Defendant, David Mark Russo was indicted on 18 counts of criminal offenses including murder (death penalty), weapons and other related offenses. He brings this motion seeking dismissal of the indictment based upon the claim that he was *224indicted by a grand jury which was unlawfully and unconstitutionally selected. He further requests the entry of an order staying his jury trial or re-indictment until the appropriate defects are cured. Defendant alleges four separate defects in the Gloucester County jury selection process which he claims entitles him to the requested relief: (1) there exists a substantial underrepresentation of blacks in the qualified jury pool; (2) the statutorily mandated use of driver license lists has been abrogated as to 24.3% of the eligible jury population of the county; (3) the voter and driver lists are merged with a computer program that fails to eliminate obvious duplicates; and (4) the grand jury was improperly empaneled.
Evidentiary hearings were held on April 1, 2 and 7, 1986. The prosecution opposed point (1) of defendant’s allegations but offered no opposition to points (2), (3) and (4). The court stayed the trial date of April 15, 1986 and reserved oral decision on all issues in order to submit a written opinion.
FACTS.
Selection Procedure in Gloucester County.
Jurors in Gloucester County are selected from a source list comprised of a merged list of the county’s registered voters and licensed drivers. Names appear on the source list alphabetically by zip code and street name, and numerically by house number. Jurors are randomly selected from the source list as needed. Random selection is accomplished by the systematic extraction of names from the list according to a predetermined fixed interval. The interval is determined by dividing the number of names on the list by the number of questionnaires that the jury commission sends out. The computer uses the interval to select names from the total list which guarantees the selection of jurors from each municipality and each street without selecting more than one person from every household. Anyone who has served on a jury in the past four years or has received a qualification questionnaire in the past three years is *225excluded from the selection process. The computer also keeps a record of persons who are designated permanently or temporarily ineligible to serve by the jury commission.
The questionnaires are mailed to the prospective jurors who should complete and return them to the jury commission office where they are screened for eligibility. All those deemed eligible are re-entered into the computer. The random number generator method is used to select a qualified pool of grand jurors from the list of eligible jurors. Those who are not selected for grand jury service are designated as petit jurors. The jurors on the lists of grand and petit jurors are then given random numbers and are arranged which consists of placing jurors in an ascending sequence according to their random numbers.
An order is received from the assignment judge informing jury control of the number of panels, the number of people per panel, and the reporting date for each panel. The petit jury qualified list is divided into panels which conforms to the order of the assignment judge. Each panel is divided into subpanels consisting of 50 jurors listed alphabetically. The grand jury qualified list is divided into panels but these panels are left in random sequential order.
DISCUSSION.
Defendant claims that an underrepresentation of constitutional proportions is revealed when the number of blacks, ages 18 to 74, in the population of Gloucester County is compared with the number of blacks in the qualified jury pool. Defendant conducted a telephone survey of 363 members of the qualified jury pool for the February 1985 mailing period and found that 4.63% of the pool consisted of blacks.1 This figure was compared to census bureau statistics for Gloucester County which determined that blacks, ages 18 to 74, constituted 8.58% of the *226population. It is this disparity that forms the basis for defendant’s claim of underrepresentation of blacks in the qualified jury pool. Before addressing the merits of this claim, a brief discussion of defendant’s other allegations follows.
In point (2) defendant asserts that the statutorily mandated use of driver license lists has been abrogated as to 24.3% of the eligible jury population of Gloucester County. Specifically, the names of Gloucester County licensed drivers residing in zip codes 08012, 08020, 08032, 08081, 08094, 08343, 08344 and 08360 are omitted from the driver list that is merged into the source list, as mandated by N.J.S.A. 2A:70-4. People residing in these zip-code areas who are licensed drivers but not registered voters will not be included on the jury source list and will be excluded from jury service. The reason for the exclusion of these zip codes is that all of them, except 08020, contain names of persons who do not reside in Gloucester County.2 Before these zip codes were removed, a great deal of time and money was expended in mailing, receiving and sorting questionnaires to and from persons who did not reside in Gloucester County. Though the decision to remove these zip codes was made in good faith and based on cost-benefit analysis, it was not in compliance with N.J.S.A. 2A:70-4 and must be corrected.3 See State v. Wagner, 180 N.J.Super. 564, 567-568 (App.Div. 1981).
The failure to include licensed drivers in all zip-code areas assigned to residents of Gloucester County decreased or eliminated the opportunity of those affected residents to be included in the jury source list. Conversely, such failure improperly increased the opportunity of those otherwise eligible residents of this county to be .included in the jury source list by virtue of the fact that their names were registered on both voter and *227driver license lists. The grand jury and petit jury panels so drawn were improperly empaneled and the system must be corrected, immediately, to include in the jury source list those persons enumerated by N.J.S.A. 2A:70-4.
____ registry lists of the several municipalities and election districts of their county, and lists which shall be compiled by the Division of Motor Vehicles, of the names and addresses of the holders of motor vehicle driver licenses who are residents of their county. The commissioners shall use these lists to compile a single list from which all jurors shall be selected.
In the legislative statement appended to L. 1979, e. 271 the Assembly Judiciary, Law, Public Safety and Defense Committee provided:
Juries chosen from a representative community sample are a fundamental component in our system of justice. However, the limitations on sources which jury commissioners may use in compiling prospective jury lists and the reluctance of citizens to serve on juries because of the financial sacrifice involved have greatly hampered the justice system’s ability to get juries which truly reflect community standards. The purpose of this bill is to improve this situation by requiring the jury commissioners to select jurors from a single list composed of all registered voters and all holders of motor vehicle drivers licenses in the county.
Defendant suggests that additional areas may be utilized in order to supplement the jury source list; however, it is apparent that the law makers were concerned with enabling jury commissioners to provide juries chosen from a representative community sample, and they were also aware of the limitations on sources to be utilized in compiling jury source lists which truly reflect community standards. They mandated that only two lists shall be so utilized. The statute is mandatory, not discretionary. The law is clear, and we are bound to comply. Only legislative action can modify or amend the statute to include other sources to supplement the jury source list. Compliance with the statute is ordered.
In point (3) defendant attacks the procedure that is used to merge the voter and driver lists into the jury source list. The computer program that is utilized for this task only eliminates exact duplicates and fails to eliminate other obvious duplicates. This affects the randomness of the jury selection *228process. For example, if an individual in Gloucester County records his name as William J. Legal for the purpose of driver license registration and as William Legal for the purpose of voter registration, when these lists are merged, this obvious duplication will not be recognized and his name will appear twice on the jury source list. Similarly, if an individual records one name with two different addresses (one as a licensed driver and one as a registered voter) duplication will occur. Under these facts, Mr. Legal has twice the chance of being selected for jury service as an individual who is listed only once on the source list. As previously stated, 24.3% of the eligible jurors of the county will not find their names on the driver license list that is merged into the source list; therefore, duplications on the source list of residents in these zip codes are virtually non-existent (1.1% duplication rate), while duplications in the rest of the county are much more prevalent (12.2% duplication rate). The net result is that individuals residing in these zip code areas are less likely to be summoned as jurors as compared to the rest of the county. This defect has a negative impact on the randomness of the jury selection process and must be corrected. See State v. Wagner, supra.
In point (4), defendant asserts that the assignment judge failed to comply with the statutory requirements when selecting the foreperson and deputy foreperson for the grand jury. The assignment judge examined the questionnaires of all grand jury panel members present, then first selected the foreperson and deputy foreperson; the remaining members of the panel were then randomly selected. The assignment judge and his predecessors employed this procedure in good faith, in compliance with the spirit of the law and with the intention of selecting from the entire panel, the most qualified foreperson and deputy foreperson to serve on the grand jury; however, it is a matter which requires adjustment. N.J.S.A. 2A:73-1 and R. 3:6-4 mandate that the grand jury of 23 be randomly selected first, then the foreperson and deputy foreperson to be *229selected from among the total of 23 members. The statutory mandate has been acknowledged and will be satisfied.
The finding of a defective grand jury selection procedure has prospective application only. The approach to retroactivity has no constitutional mandate. Rather the competing considerations in each case are weighed. State v. Nash, 64 N.J. 464 (1974). There are basically three factors to be considered: (1) the purpose of the rule and whether it could be furthered by retroactive application; (2) the degree of reliance placed on the old rule by those who administered it; and (3) the effect retroactive application would have on the administration of justice. Id. at 471.
As to (1), no purpose would be served by dismissing indictments and forcing the prosecutor to recall hundreds of witnesses and victims for the usually routine formality of a “true bill” vote. A court’s power to dismiss an indictment is not to be exercised except on the clearest and plainest grounds. State v. Long, 204 N.J.Super. 469 (App.Div.1985); State v. Porro, 158 N.J.Super. 269 at 281 (1978). Every criminal case in this county is affected, not just those with specific fact patterns or particular offenses. The purpose of this ruling is to give stability to the process and to mandate future compliance by the jury commissioners. In order to insure compliance it is not necessary to dismiss any indictments. All those persons indicted by the present and prior grand juries shall not have indictments dismissed as a result of this decision. The present grand jury may continue to sit and perform its duties and functions as heretofore.
As to (2), the court is satisfied that those persons who assumed and performed their duties in the jury selection process, exercised their best efforts and good faith in order to comply with the statutory requirements. The shortcoming does not represent an invidious discrimination based upon race, color, creed, national origin, ancestry, matrimonial status or sex. State v. Porro, supra.
*230This court further concludes that those persons who administered the selection process did sincerely believe that the process did properly comply with the law. It is only now, upon discovery of the defects that the process is acknowledged to be improper, but not to the degree which permits a retroactive application.
As to (3), to afford anything other than prospective application to this decision would create chaos, wreak havoc and cause pandemonium in the criminal justice system. It would result in glaring inconsistencies and disparate treatment of defendants. For example, should capital cases be treated differently from non-capital cases?
If the rule is applied only to those who in the future make timely application, then every indictment handed up is subject to dismissal, but those indicted perhaps by the very same grand jury who have had their case disposed of or have entered their plea more than 30 days ago have no recourse. Do we let the bar down completely for all except those whose cases have been fully litigated including the appellate process?
The mutations are endless and defy rational, consistent resolution if any application other than prospective is granted. No further procedure is developed and implemented. What will be the effect on the state grand jury that draws its members in part from Gloucester County? The prognosis for the criminal justice system stricken with retroactivity is total paralysis. The effect on the administration of justice would be disastrous. State v. Long, supra, 204 N.J.Super. at 487.
The finding of a defective petit jury selection procedure has prospective application only. The principles of State v. Nash, supra, govern this aspect of the challenge as well. Even though retroactivity will not cause the same inconsistencies and disparate results related to grand jury retroactivity, the detrimental consequences are substantial.
Even if retroactivity were limited to cases tried in the future by juries selected in the past by the tainted process, the delay *231would be devastating. It is true that defendants in non-capital cases may waive a trial by jury, R. l:8-l(a) and, therefore, impliedly may waive any defect in that jury. State v. Kociolek, 23 N.J. 400 (1957). However, to leave the trial courts at the mercy of defendant’s election, while new jury lists are being prepared, questionnaires mailed and a new computer program designed, is untenable.
This court was left with unavoidable conclusions. Although the system did not rise to the task of providing juries in which all had an equal chance of selection, the nullification of past and existing juries would collapse the criminal justice system in Gloucester County.
The sensible test of weighing competing considerations so succinctly and clearly set forth in State v. Nash, supra, provides the court with a viable solution to this perplexing problem amply supported by the law of the land. Trials may proceed while the jury commission moves with expedition and dispatch to comply with the mandate of the law. New jury lists shall not be prepared under the existing system.
A challenge to the array of the grand jury and petit jury must be timely made. In the case of this defendant, the challenges to the grand and petit jury array were timely made. In all other cases any motions challenging or attacking the validity of the grand and petit jury array are subject to the time periods set forth in R. 1:8-3, R. 3:6-2, R. 3:10-1 et seq. and other appropriate rules. Each case shall be decided on an individual case by case basis by the appropriate judge.
The court now turns its attention to point (1) of defendant’s motion.
In point (1) defendant pursues his challenge to the jury selection process under the Equal Protection Clause of the Fourteenth Amendment and the Sixth Amendment of the United States Constitution. In order to prove an equal protection violation, defendant must show that “the procedure employed *232resulted in a substantial underrepresentation of an identifiable group. The prima facie case consists of a showing that a recognizable, distinct class has been singled out for different treatment under the law by its substantial underrepresentation in the jury system.” State v. Ramseur, 197 N.J.Super. 565, 574 (Law Div.1984); see also Castaneda v. Partida, 430 U.S. 482, 484, 97 S.Ct. 1272,1275, 51 L.Ed.2d 498 (1977). To prove a Sixth Amendment violation, defendant must demonstrate: (1) that the group alleged to be excluded is a “distinctive group in the community; (2) that the representation of this group in venires from which the juries are selected is not fair and reasonable in relation to the number of persons in the community; and (3) that this underrepresentation is due to systematic exclusion of the group in the jury selection process.” Duren v. Missouri, 439 U.S. 357, 364, 99 S.Ct. 664, 668, 58 L.Ed.2d 579 (1979). A Sixth Amendment challenge differs from a Fourteenth Amendment challenge in that the challenger need not prove purposeful or intentional discrimination. The Sixth Amendment challenge focuses upon the impact which the jury selection system has on any particular cognizable group. State v. Ramseur, supra 197 N.J.Super. at 575; Duren v. Missouri, supra 439 U.S. at 368, n. 26, 99 S.Ct. at 670, n. 26. Blacks are a cognizable group for the purpose of a Fourteenth and Sixth Amendment challenge to a jury selection system. State v. Ramseur, supra 197 N.J.Super. at 579. The prima facie case may only be overcome by a showing “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 select group.” Duren v. Missouri, supra 439 U.S. at 367-368, 99 S.Ct. at 670.
The initial focus of both a Fourteenth and Sixth Amendment challenge to the jury selection process is on the degree of underrepresentation of the particular group. In this case, defendant produced evidence that 8.58% of Gloucester County’s eligible jury population is composed of blacks but only 4.63% of the qualified jury pool consists of blacks.
*233Before a determination is made as to whether these figures amount to a significant enough underrepresentation of blacks to enable defendant to establish a prima facie case, an analysis of defendant’s survey and the results therefrom is warranted. Defendant’s survey of the jury pool was conducted from the February 1985 mailing period. In that period, 17,000 qualification questionnaires were sent out to members of the source list; 5,190 (31%) did not respond and 1024 (6%) were undeliverable. From the remaining 10,786, 2604 (15%) were disqualified pursuant to N.J.S.A. 2A:69-1 and 1694 (10%) were granted exemptions pursuant to N.J.S.A. 2A:69~2; therefore, the total number of qualified questionnaires was 6,488 (38%). Of the 6,488, a random sample of 499 was extracted and defendant attempted to contact these people telephonically to determine racial composition. 363 people were reached (72.8%), and it was found that 17 (4.63%) were black. The following tables summarize these results:
TABLE I.
QUALIFICATION_NUMBER PERCENT NUMBER PERCENT
A. Total Number Qualification 17,000 100%
Questionnaires Mailed_
B. No Response 5,190 31%
C. Undeliverable 1,024 6%
D. Disqualified 2,604 15%
E. Exempt 1,694 10%
F. Total Number Non-Qualified = B+C+D+E 10,512 62%
G. Total Number Qualified = A-F 6,488 38%
TABLE II.
DEFENDANT’S SURVEY.
A. Sample Size 499—7.7% of Qualified Pool
B. Response Rate 363—5.6% of Qualified Pool; 72.8% of Sample
C. Number of Blacks 17— .3% of Qualified Pool; 4.63% of Sample
One of the criticisms of this survey which affects its credibility is the nonresponse rate. Of the 499 members of the quali*234fied jury pool randomly selected for analysis, 136 (27.2%) did not respond to the survey. When there is a relatively high nonresponse rate, as in this survey, statisticians and social scientists look for nonresponse bias. Often times people who do not respond to surveys differ from those who do. Freedman, Pisani & Purves, Statistics (1978) 304-314. From the evidence presented and the testimony adduced for this motion, it is impossible to ascertain the racial composition of the nonresponse group, and it does not automatically follow that it is representative of the entire sample.
Defendant has produced no evidence or testimony regarding the characteristics of this nonresponse group. A geographic inference analysis such as the one discussed in Ramseur was not conducted and is one of the “missing links” in this survey, and without such analysis, the disparities presented by defendant are subject to question and doubt. This problem is further compounded by the inherent difficulty of conducting reliable telephone surveys because of the likelihood that those people who were not reached do not have workable telephone service, are of a lower economic class, and correspondingly are composed of a large percentage of blacks. As set forth in Graham Kalton’s Introduction to Survey Sampling (1983),
In considering a telephone survey, it should nevertheless be recognized that some 7% of households are not covered, and that these households are more heavily concentrated among low income households, households in which the head is non-white and under 35, and households located in the South, (citations omitted) In cases in which the survey’s objectives require good representation of these subgroups, a telephone sample may need to be augmented by another sample—perhaps an area sample—in a dual frame design, [at 87]
This scenario presents the probability that the actual percentage of blacks in the qualified jury pool was underestimated by defendant’s survey. It is impossible to determine how seriously the 1985 survey was affected by the question of telephone ownership. However, since the means of producing evidence necessary to make such a determination was exclusively within the power of defendant, the failure of production must give rise to an inference adverse to the defense. State v. Clawans, 38 *235N.J. 162 (1962). The facts of nonresponse bias and defendant’s failure to address this factor cast doubt upon the credibility of defendant’s survey and the results.
There are two measurements which have been used in cases essentially similar to the one at bar in analyzing jury representativeness; the absolute disparity and the comparative disparity measurements. The absolute disparity measurement is determined by simply subtracting the percentage of the group which appears on the list from the percentage of that group in the population. The comparative disparity measurement is calculated by dividing the absolute disparity figure by the percentage of the group in the population, and the accompanying statistical significance test gives the likelihood that these results occurred by chance and is measured by standard deviations. According to defendant's survey, the absolute disparity of blacks is 3.95% and the comparative disparity measurement is 46% with a statistical significance of 2.44 standard deviations. Defendant claims that the comparative disparity measurement should be employed, Castaneda v. Partide, supra 430 U.S. at 496, n. 17, 97 S.Ct. at 1281, n. 17; Alexander v. Louisiana, 405 U.S. 625, 630, n. 9, 92 S.Ct. 1221, 1225, n. 9, 31 L.Ed.2d 536 (1972); Whitus v. Georgia, 385 U.S. 545, 552, n. 2, 87 S.Ct. 643, 647, n. 2, 17 L.Ed.2d 599 (1967); while the prosecutor contends that the absolute disparity measurement is more representative/ accurate. Swain v. Alabama, 380 U.S. 202, 85 S.Ct. 824, 13 L.Ed.2d 759 (1965); State v. Porro, 158 N.J.Super. 269 (App. Div.1978), cert. den. 439 U.S. 1047, 99 S.Ct. 724, 58 L.Ed.2d 706 (1978); State v. Ramseur, supra. The United States Supreme Court has resolved the issue of the proper standard of representativeness by adopting the absolute disparity measurement as the favored standard. Duren v. Missouri, supra; Castaneda v. Partide, supra; Swain v. Alabama, supra; State v. Ramseur, supra. Though the comparative disparity test does add some insight into the degree of representativeness in the jury pool, it often artificially inflates the numbers and leads to absurd results. See Smith v. Yeager, 465 F.2d 272, 279, n. 18 *236(3 Cir.1972), cert. den. 409 U.S. 1076, 93 S.Ct. 685, 34 L.Ed.2d 665 (1972). The cases cited by defendant as support for the utilization of the comparative disparity measurement discuss this measurement only tangentially in footnotes. While the comparative disparity standard is mentioned in these decisions, it is the absolute disparity method that engages the thrust of the court’s discussion and is the more significant measurement for analyzing jury representativeness. In fact, when the Ramseur court discussed the different measurements of disparity, it listed the Castaneda decision as supportive of the absolute disparity method. Ramseur, supra 197 N.J.Super. at 573, n. 3.
In order to establish a prima facie case under the absolute disparity measurement, a disparity of at least 10% must be proven. Swain v. Alabama, supra; State v. Porro, supra; State v. Ramseur, supra. Accepting defendant’s statistics to be valid, an absolute disparity of only 3.95% has been proven. Therefore, no basis for constitutional relief is present.
Assuming, arguendo, that defendant has shown a significant underrepresentation of blacks, he has still failed to make out a prima facie case because no systematic exclusion has been proven. Duren v. Missouri, supra. It is not enough that defendant demonstrate that the percentage of blacks on the qualified jury pool is significantly lower than the percentage of blacks ages 18 to 74. There are various “weeding out” procedures that occur between the formation of the jury source list and the composition of the qualified jury pool which are designed to remove individuals from jury service who are unable or unqualified to serve. (See Table I supra). To show that the exclusion of blacks was systematic, the removal must be “inherent in the particular jury selection process utilized.” Id. 439 U.S. at 366, 99 S. Ct. at 669; Taylor v. Louisiana, 419 U.S. 522, 95 S.Ct. 692, 42 L.Ed.2d 690 (1975). Defendant claims that blacks were systematically excluded from jury service in three ways: (1) the voter and driver lists are not representative *237of blacks in Gloucester County; (2) the duplication problem discussed supra; and (3) the elimination of eight zip codes discussed, supra. It is the decision of this court that defendant has not satisfactorily demonstrated these factors systematically exclude blacks from jury service.
Defendant alleges that the voter and driver lists that are merged into the source list are not representative of blacks in the county. In support of this contention defendant offers no statistical evidence. Without statistical support showing that the percentage of blacks on these lists is lower than the actual percentage of blacks ages 18 to 74 in the county, this court will not question the wisdom of the Legislature in deciding that potential jurors will be drawn from voter and driver lists only. N.J.S.A. 2A:70-4. Defendant alleges systematic exclusion of blacks because the duplication problem discussed supra, results in whites and people of middle- to upper-income having a greater chance of being selected as jurors than blacks. Defendant claims that people of middle- to upper-income are more likely to be both registered voters and licensed drivers, and are therefore more likely to have their names listed twice on the jury source list. Though this argument has some intuitive merit, no statistical evidence is given in support thereof. Similarly, defendant’s allegation that the zip-code removal process discussed supra systematically excludes blacks is unsupported by any statistical evidence. There is no demonstration that the municipalities affected by the zip-code exclusions are composed of a high percentage of blacks. Without such proof, defendant’s claims on these issues must fail. See State v. Clawans, supra.
CONCLUSIONS.
The process in composing the jury source list in Gloucester County is defective because it is not random.
In accordance with N.J.S.A. 2A:70-4, all residents of Gloucester County who own a driver license shall be included in the jury source list.
*238Duplications and exclusions of eligible residents of Gloucester County from the jury source list shall be eliminated by the jury commissioners.
The grand jury shall be empaneled in accordance with N.J. S.A. 2A:73-1 and R. 3:6-4.
All rulings in the case at bar shall have prospective application only.
The existing grand jury may continue to function and perform its duties without interruption.
The petit jury shall continue to function and perform its duties without interruption.
Trials shall be scheduled without interruption. Trials are neither discontinued nor suspended.
Motions must be filed in a timely manner in accordance with the appropriate rules, including but not limited to R. 1:8-3, R. 3:6-2, R. 3:10-1 et seq. and R. 1:1-2.
Each case shall be considered individually on a case by case basis by the appropriate judge.
New jury source lists shall not be prepared until the defects in the present system are corrected.
The motions by defendant were timely made.
The motion to stay the trial is granted.
The motion to dismiss indictment no. 395-4-85 is granted.
A detailed analysis of the survey and its results will be made, infra.
Zip-code analysis is used to locate licensed drivers in Gloucester County.
No apparent reason exists for the removal of 08020 which is entirely within the county.