JUSTICE ERICKSON delivered the Opinion of the Court.

We granted certiorari and consolidated for argument and decision People v. Cerrone, 867 P.2d 143 (Colo.App. 1993) (Cerrone III) and People v. Goetz, No. 90CA0514 (Colo.App. Apr. 14, 1994) (not selected for publication) (Goetz II). In Cerrone III, the court of appeals found the trial court did not deliberately exclude persons from the 1985-86 Colorado state grand jury on the basis of economic status and held that "the use of hourly wage earner status as one factor in trying to impanel grand jurors who would consistently attend the scheduled sessions during the year was not error." Cerrone III, 867 P.2d at 147. The court of appeals adopted the same reasoning in Goetz II. In contrast to the court of appeals analysis, we conclude that the state's asserted reasons for excluding hourly wage earners are inherently discriminatory and violate the mandate of section 13-71-103, 6A C.R.S. (1987). We hold, however, that this violation does not require that the defendants' convictions be overturned. We therefore affirm the judgments of the court of appeals, while not subscribing to the reasoning that led to those judgments.

I Procedural History

The petitioners, Cerrone and Goetz (defendants) were indicted by the 1985-86 state grand jury. The defendants filed several pre-trial motions challenging the composition of the grand jury. Following hearings, those challenges were rejected by the trial court. The defendants' cases were severed for trial. Each defendant was convicted of one count of pandering and several violations of the Colorado Organized Crime Control Act. On separate appeals, each defendant raised various issues, including an alleged Fourteenth Amendment equal protection violation based on racial discrimination in the selection of the 1985-86 state grand jury.

In People v. Cerrone, 829 P.2d 468, 472 (Colo.App. 1991) (Cerrone I), the court of appeals found "there was no evidence from which the trial court could conclude that the showing of discrimination had been rebutted." The court reversed Cerrone's convictions and did not address the other issues raised on appeal. Id. Several months later, the court of appeals reversed Goetz' conviction based on Cerrone I. People v. Goetz, No. 90CA0514 (Colo.App. Feb. 13, 1992) (not selected for publication) (Goetz I).

The state appealed both decisions to this court, which joined the cases for argument and opinion. In People v. Cerrone, 854 P.2d 178 (Colo. 1993) (Cerrone II), we reversed Cerrone I. We concluded "that the trial court did not err when it found . . . that the defendants did not sustain their burden of proving purposeful racial discrimination under the Fourteenth Amendment." Id. at 193-94. Both cases were remanded to the court of appeals to address the remaining issues raised on appeal, including the allegations of discrimination in the grand jury selection based on economic status. See id. at 194.

In Cerrone III, 867 P.2d 143 (Colo.App. 1993), the court of appeals upheld Cerrone's conviction. The court found that during the selection of the state grand jury there was no discrimination on the basis of economic status. The Goetz II panel adopted the reasoning of Cerrone III and affirmed Goetz' convictions. Goetz II, No. 90CA0514 (Colo.App. Apr. 14, 1994) (not selected for publication). The judgments should be affirmed in both cases although for different reasons than articulated by the court of appeals.

II Facts Regarding Alleged Discrimination Based On Economic Status

The selection of the 1985-86 state grand jury was initiated on January 28, 1985, when the Colorado Attorney General petitioned the Chief Judge of the Denver District Court for an order impaneling a state grand jury. The chief judge granted the petition and ordered that the 1985-86 state grand jury consist of twelve jurors to be selected from the counties of Denver, Adams, Arapahoe, Boulder, and Jefferson. The state court administrator's office compiled a list of 375 prospective grand jurors, seventy-five from each of the five counties specified by the chief judge. Each of the 375 prospective grand jurors was mailed a summons to appear in court, and each was mailed a "Juror Selection Questionnaire" that he or she was instructed to fill out and return to the court immediately. The questionnaire asked about the age and occupation of immediate family members living with the prospective juror and whether the prospective juror had close friends or relatives practicing criminal law or employed in law enforcement work. Additionally, the questionnaire stated: "The Grand Jury generally meets one full day per week for a year. Allowances will be made for vacations, illnesses, and essential business trips. Do you have any problems, including health problems, that would interfere with your serving on the Grand Jury? Yes ___ No ___ If yes, please describe."

Most of the 375 questionnaires were returned to the court during February and early March. Based on answers to the questionnaires and with advice from a deputy attorney general and several assistant attorneys general, the chief judge settled on a list of forty-one prospective jurors whom he required to appear in court on March 15, 1985, for a day of oral voir dire at the close of which the twelve-member 1985-86 grand jury was to be impaneled. Those persons who returned questionnaires to the court but who were not chosen to appear for voir dire on the impanelment date received letters from the court notifying them that they were no longer required to appear in court on March 15. A small number of prospective jurors who failed to return their questionnaires also were present on March 15. The chief judge asked each of them to fill out a questionnaire at that time and subsequently excused all but one of them. This meant that the original pool of 375 prospective jurors had been narrowed down to forty-two for the purpose of oral voir dire and impanelment on March 15, 1985.

The defendants have not challenged the manner in which the original pool of 375 prospective jurors was selected. Instead, the defendants claim that persons of low economic status were excluded from the venire of forty-two prospective jurors chosen to appear for oral voir dire on the impanelment date. On September 26 and October 30, 1986, the trial court conducted hearings on the defendants' motions, which challenged, among other things, the exclusion of persons of low economic status from the venire.

Testimony from these hearings established that the chief judge relied on a coded numbering system. The code numbers corresponded to reasons for postponement or excusal from grand jury service. For example, if a questionnaire was marked with the number "12," it meant that there was a medical reason for excluding that prospective juror, while the number "8" meant that the prospective juror was excluded because he or she no longer resided in Colorado. The number "11" signified unspecified "other" reasons for excusal from the venire.

Approximately sixty-eight questionnaires were marked with the number "11" (category 11 jurors). Of that group of sixty-eight, approximately fifty answered "No" to the question "Do you have any problems, including health problems, that would interfere with your serving on the Grand Jury?" One defense witness testified that no category 11 juror without a high school education was allowed to be part of the venire. The same witness also testified that category 11 jurors tended to be laborers or wage earners and were not as well educated as those selected to the venire.

Further testimony was offered by another defense witness, who qualified as an expert in labor economics and statistics. The statistician compared the category 11 jurors with the venire. She found that 57% of the venire were college graduates, while college graduates constituted only 12% of category 11 jurors. Zero percent of the venire had less than a high school education compared to 20% in the category 11 group. The statistician concluded that the venire members were significantly better educated than category 11 jurors. The statistician also opined that there is a "very high correlation between income and education." Concluding the two groups were "distinctly different," the statistician stated with a 99.9% confidence level that "a systematic selection process . . . caused the difference. . . ." On December 5, 1986, the trial court issued a written order finding that the defendants established a prima facie case of discrimination based on "race, national origin or economic status."

On January 30, 1987, to rebut the defendants' prima facie case of discrimination based on economic status, the state called as witnesses a deputy attorney general and one assistant attorney general who advised the chief judge in selecting the 1985-86 state grand jury. The state also called two other assistant attorneys general who assisted the chief judge in selecting statewide grand juries in other years. The state's primary response consisted of testimony describing the "screening process."

The deputy attorney general testified that members of the attorney general's staff divided the questionnaires into two stacks: those they recommended for the venire and those they recommended be excluded. The stacks then went back to the chief judge for the final decision. In the selection of grand jurors, the deputy attorney general testified that the most significant consideration was the prospective grand juror's ability to attend the grand jury sessions, which occur every Friday for approximately forty weeks. The deputy attorney general stated the staff looked for persons who had the ability to understand complex legal cases.

The deputy attorney general testified that the selection staff considered such factors as education level and whether the potential grand juror was an hourly wage earner. They rarely selected someone who lacked a high school education and preferred persons with an understanding of accounting, financial, or medical issues. In the deputy attorney general's view, additional education would help a grand juror understand the issues that came before the grand jury.

The deputy attorney general acknowledged that persons were excluded from the venire who stated that grand jury duty would not be a hardship. However, the deputy attorney general stated that many persons who claimed no hardship on the questionnaire discover during oral voir dire that they have problems preventing them from serving on the grand jury. Testimony showed that the selection staff tries to anticipate these occurrences, so they consider whether the potential grand juror is an hourly wage earner. Hourly wage earner status was a factor the selection staff used to determine whether grand jury duty would create a hardship.

The trial court found:

The question of economic status is a bothersome one. . . . [T]he problem of economic status and what economic status means under our statute has never been clearly articulated. By making a deliberate attempt to get certain kinds of people on the jury, the result was that persons of lower income may not have been included. However, the Court is going to find that as it relates to economic status that . . . there was no deliberate effort to exclude persons on the basis of their economic status and in fact from the information that was available on the jury questionnaire forms it is not readily apparent what the economic status was.

While the statistician extrapolated economic status from the amount of education, that those particular extrapolations or facts were certainly not necessarily known to the persons who made the recommendations or to the Judge who made the selection at the time that the selections were made.

The court of appeals affirmed the trial court's findings and affirmed Cerrone's conviction. See Cerrone III, 867 P.2d 143, 147 (Colo.App. 1993); see also Goetz II, No. 90CA0514 (Colo.App. Apr. 14, 1994) (not selected for publication) (adopting the reasoning of Cerrone III and affirming Goetz' conviction).

III

Section 13-71-103, 6A C.R.S. (1987), in effect during the time at issue in the present case, provided that "[a] citizen shall not be excluded from jury service in this state on account of race, color, religion, sex, national origin, or economic status." § 13-71-103, 6A C.R.S. (1987). The crux of the present appeal is a determination of the appropriate standard to be used to evaluate whether a violation of section 13-71-103 has occurred. The statute neither defines the term "economic status" nor provides guidance on the standard to be used.

A

We note that section 13-71-103 must be read in light of section 13-71-102, also in effect during the time at issue here, which states:

It is the policy of this state that all persons selected for jury service shall be selected at random from a fair cross section of the population of the area served by the court, and that all qualified citizens shall have the opportunity in accordance with the provisions of this article to be considered for jury service in this state and an obligation to serve as jurors when summoned for that purpose.

§ 13-71-102, 6A C.R.S. (1987). While section 13-71-102 articulates the policy underlying article 71, the "Uniform Jury Selection and Service Act" (UJSSA), it neither defines "economic status" nor contains sufficient direction to establish a proper test for determining a violation of section 13-71-103.

However, these statutes, read together, support well-established principles of jury service. Jury service is both a duty and a privilege of citizenship. Thiel v. Southern Pac. Co., 328 U.S. 217, 224 (1946). Broad participation in the justice system is desirable because it reinforces public confidence in the system's fairness. See Balzac v. People, 258 U.S. 298, 310 (1922) ("The jury system postulates a conscious duty of participation in the machinery of justice. . . . One of its greatest benefits is in the security it gives the people that they, as jurors actual or possible, being part of the judicial system of the country, can prevent its arbitrary use or abuse."). In addition, jury service provides individuals with an opportunity to participate in the civic life of our nation. See Powers v. Ohio, 499 U.S. 400, 407 (1991) ("[W]ith the exception of voting, for most citizens the honor and privilege of jury duty is their most significant opportunity to participate in the democratic process."). Discrimination during jury selection undermines these important values. Moreover, discrimination deprives individual defendants of a central right in our system of justice, the right to be judged by a jury of their peers. Strauder v. West Virginia, 100 U.S. 303, 308 (1879).

The General Assembly enacted section 13-71-103 to ensure that all qualified citizens have the opportunity to serve as jurors. No doubt the legislature intended to preserve the values of participatory democracy and the public confidence in the jury system by ensuring that individuals would not be excluded from jury service on the basis of invidious criteria.

The defendants claim the exclusion of wage earners from jury service should be evaluated under two grounds: (1) the right to a jury selected from a fair cross-section of the community provided by the Sixth Amendment; and (2) the UJSSA and its counterpart, the Federal Jury Selection Act (FJSA). The defendants claim that the test for establishing a violation under these two grounds is similar. Neither of these tests, the defendants maintain, requires a showing of a discriminatory purpose by the state. The state claims the three-part Batson test adopted by this court in Cerrone II should apply to any claim which implicates either section 13-71-103 or its constitutional underpinnings. Under that analysis, the defendants must establish the existence of purposeful discrimination by the state. We agree with the state.

B

The defendants contend that the test for a violation of section 13-71-103 is essentially the same as that for a Sixth Amendment challenge. We must determine whether a violation of section 13-71-103 should be evaluated under criteria applicable to a Sixth Amendment challenge or a Batson equal protection challenge. A challenge of discriminatory selection of grand juries in state courts may be brought under the Equal Protection Clause of the Fourteenth Amendment. Castaneda v. Partida, 430 U.S. 482, 492 (1977). A traverse or petit jury challenge may be brought under the Fourteenth Amendment or under the fair cross-section requirement of the Sixth Amendment. Taylor v. Louisiana, 419 U.S. 522, 525-26 (1975). However, it is unclear whether a challenge to the composition of a state grand jury may also be brought under the Sixth Amendment. See, e.g., Machetti v. Linahan, 679 F.2d 236, 239 (11th Cir. 1982) (allowing a Sixth Amendment challenge of a state grand jury), cert. denied, 459 U.S. 1127 (1983). But see Castaneda, 430 U.S. at 510 (Powell, J., dissenting) ("[I]n a state case in which the challenge is to the grand jury, only the Fourteenth Amendment applies, and the defendant has the burden of proving a violation of the Equal Protection Clause.").

The fundamental difference between the two challenges is that under an Equal Protection Clause claim based upon Batson, a defendant must prove purposeful discrimination. See Batson v. Kentucky, 476 U.S. 79, 93 (1986); Machetti, 679 F.2d at 241 n. 6. "`Discriminatory purpose' . . . implies more than intent as volition or intent as awareness of consequences. It implies that the decisionmaker . . . selected . . . a particular course of action at least in part `because of,' not merely `in spite of,' its adverse effects upon an identifiable group." Hernandez v. New York, 500 U.S. 352, 360 (1991) (plurality opinion) (citation and footnotes omitted) (quoting Personnel Adm'r v. Feeney, 442 U.S. 256, 279 (1979)).

In Cerrone III, the court of appeals found that the phrase "on account of" in the statute means, among other things, "because of" and that section 13-71-103 implies "affirmative conduct." See Cerrone III, 867 P.2d 143, 146 (Colo.App. 1993); see also supra text accompanying note 8. We agree. Even if a Sixth Amendment challenge is legally possible in the present case, the language of section 13-71-103 implies affirmative conduct and not mere acquiescence to a disproportionately composed grand jury. The implication of affirmative conduct found in section 13-71-103 requires a showing by a defendant of purposeful discrimination by the state. Because the statute requires a defendant to establish purposeful discrimination, a Batson analysis is the appropriate standard for evaluating violations of section 13-71-103.

IV

In Batson, the Court reviewed an Equal Protection challenge to a prosecutor's use of peremptory strikes during selection of a petit jury. The Court developed a three-part analysis for evaluating claims of purposeful discrimination on account of race during the jury selection process. Batson v. Kentucky, 476 U.S. 79, 94, 96. Under this analysis, a defendant first must make a prima facie showing that a purposeful discrimination on account of race occurred during the jury selection process. Id. at 94, 96. Once a prima facie showing has been made, the burden shifts to the state to articulate a neutral explanation for its behavior. Id. at 97. Although the burden of production shifts to the state, the burden of persuasion remains with the defendant. If the state fails to meet its burden of production, the defendant prevails. See Castaneda v. Partida, 430 U.S. 482, 501 (1977). However, once the state articulates a neutral explanation for its jury selection, the trial court must weigh the evidence to determine if the defendant has satisfied his burden of persuasion. Batson, 476 U.S. at 98.

Section 13-71-103 prohibits purposeful discrimination, and the Batson test provides a useful analytic framework with which to evaluate allegations of purposeful discrimination. It must be noted, however, that section 13-71-103 is a legislatively created prohibition on discrimination during jury selection and not a constitutional one. Although utilizing a similar standard, matters regarding a violation of section 13-71-103 and the Equal Protection Clause of the United States Constitution remain separate and distinct.

A

In Cerrone II, we articulated the elements of a prima facie case of purposeful discrimination in jury selection:

Under Batson, all that is required is that the defendant show that (1) the venire in question was selected under a practice providing the opportunity for discrimination, (2) members of a cognizable . . . group were substantially underrepresented on the venire, and (3) the defendant is a member of that underrepresented . . . group.

Id. at 188. We noted in Cerrone II that a defendant is no longer required to show membership in the same group that is underrepresented on the venire. Id.; see also Powers v. Ohio, 499 U.S. 400, 402 (1991). In determining whether the defendant has established a prima facie case of purposeful discrimination, "the trial court must determine whether `the totality of the relevant facts gives rise to an inference of discriminatory purpose.'" Cerrone II, 854 P.2d at 188 (quoting Batson v. Kentucky, 476 U.S. 79, 93-94 (1986)).

In the present case, the defendants must first establish that individuals of a low economic status constitute a "cognizable group." The jury questionnaire did not specifically inquire about the prospective juror's economic status. However, by using the phrase "economic status" in section 13-71-103, the General Assembly intended to refer to a specific group. See Florence v. Board of Waterworks, 793 P.2d 148, 151 (Colo. 1990) ("In interpreting a statute, whenever possible, each word should be given effect and each provision construed in harmony with the overall statutory scheme."). Although a certain economic class may not illustrate a classic example of a cognizable group, the language of the statute shows the General Assembly's intent that discrimination based on economic status not be sanctioned.

The record reflects that members of a particular part of society were excluded. The exclusion was systematic and not random. The defendants extrapolated from the statistical evidence that more persons with lower incomes were excluded than proportionally exist in the general community. The defendants established an underrepresentation of persons with lower education levels. The record shows, based on some correlation between education and income, that the manner of selecting the venire provided the state with an opportunity to discriminate on the basis of economic status. The evidence supports the finding of a prima facie case of discrimination on the basis of economic status.

B

Once a defendant has made a prima facie case of discrimination, the second prong of the Batson analysis requires the state to articulate a non-discriminatory or neutral reason for its jury selection. Batson, 476 U.S. at 97. At this second step in the inquiry, the issue is the facial validity of the state's explanation. The state may not rebut a prima facie case of discrimination through mere denials of a discriminatory motive or protestations of good faith. Id. at 94, 98. However, unless a discriminatory intent is inherent in the state's explanation, the reason offered will satisfy the state's burden of production. See Purkett v. Elem, 115 S.Ct. 1769, 1771 (1995) (per curiam); Hernandez v. New York, 500 U.S. 352, 360 (1991).

In the context of this case, in order to rebut the defendants' prima facie case of discrimination based on economic status, the state was required to articulate an economically neutral explanation for its exclusion of hourly wage earners. In particular the state was required to offer an explanation based on something other than the economic status of the juror.

In this case, the deputy attorney general responsible for coordinating the selection of the statewide grand jury provided reasons for the disproportionate exclusion of hourly wage earners from the venire. The deputy attorney general explained that the state desired jurors who could make the considerable time commitment necessary for grand jury service. The deputy attorney general believed that the time commitment involved would create a greater economic hardship for hourly wage earners than for salaried workers. In particular, the deputy attorney general believed that because of this economic hardship, hourly wage earners would make less stable grand jurors and would be either unwilling or unable to serve on the grand jury. Thus, the selection staff systematically recommended dismissal of hourly wage earners from grand jury service.

As an example, when explaining why a particular person may have been excluded from consideration, the deputy attorney general stated:

This juror's credentials. It indicates she's an employee of Taco John's in Longmont and is in the cooking and food preparation business there. For that reason she's probably paid by the hour and for that reason would not receive compensation from her employer and it's just one of those things that we try to keep this kind of individual from having to serve on [the] Grand Jury.

When explaining why another individual may have been excluded, the deputy attorney general further testified:

It [the questionnaire] indicates he's worked there for a little over two years. That he runs a printing press. That he has a high school degree but no education beyond that. And the advice that would routinely be given on a questionnaire like this is that he's probably employed on an hourly basis, which would make his appearance every Friday difficult for him. That, in combination with the fact he wouldn't appear to have any experience that would be particularly helpful to the Grand Jury, we would have asked — we would have suggested that he be excused from appearance on the impanelment date.

It is clear from the testimony in the record that the deputy attorneys general reviewing the applications systematically rejected individuals who appeared to be hourly wage earners.

The deputy attorneys general and presumably Chief Judge Flowers surmised that hourly wage earners, as a group, would be less able to serve on the grand jury because of their economic status. They believed that the time commitment involved in grand jury service would create a greater economic hardship on hourly wage earners than on salaried workers. Implicit in this belief is an assumption about the economic status of hourly wage earners and the effect that economic status would have on their ability to serve as grand jurors. Thus, the state excluded an economic class from the opportunity to participate in grand jury service based on generalized assumptions. This is inherently discriminatory.

Courts may excuse a potential juror from jury service upon a finding of undue hardship. § 13-71-112(2), 6A C.R.S. (1987). A finding of undue financial burden may constitute an undue hardship. People v. Reese, 670 P.2d 11, 14 (Colo.App. 1983). While it is clear that jurors may be excused from jury service because of actual undue hardship, it does not follow that all hourly wage earners may be excluded from jury service because of a supposed hardship. The state may not make a generalized assumption that all hourly wage earners will undergo too great an economic hardship to be able to serve on a grand jury. Some hourly wage earners might choose to serve despite the economic hardship. Others may be able to adjust their employment hours in order to accommodate grand jury service. Still others may have sources of income in addition to their wages. In this case, many hourly wage earners excluded from service indicated on their jury questionnaires that they did not believe that grand jury service would be a hardship for them. Instead of making an individualized finding of actual hardship for these hourly wage earners, the state assumed they would be unable to serve because of their economic status. To allow the state to exclude individuals on this basis would "breathe life into any latent tendencies to establish the jury as the instrument of the economically and socially privileged." Thiel, 328 U.S. at 224. The result would be that no statewide grand jury would have the benefit of the perspective of someone of a lower economic status. The state's actions were discriminatory and violated the mandate of section 13-71-103.

We recognize that the state in all likelihood was not motivated by animus against hourly wage earners. On the contrary, the state's systematic exclusion of hourly wage earners was most likely motivated by concerns of administrative convenience and an intent to protect hourly wage earners from economic hardship. Despite their well-intentioned motives, the state rejected jurors "on account of" their economic status. This is exactly what section 13-71-103 prohibits. Discrimination may not be justified by administrative convenience. Moreover, the protection may also be seen as paternalistic exclusion. Section 13-71-103 mandates that all individuals have the opportunity to serve on a jury without regard to race, color, religion, sex, national origin, or economic status. Excluding individuals on these bases even with the best of intentions cannot be countenanced.

Because the state has not articulated a non-discriminatory economically neutral explanation for its exclusion of hourly wage earners from the statewide grand jury, it is unnecessary to reach the third stage of the Batson analysis. The state has purposefully discriminated on the basis of economic status and has violated section 13-71-103, 6A C.R.S. (1987).

V

Although we conclude that the state has violated section 13-71-103, the defendants' convictions should not be overturned. We have never articulated the proper standard for reviewing selection of a statewide grand jury under section 13-71-103. Until now, the state was unaware that its grand jury selection practices were illegal. While not ruling on the proper remedy should the state continue to select jurors in a discriminatory manner, it is unnecessary to overturn the defendants' convictions in this case.

Both defendants were afforded the full procedural protections of a trial before a properly constituted petit jury. After their respective trials, each was found guilty beyond a reasonable doubt of one count of pandering and several violations of the Colorado Organized Crime Control Act, section 18-17-104, 8B C.R.S. (1986). Although the grand jury was selected in violation of the law, there is no indication that the defendants suffered any prejudice. In addition, any prejudice resulting from the determination of probable cause underlying the indictments by the improperly constituted grand jury was dispelled when the petit jury found the charges sustained by proof beyond a reasonable doubt. Therefore, the convictions of the defendants should not be overturned.

VI

In conclusion, the state discriminated on the basis of economic status in violation of section 13-71-103, 6A C.R.S. (1987), when it selected the 1985-86 statewide grand jury. Despite this violation, under the circumstances in this case the defendants' convictions should not be overturned. Accordingly, the judgments of conviction in Cerrone III, 867 P.2d 143 (Colo.App. 1993) and Goetz II, No. 90CA0514 (Colo.App. Apr. 14, 1994) (not selected for publication) are affirmed but for reasons different than those set forth by the court of appeals.

JUSTICE SCOTT specially concurs.

JUSTICE VOLLACK does not participate.