concurring in the judgment.
Though I remain persuaded that, on the facts of this case, the student editors were public actors acting under color of state law, further examination of the record during en banc review leads me to conclude that this issue is not material to our decision and that the district court properly entered judgment in favor of defendants. I write separately for two reasons. First, because this case is easily resolved on statutory grounds — lack of evidence from which a factfinder could conclude that the defendants “caused” the constitutional violation complained of — I believe the majority’s central state action ruling to be an unfortunate breach of the “fundamental rule of judicial restraint ... that [courts] will not reach constitutional questions in advance of the necessity of deciding them.” Three Affiliated Tribes v. Wold Engineering, P.C., 467 U.S. 138, 157, 104 S.Ct. 2267, 2278-79, 81 L.Ed.2d 113 (1984). Second, I believe the majority’s unnecessary constitutional ruling to be wrong on the merits.
I.
Defendants cannot be liable to plaintiff for damages unless, among other things, they subjected plaintiff to, or caused plaintiff to be subjected to, a deprivation of federal rights. See 42 U.S.C. § 1983. Because the Supreme Court has made it clear that § 1983’s causation language is to be narrowly construed, see Monell v. New York City Dep’t of Social Servs., 436 U.S. 658, 691-94, 98 S.Ct. 2018, 2036-38, 56 L.Ed.2d 611 (1978) (interpreting the language of § 1983 to preclude the imposition of vicarious liability), the question here quickly reduces to whether one or more defendants can, at the least, be seen as having caused the rejection of the advertisements within the narrow meaning of the statute. See id.
As the majority notes, plaintiff has not named as defendants those persons — the student editors of the newspaper and yearbook — who may most readily be seen as having directly subjected him to the alleged deprivation of his First Amendment and equal protection rights. Nor has plaintiff argued that the acts and omissions for which the named defendants may most readily be seen as responsible — delegating decision-making authority to the students and failing to override the students’ decisions — caused him to be subjected to constitutional harm. See, e.g., Febus-Rodriguez v. Betancourt-Lebron, 14 F.3d 87, 91-92 & n. 4 (1st Cir.1994) (discussing this circuit’s standard for establishing supervisory liability under § 1983). As a result, no trial is warranted unless there is a genuine issue of material fact as to whether (1) at least one individual defendant actually colluded with the students in the decisions to reject the advertisements; or (2) the rejection was pursuant to a policy or custom of the Town of Lexington. See, e.g., Board of County Commissioners v. Brown, — U.S. -, -, 117 S.Ct. 1382, 1388, 137 L.Ed.2d 626 (1997) (reiterating that, for purposes of § 1983, a municipality causes one to be subjected to a deprivation of federal rights only through its duly-enacted policies or widespread customs having the force of law). In my view, there is insufficient evidence to warrant a trial against any of the named defendants under either of these theories.
Though I continue to disagree with the majority’s conclusion that defendants Kafris-sen and Mechem did not influence the students’ decisions to reject the advertisements, I now concur that the summary judgment record permits only one inference: the students made the ultimate decisions. In the end, I am constrained to agree that, in the face of largely uncontradicted testimony to *257contrary effect, Kafrissen and Meehem’s use of the term “we” in message-relaying correspondence with plaintiff, and Kafrissen’s pri- or threat to resign, see ante at 251 n. 9, do not allow a determination that defendants Kafrissen or Mechem can be held liable for the decisions made. And with respect to the other individually-named defendants, so also do I agree that there is no basis for concluding that they participated in the rejections of the advertisements. Thus, there is insufficient evidence of statutory causation for plaintiff to proceed to trial against any of the individual defendants.
With respect to the Town, I agree with the majority’s conclusion that there is no basis for attributing to it the conduct of the students. See ante Part III, Sections A and C. I take issue, though, with the method by which the majority reaches its conclusion. Specifically, I disagree with its direct (though reverse) application to this case of those Supreme Court state action eases which look for state action in private conduct. See generally id. (applying, in order, Terry v. Adams, 345 U.S. 461, 73 S.Ct. 809, 97 L.Ed. 1152 (1953); Morse v. Republican Party of Virginia, 517 U.S. 186, 116 S.Ct. 1186, 134 L.Ed.2d 347 (1996); Burton v. Wilmington Parking Auth., 365 U.S. 715, 81 S.Ct. 856, 6 L.Ed.2d 45 (1961); Flagg Bros., Inc. v. Brooks, 436 U.S. 149, 98 S.Ct. 1729, 56 L.Ed.2d 185 (1978); Jackson v. Metropolitan Edison Co., 419 U.S. 345, 95 S.Ct. 449, 42 L.Ed.2d 477 (1974); Rendell-Baker v. Kohn, 457 U.S. 830, 102 S.Ct. 2764, 73 L.Ed.2d 418 (1982); and Edmonson v. Leesville Concrete Co., Inc., 500 U.S. 614, 111 S.Ct. 2077, 114 L.Ed.2d 660 (1991), to determine that the defendants to this lawsuit, all of whom are public actors, are not liable for the students’ conduct). While these cases might, by analogy, shed light on what will not be considered constitutionally tortious municipal conduct under § 1983, see, e.g., Rendell-Baker, 457 U.S. at 840-41, 102 S.Ct. at 2770-71 (strongly supporting an argument that neither municipal funding nor municipal regulation of a private entity constitutes municipal “policy” for purposes of § 1983), not one of them speaks to the question directly posed here: can conduct by non-legislative and non-policy making actors be deemed to have been sufficiently caused by municipal policy or custom for liability to attach to the Town? Nor does any one of these cases involve harm-causing conduct which can be seen as having been committed by public actors. For these reasons, I fear that the majority opinion confuses more than it clarifies.
Moreover, the majority has entered an area it could and should avoid. The Town’s freedom from liability flows less from the fact (if it be fact, see infra Part II) that the students are private actors than from the fact that the students’ actions were not caused by Town policy or custom. Again, plaintiff has not advanced as a theory of liability the Town’s decision to let the students decide — the only municipal policy or custom which can arguably be seen at play here. And even if plaintiff had so argued, it seems obvious that, as an action taken in what appears to have been good faith reliance upon state law, see Mass. Gen. L. e. 71, § 82, this policy cannot give rise to municipal liability under § 1983. See Surplus Store & Exchange, Inc. v. City of Delphi 928 F.2d 788, 791-92 (7th Cir.1991).
In the end, defendants are entitled to judgment because they did not, under § 1983, ultimately cause the conduct of the non-party students. We should not go beyond this simple fact to decide the case.
II.
By resolving this dispute through application of those cases which look for state action in private conduct, the majority proceeds from the premise that the students were private actors. I not only find this implicit holding to be unnecessary, I believe it to be incorrect on the merits. In my view, had plaintiff sued the student editors directly, we would have been obliged to rule that they were, in fact, public actors insofar as they solicited and published advertisements from paying third parties.
Whether a person or entity is a private or a public actor obviously cannot be resolved through application of cases which presume that the actor is private; it is resolved by a fact-specific inquiry into whether the person or entity is, in context, acting “under color of *258state law.” See Polk County v. Dodson, 454 U.S. 312, 322 n. 12, 102 S.Ct. 445, 452 n. 12, 70 L.Ed.2d 509 (1981) (noting the distinction). Although the Supreme Court has sometimes stated that the state action and under color of state law questions are coextensive, see, e.g., United States v. Price, 383 U.S. 787, 794 and n. 7, 86 S.Ct. 1152, 1156-57 and n. 7, 16 L.Ed.2d 267 (1966), it also has recognized that they are not invariably the same. That the inquiries sometimes diverge is clear in Dodson where, without reference to a single state action case, the Court concluded that a state public defender does not act under color of state law while acting as counsel to an indigent defendant in a state criminal proceeding. 454 U.S. at 320-24, 102 S.Ct. at 450-53.
Here, as in Dodson, the question (had plaintiff raised it) would not have been whether private conduct should be attributed to the Town; rather, it would have been whether the conduct was, as an initial matter, public or private. Cf. Blum v. Yaretsky, 457 U.S. 991, 1003-04, 102 S.Ct. 2777, 2785-86, 73 L.Ed.2d 534 (1982) (contrasting “those cases in which the defendant is a private party and the question is whether his conduct has sufficiently received the imprimatur of the State so at to make it ‘state’ action for purposes of the Fourteenth Amendment” with “eases in which the challenged conduct consists of enforcement of state laws or regulations by state officials who are themselves parties in the lawsuit”). And the state action tests the majority relies upon,.designed as they are to determine whether private conduct is attributable to the state, would not have helped answer the question.
To illustrate, when an on-duty municipal police officer misuses the power of the office to carry out a personal vendetta, we do not decide whether he was acting under color of state law by reference to whether the municipality is itself liable for the conduct. See, e.g., Martinez v. Colon, 54 F.3d 980 (1st Cir.), cert. denied, — U.S. -, 116 S.Ct. 515, 133 L.Ed.2d 423 (1995). We decide it on a more contextually-appropriate inquiry into whether the officer has “exercised power ‘possessed by virtue of state law and made possible only because the wrongdoer is clothed with the authority of state law.’ ” Id. at 986 (quoting West v. Atkins, 487 U.S. 42, 49, 108 S.Ct. 2250, 2255, 101 L.Ed.2d 40 (1988)). So here should we decide whether the students are public or private actors by reference to criteria other than those we would use to decide whether the Town must pay for the students’ acts.
What criteria should be used? A helpful starting point is Dodson, where the Court’s ruling was informed by two primary considerations: (1) “a public defender is not amenable to administrative direction in the same sense as other employees of the State,” 454 U.S. at 321, 102 S.Ct. at 451; and (2) “it is the constitutional obligation of the State to respect the professional independence of the public defenders whom it engages,” id. at 321-22, 102 S.Ct. at 451. Here, both factors militate in favor of finding that, insofar as they solicited and published (or declined to publish) advertisements from paying third parties, the students acted under color of state law. Certainly, the power of school officials to regulate the content of student publications and the acts of their student editors, see Hazelwood Sch. Dist. v. Kuhlmeier, 484 U.S. 260, 266-70, 108 S.Ct. 562, 567-70, 98 L.Ed.2d 592 (1988), is near its apex where the subject of the regulation involves the students’ commercial interactions with third parties. And where these interactions arguably implicate the constitutional rights of those third parties, cf. Dodson, 454 U.S. at 321-22, 102 S.Ct. at 451-52, and hold out the prospect of monetary benefit to the Town, see, e.g., Burton, 365 U.S. at 724, 81 S.Ct. at 861 (indicating that conduct which leads to monetary benefits for the State will often be deemed action on behalf of the State), the question is less whether the students may be regulated and more whether the students must be regulated.
My position is narrow. I have never doubted that the student writers are private actors with respect to reporting and editorializing. A contrary holding would, after all, effectively spell the end of public school student publications; one would be hard-pressed to report and could never editorialize without violating the First Amendment’s mandate of viewpoint neutrality. See gener*259ally R.A.V. v. City of St. Paul, 505 U.S. 377, 112 S.Ct. 2538, 120 L.Ed.2d 305 (1992). I only suggest that, to the extent public school students solicit funds to support a public enterprise in their capacities as officials of that enterprise, they act under color of state law. See Dodson, 454 U.S. at 324-25, 102 S.Ct. at 453-54 (making clear that an actor can act under color of state law in one capacity but not in another); see also ante at 254.
At the very least, that the students are private actors is not such an open and shut matter that it should be assumed sub siten: tio. If the student editors of the Musket determined to run the paid political advertisements of Democratic candidates for Town office but not those of Republican candidates, and if the Republican candidates sought in-junctive relief against the students in their capacities as editors of the Musket, would we summarily conclude that the challenged action was not undertaken under color of state law? I would like to think not. Though the facts of the present case are less egregious, the underlying question — not presented here because of plaintiffs pleading decisions — is the same.
III.
For the reasons stated, I concur in the majority’s conclusion that judgment' was properly entered in favor of defendants. I do not, however, concur in the reasoning that leads it to this conclusion. I would instead resolve the ease under well settled law that precludes a finding, under § 1983, against any of the defendants named in the com-, plaint for the theories of recovery plaintiff has advanced.