concurring.
I cannot agree that the Board’s disqualification of the signatures gathered by all the circulators affiliated with the Stars & Stripes operation can be squared with the demands of the First Amendment or brought within our decided cases. The Board’s ultimate determination is nonetheless sustainable on a narrower ground subsumed in the Board’s opinion.
Notwithstanding generalized allegations, the evidence of specific misconduct was limited to a third of the seventy-nine circu-lators affiliated with Stars & Stripes. The Board had testimonial evidence of wrongdoing by thirteen circulators,1 and documentary evidence of wrongdoing by another eight circulators.2 The Board could also properly draw inferences of wrongdoing as to six additional circulators who were subject to allegations of wrongdoing, served with subpoenas, and did not appear.3 Because of the reliance placed on the circulators’ affidavits to ensure the validity of the signatures, the Board was justified in excluding all the signatures gathered by these twenty-seven circulators whose affidavits were fraudulent or facially unreliable.4 See Williams v. District of *834Columbia Bd. of Elections & Ethics, 804 A.2d 316, 318 (D.C.2002) (per curiam). In Williams we upheld the Board’s decision to disqualify all the signatures gathered by three circulators based on the Board’s finding that they had engaged in pervasive fraud, as shown by substantial evidence that was specific to those three circulators: errors and forgeries evident on the face of the petitions themselves as well as an inference from the circulators’ decision not to appear after they were challenged and served with subpoenas. See id. at 320. Thus, we held that the Board “could disallow all of the signatures affected by the wrongdoing.” Id. at 318 (emphasis added). In contrast, there is no comparable evidence here to disqualify signatures gathered by fifty-one of the seventy-nine circulators affiliated with Stars & Stripes.5 “At most” there is “some suspicion” that perhaps some of these circulators might have engaged in wrongdoing — associative suspicion which the Board has previously found too tenuous to disqualify presumptively valid votes, even when the persons suspected of wrongdoing were summoned, served, and did not appear. Allen v. District of Columbia Bd. of Elections & Ethics, 663 A.2d 489, 496 (D.C.1995) (noting presumption that voters’ affidavit of residence is truthful and refusing to draw negative inference from failure to appear in response to subpoena because “there might be a variety of reasons for not attending, and the notion that a voter stayed away because he or she had no response to petitioners’ allegations is altogether speculative.”). The Board’s “totality of the circumstances” rationale to discredit the affidavits of all circulators merely because they were affiliated with Stars & Stripes stretches Williams past the breaking point and is not supported by substantial evidence. See Williams, 804 A.2d at 318 (noting that court must accept Board’s findings of fact if supported by substantial evidence).
The most concrete example that the Board’s sweeping remedy of disqualifying all Stars & Stripes circulators does not reasonably flow from the evidence is provided by the evidence of record concerning two D.C. residents, Bobbie Diggs and Margol Inabinet, who participated in the Stars & Stripes petition drive. The Board credited their testimony and found that they had substantially complied with the statutory requirement that signatures be gathered “in the presence” of a D.C. resident. D.C.Code § l-1001.16(h)(3). Yet the Board’s decision invalidated the 890 signatures they gathered.6 By focusing on deficiencies and irregularities in the Stars & Stripes operation overall (and it appears it was far from exemplary), the Board shifted its attention away from the “signatures affected by the wrongdoing” of particular circulators — the basis for our affirmance in Williams — and arbitrarily struck signatures gathered by circulators it found to be reliable.7 804 A.2d at 318. As we have *835previously emphasized, “consistent with the overall tenor of the Initiative Act which prevents ‘harmless error’ in the signature collection process from vitiating the validity of the petitions ... the paramount concern is with the validity of the signatures.” Dankman, swpra note 4, 443 A.2d at 515. The Board undoubtedly has constitutional space and statutory authority to oversee the petitioning process to ensure its integrity, see Buckley v. American Constitutional Law Found., 525 U.S. 182, 191 n. 10, 119 S.Ct. 636, 142 L.Ed.2d 599 (1999), but the overarching interest lies in protecting the franchise of those D.C. residents who signed the petitions. Because this is “core political speech” protected by the First Amendment, any remedy that infringes upon it must be narrowly tailored to the evidence of wrongdoing. See id. at 192, 119 S.Ct. 636. Where First Amendment concerns are implicated we “must make an independent examination of the whole record in order to make sure that the judgment does not constitute a forbidden intrusion on the field of free expression.” Guilford Transp. Indus, Inc. v. Wilner, 760 A.2d 580, 592 (D.C.2000). Although the Board expressly recognized the important First Amendment rights at stake, as I have discussed its chosen remedy went beyond what was supported by the evidence — indeed beyond what was necessary to decide the precise question before it. Particularly where the signatures collected do not decide an election, but merely determine whether the issue is to be presented to the full electorate for a vote, the First Amendment balance should be struck in favor of speech. See Citizens Against Legalized Gambling v. District of Columbia Bd. of Elections & Ethics, 501 F.Supp. 786, 789 (D.D.C.1980). In the words of Justice Brandéis, “If there be time to expose through discussion the falsehood and fallacies, to avert the evil by the processes of education, the remedy to be applied is more speech, not enforced silence.” Whitney v. California, 274 U.S. 357, 377, 47 S.Ct. 641, 71 L.Ed. 1095 (1927) (Brandéis, J., concurring). The Board’s overly-broad remedy to address what it *836sincerely perceived to be serious shortcomings in the way the petition circulators were managed, placed greater emphasis on regulating the process than in ascertaining whether there were the requisite number of signatures to allow the initiative to enter the sphere of public debate.
My conclusion that the evidence does not support the Board’s decision to invalidate all the signatures gathered by circula-tors identified with Stars & Stripes, does not mean, however, that we cannot sustain its ultimate determination that petitioner failed to gather the number of signatures necessary to place the slots initiative on the November ballot. As my colleagues in the majority recognize, there is substantial evidence in the record and analysis in the Board’s opinion to justify affirmance on a narrower basis. See Securities & Exchange Comm’n v. Chenery, 318 U.S. 80, 95, 63 S.Ct. 454, 87 L.Ed. 626 (1943) (holding that “an administrative order cannot be upheld unless the grounds upon which the agency acted in exercising its powers were those upon which its action can be sustained.”). The Board’s Clarification Memorandum Opinion following remand notes that there were documented irregularities in the affidavits of thirteen circula-tors who were not linked to Stars & Stripes.8 Because the decisive inquiry consistent with the First Amendment is not the quality (or lack thereof) in the training and supervision provided by the umbrella organization, but the validity of the signatures as attested by reliable cir-culators’ affidavits, the 996 signatures gathered by those thirteen circulators— whatever their affiliation — should have been excluded by the Board. This is not a matter for administrative discretion, as the Board must be satisfied that the proffered signatures are valid either by reliance on the circulators’ affidavits or through its own random and statistical sampling. See D.C.Code § 1 — 1001.16(h) & (o)(l). When these 996 signatures are added to the 3186 signatures gathered by Stars & Stripes circulators as to whom there was sufficient evidence of wrongdoing on the record, see footnotes 1, 2 & 3, supra, the number of signatures fell to 17,097, below the requisite 17,599 signatures. It is on that narrower ground subsumed within the Board’s Clarification Memorandum Opinion that I affirm the Board’s conclusion that petitioners presented an insufficient number of signatures to place the initiative on the ballot.
. See Board’s Clarification Memorandum Opinion, Attachment A, Items A(l) and B(l) & (2). These circulators, who gathered a total of 860 (296 + 418 + 146) signatures, submitted petitions with crossed-out names and addresses or petitions in which all signatures were in the same handwriting. The petitioner argues that alterations in the affidavits of cir-culators listed in Item B(2) could have innocent explanations and the Board therefore unjustifiably drew an inference of wrongdoing based simply on their failure to appear even though they were not served. A review of the petitions sheets submitted by these circulators shows that petitioners’ argument is farfetched and that the Board had sound documentary evidence to support its decision to reject the signatures gathered by these circu-lators.
. See Attachment A, Item C(l). These circu-lators gathered a total of 553 signatures. In supplemental filings, petitioner notes that Sheila Washington, one of the circulators who the Board claims was subpoenaed and failed to appear, did in fact testify before the Board. Because only forty signatures at issue are attributable to Ms. Washington, even taking this assertion as true, for the reasons described infra the proposed initiative would still fail.
.Even though the Board no longer has a rule that expressly so provides, First Amendment concerns would caution that such a remedy is proper in the absence of countervailing proof that notwithstanding the deficiencies in the circulators’ affidavits, the signatures these cir-culators collected were, in fact, valid. Cf. Dankman v. District of Columbia Bd. of Elections & Ethics, 443 A.2d 507, 513 (D.C.1981). There was no such evidence presented to the Board. Before this court, seventeen D.C. residents who signed the petitions and wanted to have their votes counted sought to intervene, but their motion was denied by my colleagues in the majority.
. See Attachment A, Items C(2) and D. These circulators gathered a total of 2518 (1540 + 978) signatures.
. It could be assumed that the Board’s decision to invalidate the signatures gathered by Diggs and Inabinet in the original order rested on the Board’s view that all the Stars & Stripes circulators had engaged in "false advertising.” Even after the Board has eschewed reliance on that rationale, however, it perseveres on the identical remedy, which as indicated in the text, is flatly contradicted by the evidence.
.In its opinion, the Board effectively shifted the burden to the petitioner to refute allegations that even those circulators who were not implicated and whose affidavits were sufficient on their face were tainted by misconduct. Because I sustain the Board’s ultimate decision on a narrower ground I need not decide on the extent, if any, to which the Board properly could rely on the fact that *835petitioner did not present evidence to counter allegations of a pattern of fraud in the Stars & Stripes operation. If required to do so, I would think it difficult, on this record, to sustain the broad "missing witness” inference made by the Board and that my colleagues condone. That inference is valid where the witness is "peculiarly within the control” of a party who has reason to believe that the witness would have evidence "relevant and material to a disputed issue in the case.” Thomas v. United States, 447 A.2d 52, 57 (D.C.1982). But it might not have been apparent to the petitioner during the hearing that following a remand by this court the Board would in this case for the first time decide to invalidate a great number of signatures based on a "totality of the circumstances” approach that extrapolated known misconduct by twenty-seven circulators to taint the efforts of fifty-one others based on their common affiliation with an umbrella organization. Moreover, although the circulators involved in the petition drive were paid by the Stars & Stripes umbrella organization, they were not "peculiarly within its control” as they were not employees in the traditional sense. As is apparent from the record, a number of the cir-culators (some of whom lived in homeless shelters and halfway houses and were likely un- or under-employed) formed only a loose association with the organization when they took advantage of an opportunity to make some money by signing up on short notice to work on the petition drive for a few days. There undoubtedly were others, the professional non-residents, who are more permanently associated with Stars & Stripes, notably Michael Jones. In light of the testimony directly linking Mr. Jones to gross improprieties, the Board could well burden petitioner’s cause with a negative missing witness inference when he failed to appear at the Board’s hearing. But even that inference could take the Board only so far as there was little evidence of how many circulators were in fact tainted by his methods, and whether they were in addition to the ones already implicated.
. See Board’s Attachment B, items A & B. These circulators gathered a total of 996 (358 + 218 + 420) signatures.