hispanic persons. The expert concluded that the absence of Hispanic persons from the list of grand jurors in the years at issue constituted strong evidence of a nonrandom selection procedure and the existence of bias against members of the Hispanic population.
*498 Dr. Jeffrey W. Eiseman, the Commonwealth's expert witness, utilized a different test for analyzing the data on Hispanic persons in Essex County. Dr. Eiseman calculated underrepresentation on a community-by-community basis as opposed to a county-wide basis, since the established practice seems to have been that each town or city in Essex County sends, in response to writs of venire facias, an equal number of veniremen to a panel. In determining the minimum number of Hispanic persons expected to be drawn, based on their proportional representation in Essex County, the expert adjusted his figures by estimating the percentage of Hispanic persons who would not be eligible to serve on a grand jury. Dr. Eiseman thus concluded that if the statistical analysis was conducted on a community-by-community basis and took into account only those Hispanic persons who presumably were citizens or could speak English sufficiently, the defendants could not demonstrate that the observed underrepresentation resulted from nonrandom selection. On cross-examination, however, Dr. Eiseman did admit that if the analysis was performed on a county-wide basis there would be an inference that a nonrandom, discriminatory selection procedure was utilized.
The court-appointed expert, Dr. Michael Malec, was qualified in statistical analysis in the social sciences. Although *499 Dr. Malec recognized that the tests utilized by both parties' experts were common in social science research, he concluded that neither he nor any statistician could tell the court precisely which standard to apply, or whether any standard could determine bias in a jury selection procedure. He testified, however, that Dr. Marx's formula was appropriate, and that the probability of obtaining zero Hispanic persons from a random sample of 321 persons, given a percentage of 1.98% Hispanic persons in the county population, was slightly over one in 1,000, or sixteen in 10,000 (. 0016) exactly.
B. The Essex County grand jury selection system. The defendants were prepared to present detailed evidence in support of their contentions that the grand jury selection system in Essex County was susceptible of abuse, or was nonneutral. The defendants, however, chose not to present such evidence when the judge ruled that their burden on this element was satisfied by the Commonwealth's concession that the key man system of jury selection was in operation in Saugus and Lynnfield. See note 17, infra.
The Commonwealth presented testimony by public officials of ten communities in Essex County who described the procedures utilized by them in their communities to compile the master lists of prospective grand jurors during the relevant period. The ten communities from which these witnesses were drawn represent the communities in which approximately 94.9% of the county's Hispanic population resided. In the communities of Lawrence, Gloucester, and Peabody, the official responsible for drawing up jury lists would choose names at random out of a street listing book and send a questionnaire to each person so selected. These officials testified that they could use their discretion in picking individuals from the street list, and the Lawrence and Peabody officials stated that it was permissible to pick persons known to them from the street listing. These three communities, combined with Saugus and Lynnfield, contained 66.6% of the county's Hispanic population.
Town or city officials testified that in the communities of Salem, Methuen, Beverly, and North Andover they would *500 choose names at specific intervals, e.g., every fifth or tenth name, from a list of residents, or one name after another, in equal numbers from each sector of the city. In compiling the lists, these officials would observe the surnames of the residents from the street listings or census records. These four communities accounted for approximately 11% of the county's Hispanic population. The other three communities of Lynn, Haverhill, and Andover utilized a computer which randomly drew names of adults from street lists. The officials of these communities sent out jury questionnaire to individuals on these computerized lists. Eighteen and two-tenths percent of the Hispanic population of Essex County resided in these communities.
Most of the witnesses testified that persons chosen from the street or census lists as prospective jurors from 1976 through 1981 were not screened to determine their eligibility to become jurors but merely were sent the standardized jury questionnaire. The questionnaire was used to determine whether prospective jurors had a characteristic which could render them ineligible for service, e.g., alien status, physical or mental incapacity, holding certain public employment, or having a particular criminal record. See G.L.c. 234, §§ 1, 1A, & 4. Only when the questionnaires were returned would persons who were found ineligible to serve on a jury be eliminated from the list.
*501 As grand jurors were needed for an Essex County panel, the city councillors or selectmen from each community would conduct a blind draw from the master list of a sealed envelope or capsule containing a juror's name.
C. The order by the Superior Court judge. The judge decided that the defendants' underrepresentation claim should be evaluated by reference to the percentage of Hispanic and white adults in the entire Essex County population, as opposed to estimating these percentages on a community-by-community basis. In support of his decision, the judge cited the enabling statute which provides for the recruitment of grand jurors from the county as a whole. G.L.c. 277, § 2G. Furthermore, the judge reasoned that, even if the grand jury selection system discriminated against residents in larger urban centers because it took roughly the same number of jurors from each community, this fact did not rebut the inference of county-wide systematic underrepresentation, especially if the system reduced the chances that Hispanic persons would be selected.
The judge found unpersuasive the Commonwealth's claim that a county-wide analysis distorted the statistics because it did not consider the percentage of ineligible Hispanic persons in the population. The judge reasoned that the Commonwealth failed to rebut the inference of discrimination which was established by the defendants' showing of systematic underrepresentation of Hispanic persons. The Commonwealth failed to present credible, cogent evidence of the percentage of Essex County Hispanic persons who were ineligible for grand jury service. See Castaneda v. Partida, 430 U.S. 482, 498-499 (1977).
The judge proceeded to determine whether the defendants had made out a prima facie case of discrimination under the equal protection clause of the Federal Constitution and art. 12 of the Massachusetts Declaration of Rights. The judge found that the Hispanic defendants belonged to a "recognizable, distinct class, singled out for different treatment under the laws, as written or as applied." Id. at 494. The judge Concluded that the defendants, as required by the Federal Constitution, showed both total exclusion and disproportionate underrepresentation *502 of Hispanic persons over a significant period of time. The judge also decided that under art. 12 the defendants demonstrated a "substantial, identified exclusion of [a] distinct, qualified segment of society." Commonwealth v. Bastarache, 382 Mass. 86, 102 (1980).
Based on the parties' stipulation that 321 of the 328 members of the Essex County grand jury venires between 1976 and 1981 were non-Hispanic persons, and based on evidence that the chances of finding three Hispanic persons among the missing seven members were three in 10,000 (. 0003), and that the probability of drawing no Hispanic persons from a random sample of 321 adults in Essex County was sixteen in 10,000 (. 0016), the judge decided that underrepresentation of Hispanic persons was statistically significant. On the basis of this evidence, the judge concluded that both disproportionate underrepresentation and total exclusion of Hispanic persons from the Essex County grand jury venires had taken place from 1976 to 1981.
The defendants satisfied the second and final prerequisite of a prima facie case under a Federal equal protection claim, the judge held, because stipulations and evidence presented by the Commonwealth conclusively showed that the Essex County grand jury selection procedure was "susceptible of abuse or ... not racially neutral." See Castaneda v. Partida, supra at 494.
The judge then investigated whether the Commonwealth had met its burden of rebutting, under the Federal Constitution, the inference of intentional discrimination demonstrated by the *503 defendants' prima facie case. The judge invited the Commonwealth to offer evidence which would show that the absence of Hispanic persons from the grand jury venires resulted from a neutral nondiscriminatory factor or from factors such as statutory ineligibility, the greater mobility of the county's Hispanic population, or their failure to receive or to reply to a summons. After reviewing the Commonwealth's evidence, which consisted entirely of the previously described testimony concerning the operational procedures of the jury selection system in ten Essex County communities, the judge determined not only that the Commonwealth failed to rebut the prima facie case but also that the system was conclusively shown, by the Commonwealth's evidence, to be susceptible of abuse and not ethnically neutral. Specifically, he found that 66.6% of the county's Hispanic persons resided in communities using the key man system to compile the initial list of prospective grand jurors, and 73.8% of the Hispanic persons lived in communities where the initial selection was either a key man or manual system. The judge noted that a computerized system of compiling lists of prospective jurors was in effect in communities which contained only 22% of the county's Hispanic population.
The judge also decided that the Commonwealth's evidence failed to rebut the prima facie case of discrimination under art. 12. The judge interpreted Commonwealth v. Bastarache, supra, as proscribing, under art. 12, any jury selection system which results in intentional or unintentional discrimination by a significant underrepresentation or total exclusion of one of the protected groups referred to in art. 1. See id. at 101; art. 1 of the Massachusetts Declaration of Rights, as amended by art. 106 (equality under the law shall not be abridged based on sex, race, color, creed, or national origin). The judge therefore *504 ordered the dismissal of the indictments against the defendants.
2. The claims of error. A. Sufficiency of evidence. In reviewing a judge's decision, before trial, to dismiss the indictments against the defendants, this court must determine whether the evidence supported the judge's factual findings and whether the findings warranted the rulings of law. See Commonwealth v. Botelho, 369 Mass. 860, 868 (1976).
We find unpersuasive the Commonwealth's claim that the defendants did not show disproportionate underrepresentation in the Essex County grand jury venires during the years 1976 through 1981. Our General Laws require the drawing of juries on a county-wide basis. The judge correctly determined the percentage of Hispanic persons by reference to the entire county and not on an individual community basis. See G.L.c. 277, § 2G. Even assuming that the Essex County selection system discriminates against all residents of the larger urban areas by drawing approximately the same number from each community regardless of size, this would not refute the existence of systematic county-wide discrimination.
The judge also correctly refused to comply with the Commonwealth's request that, in determining underrepresentation, he consider only the percentage of Essex County Hispanic persons who were eligible to serve on a grand jury. In the challenged selection procedure, it appears that local officials compiled lists of prospective grand jurors only, and did not investigate whether a resident was a United States citizen or was able to speak and understand English, both of which could ultimately lead to exemption from jury service. See G.L.c. 234, §§ 1 & 4. As the local officials testified, statutory ineligibility could be determined only by the answers to the standardized juror questionnaires which the residents selected were required to complete and return. Furthermore, the burden was on the State to present evidence of ineligibility once the defendants established underrepresentation based on the percentage *505 of their minority group in the general population as compared with their representation in the grand jury venire. See Castaneda v. Partida, 430 U.S. at 498-499. There was no such competent evidence offered.
Based on the parties' stipulations and the evidence, the judge properly could find that Hispanic persons were both totally excluded from, and substantially underrepresented in, the Essex County grand jury venires during the years in question.
The Commonwealth claims that the defendants did not show that the Essex County grand jury selection procedure made significant use of a key man system merely because the Commonwealth conceded that a key man system existed in the towns of Lynnfield and Saugus. Hispanic persons residing in these two communities comprised only .9% of the Hispanic population in Essex County. The judge found that a key man system was also utilized in Lawrence, Peabody, and Gloucester.*506 Evidence of selection procedures in Lawrence, Peabody, and Gloucester, together with Saugus and Lynnfield, clearly established the existence of a key man system in communities wherein resided approximately 66.6% of the county's Hispanic population.
We conclude that the record in this case amply supports the findings of the judge. We turn now to examine whether the judge properly ruled that the indictments be dismissed under State constitutional law.
B. The claim under art. 12. The Commonwealth claims that the judge incorrectly decided that any unintentional discrimination against Hispanic persons resulting from the key man system, or any other nonrandom jury selection system, violates art. 12 of the Massachusetts Declaration of Rights.
We have interpreted sections of our Declaration of Rights so as to provide broader protection to criminal defendants than is available under corresponding provisions in the United States Constitution. See Attorney Gen. v. Colleton, 387 Mass. 790, 796, 801 (1982); Commonwealth v. Soares, 377 Mass. 461, 475-477, cert. denied, 444 U.S. 881 (1979). Two recent decisions of this court indicate through their analysis of State constitutional and statutory provisions that art. 12 safeguards defendants against systematic, albeit unintentional, discrimination against their protected class resulting from a nonrandom jury selection system. See Commonwealth v. Bastarache, 382 Mass. 86, 101-103 (1980); Commonwealth v. Soares, supra at 478-479, 481-482. In Soares we held that the use *507 of peremptory challenges by a prosecutor to exclude black persons from a petit jury denied the defendants their rights under art. 12 to be tried by a jury "fairly drawn from the community." Id. at 463. Although Soares involved peremptory challenges in the empanelment of a petit jury, we analyzed and set forth in that decision the essential safeguards of art. 12 with respect to jury selection in general. Id. at 478-479, 481-482. We reasoned that under art. 12 a "defendant is constitutionally entitled to a jury selection process free of discrimination against his grouping in the community," and that a fair jury is one that represents a cross section of individuals and ideas in the community. Id. at 478, quoting Commonwealth v. Rodriquez, 364 Mass. 87, 92 (1973). See Commonwealth v. Ricard, 355 Mass. 509, 512 (1969). The groups which art. 12 protects and which cannot form the basis for juror exclusion are the same classes referred to in art. 1 of the Massachusetts Declaration of Rights. Soares, supra at 486 n. 29, 488-489. We noted in Soares that any jury selection system in the Commonwealth, including the compilation of master lists from which venires are drawn, must be carried out on a random basis and must represent a fair cross section of the relevant community. Soares, supra at 478 n. 13, citing G.L.c. 234A, § 1; 482 & n. 21, citing G.L.c. 234, §§ 1-4. We recognized, however, that absolute proportionality would be an impossible objective given the statutorily prescribed methods of jury selection and provisions allowing for the removal of any juror who expressed an interest, bias, or prejudice relating to a particular case. Soares, supra at 481-482.
Following the interpretation of art. 12 presented in Soares, we stated in Commonwealth v. Bastarache, supra, that we express "particular sensitivity in analyzing jury selection practices to discrimination against those groupings in the community that are set out in art. 1 of the Declaration of Rights." Id. at 101. In Bastarache, we decided that a classification *508 comprised of younger persons (between the ages of eighteen and thirty-four) did not constitute a protected group for purposes of challenging grand jury selection procedures under the Federal Constitution. Because the defendant did not present in a timely manner his State constitutional challenge to the selection procedure, the court did not directly resolve the State issue. Although we intimated that the defendant's discrimination claim based on underrepresentation of younger persons would not prevail since this group was not a protected class under art. 1, we stated that substantial, albeit unintentional, discrimination through a jury selection procedure against any group protected under art. 1 would raise a State constitutional question. Id. at 101, 102.
The Bastarache court also determined that the key man system of choosing juries presented a possibility of abuse by allowing for the selection of jurors "based on subjective considerations." See id. at 102-103. See generally G.L.c. 234, § 4 (resident's name shall be placed on jury list only upon finding by public official, upon personal knowledge, interrogation, or questionnaire, that resident is of good moral character, of sound judgment, and free from all legal exceptions). To effectuate the compilation of jury lists by random selection, in conformance with art. 12, we suggested, under our superintendence power, G.L.c. 211, § 3, that the Attorney General prescribe random selection procedures for all counties or, alternatively, that the Legislature extend the random procedure provided for in G.L.c. 234A (then applicable to Middlesex County only). Id. at 103. In apparent response to the Bastarache case, the Legislature amended c. 234A by making the purely random selection procedure applicable to any county designated by this court as a "participating county," which would result in the abandonment of the key man system, or any other nonrandom system then in effect in those counties. See G.L.c. 234A, § 1, as appearing in St. 1982, c. 298, § 1.
In accordance with the Soares and Bastarache cases, the Superior Court judge was justified in concluding that the defendants made out a prima facie case of discrimination under art. 12. The Hispanic defendants possessed the rights, under *509 arts. 12 and 1, to have prospective grand jurors selected by a process which did not systematically discriminate against their protected class, and to have a grand jury venire which represented a fair cross section of the community. Moreover, the defendants were entitled, under the same constitutional provisions, to a jury selection procedure which precluded the possibility of even an unintentional exclusion of prospective jurors based on national origin. The judge properly found that the defendants were all Hispanic persons, that Hispanic persons comprised 1.98% of the Essex County population, and that there was a total exclusion of Hispanic persons from lists of prospective grand jurors in Essex County over the period at issue. The judge also found that during this period a key man system or other nonrandom system of selection existed in Essex County cities and towns where 73.8% of the county's Hispanic residents lived. Taken as a whole, these findings warrant the judge's conclusion that the Essex County juror selection system resulted in systematic, albeit unintentional, discrimination against Hispanic persons in violation of art. 12.
The judge invited the Commonwealth to rebut the defendants' prima facie case. There was no credible evidence presented to rebut the prima facie case. Instead, the Commonwealth's evidence clearly demonstrated that either a key man system or a nonrandom selection procedure was utilized in communities accounting for more than three-quarters of the county's Hispanic population. The evidence thus justified the judge's conclusion that the Commonwealth had not rebutted the prima facie case of discrimination under art. 12.
Since the judge correctly concluded that the previous Essex County jury selection system violated the defendants' rights *510 under art. 12, we affirm his order dismissing the indictments returned against the defendants.
 Against Jorge J. Barbosa, Jose Cayas, Jose M. Contreras, Felix DeJesus, Felipe Diaz, Moses Felix, Michael Knikker, Jose Pichardo, Otero EdgardoSuarez, Hiram Zaragosa, Julio R. Martinez, Carlos Sanchez, Cesar Rios, and Kelley Ferrer.
 The original motion was also filed on behalf of several black defendants. Following evidentiary hearings, the Superior Court judge denied the claims of the black defendants in his interim order of June 11, 1982. Also, the district attorney has nol prossed the indictments of one of the Hispanic defendants. These latter individuals are no longer parties to the present action.
 United States Department of Commerce, 1980 Census of Population and Housing, Advance Reports 3 (1981). The report defined persons of Spanish origin as those individuals who indicated on the census questionnaire that they were of Spanish or Hispanic origin, or those persons who classified themselves as Puerto Rican, Mexican, or Cuban.
 The equal protection clause of the Fourteenth Amendment to the Federal Constitution provides in relevant part that no "State [shall] ... deny to any person within its jurisdiction the equal protection of the laws." U.S. Const. amend. XIV, § 1.
 Article 12 of the Massachusetts Declaration of Rights provides in relevant part: "No [person] shall be ... deprived of ... life, liberty, or estate, [except] by the judgment of his peers...."
 Lawrence contained 60.84% of the county's total Hispanic population, and 11.8% of the Hispanic population in Essex County lived in Lynn. The other towns and cities containing the bulk of the Hispanic population were Haverhill, Salem, Methuen, Peabody, Gloucester, Beverly, Andover, and North Andover.
 Dr. Marx utilized a binomial distribution test at a .05 level of significance, which posits that if fewer than the "minimum" number of a group, i.e., less than its proportional representation in the studied population, appear in a sample, then the probability of nonrandom selection is statistically significant. The binomial distribution test at a .05 level of significance is a standard commonly used by researchers in the social and behavioral sciences to determine whether the results of their studies are statistically significant or if it is more likely that they occurred by chance. See generally Finkelstein, The Application of Statistical Decision Theory to the Jury Discrimination Cases, 80 Harv. L. Rev. 338, 353-355, 359, 364 (1966). Neither the United States Supreme Court nor this court has accepted any precise statistical formula for analyzing jury discrimination cases. See Castaneda v. Partida, supra at 494-496; Commonwealth v. Bastarache, 382 Mass. 86, 95-96 (1980). Because we agree with the judge that the statistics reveal a total exclusion of Hispanic persons from the list of prospective grand jurors, we do not consider the appropriateness of the .05 test in the present case.
 Dr. Eiseman suggested the use of a higher standard, accepting the sample as a random selection unless the odds of that result occurring by chance exceeded forty-four to one. Under this test and a community-by-community analysis, Dr. Eiseman concluded that the defendants could not show in a statistical sense that the absence of Hispanic persons from the grand jury venires was due to a biased selection procedure.
 General Laws c. 277, § 2G, provides the times for selection of, and numbers of, grand jurors in Essex County. Under G.L.c. 277, § 3, as amended by St. 1977, c. 415, § 12, grand jurors shall be "drawn, summoned and returned in the same manner as traverse jurors." General Laws c. 234 sets forth, inter alia, the procedures for compiling jury lists in the Commonwealth and for drawing and summoning jurors from the master lists to serve on grand and petit juries. G.L.c. 234, §§ 4-31. The Essex County practice of having equal numbers of grand jurors from each city and town in the county appears to be in violation of G.L.c. 234, § 10, which requires that the "number of jurors required by such writs [of venire facias] shall be apportioned among the cities and towns, as nearly as may be, according to their respective populations" (emphasis supplied).
 The city of Lynn, however, had a system whereby a questionnaire would be sent only to every third name on the computerized list. The court-appointed expert testified that this was still an appropriate method of random selection.
 In Beverly, North Andover, and Methuen the community officials stated that they might not send a questionnaire to a prospective juror if an exemption was obvious from a listing of occupation or a designation of age. In Gloucester, the street listing would indicate whether a resident was also a United States citizen. Testimony from the Lawrence official indicated that the standard juror questionnaire was only printed in English.
 Volunteers for service on a grand jury were also commonly accepted in all of the communities except Haverhill, Beverly, Methuen, and Andover. The evidence as to Lawrence, Lynn, and Peabody indicates that an average of twentyfive to fifty volunteers were accepted as prospective grand jurors on a yearly basis in each of these communities. In North Andover advertisements were placed in newspapers soliciting volunteers. The Commonwealth concedes that this practice, at least, was in violation of G.L.c. 234, § 38. We need not decide whether the practice in Lawrence, Lynn, and Peabody also violated § 38.
 In analyzing the evidence of disproportionate representation, the judge relied on the .05 level of significance test proposed by the defendants' expert. The judge recognized that the Supreme Court has never accepted one level of significance test as binding and that the Court consistently had used an arithmetical standard. See Castaneda v. Partida, supra at 496. Yet the judge read a footnote in the Castaneda case, where the Court indicated in dictum that a random sample could be evaluated by using a forty-four to one standard, as signaling a departure from the arithmetic disparity rule and a reliance on a more precise statistical test. See Castaneda v. Partida, supra at 496 n. 17. Moreover, the judge advocated using the .05 level of significance standard to evaluate the art. 12 claim based on recent case law which required application of "the closest scrutiny" to any deviation from a purely random method of juror selection which has adverse impact on ethnic or racial groups. See Commonwealth v. Bastarache, supra at 101-102.
 The judge also found that 83.5% of the county's Hispanic population lived in communities where volunteers regularly were accepted for jury service. The judge noted that four of the public officials who compiled jury lists testified that they did not consider a resident's ethnicity in performing their tasks. The judge concluded that mere affirmations of good faith by these workers could not rebut an inference of intentional discrimination in the selection procedure. See Castaneda v. Partida, supra at 498 n. 19.
 The one exception was Gloucester, where the street listing would indicate whether a resident was a United States citizen.
 The judge properly refused to allow questioning by counsel on the 1970 census since the Commonwealth failed to provide a foundation concerning its reliability as an accurate source of estimating the present Hispanic population in Essex County. Excerpts from a 1980 study of language characteristics among Hispanic adults, published by the National Center for Education Statistics, was offered in evidence presumably to show the percentage of Hispanic persons who could not speak English. The sample was never identified as drawn from an Essex County or Massachusetts population.
 The Castaneda Court described a key man system as a procedure whereby jury commissioners select prospective jurors from among the citizenry based on their belief that the individuals chosen will meet certain statutory qualifications. Each person selected must be of good moral character, sound mind, literate, and must not be under a pending indictment or have a felony record. Castaneda v. Partida, 430 U.S. 482, 484-485 (1977). This court has decided that the key man system exists in Massachusetts because G.L.c. 234, which prescribes the method of preparing lists of prospective jurors, relies on the public employees or officials in each community to select such individuals based, in part, on knowledge of their "good moral character" or "sound judgment," derived from personal familiarity, direct interrogation, or by answers to a standardized questionnaire. See Commonwealth v. Bastarache, 382 Mass. 86, 92, 102 (1980); G.L.c. 234, § 4. See also G.L.c. 234, § 1 (other exemptions from jury service due to position as public servant, elected representative, etc.).
 The Commonwealth appears to concede that the key man finding as to Gloucester is correct and hence challenges the findings as to the use of the key man system in Lawrence and Peabody only.
 Although the judge carefully examined both Federal and State constitutional authority, we need not express our views as to Federal constitutional matters since, in our view, determination of the State constitutional claim is dispositive of the Commonwealth's appeal.
 It is likely that the distinction between equal protection analysis (grand juries) and Sixth Amendment analysis (petit juries) found in Federal law (see Commonwealth v. Bastarache, supra at 89 n. 1, 96-99) is not relevant to State law analysis under art. 12. See id. at 101-103.
 Previously, this court has interpreted challenges to a jury selection procedure under Federal law and held that only purposeful discrimination against a discrete group constitutes a violation of the equal protection clause. See Commonwealth v. Peters, 372 Mass. 319, 323 (1977); Brunson v. Commonwealth, 369 Mass. 106, 113 (1975). In dictum, the Brunson court noted that, under Federal law, a showing of actual prejudice probably would be required to substantiate an underrepresentation claim based on a grand jury selection procedure since the grand jury has the power to indict a defendant by a majority vote, and the inclusion of more members of the discrete group on the panel might not affect the outcome. Id. at 120.
 General Laws c. 234A, §§ 1, 3, and 4, now require random selection of prospective jurors in participating counties to be designated by the Supreme Judicial Court. General Laws c. 234A, § 3, as appearing in St. 1982, c. 298, § 1, provides in relevant part: "All persons selected for juror service on grand and trial juries shall be selected at random from the population of the judicial district in which they reside. All persons shall have equal opportunity to be considered for juror service.... No person shall be exempted or excluded from serving as a grand or trial juror because of race, color, religion, sex, national origin, economic status, or occupation" (emphasis supplied).
We designated Essex County as a "participating county" pursuant to G.L.c. 234A, § 1, with the new grand jury selection procedure to take effect as of January 1, 1984.
 On this record it appears that there is nothing to preclude the defendants from being reindicted by a properly composed grand jury. See G.L.c. 277, § 63 (statute of limitations).
The principle we announce today is not to have retroactive application. That principle will be open on direct appeal only to such defendants who properly have raised, or may in the future raise, similar claims by a pretrial motion to dismiss. Commonwealth v. Rodriquez, 364 Mass. 87, 91 (1973). Mass. R. Crim. P. 13 (a) (c), 378 Mass. 871 (1979). Cf. Reddick v. Commonwealth, 381 Mass. 398, 399-404 (1980) (Soares rule will not be applied retroactively).