This case presents the question whether appellants, who demonstrated in support of more humane treatment of homeless persons by unlawfully refusing to leave a Metro station when ordered to do so, have proffered sufficient evidence to entitle them to discovery and to an evidentiary hearing on their claim of unconstitutional selective prosecution. They claimed they were deemed ineligible for diversion from prosecution as first offenders solely because they exercised their rights protected by the First Amendment. We agree that appellants have proffered sufficient evidence to support their allegation. We therefore reverse and remand for further proceedings.
I.
A. Procedural History
These consolidated appeals arise out of a series of demonstrations at the Farragut West Metro station in November and December of 1987. The demonstrations were held to protest the installation by the Washington Metropolitan Area Transit Authority (Metro or WMATA) of a locked gate designed to bar homeless persons from taking shelter in the station after closing time. The demonstrators positioned themselves in an area near the gate in order to prevent the gate from being closed at closing time. The three appellants, each of whom was a student at George Washington University at the time, participated in one of these demonstrations. Shortly after closing time, the demonstrators were warned that they would be arrested if they remained in the area near the gate. Each appellant refused to leave when so directed, and each was arrested and charged with unlawful entry, in violation of D.C.Code § 22-3102 (1989).
All of the appellants requested admission to the United States Attorney’s pretrial diversion program, which provides an alternative to prosecution for first offenders charged with certain misdemeanors. When an eligible offender is so “diverted,” the United States Attorney dismisses the charges, usually in exchange for the successful completion by the offender of a specified number of hours of community service. The offender must also refrain from the commission of an offense during the diversion period. If the offender successfully completes the diversion program, the stigma of a criminal conviction is avoided.
All of the appellants were eligible for the diversion program. Each was a first offender, and it is undisputed that unlawful entry is a divertable misdemeanor offense. *603Nevertheless, all three were denied admission into the program. Reserving their appellate rights with respect to the denial of diversion, appellants entered pleas of guilty, and each received a suspended sentence and a term of probation. These appeals followed.
B. Ms. Federov and Ms. Donne
The earlier of the cases addressed in these appeals involved Veronica Federov and Stephanie Donne. In proceedings before Judge Salzman, these appellants filed a motion to dismiss the charges against them or, in the alternative, to order the government to admit them to the pretrial diversion program. They argued, among other things, that the prosecutor’s decision not to divert them constituted selective prosecution, denying them liberty without due process of law, in violation of the Fifth Amendment, and punishing them for the exercise of their right to free speech, in violation of the First Amendment. Ms. Federov and Ms. Donne asserted that the United States Attorney had a policy of denying diversion to otherwise eligible persons charged with unlawful entry whose conduct occurred in the context of a political demonstration. They claimed that similarly situated individuals, charged with unlawful entry, whose activities did not arise in the context of the exercise of First Amendment rights had been admitted to the diversion program. In support of their motion, appellants recited the following specific facts: (1) At a pretrial diversion conference, Ms. Deborah Jones of the United States Attorney’s office informed Ms. Donne and her student counsel that diversion had been denied. Ms. Jones provided no explanation but stated that the decision had been made by her supervisor, Ms. Katherine Winfree, Chief of the Misdemeanor Trial Section of the United States Attorney’s office. (2) Ms. Katherine Ellis of the United States Attorney’s office subsequently informed Ms. Donne’s student counsel that a decision had been made not to grant diversion to anyone who had participated in the Farragut West demonstrations. (3) Ms. Jones told the same student counsel that it would be futile to bring Ms. Federov in for a diversion conference because diversion would be denied. (4) Finally, Ms. Winfree informed student counsel for Gregory Kandt, a Farragut West demonstrator who had been charged in a related case which is not before us on this appeal, that her client had been denied diversion because he had engaged in a political demonstration.
Ms. Federov and Ms. Donne also made an extensive discovery request. When the government declined to provide the material sought, their counsel filed a motion to compel discovery. In support of that motion, these appellants alleged that the United States Attorney had singled them out for prosecution because their conduct was related to expressive activity protected by the First Amendment. They based this contention in part on the preliminary results of an investigation, by D.C. Law Students in Court, into the United States Attorney’s disposition of all unlawful entry cases in the District of Columbia in 1985, 1986, and 1987. Counsel represented that the investigation had identified eight persons charged with unlawful entry who had been diverted. None of these eight had been arrested in the context of a political demonstration.
At a hearing on January 28, 1988, Judge Salzman announced in open court that he would deny appellants’ motions to dismiss and to compel discovery, finding them to be “utterly without legal merit.” Counsel for Ms. Federov and Ms. Donne requested permission to make an offer of proof, but Judge Salzman denied their request. On February 4, 1988, Judge Salzman followed up his oral decision with a written opinion in which he formally denied appellants’ motions. He held that, in order to make out a claim of selective prosecution, appellants were required to demonstrate (1) that others “similarly situated” were not prosecuted and (2) that appellants were singled out for prosecution as a result of some improper motivation on the part of the United States Attorney. See (Elizabeth) Smith v. United States, 460 A.2d 576, 578 (D.C.1983) (per curiam). He concluded *604that they had failed to meet their burden in relation to both of these issues.
With respect to the first prong, Judge Salzman construed appellants’ motion as suggesting that they should be considered similarly situated to “all other persons charged with the offense of unlawful entry.” 1 He found that this proposed class of similarly situated persons was “painted with too broad a brush.” In Judge Salzman’s view, the appropriate class of persons with whom appellants should have been compared for selective prosecution purposes consisted of “all demonstrators on the homeless’ behalf at the Farragut Metrorail station.” He reasoned that an individual prosecuted for failing to leave a department store after previously being barred, in lieu of prosecution for shoplifting, was not ‘similarly situated’ with these defendants, even though that individual, too, was charged with unlawful entry. Because appellants had not alleged that they were treated differently from other Farragut West demonstrators, Judge Salzman held that their selective prosecution claims must fail.
Judge Salzman also held that appellants had failed to satisfy the second prong of the test in (Elizabeth) Smith, supra, because they had not shown that the decision to prosecute them was improperly motivated. Citing Adderley v. Florida, 385 U.S. 39, 47-48, 87 S.Ct. 242, 247-48, 17 L.Ed.2d 149 (1966), and Leiss v. United States, 364 A.2d 803, 807-09 (D.C.1976), cert. denied, 430 U.S. 970, 97 S.Ct. 1654, 52 L.Ed.2d 362 (1977), the judge concluded that appellants had made “no colorable showing that the government violated their First Amendment rights by removing them from the closed Metro station.” Acknowledging that appellants’ conduct had involved expressive activity, Judge Salzman found that the government had a legitimate interest in closing the Metro at midnight “for all purposes,” including the conducting of a demonstration. (Emphasis in original.) He concluded that appellants had been arrested and prosecuted for their “willful and intentional refusal to vacate WMATA property” after closing hours and not for their expressive activity.2
Judge Salzman also stated that the United States Attorney's diversion program was reserved for persons who genuinely regretted their illegal conduct and who were willing to comply with an educational program intended to instill respect for the law. He found that appellants were accused of “knowingly violating the law,” that they had intended to be arrested in order to draw attention to their views on the needs of the homeless, and that they had given no indication that they would refrain from such conduct in the future “if they believe it necessary to make another such ‘statement.’ ” He held that “in light of [appellants’] deliberate conduct and their apparent unwillingness to agree not to trespass similarly on Metro property in the future,” it could not be said that the prosecutor’s refusal to admit them to the diversion program was arbitrary or malicious.
Finally, Judge Salzman denied appellants’ motion to compel discovery. He found that they were not entitled to any discovery because they had failed to make out a colorable claim of selective prosecution.
C. Mr. Mellecker
The case of the other demonstrator whose appeal is before us, Dana Mellecker, was assigned to Judge Burnett. Like Ms. Donne and Ms. Federov, Mr. Mellecker alleged that the decision to prosecute him was made pursuant to a policy of denying diversion to all persons charged with unlawful entry whose conduct occurred in the context of the exercise of First Amendment rights. Like his co-appellants, he asked the *605trial judge to dismiss the charges against him or, in the alternative, to order the government to admit him to the diversion program. Judge Burnett held a hearing on the motion on March 29,1988, approximately two months after the decision by Judge Salzman.3
At the hearing on the motion, counsel for Mr. Mellecker proffered the testimony of two witnesses whom he proposed to present in support of his allegation that the United States Attorney had a policy of not considering for diversion otherwise eligible first offenders who were charged with unlawful entry in the context of a political demonstration. The first witness, the student attorney for Gregory Kandt, was said to be prepared to testify that she had been told by the Chief of the Misdemeanor Trial Section that such a policy was in effect. Counsel further represented that this witness was prepared to describe an investigation by D.C. Law Students in Court into the United States Attorney’s disposition of unlawful entry cases in the District of Columbia from January 1, 1985 through January 1, 1988.4 Counsel explained that the students who conducted the investigation had examined the court jacket in “every” unlawful entry case which was not immediately “no-papered” (i.e., dismissed) by the government, in an effort to determine whether first offenders charged with unlawful entry had been treated less favorably where it appeared that their conduct was politically motivated.
The second witness proffered by Mr. Mellecker’s counsel, a professor of statistics at Howard University, was said to be prepared to present the result of his analysis of the data collected by D.C. Law Students in Court. Counsel represented that over a three-year period, not a single political demonstrator charged with unlawful entry had been granted diversion. In contrast, approximately 27% of the persons charged with unlawful entry who were not political demonstrators and who did not have prior records or other pending cases in the District of Columbia were admitted to diversion.
Counsel also represented that the diversion interviews given to her clients were “extremely perfunctory.” Counsel stated, “My clients were never asked a single question. We were told absolutely not. When I asked what is the reason, I was advised to speak with Ms. Winfree, the head of the Misdemeanor Trial [Sjection of the United States Attorney’s office....” Counsel said that she subsequently spoke to Ms. Winfree, who advised her that Mr. Mellecker would not be granted diversion because of the nature of the offense with which he was charged.
The government opposed Mr. Mellecker’s request for an evidentiary hearing on his motion. In response to repeated inquiries from Judge Burnett, the prosecutor would not categorically confirm or deny that the government’s policy was as represented by Mr. Mellecker, although he did claim that diversion decisions were made on an individual basis. Judge Burnett declined to hold an evidentiary hearing, but said he would assume for purposes of ruling on the motion to dismiss that the United States Attorney had a policy against diverting political demonstrators, as Mr. Mellecker alleged. The judge stated that he would consider the “pure legal issue” whether such a policy was constitutional.
On April 14, 1988, Judge Burnett issued a memorandum opinion denying the motion to dismiss. With respect to the selective prosecution claims, he essentially adopted the reasoning of Judge Salzman. He found that appellant’s proposed class— those charged with unlawful entry who were otherwise eligible for pretrial diversion — was too broad. He agreed with Judge Salzman’s finding that the appropriate comparison group for Mr. Mellecker’s selective prosecution claim consisted of per*606sons who, like Mr. Mellecker, had refused to leave the Farragut West Metro station as directed. Since the other members of that group did not receive more favorable treatment than Mr. Mellecker did, Judge Burnett held that the decision to deny diversion to Mr. Mellecker was not based on invidious discrimination. He agreed with Judge Salzman’s analysis and concluded that Mr. Mellecker had been arrested and prosecuted for refusing to leave the Farragut West Metro station after closing time, and not because he had exercised rights protected by the Constitution.
Judge Burnett went on to reject Mr. Mel-lecker’s separate First Amendment claims on similar grounds. He held that appellant had been prosecuted “because of” his failure to leave the station after closing time and not “because of” his protest activities. (Emphasis in original.) He also thought that Mr. Mellecker had made no showing that the United States Attorney’s policy with regard to diversion “is intended to keep the defendants from protesting on behalf of the homeless or that it is in any way based on the content or nature of the message communicated in the demonstration.”
II.
Appellants frankly acknowledge that they have violated a criminal statute. They seek reversal of their convictions not on the basis of a claim of innocence but rather because, they say, other persons who have committed similar crimes have been diverted rather than prosecuted. The burden they must carry to prevail on such a claim is, and should be, a heavy one.
A. Prosecutorial discretion and its limits
Although not limitless, the boundaries of prosecutorial discretion are quite wide. See Oyler v. Boles, 368 U.S. 448, 456, 82 S.Ct. 501, 505-06, 7 L.Ed.2d 446 (1962). Conversely, the scope of judicial review of a decision whether to prosecute is narrow. “The power to decide when to investigate, and when to prosecute, lies at the core of the Executive’s duty to see to the faithful execution of the laws; when reviewing the exercise of that power, the judicial authority is, therefore, at its most limited.” Community for Creative Non-Violence v. Pierce, 252 U.S.App.D.C. 37, 39, 786 F.2d 1199, 1201 (1986). As the Supreme Court explained in Wayte v. United States, 470 U.S. 598, 607, 105 S.Ct. 1524, 1530-31, 84 L.Ed.2d 547 (1985):
This broad discretion rests largely on the recognition that the decision to prosecute is particularly ill-suited to judicial review. Such factors as the strength of the case, the Government’s enforcement priorities, and the case’s relationship to the Government’s overall enforcement plan are not readily susceptible to the kind of analysis the courts are competent to undertake. Judicial supervision in this area, moreover, entails systemic costs of particular concern. Examining the basis of a prosecution delays the criminal proceeding, threatens to chill law enforcement by subjecting the prosecutor’s motive and decisionmaking to outside inquiry, and may undermine prosecutorial effectiveness by revealing the Government’s enforcement policy.
Because the discretion conferred upon the prosecutor is broad, a defendant who alleges discriminatory enforcement of a valid statute “carries a heavy burden of proof.” (Elizabeth) Smith, supra, 460 A.2d at 578. The defendant is obliged to make an initial showing that he or she has been singled out for prosecution among others similarly situated, and that his or her prosecution was improperly motivated, i.e., that it was based upon an impermissible consideration, such as race, religion, or the exercise of constitutional rights. Id.; Davis v. United States, 390 A.2d 976, 980 (D.C.1978); see also United States v. Berrios, 501 F.2d 1207, 1211 (2d Cir.1974). To support a defense of selective prosecution, both of these propositions must be clearly and separately established. Attorney General of the United States v. The Irish People, Inc., 221 U.S.App.D.C. 406, 410, 684 F.2d 928, 932 (1982), cert. denied, 459 U.S. 1172, 103 S.Ct. 817, 74 L.Ed.2d 1015 (1983). This test is a “rigorous” one, Unit*607ed States v. Mangieri, 224 U.S.App.D.C. 295, 298, 694 F.2d 1270, 1273 (1982), for “[t]he presumption is always that a prosecution for violation of a criminal law is undertaken in good faith and in a nondiscriminatory fashion for the purpose of fulfilling a duty to bring violators to justice.” United States v. Falk, 479 F.2d 616, 620 (7th Cir.1973) (en banc).
The United States Attorney’s broad discretion as to who should be prosecuted and who should not applies with full force to management of the pretrial diversion program. See Chavez v. United States, 499 A.2d 813, 814 (D.C.1985); Baxter v. United States, 483 A.2d 1170, 1172 (D.C.1984). Nevertheless, there are limits to prosecutorial discretion. As the Supreme Court aptly put it many years ago, the law may not be enforced “with an evil eye and an unequal hand.” Yick Wo v. Hopkins, 118 U.S. 356, 373-74, 6 S.Ct. 1064, 1073, 30 L.Ed. 220 (1886). The prosecutor’s decision whether to grant diversion is subject to judicial review for conformity with equal protection principles, and a deni al of diversion may not be upheld if it “has been based upon some form of invidious or otherwise impermissible form of discrimination, or is arbitrary and capricious.” Irby v. United States, 464 A.2d 136, 141 (D.C.1983) (citing United States v. (James) Smith, 354 A.2d 510, 512-13 (D.C.1976)); see also Chavez, supra; Baxter, supra.
We think it is indisputable that it would not be a proper exercise of discretion for the United States Attorney to prosecute all black participants in a demonstration while admitting their white counterparts to diversion; nor could the United States Attorney lawfully draw such a distinction between men and women. Similarly, although the point is perhaps less obvious, the government may not draw invidious distinctions between those who exercise rights protected by the First Amendment and those who do not; specifically, it may not treat the former more harshly than the latter. See Falk, supra, 479 F.2d at 619-20. This court has held that “ ‘[a] policy which results in the closing of the Rotunda when people exercise rights protected by the First Amendment, but [which results] in the Rotunda remaining open when such rights are not being exercised, penalizes and chills constitutionally protected activity....’” Wheelock v. United States, 552 A.2d 503, 509-10 (D.C.1988) (quoting United States v. Murphy, 114 Daily Wash.L. Rptr. 2149, 2158 (October 20, 1986)); accord United States v. Nicholson, 97 Daily Wash.L.Rptr. 1213 (July 17, 1969), aff'd, 263 A.2d 56 (D.C.1970).5 This reasoning applies in equal measure to a policy under which those charged with unlawful entry who are seeking to exercise the right to free speech are prosecuted while those who are not seeking to exercise that right, receive pretrial diversion.
B. The threshold necessary for entitlement to discovery and to an eviden-tiary hearing
The test for evaluating selective prosecution claims is well-established. A defendant claiming selective prosecution must show that other persons similarly situated were not prosecuted and that the decision to single out the defendant for prosecution was improperly motivated. (Elizabeth) Smith, supra, 460 A.2d at 578; see also Wayte, supra, 470 U.S. at 608, 105 S.Ct. at 1531 (to sustain a claim of selective prosecution, defendant must show that prosecutor’s enforcement policies had both a discriminatory effect and a discriminatory purpose); Reiss, Prosecutorial Intent in Constitutional Criminal Procedure, 135 U.Pa.L.Rev. 1365, 1368-74 (1987). However, a defendant is not required to prove each element of the claim conclusively in order to obtain an evidentiary hearing or to be entitled to discovery. On the other hand, the precise showing to be made is not quite clear from the cases.6 In Berrios, *608supra, 501 F.2d at 1211, a ease which has been “widely followed,” Reiss, supra, 135 U.Pa.L.Rev. at 1370 & n. 15 and cases there cited, the court stated that while the defendant must establish a prima facie showing of selective prosecution in order to prevail on that defense, only a “colorable basis” for the claim must be shown before being entitled to discovery with respect to this defense. In (Elizabeth) Smith, supra, 460 A.2d at 579, however, this court sustained the trial judge’s refusal to grant discovery (even though she had held an evidentiary hearing) on the ground that the defendants failed to establish a prima facie case of selective prosecution. This is less than a clear holding that a prima facie case is required, for the court relied largely on Irish People, supra, 221 U.S.App.D.C. at 410 & n. 8, 684 F.2d at 932 & n. 8, in which the United States Court of Appeals for the District of Columbia Circuit opined that a “colorable showing” is sufficient to obtain discovery.
Except to note that the threshold should not be set too low, we need not resolve in this case the precise contours of the showing which a defendant must make to obtain discovery or an evidentiary hearing on a claim of selective prosecution. Assuming, as the government contends, that a prima facie showing is required, we are satisfied that appellants have made one and that they are therefore entitled both to reasonable discovery and to an evidentiary hearing.
C. Selecting the relevant group for comparison: Who is “similarly situated”?
As we have noted above, appellants offered to prove by statistical evidence, by their own experiences, and by extrajudicial statements of a comparatively high-ranking official of the United States Attorney’s office, that they received less favorable treatment than others similarly situated because their trespass was connected with their exercise of First Amendment rights. The trial judges refused to consider this evidence on the ground, among others, that appellants’ proposed class was overbroad. They held that the appropriate class for selective prosecution purposes consisted of those persons who had participated in a demonstration at the Farragut West Metro station. We conclude, however, that this analysis begs the question. By so narrowing appellants’ proposed class, the trial judges avoided consideration of the critical issue in this ease.
As one commentator has noted, the first prong of the selective prosecution test is “reasonably straightforward.” Reiss, supra, 135 U.Pa.L.Rev. at 1371. To satisfy this prong, a prosecuted defendant must show that others “similarly situated” were not prosecuted. Offenders who have not been prosecuted will generally be regarded as similarly situated where (1) they have engaged in “essentially” the same conduct as the defendant; (2) “as a result of which” they have been accused of violating the same statutes; and (3) the magnitude of the nonprosecuted offenders’ conduct was not “materially different” from that of the defendant. Id. at 1371 n. 19 (citations omitted). If the first prong is satisfied, the defendant may then move on to the second prong and attempt to prove he or she was prosecuted for an impermissible reason. Id. at 1372.
Here, appellants have offered to prove that they are similarly situated to other persons charged with unlawful entry who were found to be eligible for diversion. All such persons were alleged to have trespassed on property not their own, all were consequently charged with violating the unlawful entry statute, and all were first *609offenders. Those other persons differed from appellants in at least one crucial respect: their alleged conduct was not associated with the expression of views protected by the First Amendment. However, such a distinction between persons accused of the identical statutory offense of unlawful entry cannot render such persons “dissimilarly situated” for selective prosecution purposes. To conclude otherwise is to maintain that persons whose trespass had First Amendment ramifications have, as a result of that fact alone, committed a crime of greater “magnitude” than persons whose trespass was not politically motivated, a proposition which runs afoul of Wheelock, Murphy, and Nicholson, supra. To proclaim, without analysis, that persons charged with unlawful entry whose conduct was associated with expressive activity are intrinsically more culpable than those whose conduct was unrelated to First Amendment concerns assumes the critical conclusion on which the outcome of these cases depends.
In holding that an individual prosecuted for failing to leave a department store after previously being barred, in lieu of prosecution for shoplifting, is not similarly situated to Ms. Federov and Ms. Donne, Judge Salzman apparently found it to be self-evident that the political context of the appellants’ conduct rendered appellants different for purposes of selective prosecution analysis from other persons charged with unlawful entry. But department store first offenders are similarly situated to appellants in relation to the kind of wrong that they have done; they differ from appellants in that their conduct is not related to the exercise of First Amendment rights. The issue whether the United States Attorney may administer the pretrial diversion program in a manner that discriminates against persons whose conduct was so related must be confronted directly. It may not be avoided by defining those exercising First Amendment rights out of the class of first offenders who have been charged with unlawful entry and who are thus eligible for diversion.
The trial judges focused their inquiry on whether there was disparate treatment within the class of alleged individual victims of discrimination, rather than on whether those victims, taken as a class, were treated less favorably than others who were similarly situated except for the exercise of protected rights. To use an imperfect but revealing analogy, under the analysis used by the trial judges, a claim of racial disparities between a school in a predominantly white neighborhood and one in a predominantly black neighborhood would be resolved without comparing the opportunities available, respectively, to the students in each of the two schools. Rather, the court would confine itself to an inquiry as to whether there was discrimination among the students at the black school.
At oral argument, however, counsel for the United States suggested for the first time that the exercise of First Amendment rights, as such, may not have been the only basis for distinguishing appellants from other unlawful entrants who were found eligible for the diversion program. Counsel argued, for example, that diversion may have been denied in these and other demonstration cases because such cases receive more publicity than other prosecutions for unlawful entry, and because admission of demonstrators to diversion would impede a policy of deterrence by creating a public perception that unlawful entry is not a serious offense and will be punished only by a slap on the wrist. Cf. Wayte, supra, 470 U.S. at 613, 105 S.Ct. at 1533-34. It may also be that demonstrations, which often involve concerted action by significant numbers of persons, require a greater police presence and are more disruptive than other varieties of unlawful entry, and that it is this aspect of such activities, rather than their expressive character, which has led the United States Attorney to treat demonstrators differently.
Wayte certainly suggests that such possibilities could foreclose appellants from meeting the second, “improper motivation” prong of their burden to show selective prosecution. However, in view of the importance of First Amendment rights and of appellants’ proffered evidence of an extrajudicial admission by an Assistant United *610States Attorney that diversion had been denied automatically because appellants were political demonstrators, we are unpersuaded that these other possible but unproved explanations first asserted on appeal can automatically defeat the required showing that appellants are “similarly situated” with other offenders who were offered diversion. Establishment of these alternative explanations for the government’s administration of its diversion policy is a matter for discovery and an evidentiary hearing, not for resolution on appeal.
Our approach to this issue — defining “similarly situated” by reference to all first offenders charged with unlawful entry— follows from the Supreme Court’s decision in Wayte. In that case, the defendant was indicted for failing to register for selective service. He moved to dismiss the indictment on the ground, among others, that he was being selectively prosecuted. He alleged that “vocal” nonregistrants had been singled out for prosecution on the basis of their exercise of First Amendment rights. Id. at 604, 105 S.Ct. at 1528-29. In analyzing Wayte’s selective prosecution claim, the Court treated all nonregistrants as similarly situated, even though the class consisted of an estimated 674,000 individuals.7 The Court determined that “the prosecution pool consisted of all reported nonregis-trants, not just ‘vocal’ nonregistrants.” Id. at 609 n. 10, 105 S.Ct. at 1531-32 n. 10. It rejected the selective prosecution claim, however, on the ground that petitioner “ha[d] not shown that the enforcement policy selected nonregistrants for prosecution on the basis of their speech.” Id. at 609, 105 S.Ct. at 1531. Rather, reported nonre-gistrants had been prosecuted largely because they were the easiest violators to identify under a passive enforcement policy. Id. at 612-13, 105 S.Ct. at 1533-34.
If the Court in Wayte had accepted an analysis of the kind used by the trial judges here, it could conceivably have disposed of the selective prosecution claim by dividing the class of nonregistrants into sub-classes. Indeed, the Court’s opinion reveals that the individuals who had failed to register were anything but a homogeneous group. Some of them had actively protested against the Selective Service laws. Some had submitted letters to the government declaring their refusal to register. Some had been reported by others. Some had neither reported themselves nor been reported by anyone else. Eschewing subclassifications predicated on these dissimilarities, however, the Court treated all those who refused to register as members of a single class and considered whether those selected from the class for prosecution were selected on an impermissible basis. A similar approach should have been followed here.
D. The sufficiency of the evidence
In substance, appellants asserted that each of them was denied diversion with little or no inquiry into his or her personal circumstances. As to discriminatory effect, they claimed to be ready to prove that no political demonstrator charged with unlawful entry had received diversion during the period of the D.C. Law Students in Court study, whereas a substantial number of other alleged first offenders of the unlawful entry statute had been diverted. As to discriminatory purpose, they asserted that the Chief of the Misdemeanor Trial Branch — the official in the United States Attorney’s office to whom they were referred as the supervisor in charge — had acknowledged that political demonstrators do not receive diversion, evidently without regard to their individual circumstances.
In assessing the sufficiency of the evidence, at least as to discriminatory effect upon the exercise of rights protected by the First Amendment, the modes of analysis used by courts seeking to evaluate alle*611gations of discrimination based on race are instructive. “ ‘In the problem of racial discrimination, statistics often tell much, and [e]ourts listen.’ ” Harris v. District of Columbia Comm’n on Human Rights, 562 A.2d 625, 632 (D.C.1989) (quoting Alabama v. United States, 304 F.2d 583, 586 (5th Cir.), aff'd, 371 U.S. 37, 83 S.Ct. 145, 9 L.Ed.2d 112 (1962)). “[N]othing is as emphatic as zero.” United States v. Hinds County School Bd., 417 F.2d 852, 858 (5th Cir.1969), cert. denied, 396 U.S. 1032, 90 S.Ct. 612, 24 L.Ed.2d 531 (1970). The zero in the present case is especially meaningful if one considers that appellants, who were young college students, were allegedly denied diversion without any appreciable inquiry into their particular circumstances. According to the D.C. Law Students in Court study, by contrast, more than a quarter of the eligible defendants in unlawful entry cases who were not engaged in expressive activity had been admitted to diversion.
As to discriminatory purpose, appellants have proffered an extrajudicial admission by the responsible government official which, if credited, appears to explain the zero as intentional discrimination against political demonstrators. Such admissions, for example, have been powerful evidence in cases of alleged racial discrimination. See, e.g., United States v. Real Estate Dev. Corp., 347 F.Supp. 776, 783, 785 (N.D.Miss. 1972) (“statements which come close to admissions of a [racially] discriminatory policy, especially when bolstered by explicit admissions, are persuasive evidence of discrimination”). But the extrajudicial admission reflects a discriminatory purpose only if it shows that the government prosecuted appellants, rather than considered them for the diversion program, “because of [their] protest activities.” Wayte, supra, 470 U.S. at 610, 105 S.Ct. at 1532 (emphasis in original). A discriminatory purpose in selective prosecution cases cannot be established merely by showing the government was aware that administration of its diversion policy would result in prosecutions, not diversions, of all political demonstrators charged for the first time with unlawful entry. See id.
In this case it is clear that the Wayte purpose test is met. Appellants have presented a prima facie case that, in contrast with all other unlawful entrants, the government intended to punish all demonstrators, whatever their collective message on any particular occasion. This apparently purposeful chill on First Amendment rights ostensibly violates the Constitution, see Wheelock, supra, and is a sufficient showing of discriminatory purpose to establish a prima facie ease of selective prosecution under Wayte. It remains for the government, in response to discovery or at a hearing, to establish that any discriminatory purpose was not aimed at political discourse. Such a showing would rebut not only the claim of unlawful discriminatory purpose but also the claim that appellants are “similarly situated” with other alleged offenders deemed eligible for the diversion program. But, on this record and absent discovery, we cannot say appellants have failed to make a sufficient showing.
It bears noting, for purposes of the proceedings on remand, that the government’s burden to show a nonpolitical discourse focus in the exercise of prosecutorial discretion is not necessarily heavy here. If, as the government contended for the first time at oral argument on appeal, the policy against diverting any first-time political demonstrator was motivated by a concern about public disruption or about the need to use large numbers of police — and not about the exercise of First Amendment rights — the government could respond to discovery directed at motivation and, by appropriate motion, obtain a court ruling on the true purpose underlying the government’s policy. The trial court can control discovery and limit a hearing, as needed, to avoid undue intrusion into prosecutorial discretion. In sum, the problem here is that the government has presented no evidence, aside from counsel’s oral argument on appeal, which would justify an inference, in rebuttal of appellants’ prima facie case, that the motivation underlying a refusal to make diversion available for political demonstrators was not directed at First Amendment rights.
*612 E. Other Issues
Both trial judges apparently accepted the view that, because appellants had violated the law, the communicative nature of their conduct did not immunize them from prosecution. We have no quarrel with this proposition in the abstract, see United States v. O’Brien, 391 U.S. 367, 88 S.Ct. 1673, 20 L.Ed.2d 672 (1968) but it does not have application here. The issue is not whether appellants violated the law and thus rendered themselves subject to arrest and prosecution. They have never contended that their participation in a political demonstration precluded application of the unlawful entry statute to their conduct. Appellants claim only that the decision to prosecute them for their acknowledged violation of the law, rather than to grant them diversion, was in reprisal for their exercise of constitutional rights. Although the trial judges correctly recognized that appellants’ First Amendment activities did not immunize them from arrest and prosecution, those . activities cannot legitimately serve as a basis for treating appellants more severely than those persons charged with unlawful entry whose conduct did not implicate the First Amendment.
Judge Salzman also found that the United States Attorney's diversion program is reserved for persons who genuinely regret their unlawful conduct and who are receptive to rehabilitation efforts. He held that Ms. Donne and Ms. Federov did not meet these criteria because they had knowingly violated the law and had shown no remorse for their actions. This holding was, at best, premature. The judge’s assumption that the United States Attorney grants diversion only to those accused defendants who are remorseful was based on two opinions of this court which described the goals of the diversion program. See Williams v. United States, 427 A.2d 901, 903-04 (D.C. 1980), cert. denied, 450 U.S. 1043, 101 S.Ct. 1763, 68 L.Ed.2d 241 (1981); (James D.) Smith, supra, 354 A.2d at 511. These opinions, however, do not purport to describe how the United States Attorney actually administers the diversion program.8 It is the prosecutor’s implementation of the program, rather than its formulation, which is at issue here. Ironically, in denying appellants’ offer of proof, Judge Salzman precluded any inquiry into whether appellants’ actual or supposed lack of contrition played any part in the prosecutor’s decision. We do not suggest that the United States Attorney may not consider remorse or the lack thereof in determining appellants’ eligibility for diversion, if this is the general practice. The record, however, is barren of any evidence that this is what occurred here.
According to defense counsel’s offer of proof, there was no inquiry made by Ms. Winfree or by her subordinates as to whether any appellant was remorseful. The United States Attorney could surely make such a determination only on a case-by-case basis. Appellants offered to show, however, that there was a blanket rule denying diversion to all defendants whose arrests for unlawful entry arose in connection with activities related to the exercise of First Amendment rights. It appears that Judge Salzman simply assumed that Ms. Donne and Ms. Federov were not remorseful because they were political demonstrators. We discern no factual basis for making such an assumption with respect to them or to all political demonstrators as a group. A college student who has not previously engaged in acts of civil disobedience, for example, might well regret his or her involvement in an unlawful demonstration, while a seasoned political activist might not.
Judge Burnett predicated his decision, in part, on the ground that Mr. Mellecker had failed to show that the United States Attorney’s policy with respect to diversion was based on the content of the message the demonstrators and others similarly situated were attempting to communicate. They do not claim, however, that the diversion policy as implemented gives preference to one viewpoint over another. Rather, they say *613that the United States Attorney discrimina-torily denies diversion to all demonstrators who trespass in order to convey a First Amendment message, no matter what that message may be, while providing more favorable treatment to trespassers who have no message at all. This kind of discrimination is also proscribed. Just as the government may not pursue an otherwise valid regulatory interest in a manner that discriminates against a particular viewpoint or message, see, e.g., Carey v. Brown, 447 U.S. 455, 460-63, 470-71, 100 S.Ct. 2286, 2289-91, 2295-96, 65 L.Ed.2d 263 (1980); Erznoznik v. City of Jacksonville, 422 U.S. 205, 215, 95 S.Ct. 2268, 2275-76, 45 L.Ed.2d 125 (1975), so it may not prosecute unlawful entry cases in a manner which discriminates between speakers and non-speakers. See Wheelock, supra.9
F. The Dissent
Judge Schwelb argues in dissent that appellants have not made a prima facie showing that they are similarly situated with other unlawful entrants who have received diversion and, further, that the have not made a prima facie showing of discriminatory purpose. We find neither argument persuasive. Judge Schwelb agrees that “the trial judges were fatally under-inclusive when they defined the appropriate group, for purposes of comparison, as consisting solely of participants in the demonstrations at the Farragut West Metro station.” Post at 613-14. But he is only a tad more inclusive; he perceives only political demonstrators as such — Farragut West and otherwise — as the similarly situated group. See post at 615-16. He does so for two reasons. First, demonstrators, in contrast with other unlawful entrants, pose a more significant law enforcement problem; they draw the media and require more police. See post at 616-17. Second, “unlawful entry prosecutions are a varied lot,” post at 617, in contrast (as he sees it) with the group of nonregistrants for the draft in Wayte.
As to the first, we do not understand why, after appellants have made a prima facie showing of non-diversion of political demonstrators, this court should decide that the reason the prosecutor refused to divert any political demonstrator was a desire to discourage crowds rather than political expression. The United States Attorney’s office has never said so. Moreover, as Judge Schwelb himself notes, “appellants’ conduct differed from most ‘non-demonstrator’ cases in that they were exercising rights protected by the First Amendment.” Post at 616. If the prosecuting authorities had a reason for not diverting political demonstrators altogether unrelated to political expression, then the policy could have been expressed in a way that does not implicate expression, and the group automatically precluded from diversion could have been defined differently. As it is, for purposes of this appeal, the government has been willing to have the class of unlawful entrants at issue here defined as political demonstrators — a description that inherently describes persons actively conveying a message. The class is not defined differently as a crowd requiring more police resources than the usual unlawful entry case.
Nor do we find significance in Judge Schwelb’s effort to distinguish unlawful entrants as a group from non-registrants for the draft on the ground that the former are “varied” while the latter are deemed monolithic. Whether one violates the unlawful entry statute or a draft registration statute, one violates a particular command of the law. Judge Schwelb himself, while arguing that “this court’s decisions in unlawful entry cases reveal more variety than similarity between them,” post at 615, also acknowledges the variety of situations in which men have refused to register for the draft; he agrees that the contexts and *614styles for refusing to register differed from person to person. See post at 617 n. 6. In short, the ways in which both statutes have been violated are varied.
What we believe Judge Schwelb overlooks is appellants’ prima facie case, shifting the burden of production to the government to show — without basing that showing on the exercise of First Amendment rights — that appellants’ conduct is not essentially the same as other unlawful entrants who were eligible for diversion and that appellants’ conduct was of a different magnitude.
As to discriminatory purpose, we reject once again Judge Schwelb’s distinction. He agrees that appellants have presented “a prima facie case that demonstrators are not considered for diversion simply because they are demonstrators.” Post at 619. And yet he finds no “basis for concluding that appellants were denied diversion because they had a message.” Post at 619 (emphasis in original). That is a non sequitur if one accepts, as we do, that by definition a “demonstrator” in the context here — a political demonstrator — is one who seeks to convey a political message. On this premise, appellants have presented a prima facie case that persons arrested for unlawful entry while seeking to convey a political message are not considered for diversion because they sought to convey that message.
Accordingly, we remand the cases to the trial courts with directions to permit appellants to conduct reasonable discovery and to conduct evidentiary hearings on their claims of selective prosecution and First Amendment violation.10
Reversed and Remanded.
. In fact, appellants had argued that they should be regarded as similarly situated with other persons who were charged with unlawful entry and who were first offenders and hence eligible for diversion.
. Judge Salzman held that appellants’ failure to show a violation of their First Amendment rights was fatal to the second prong of their selective prosecution claims. He apparently, and reasonably, viewed their First and Fifth Amendment claims as coextensive in this respect.
. Mr. Mellecker’s motion was heard in conjunction with that of another Farragut West demonstrator, Lauren Fischel. Ms. Fischel has since moved to withdraw her appeal.
. As noted earlier, the preliminary results of this investigation had been proffered by counsel for Ms. Donne and Ms. Federov in their motion to compel discovery. By the time of the hearing before Judge Burnett, the investigation had apparently been completed.
. Nicholson is appended to Dellums v. Powell, 184 U.S.App.D.C. 275, 566 F.2d 167 (1977), cert. denied, 438 U.S. 916, 98 S.Ct. 3146, 57 L.Ed.2d 1161 (1978).
. One might reasonably conclude that the test for granting discovery on a claim of selective prosecution is somewhat less stringent — and certainly no more stringent — than the test for granting an evidentiary hearing. See, e.g., Unit*608ed States v. Schmucker, 815 F.2d 413, 418-19 (6th Cir.1987) (defendant may be entitled to discovery upon introducing some evidence tending to show the existence of the essential elements of the defense, but must establish prima facie case to gain entitlement to evidentiary hearing); United States v. Kerley, 787 F.2d 1147, 1150 (7th Cir.1986) (defendant need not make out prima facie case to obtain discovery but need only show a colorable basis for the claim); but see Wayte, supra, 470 U.S. at 621 n. 1, 105 S.Ct. at 1538 n. 1 (Marshall, J. dissenting) (discussing unresolved conflict in the federal circuits about standards governing decision whether to grant discovery); Reiss, supra, 135 U.Pa.L. Rev. at 1373 n. 31 (same).
. Of 674,000 nonregistrants, only the names of those men "who [explicitly] advised [the Selective Service] that they had failed to register or who were reported by others as failing to register," Wayte, supra, 470 U.S. at 601, 105 S.Ct. at 1527, were subjected to enforcement procedures. The Selective Service eventually referred the names of 286 men to the Department of Justice for prosecution. Id. at 604 n. 3, 105 S.Ct. at 1529 n. 3. Sixteen of these 286 were indicted. Id.
. Williams simply quotes from (James) Smith, which is now thirteen years old. The record in the present case does not reveal whether the court's observations more than thirteen years ago correctly describe current practice.
. Decisions such as Regan v. Taxation with Representation of Washington, 461 U.S. 540, 546, 103 S.Ct. 1997, 2001, 76 L.Ed.2d 129 (1983), and Cammarano v. United States, 358 U.S. 498, 513, 79 S.Ct. 524, 533-34, 3 L.Ed.2d 462 (1959), do not require a different result. These cases stand for the proposition that the government is not constitutionally required to subsidize lobbying or similar activities. The present case involves allegedly harsher treatment in the criminal justice process for those who seek to exercise constitutionally protected rights.
. Because the selective prosecution claims in these cases are based on alleged discrimination against those who assert rights secured by the First Amendment, they are closely related to appellants' other claims directly attributable to the First Amendment, and we therefore conclude that no separate discussion of the latter is required.