In Puerto Rico, organizations that seek to be recognized as political parties must gather roughly 100,000 endorsing petitions, each signed by a registered voter and sworn to before a notary public. Since only a lawyer can become a notary in Puerto Rico, there are fewer than 8,000 notaries in the entire commonwealth — and notarial services do not come cheap. Chafing under these restrictions, a nascent political party — the Partido Acción Civil *231(the Party) — challenged various aspects of the law, including the lawyer-notarization requirement, in the local courts. The Party lost.
Plaintiff-appellee José Emilio Pérez-Guzmán (Pérez), a member of the Party who had not participated in the earlier suit, remained dissatisfied with the lawyer-notarization requirement. He sued the members of the Puerto Rico State Elections Commission (the Commission) in the federal district court. The district court rejected a proffered res judicata defense and found that the lawyer-notarization requirement violated the plaintiffs First Amendment rights. Pérez-Guzmán v. Gracia, 260 F.Supp.2d 389 (D.P.R.2003). The Commission and the Commonwealth (which had intervened in the proceedings below) appeal from this ruling.
The questions raised by these appeals are novel and important. The res judicata issue involves the extent to which a judgment against an association can preclude a later action by a member of that association. The constitutional issue pits the government’s interest in regulating elections against an individual’s interests in electoral participation and freedom of association. Having worked our way through both issues, we conclude, as did the district court, that the instant action is not pretermitted by the prior judgment and that the lawyer-notarization requirement unduly burdens First Amendment rights. Consequently, we affirm the judgment below.
I. BACKGROUND
The Party is an unincorporated association seeking to be registered by petition as a political party, and Pérez is among its members. Under commonwealth law, a “Party by Petition”
— Shall be any group of citizens who, desiring to appear on the electoral ballot of a general election, shall register as a political party, on or before June 1 of the election year, by filing with the Commission sworn petitions to such effect, before notary publics duly admitted to the practice of notary, pursuant to the provisions of the Notary Act in effect, who shall collect from the Electoral Commission a fee of one (1) dollar for each valid, notarized petition signed by a number of electors of no less than five (5) percent of the total votes cast for all candidates for the office of Governor in the preceding general election.
P.R. Laws Ann. tit. 16, § 3101(3) (2000). Because more than 2,000,000 votes were cast in the 2000 gubernatorial election, a group that currently desires to register a political party must amass in excess of 100,000 notarized petitions. Each petition must be signed and sworn to before a notary public and filed with the Commission within seven days after notarization. Id. § 3102.
There is nothing wrong with a state demanding that a would-be political party demonstrate a “significant modicum of support” before gaining access to the ballot,1 Jenness v. Fortson, 403 U.S. 431, 442, 91 S.Ct. 1970, 29 L.Ed.2d 554 (1971), and, thus, petitioning requirements are commonplace in such situations. But Puerto Rico’s obviously burdensome rule — a rule that requires each signature to be separately notarized — is quite uncommon. See, e.g., Am. Party of Tex. v. White, 415 U.S. 767, 775 n. 6, 94 S.Ct. 1296, 39 L.Ed.2d 744 *232(1974) (noting that, under Texas law, a single notarial certificate “may be so made as to apply to all [signatories] to whom [the oath] was administered”); Ga.Code Ann. § 21-2-170(d) (Supp.2002) (providing that each sheet of a nominating petition must bear the circulator’s affidavit pertaining to the voter signatures reflected thereon). Puerto Rico’s rule that only attorneys can serve as notaries, P.R. Laws Ann. tit. 4, § 2011 (2000), is also rare. The combination of the two rules is, insofar as we can tell, unique.
Given this peculiar collocation of circumstances, it is not surprising that controversy has surrounded Puerto Rico’s notarization requirement. We reviewed the procedural history of the Party’s challenge to it in an earlier opinion, see Cruz v. Melecio, 204 F.3d 14, 17-18 (1st Cir.2000), and we briefly rehearse that history in order to put the appellants’ res judicata defense into perspective.
On October 6, 1998, the Party filed an action in the Puerto Rico Court of First Instance against the Commission and others. Its complaint averred, inter alia, that the lawyer-notarization requirement transgressed the Constitution. The court granted summary judgment in favor of the defendants; the Puerto Rico Circuit Court of Appeals upheld the judgment; the Puerto Rico Supreme Court also affirmed, see Civil Action Party v. Commonwealth, 2000 TSPR 29, 2000 WL 228543 (P.R. Feb.25, 2000) (CAP I), reconsideration denied per curiam, 2000 TSPR 61, 2000 WL 462276 (P.R. Apr.25, 2000) (CAP II); and the United States Supreme Court denied certiorari, 531 U.S. 920, 121 S.Ct. 283, 148 L.Ed.2d 204 (2000).
Just two days before the intermediate appellate court ruled, fourteen Party members filed an action for declaratory and injunctive relief in Puerto Rico’s federal district court. The action raised essentially the same federal constitutional claims, including the claim that the lawyer-notarization requirement violated the plaintiffs’ rights to free speech and association, to participate in the political process, to vote, and to enjoy equal protection of the laws. Cruz, 204 F.3d at 17. The district court dismissed the action on the merits.
An appeal ensued. In it, we first addressed the potential applicability of res judicata. We held that the defense did not apply because the commonwealth court proceedings were, at that point, still in progress. Id. at 20-21. We then determined that the district court had erred in dismissing the action for failure to state a potentially viable claim. Id. at 22. The complaint had alleged facts which, if true, “tend[ed] to support the appellants’ claims that the notarization requirement and seven-day [filing] deadline unduly burden ballot access.” Id. Thus:
If ... the appellants can prove that notarization is prohibitively expensive or otherwise difficult to achieve (as the complaint avers), then the Commission will have to show that the notarization requirement is narrowly drawn to advance a compelling governmental interest. This showing requires the Commission to come forward with proof. Whether it ultimately can succeed in this endeavor is a sufficiently open question that we cannot conclude, on the pleadings, that no set of facts exists under which the appellants might prevail.
Id. (footnote and internal citation omitted). Accordingly, we vacated the order of dismissal.
Still, we did not allow the case to proceed unabated, but, rather, instructed the district court to stay further proceedings pending the Puerto Rico Supreme Court’s decision. Id. at 25. Among the factors we found “highly relevant to the calculus of abstention” was our belief that “the appel*233lants [had] filed the present suit in an effort to detour around an unfavorable judgment of the commonwealth trial court.” Id. at 24. After the Puerto Rico Supreme Court affirmed the judgment against the Party and the United States Supreme Court denied certiorari, the district court dismissed Cruz on res judicata grounds. The Cruz plaintiffs did not appeal, and that appeared to be the end of the matter.
Appearances can be deceiving. The next year, Pérez filed this action challenging the lawyer-notarization requirement (but not the seven-day filing requirement). In relevant part, the complaint sought a declaration that the lawyer-notarization requirement, P.R. Laws Ann. tit. 16, § 3101(8), violated Pérez’s rights to freedom of speech and equal protection. It also prayed for an injunction prohibiting the defendants from enforcing section 3101(3) and the corresponding regulations.
After an evidentiary hearing, the district court entered an order “declaring the requirement ... that petitions for new political parties be notarized by a lawyer to be in violation of the First Amendment of the Constitution.” Pérez-Guzmán, 260 F.Supp.2d at 394. The court’s determination rested in large part on the idiosyncratic nature of Puerto Rico’s lawyer-notarization requirement. In that regard, the court made three critical findings. First, it found as a fact that the cost of notarizing 100,000 petitions would be at least $1,500,000. Given this substantial outlay, the lawyer-notarization requirement imposed a severe financial burden upon a citizen’s “right to associate and to form political organizations which advance common goals and ideals.” Id. at 392 (footnote omitted). Second, the court found that the appellants had failed to show that the lawyer-notarization requirement was narrowly tailored to advance a compelling governmental interest. Id. at 393. In making this point, the court emphasized that in other electoral contexts the Commission had authorized ordinary voters to verify signatures despite the fact that they were not notaries, but had refused to sanction any comparable arrangement for witnessing party-registration petitions. Id. at 391, 393. There was no evidence that lawyer-notaries perform this attestation function better than'authorized ad hoc notaries. Id. at 393. Finally, the court found that other existing safeguards, such as the Commission’s practice of independently verifying each petition submitted to it, adequately protect Puerto Rico’s legitimate interest in the integrity of its electoral processes. Id. at 390, 393.
In a separate (unpublished) order, the court rejected the asserted res judicata defense. The court found no privity between Pérez and the Party because the record did not show that Pérez had in any way participated in or controlled the Party’s case. Indeed, the court found “no evidence” that Pérez was even a Party member during the currency of the earlier litigation.
Following the entry of a judgment declaring section 3101(3) unconstitutional, these timely appeals eventuated. In an abundance of caution, we stayed the execution of the judgment and expedited appellate review. We now confront the substance of the appeals. Because the res judicata defense, if successful, would end our inquiry, we start there.
II. RES JUDICATA
The applicability of the doctrine of res judicata presents a question of law over which we exercise plenary review, See Gonzalez v. Banco Cent. Corp., 27 F.3d 751, 755 (1st Cir.1994). Inasmuch as we are called upon here to determine the pre-clusive effect of a judgment entered by the *234commonwealth courts, Puerto Rico law supplies the rule of decision. Kremer v. Chem. Constr. Corp., 456 U.S. 461, 466, 102 S.Ct. 1883, 72 L.Ed.2d 262 (1982); Cruz, 204 F.3d at 18-19.
In Puerto Rico, the doctrine of res judicata is embedded in the Civil Code:
In order that the presumption of the res adjudicata may be valid in another suit, it is necessary that, between the case decided by the sentence and that in which the same is invoked, there be the most perfect identity between the things, causes, and persons of the litigants, and their capacity as such.
P.R. Laws Ann. tit. 31, § 3343 (1990). The issue here involves the identity of the parties in the two actions. We know that the requirement of “perfect identity” cannot be taken literally; mere nominal differences will not undermine the preclusive effect of an earlier judgment. Cruz, 204 F.3d at 19. Hence, parties who are in privity with each other are considered identical for res judicata purposes. Banco Cent., 27 F.3d at 756-57.
In Cruz, we determined that privity existed between the plaintiffs and the Party (the plaintiff in the original commonwealth court action), stating:
Although the present appellants are not named parties in the commonwealth court proceedings, they are members of the organization that is the plaintiff there and they control that litigation. This type of privity suffices for res judi-cata purposes.
204 F.3d at 19. This determination depended on a combination of two factors: the plaintiffs were members of the Party and they controlled its conduct of the litigation. In turn, the finding of control derived from the Cruz plaintiffs’ admissions. They made no bones about the fact that they were the prime movers behind the Party’s case, “vigorously asserting] that they were denied a full and fair opportunity to litigate their federal claims in the commonwealth proceedings.” Id. (emphasis supplied). Because the Cruz plaintiffs claimed ownership of the Party’s case, we held them to the natural consequences of that claim.
Before us, the appellants engage in a largely didactic exercise, asserting that Cruz stands for the proposition that any member of a political party has control over litigation brought by that party (and, therefore, that the party is in privity with every one of its members). But this assertion rests on a porous foundation—the notion that Cruz means something other than what it says. We reject that notion. Cruz required proof of control as a building block in the showing of privity, and without such proof there would have been no privity in the circumstances of that case.
This action is at a considerable remove. Unlike the Cruz plaintiffs, Pérez steadfastly disclaims any exercise of control over the original case. In the absence of an admission, the appellants, as the proponents of the res judicata defense, have the burden of establishing a factual basis for that defense. Banco Cent., 27 F.3d at 759, But they have not adduced any evidence sufficient to carry their burden (indeed, they have made no discernible effort to do so). They have shown only that Pérez’s attorney also represented the Party—and that is not a sufficient predicate for an inference of control. See id. (noting that courts regularly “have refused to find substantial control merely because a nonparty retained the attorney who represented a party to the earlier action”). The district court’s finding that Pérez did not control the prior litigation is, therefore, inexpugnable.
The appellants’ first fallback argument is that, even absent control, Pérez’s mem*235bership in the Party, his commonality of interests with the Party and its other members, and his desire to register the Party combine to establish privity as a matter of law. The Party, so this argument goes, is in privity with every one of its members because it sued to advance their associational rights. To a person, the members stood to gain if the Party had prevailed; thus, all of them should be bound by the Party’s loss.
This argument is misguided. Concepts of privity do not grow more expansive when a political party or its members raise First Amendment claims. After all, the fact that First Amendment rights are at stake ought to weigh against, not for, a finding of preclusion. Cf. Richards v. Jefferson County, 517 U.S. 793, 797, 116 S.Ct. 1761, 135 L.Ed.2d 76 (1996) (noting that “extreme applications of the doctrine of res judicata may be inconsistent with a federal right that is ‘fundamental in character’ ”) (quoting Postal Teleg. Cable Co. v. Newport, 247 U.S. 464, 476, 38 S.Ct. 566, 62 L.Ed. 1215 (1918)); Williams v. Rhodes, 393 U.S. 23, 30-31, 89 S.Ct. 5, 21 L.Ed.2d 24 (1968) (acknowledging that First Amendment rights “rank among our most precious freedoms”).
At any rate, the question of whether an individual should be bound by an association’s actions in a prior suit is distinct from whether, in practice, the right at issue has an associational component. As we explained in Cruz, “[i]ndividuals have constitutionally protected interests in free association and electoral participation, including the formation of new political parties.” 204 F.3d at 22 (emphasis supplied). Joining a political party is an exercise of associational rights, but forging such a link does not automatically consign the defense of those rights to the party. Although political parties may “derive rights from their members,” FEC v. Colo. Repub. Fed. Campaign Comm., 533 U.S. 431, 448 n. 10, 121 S.Ct. 2351, 150 L.Ed.2d 461 (2001), the members’ own rights are not automatically subsumed in a judgment against the party.2
This might have been a closer case if the appellants had proven that the Party, in the manner, say, of certain labor unions or trade associations, served generally as the duly constituted representative of its members in litigation affecting common interests. See, e.g., Gen. Foods Corp. v. Mass. Dep’t of Pub. Health, 648 F.2d 784, 787-88 (1st Cir.1981); see generally 18A Charles Alan Wright, Arthur R. Miller & Edward H. Cooper, Federal Practice and Procedure: Jurisdiction 2d § 4456, at 502 (2002) (citing cases). But the record here is bereft of anything showing that either political parties in general or this Party in particular are invested with authority to represent members. In the absence of any evidence anent the Party’s charter, bylaws, or membership terms, there is simply no basis for a finding that Pérez ever authorized the Party, expressly or by fair implication, to sue on his behalf.3
*236This lack of authorization is especially significant because the Party is a jural entity entitled to sue in its own right. Unlike the trade association in General Foods, 648 F.2d at 787, its standing to challenge the lawyer-notarization requirement did not “depend[ ] on [any] claim to represent its members as the real parties in interest.” And unlike the trade association in Expert Electric, Inc. v. Levine, 554 F.2d 1227, 1229 (2d Cir.1977), the Party was not “established to jointly represent its [members]” in all phases of a particular activity. Without such indicia, mere membership cannot allow the Party’s actions to bind Pérez. See Restatement (Second) of Judgments § 61(2) (1982) (noting that if an unincorporated association is a jural entity distinct from its members, a judgment against the association binds its members only to the same extent as a judgment against a corporation binds its shareholders); id. § 59 (explaining that, with exceptions not relevant here, a judgment in an action in which a corporation is a party does not bind corporate shareholders).
From what we have said, it should be evident that the commonality of First Amendment interests between a political party and its members is not itself enough to support a finding of privity. See Griffin v. Burns, 570 F.2d 1065, 1071 (1st Cir.1978) (explaining, in a case involving First Amendment rights, that “[m]ere similarity of interest and a quantum of representation in the earlier suit do[ ] not suffice to bar a non-party”). Without evidence of either control or authorization, the appellants’ proposition is untenable.
The appellants next argue for a determination of privity based on the doctrine of virtual representation — a term that is used as a synonym for “de facto representation” based on an identity of interests between a party to the earlier suit and a nonparty, Banco Cent., 27 F.3d at 758 n. 5. That possibility need not occupy us for long. The appellants did not rely on this doctrine below and, accordingly, they have forfeited the right to raise it in this venue. See, e.g., Teamsters Union, Local No. 59 v. Superline Transp. Co., 953 F.2d 17, 21 (1st Cir.1992) (“If any principle is settled in this circuit, it is that, absent the most extraordinary circumstances, legal theories not raised squarely in the lower court cannot be broached for the first time on appeal.”); Clauson v. Smith, 823 F.2d 660, 666 (1st Cir.1987) (holding that new legal theories “cannot be surfaced for the first time on appeal”).
Even absent this procedural default, the argument would be unavailing. The record here is barren of any evidence that would support a finding of virtual representation. In particular, there is no proof that either Pérez or the Party, in the institution of this action, were engaged in “tactical maneuvering designed unfairly to exploit technical nonparty status in order to obtain multiple bites of the litigatory apple.” Banco Cent., 27 F.3d at 761. Given this dearth of evidence, we cannot find that the Party represented Pérez, virtually or otherwise, in maintaining the earlier action.
The authorities relied upon by the appellants in support of their various res judica*237ta theories are inapposite here. We briefly discuss the two principal decisions.4
In Tahoe-Sierra Preservation Council v. Tahoe Regional Planning Agency, 322 F.3d 1064 (9th Cir.2003), the plaintiff association was formed for the avowed purpose of actively representing the interests of its individual members before various regulatory agencies and had the authority to bring claims on behalf of its members. Id. at 1083. On that basis, the court concluded, with liberal use of adverbs, that “the Association represented the interests of its member property owners sufficiently thoroughly to bind other members alleging similar wrongs arising from the same set of facts.” Id. (footnote omitted). The factors that the Tahoe court found significant, such as association for the express purpose of litigation and authorization of representation, are totally absent here.
So too Tyus v. Schoemehl, 93 F,3d 449 (8th Cir.1996), in which the aldermen-plaintiffs filed a second action while the first was pending, “simply adding new plaintiffs.” Id. at 457. The court found unmistakable evidence of “tactical maneuvering” such that not applying preclusion would have “allow[ed] various members of a coordinated group to bring separate lawsuits in the hope that one member of the group would eventually be successful, benefiting the entire group.” Id. As we have observed, however, there is no evidence that Pérez is part of a coordinated strategy to circumvent a prior judgment. The appellants simply have not proven collusion.
With the facts and the law arrayed against them, the appellants shift their emphasis to policy grounds. They contend that a rule of blanket preclusion is necessary to promote finality and forestall ma-. nipulative practices, including claim-splitting, forum-shopping, and serial litigation. But we are skeptical of the value of a mechanical rule, and there is no sign that any manipulation occurred in this case. The appellants have not shown that Pérez is a puppet dancing on a string pulled by the Party or that the Party has orchestrated a strategy of serial attacks on the lawyer-notarization requirement. In lieu of proof, the appellants apparently would have us infer guilt by association. We will not do so.
Relatedly, the appellants predict disaster if blanket preclusion is rejected. They insinuate that members of a political party would have no incentive to join a pending action, for they would benefit if the party won but would be free to litigate the point if the party lost. Endless litigation would be the norm.
This prediction is much too gloomy. For one thing, it overlooks the salutary effect of the doctrine of stare decisis on repetitive litigation. Perhaps more important, it overlooks the principle that “[t]he law does not impose upon any person absolutely entitled to a hearing the burden of voluntary intervention in a suit to which he is a stranger.” Chase Nat’l Bank v. Norwalk, 291 U.S. 431, 441, 54 S.Ct. 475, 78 L.Ed. 894 (1934); accord Martin v. Wilks, 490 U.S. 755, 763, 109 S.Ct. 2180, 104 L.Ed.2d 835 (1989) (reaffirming that “a party seeking a judgment binding on another cannot obligate that person to intervene; he must be joined”). Finally, the appellants’ prediction proves too much: many others also stood to benefit from the *238Party’s litigation (for example, other aspiring political parties and their members)— yet no one is sanguine enough to suggest that they too should be bound by its defeat. So long as we retain our historic tradition that each person is entitled to his own day in court, we cannot extend preclusion to nonparties merely because they have failed to seize an opportunity to intervene in a prior action.
That ends this aspect of the matter. As we have said, Puerto Rico law supplies the rule of decision as to the efficacy of the res judicata defense in this case. The appellants have not persuaded us that the Puer-to Rico courts would find privity in these circumstances. See, e.g., Montalvo-Huertas v. Rivera-Cruz, 885 F.2d 971, 975 (1st Cir.1989) (holding, under Puerto Rico law, that a corporation was not in privity with the plaintiff class in a prior action challenging the same business regulation merely because the two sets of plaintiffs enjoyed a “community of interest”); Pol Sella v. Lugo Christian, 107 P.R. Dec. 540, 549 (1978) (explaining that res judicata applies when the same person is the real party in interest vis-á-vis both suits); Heirs of Zayas Berrios v. Berrios, 90 P.R.R. 537, 552 (1964) (finding identity of parties where defendants in second case “actually controlled” the first case); cf. A & P Gen. Contractors, Inc. v. Association Caná, Inc., 110 P.R. Dec. 753, 10 P.R. Off. Trans. 987, 994-95 (1981) (reiterating that Puerto Rico has not “abandoned the rule of ... identity” with respect to res judica-ta). In the absence of privity, it necessarily follows that the district court did not err in refusing to honor the res judicata defense.5
III. THE MERITS
This brings us to the heart of the matter: the interaction between the lawyer-notarization requirement and the First Amendment. We review de novo the trial court’s elucidation of First Amendment standards and its application of those standards to the discerned facts. Globe Newsp. Co. v. Beacon Hill Arch’l Comm’n, 100 F.3d 175, 181 (1st Cir.1996); Kassel v. Gannett Co., 875 F.2d 935, 937 (1st Cir.1989). This protocol is not a procedural device, but, rather, “a rule of federal constitutional law” reflecting “a deeply held conviction that judges ... must exercise such review in order to preserve the precious liberties established and ordained by the Constitution.” Bose Corp. v. Consumers Union of U.S., Inc., 466 U.S. 485, 510-11, 104 S.Ct. 1949, 80 L.Ed.2d 502 (1984).
Like any other citizen, Pérez has “constitutionally protected interests in free association and electoral participation, including the formation of new political parties.” Cruz, 204 F.3d at 22 (collecting cases). Although these interests are important, they are not absolute. Fair, honest, and orderly elections do not just happen. Substantial state regulation is a prophylactic that keeps the democratic process from disintegrating into chaos. Consequently, there is a strong state interest in regulating all phases of the electoral process, including ballot access. Storer v. Brown, 415 U.S. 724, 730, 94 S.Ct. 1274, 39 L.Ed.2d 714 (1974); Libertarian Party of Me. v. Diamond, 992 F.2d 365, 370 (1st Cir.1993).
We do not mean to minimize the schizophrenic nature of election campaigns, *239which are as much a means of disseminating ideas as a means of attaining political objectives. See Ill. State Bd. of Elections v. Socialist Workers Party, 440 U.S. 173, 186, 99 S.Ct. 983, 59 L.Ed.2d 230 (1979). Given this duality, an overly stringent regulatory scheme may place an intolerably heavy burden on freedom of political expression. See id. In the last analysis, a fíne line separates permissible regulation of state election processes from impermissible abridgement of First Amendment rights.
Plotting that line calls for a careful reconciliation of competing centrifugal and centripetal forces. The rigorousness of the ensuing judicial inquiry depends upon the extent to which the challenged regulation burdens First Amendment rights. Burdick v. Takushi, 504 U.S. 428, 434, 112 5.Ct. 2059, 119 L.Ed.2d 245 (1992); Werme v. Merrill, 84 F.3d 479, 483 (1st Cir.1996). Following this prescription, we afford exacting scrutiny to severe restrictions on ballot access. Cruz, 204 F.3d at 22. That entails viewing such restrictions skeptically and requiring that they be drawn narrowly to advance compelling state interests. Norman v. Reed, 502 U.S. 279, 289, 112 S.Ct. 698, 116 L.Ed.2d 711 (1992). To this end, we start with an assessment of the severity of the restriction. We then proceed to identify the interests that the appellants believe justify its imposition, weigh the efficacy of the available alternatives, and decide where the lawyer-notarization requirement falls along the constitutional line. Timmons v. Twin Cities Area New Party, 520 U.S. 351, 358, 117 S.Ct. 1364, 137 L.Ed.2d 589 (1997).
We approach the question of severity by looking first at its theoretical underpinnings. Traditionally, the act of circulating a petition has been viewed as “a one-on-one communication.” Buckley v. Am. Const. Law Found., 525 U.S. 182, 199, 119 S.Ct. 636, 142 L.Ed.2d 599 (1999).6 One person circulates the -petition, obtains the voter’s 'signature, and verifies the origins of the signature. See, e.g., Frami v. Ponto, 255 F.Supp.2d 962, 965 (W.D.Wis.2003). In Puerto Rico, however, each of the 100,-000 petitions must be signed singly and sworn in the presence of a lawyer-notary. This requirement triangulates the normal channels of communication. Assuming that the circulator is not himself a notary, there must now be one person to initiate the conversation, a second to respond, and a third to verify that response.
This triangulation limits the efficacy of petition circulators in their own right. No matter how persuasive a circulator might be in a face-to-face conversation, he cannot seal the deal then and there. The final exchange in the interactive communication- — the voter’s official endorsement of the fledgling party — cannot occur unless and until a notary is present. The historic role of the circulator is necessarily diminished.
In this case, Pérez plainly labored under this heavy yoke. He testified without contradiction that the lawyer-notarization requirement greatly hampered his efforts to gather signatures because, in addition to convincing voters that a new political party was needed, he also had to convince them to repair to an attorney’s office. When he tried to recruit others to help in the circulation of petitions, those whom he could interest were “just common people,” not notaries. The district court credited this testimony.
*240The stringency of the restriction is aggravated because the indispensable third person in the triangulated relationship— the notary — may prove hard to find. In most jurisdictions, it is neither impractical nor burdensome for party members to become notaries so that they may verify the petitions that they circulate. See Am. Party of Tex., 415 U.S. at 787, 94 S.Ct. 1296. But Puerto Rico’s restriction of notarial status to licensed attorneys erects a high barrier to entry and virtually ensures that the supply of notaries will remain inelastic notwithstanding voters’ attempts to mount new-party registration drives.
Nor is this a purely theoretical difficulty. The district court found that Puerto Rico has only 8,000 lawyer-notaries, more or less. See Pérez-Guzmán, 260 F.Supp.2d at 391. Hence, the ratio of notaries to voters is quite small, and common sense suggests that there may be areas of the island where lawyers (and, thus, notaries) are in short supply. Even in areas where notaries are plentiful, someone must travel; either the voter must visit the notary’s office or the notary must take to the field.
Last — but far from least — someone must pay the freight.7 The district court determined that the average cost for notarizing a simple document signed at a lawyer’s office is between $15 — $20; the cost if the lawyer travels to the site of the signing escalates to $50-$60; those who would form a party by petition need to collect a minimum of 100,000 signatures; and, accordingly, the lawyer-notarization requirement adds at least $1,500,000 to the cost of a successful petition drive. Id. at 391-92. The court further found that this was “an unreasonably high price tag for an individual who is seeking to exercise his constitutional rights.” Id. at 392 (footnote omitted).
These findings are wholly supportable.8 Indeed, the court’s cost estimates may be conservative because, as a practical matter, a group attempting to form a party by petition will need to gather a surplus of signatures (as challenges likely will result in a number of invalidations). The court’s findings make manifest that the task of registering a new political party is one of daunting proportions. We hold, therefore, that the lawyer-notarization requirement imposes a severe restriction on Pérez’s ability to register a new political party, and, in turn, on his right to access the ballot.
*241The appellants’ only direct rejoinder is to remind us that the Puerto Rican Renewal Party (PRRP) managed to register for the 1984 election as a party by petition. But no other party has duplicated that feat in the quarter-century since section 3101(3) was enacted, see CAP I, Off. Trans, at 9, so this is hardly an indication that the lawyer-notarization requirement is a piece of cake. Cf. Storer, 415 U.S. at 742, 94 S.Ct. 1274 (noting evidentiary significance of a showing that particular types of aspirants have “qualified [for the general ballot] with some regularity”). The PRRP example demonstrates, at most, that the lawyer-notarization requirement is not insuperable — not that the requirement is less than severe. It follows that such a showing does not foreclose Pérez’s suit. See, e.g., Anderson v. Celebrezze, 460 U.S. 780, 791 n. 12, 103 S.Ct. 1564, 75 L.Ed.2d 547 (1983) (explaining that the ability of a few individuals to qualify as independent presidential candidates “d[id] not negate the burden imposed” by the challenged regulation); cf. Am. Party of Tex., 415 U.S. at 783, 94 S.Ct. 1296 (explaining that what is demanded as a condition to ballot access “may not be so excessive or impractical as to be in reality a mere device to always, or almost always, exclude parties with significant support from the ballot”) (emphasis supplied).
Apart from this glancing reference to the PRRP’s success, the appellants do hot seriously dispute the severe impact of the lawyer-notarization requirement. Instead, they attempt to confess and avoid. Their effort tracks three main avenues.
First, the appellants propose that a reviewing court should afford Puerto Rico wide latitude vis-a-vis the challenged regulation because the Supreme Court, in American Party of Texas and Jenness, upheld other notarization and petitioning requirements as reasonable. This proposition overlooks the fact that both cases were decided before the Supreme Court crystallized its current standard of inquiry in cases such as Burdick, 504 U.S. at 434, 112 S.Ct. 2059. More important, it overlooks the salient differences that distinguish those two cases from the case at bar. We explain briefly.
Although the Court sanctioned a notarization requirement in American Party of Texas, it gave no indication that Texas limited notarial status to members of the bar. Moreover, the relevant numerical threshold was one percent of the active electorate — not five percent — and bulk notarization seems to have been available. 415 U.S. at 775 n. 6, 94 S.Ct. 1296. The Georgia law at issue in Jenness required signatures from five percent of the active electorate but imposed no concomitant notarization requirement with respect to individual voters’ signatures. 403 U.S. at 439, 91 S.Ct. 1970.
These differences make it clear that the appellants are comparing plums to pomegranates: whereas American Party of Texas dealt with a claim that a garden-variety notarization requirement constituted a severe restriction and Jenness dealt with a claim that a petitioning requirement constituted a severe restriction, the case before us turns on the synergy among three discrete factors: the five percent numerical requirement; the requirement that each signature be individually notarized; and the restriction of notarial practice in Puerto Rico to lawyers. It is this synergy that places section 3101(3) beyond the pale, for it transforms the dynamics of participation in the electoral process.
The appellants’ second avenue for marginalizing the severity of the imposed restriction involves drawing a distinction between the act of seeking support from the voter (which anyone can do at any time and place) and the act of authenticating *242the voter’s signature. The appellants argue that the lawyer-notarization requirement comes into play only after the voter has agreed to subscribe a petition and, thus, does not have First Amendment implications. The requirement merely “regulate[s] the mechanics of the electoral process,” not “the communicative aspect of petitioning.” CAP II, Off. Trans, at 7.
This distinction is artificial. Petition cir-culators — persons like Pérez — are free only in theory “to convey their political message through whatever means they may deem convenient.” Id. The notarization requirement’s looming presence forces the circulator to structure his communications and choose his target audiences with that requirement in mind. Pérez, like most persons, is not a notary. In principle, he is free to • go from door to door soliciting support for the Party, but in practice his rate of return will suffer because he cannot gather endorsing signatures on the spot. The record makes clear not only that some voters who would be willing to sign petitions are not willing to venture to a law office but also that Pérez faces enormous difficulties in persuading notaries to accompany him in search of prospects. Given the reality of events, the lawyer-notarization requirement burdens petitioning by causing the utility of a communication to vary with its setting and with the qualifications of the circulator.
It would serve no useful purpose to dwell upon this line of argument. The short of it is that a state cannot separate petitioning into two steps, close off the second step to all but a tiny professional class, and then ignore the effects of that restriction. See Buckley, 525 U.S. at 194-95, 119 S.Ct. 636 (invalidating a registration requirement that unduly “limi[ts] the number of voices who will convey [the proponent’s] message” (quoting Meyer v. Grant, 486 U.S. 414, 422-23, 108 S.Ct. 1886, 100 L.Ed.2d 425 (1988))); Lerman v. Bd. of Elections in the City of N.Y., 232 F.3d 135, 146 (2d Cir.2000) (invalidating a residency requirement for petition verifiers because it drastically reduced the number of potential circulators); Krislov v. Rednour, 226 F.3d 851, 860 (7th Cir.2000) (invalidating a requirement that a circulator/certifier be registered to vote in particular political subdivisions because it “preclude[s] the candidate from utilizing a large class of potential solicitors to convey his message”). Accordingly, the attempted distinction fails.
The appellants’ third avenue relies upon a “causation” argument which, if credited, would make it unnecessary for us to reach the question of severity. Under this rubric, they note that Pérez collected a total of fewer than 80 signatures (only 19 of which were actually presented to the Commission). Based on this “meager” effort, they fault the district court for failing to address the issue of whether Pérez had made a diligent effort to obtain the requisite number of petitions. In their view, the absence of due diligence should bar any claim that the lawyer-notarization requirement “caused” Pérez’s inability to register the Party.
The appellants derive this due diligence condition from a misreading of Storer. In that case, the Supreme Court remanded for further factfinding as to whether, “in the context of California politics ... a reasonably diligent independent candidate [could] be expected to satisfy the signature requirements” imposed by a state statute regulating the filing of nomination papers by independent candidates. 415 U.S. at 742, 94 S.Ct. 1274. Contrary to the appellants’ importunings, this language did not portend that only those who demonstrate due diligence can mount a First Amendment challenge to a ballot access requirement. The Storer remand went to the *243burdensomeness of the challenged regulation (i.e., its severity), not to causation. While a particular plaintiffs “[p]ast experience” can have evidentiary significance in an assessment of severity, id., a showing of personal due diligence is not an element of a ballot access claim, see, e.g., Norman, 502 U.S. at 293, 112 S.Ct. 698 (holding that a state may not “require petitioners to gather twice as many signatures to field candidates in [a multidistrict subdivision] as they would need statewide” without inquiring whether petitioners had shown due diligence in trying to satisfy the challenged requirement); Anderson, 460 U.S. at 802-04, 103 S.Ct. 1564 (discussing challenge without asking whether the petitioner had shown due diligence in attempting to meet Ohio’s early filing deadline). Indeed, a rule such as that espoused by the appellants would tend to inoculate even the most blatantly unconstitutional electoral requirements from legitimate attack.9
That ends this phase of our analysis. We conclude, without serious question, that Puerto Rico’s lawyer-notarization requirement imposes a severe burden on Pérez’s rights.
We turn next to the task of “identify[ing] and evaluating] the precise interests put forward by the State as justifications for the burden imposed by its rule.” Anderson, 460 U.S. at 789, 103 S.Ct. 1564. In the course of that inquiry, we must determine the legitimacy and strength of each of those interests as well as the extent to which they dictate burdening citizens’ rights. Id.
We do not gainsay Puerto Rico’s robust interest in protecting the integrity of its election processes. See Timmons, 520 U.S. at 364, 117 S.Ct. 1364; Am. Party of Tex., 415 U.S. at 782 n. 14, 94 S.Ct. 1296. This includes an “important state interest in requiring some preliminary showing of a significant modicum of support before printing the name of a political organization’s candidate on the ballot.” Jenness, 403 U.S. at 442, 91 S.Ct. 1970. That interest is directly served by the five percent petitioning requirement. See Lubin v. Panish, 415 U.S. 709, 718, 94 S.Ct. 1315, 39 L.Ed.2d 702 (1974) (terming a signature requirement the “obvious” means of testing a candidate’s level of support). The conundrum here involves an additional safeguard — the lawyer-notarization requirement — and the interests it advances.
The appellants have identified only one such interest: preventing election fraud. They assert that notarization is necessary to “remedfy] a type of electoral fraud that the Puerto Rico Legislature found to hafye] been pervasive.” Appellants’ Reply Br. at 39. The Puerto Rico Supreme Court echoed this sentiment. It declared that notarization “serves a compelling state interest in guaranteeing the integrity of the electoral process, given the history of electoral fraud and corruption experienced in Puerto Rico.” CAP I, Off. Trans, at 15. In this regard, the court singled out “[t]he popular practice of manipulating the voting lists.” Id.
Even though we accept the legitimacy of this interest, we must nonetheless mull the extent to which that interest renders *244it necessary to burden Pérez’s rights so severely. See Anderson, 460 U.S. at 789, 103 S.Ct. 1564. The question of whether the lawyer-notarization requirement is narrowly fashioned to advance the identified state interest boils down to whether notarization by a lawyer is more likely to reduce fraud by any or all of the participants in the transaction than other available (less restrictive) alternatives. We examine this question from a variety of perspectives.
As to the voter, the appellants contend that notarization gives some meaningful assurance that voters’ signatures are authentic. “The [voter] takes oath to that fact and experience has shown that many people take their oaths seriously.” Libertarian Party of Va. v. Davis, 591 F.Supp. 1561, 1564 (E.D.Va.1984), aff'd, 766 F.2d 865 (4th Cir.1985). But this contention is disingenuous, for it is the oath that guarantees honesty — not the pedigree of the oath-giver. The appellants do not suggest that people take oaths less seriously in jurisdictions where notaries need not be lawyers or in jurisdictions where affirmations are taken by non-notary witnesses. The idea that notarization by a lawyer will significantly reduce voter fraud is utterly unsupported by the record; there is nothing to show that a voter bent on committing petition fraud will be less brazen in front of a lawyer-notary than in front of some other type of witness.10
We turn next to petition circulators. The appellants offer no developed argumentation to the effect that notarization by a lawyer will discourage circulator fraud. We assume that this is a concession. See United States v. Zannino, 895 F.2d 1, 17 (1st Cir.1990) (holding that points raised on appeal in a perfunctory fashion, without developed argumentation, are conceded). In any event, 'the district court effectively scotched any such notion, observing that “[a] lawyer will notarize a signature merely by verifying the signatory’s identification,” and that “[t]here is no reason to believe that attorneys who happen to be notaries are somehow more adept than non-lawyers at reading and verifying” identification documents. Pérez-Guzmán, 260 F.Supp.2d at 393.
This leaves the notary himself. The Puerto Rico Supreme Court suggested that the lawyer-notarization requirement “allows for the imposition of sanctions for any illegal act committed by the person before whom the petition is sworn to.” CAP I, Off. Trans, at 14 (citations and internal quotation marks omitted). But this rationale does not itself distinguish the lawyer-notary from any other person who may be authorized by the Commission to verify petitions. One need be neither a lawyer nor a notary to be fined or imprisoned for participating in the preparation of fraudulent documents. See, e.g., P.R. Laws Ann. tit. 33, § 4437 (2001).11
We recognize that “notarial practice is strictly regulated” in Puerto Rico; that notaries are “the only officers vested by *245the [Puerto Rico] Supreme Court with the authority to guarantee ... the authenticity of the documents executed before them”; and that “[t]he intervention of a notary establishes a presumption of veracity in all the documents authenticated by him or her.” CAP I, Off. Trans, at 15 (citation omitted). But "without some proof that lawyer-notarization is appreciably more effective at preventing petition fraud than non-lawyer verification — and the record contains none — all of this is beside the point. The fact that a state’s asserted interest in preventing electoral fraud is important in the abstract does not create a presumption that its chosen means of regulation will advance that interest. See Lerman, 232 F.3d at 149.
In a related vein, the appellants contend that the lawyer-notarization requirement is necessary because a successful petition effort stands to reap a munificent harvest for the Party — at least $600,000 from the Electoral Fund in an election year, $300,000 in a non-election year, and “insider” status for the next four years. See CAP I, Off. Trans, at 9 (noting that “registered parties ... share in the Electoral Fund”); CAP II, Off. Trans, at 9 (explaining that registration also allows a party to “become a part of the government body that governs the electoral process, as a result of which it acquires quasi-public functions”). These facts simply reinforce a proposition to which we already have subscribed, namely, that the state has a compelling interest in preventing the fraudulent registration of political parties. Access to special privileges says nothing about the effectiveness of lawyer-notarization as a means of advancing that interest.
In considering whether a ballot access requirement is narrowly drawn to advance the state’s interest in preventing fraud, the mechanisms that the state already has in place serve as benchmarks. Norman, 502 U.S. at 294, 112 S.Ct. 698. We think it noteworthy, therefore, that the Commission routinely allows the use of ad hoc notaries for a wide variety of analogous purposes, e.g., “for a person registering to vote, a voter changing his electoral address, petitioners seeking to have an independent candidate placed on the ballot, and petitions nominating candidates for primaries.” Pérez-Guzmán, 260 F.Supp.2d at 391.12 In addition, ad hoc notaries were authorized for use in the 1998 commonwealth-wide status plebiscite. See CAP I, Off. Trans, at 9. The record contains not one hint of election fraud or administrative error committed by, or under the auspices of, ad hoc notaries. Absent some evidence that the use of ad hoc notaries correlates with a higher incidence of corruption or unreliability, we cannot accept the Commission’s ipse dixit that it needs to insist on the use of lawyer-notaries in the new-party petition context. See Werme, 84 F.3d at 485 (explaining that “mere suspicion or paranoia is too flimsy a foundation on which to rest a claim of incipient fraud or mistake”).
Referring back to a political party’s access to the Electoral Fund and its opportunity to participate as an “insider” in the electoral system for a four-year period, the appellants posit that registration as a political party entails higher stakes than other matters (and, hence, that the use of ad hoc notaries in that context would be especially risky). By this logic, parties by petition are not situated similarly to independent candidates, candidates for major party pri*246maries, or even groups participating in the plebiscite process. We think that this vastly overstates the matter; it is not obvious to us that, say, the status plebiscite was less important to the future of Puerto Rico than the registration of a new political party. We need not pursue this point, however, because the appellants have not shown that, compared with verification by ad hoc notaries, lawyer-notarization is superior in detecting and deterring petition fraud.
We add, moreover, that the use of ad hoc notaries is not the only feasible safeguard available to the Commission (indeed, the record strongly suggests that notarization may not even be the best means currently employed by the Commission). The Commission checks every petition to ensure that the signature is valid. Pérez-Guzmán, 260 F.Supp.2d at 390. It is in the process of computerizing its voter registration records to include each registered voter’s signature. This project is slated for completion later this year, and its completion will enhance the Commission’s ability to verify petitions. To accomplish these tasks, the Commission has at its disposal a cadre of people not affiliated with any political party. These nonpartisan workers comprise the so-called Validations Unit, a unit that has the assignment of checking petitions one by one against the Commission’s records. The Secretary of the Commission, in answer to a query from the district court, testified that “it would not be difficult for the Commission to check [new-party] petitions ... against [its] records.”13
In this specialized electoral context, the Commission’s in-house verification procedure compares favorably with lawyer-notarization. As the district court found, a lawyer-notary usually verifies a signature by checking a driver’s license, a passport, or some other identification document. Pérez-Guzmán, 260 F.Supp.2d at 391. There is no indication that notaries compare signatures to those appearing on voting records or take steps to ensure that affiants are registered voters. The Commission, it appears, is better positioned than a notary to detect the species of fraud that purportedly justifies the lawyer-notarization requirement. See CAP I, Off. Trans, at 15 (offering as a justification the need to ferret out the “popular practice of manipulating the voting fists”).
Let us be perfectly clear. A state is entitled to take a “belt-and-suspenders” approach and put in place multiple mechanisms for ensuring the integrity of its electoral processes. Here, however, the record fails to show that lawyer-notarization adds anything over and above other readily available means of verification. Based in part on the absence of evidence that lawyer-notarization reduces electoral fraud and in part on the array of less restrictive alternatives available to the state (including the feasibility of using non-lawyers as ad hoc notaries and the Commission’s apparent ability to verify every petition in-house), we find the lawyer-notarization requirement broader than necessary to serve the state’s asserted interest. Consequently, the requirement embodied in section 3101(3) cannot survive a First Amendment challenge. See Norman, 502 U.S. at 293-94, 112 S.Ct. 698; Krislov, 226 F.3d at 866; Cruz, 204 F.3d at 22.
IV. CONCLUSION
We need go no further. For the foregoing reasons, we hold (1) that res judicata does not bar the maintenance of the in*247stant action, and (2) that the lawyer-notarization requirement is not narrowly drawn to advance a compelling state interest (and, thus, cannot withstand First Amendment scrutiny). We are mindful that our constitutional conclusion differs from that of the Puerto Rico Supreme Court, and we do not lightly part company with so distinguished a tribunal. But it is our responsibility to interpret and apply the Constitution of the United States, and it would be a dereliction of that duty to defer to the views of any state court.
The judgment of the district court is affirmed and the stay previously issued is dissolved.