concurring in part and dissenting in part.
The Court holds that the respondents have standing to bring this suit and to assert their own constitutional rights, if any, in an attack on Mo. Rev. Stat. § 208.152 (12) (Supp. 1975). The Court also holds that the Court of Appeals erred in proceeding to the merits of respondents’ challenge. I agree with both of these holdings and therefore concur in Parts I, II-A, and III of Justice Blackmun’s opinion, as well as in the first four sentences of Part II-B.
The Court further holds that after remand to- the District Court the respondents may assert, in addition to their own rights, the constitutional rights of their patients who would be eligible for Medicaid assistance in obtaining elective abortions but for the exclusion of such abortions in §208.152 (12). I dissent from this holding.
I
As the Court notes, ante, at 109-110, respondents by complaint and affidavit established their Art. Ill standing to invoke the judicial power of the District Court. They have performed abortions for which Missouri’s Medicaid system would compensate them directly1 if the challenged statutory section did not preclude it. Re*123spondents allege an intention to continue to perform such abortions, and that the statute deprives them of compensation. These arguments, if proved, would give respondents a personal stake in the controversy over the statute's constitutionality. See Warth v. Seldin, 422 U. S. 490, 498-499 (1975); cf. id., at 502-508; Simon v. Eastern Ky. Welfare Rights Org., 426 U. S. 26, 40-46 (1976).
II
We noted in Warth, and the Court is careful to reiterate today, ante, at 112, that the Art. Ill standing inquiry often is only the first of two inquiries necessary to determine whether a federal court should entertain a claim at the instance of a particular party. The Art. Ill question is one of power within our constitutional system, as courts may decide only actual cases and controversies between the parties who stand before the court. See Simon v. Eastern Ky. Welfare Rights Org., supra, at 41-42. Beyond this question, however, lies the further and less easily defined inquiry of whether it is prudent to proceed to decision on particular issues even at the instance of a party whose Art. Ill standing is clear. This inquiry has taken various forms, including the one presented by this case: whether, in defending against or anticipatorily attacking state action, a party may argue that it contravenes someone else's constitutional rights.2
*124This second inquiry is a matter of “judicial self-governance.” Warth v. Seldin, supra, at 509. The usual — and wise — stance of the federal courts when policing their own exercise of power in this manner is one of cautious reserve. See generally Ashwander v. TV A, 297 U. S. 288, 346-348 (1936) (Brandéis, J., concurring) . This caution has given rise to the general rule that a party may not defend against or attack governmental action on the ground that it infringes the rights of some third party, ante, at 114, and to the corollary that any exception must rest on specific factors outweighing the policies behind the rule itself.3 See Barrows v. Jack*125son, 346 U. S. 249, 257 (1953); cf. generally United States v. Richardson, 418 U. S. 166, 188-197 (1974) (Powell, J., concurring).
The plurality acknowledges this general rule, but identifies “two factual elements” — thought to be derived from prior cases — that justify the adjudication of the asserted third-party rights: (i) obstacles to the assertion by the third party of her own rights, and (ii) the existence of some “relationship” such as' the one between physician and patient. In my view these factors do not justify allowing these physicians to assert their patients’ rights.
A
Our prior decisions are enlightening. In Barrows v. Jackson, supra, a covenantor who breached a racially restrictive covenant by -selling to Negroes was permitted to set up the buyers’ rights to equal protection hr defense against a damages action by the covenantees. See Shelley v. Kraemer, 334 U. S. 1 (1948). The Court considered the general rule outweighed by “the need to protect [these] fundamental rights” in a situation “in which it would be difficult if not impossible for the persons whose rights are asserted to present their grievance before any court.” 346 U. S., at 257. It would indeed have been difficult if not impossible for the rightholders to assert their own rights: the operation of the restrictive covenant and the threat of damages actions for its breach tended to insure they would not come into possession of the land, and there was at the time little chance of a successful suit based on a covenantor’s failure to sell to them. In a second case, NAACP v. Alabama, 357 U. S. 449 (1958), an organization was allowed to resist an order to produce its membership list by asserting the associational rights *126of its members to anonymity because, as the plurality notes, ante, at 116, the members themselves would have had to forgo the rights in order to assert them. And in Eisenstadt v. Baird, 405 U. S. 438 (1972), the Court considered it necessary to relax the rule and permit a distributor of contraceptives to assert the constitutional rights of the recipients because the statutory scheme operating to deny the contraceptives to the recipients appeared to offer them no means of challenge. Id., at 446.
The plurality purports to derive from these cases the principle that a party may assert another’s rights if there is “some genuine obstacle” to the third party’s own litigation. Ante, at 116. But this understates the teaching of those cases: On their facts they indicate that such an assertion is proper, not when there is merely some “obstacle” to the rightholder’s own litigation, but when such litigation is in all practicable terms impossible. Thus, in its framing of this principle, the plurality has gone far beyond our major precedents.
Moreover, on the plurality’s own statement of this principle and on its own discussion of the facts, the litigation of third-party rights cannot be justified in this case. The plurality virtually concedes, as it must, that the two alleged “obstacles” to the women’s assertion of their rights are chimerical. Our docket regularly contains cases in which women, using pseudonyms, challenge statutes that allegedly infringe their right to exercise the abortion decision. Nor is there basis for the “obstacle” of incipient mootness when the plurality itself quotes from the portion of Roe v. Wade, 410 U. S. 113, 124-25 (1973), that shows no such obstacle exists. In short, in light of experience which we share regularly in reviewing appeals and petitions for certiorari, the “obstacles” identified by the plurality as justifying departure from the general rule *127simply are not significant. Rather than being a logical descendant of Barrows, NAACP, and Eisenstadt, this case is much closer to Warth v. Seldin, supra, in which taxpayers were refused leave to assert the constitutional rights of low-income persons in part because there was no obstacle to those low-income persons' asserting their own rights in a proper case.4 See 422 U. S., at 509-510; cf. McGowan v. Maryland, 366 U. S. 420, 430 (1961).
B
The plurality places primary reliance on a second element, the existence of a “confidential relationship” between the rightholder and the party seeking to assert her rights.5 Focusing on the professional relationships *128present in Griswold, Doe and Planned Parenthood of Missouri v. Danforth, ante, p. 52, the plurality suggests that allowing the physicians in this case to assert their patients’ rights flows naturally from those three. Indeed, its conclusion is couched in terms of the general appropriateness of allowing physicians to assert the privacy interests of their patients in attacks on “governmental interference with the abortion decision.” Ante, at 115, 118.
With all respect, I do not read these cases as merging the physician and his patient for constitutional purposes. The principle they support turns not upon the confidential nature of a physician-patient relationship but upon the nature of the State’s impact upon that relationship. In each instance the State directly interdicted the normal functioning of the physician-patient relationship by criminalizing certain procedures. In the circumstances of direct interference, I agree that one party to the relationship should be permitted to assert the constitutional rights of the other, for a judicial rule of self-restraint should not preclude an attack on a State’s proscription of constitutionally protected activity. See also Meyer v. Nebraska, 262 U. S. 390 (1923). But Missouri has not directly interfered with the abortion decision — neither the physicians nor their patients are forbidden to engage *129in the procedure.6 The only impact of § 208.152 (12) is that, because of the way Missouri chose to structure its Medicaid payments, it causes these doctors financial detriment. This affords them Art. Ill standing because they aver injury in fact, but it does not justify abandonment of the salutary rule against assertion of third-party rights.
C
The physicians have offered no special reason for allowing them to assert their patients’ rights in an attack on this welfare statute, and I can think of none. Moreover, there are persuasive reasons not to permit them to do so. It seems wholly inappropriate, as a matter of judicial self-governance, for a court to reach unnecessarily to decide a difficult constitutional issue in a case in which nothing more is at stake than remuneration for professional services. And second, this case may well set a precedent that will prove difficult to cabin. No reason immediately comes to mind, after today’s holding, why any provider of services should be denied standing to assert his client’s or customer’s constitutional rights, *130if any, in an attack on a welfare statute that excludes from coverage his particular transaction.7
Putting it differently, the Court’s holding invites litigation by those who perhaps have the least legitimate ground for seeking to assert the rights of third parties. *131Before today I certainly would not have thought that an interest in being compensated for professional services, without more, would be deemed a sufficiently compelling reason to justify departing from a rule of restraint that well serves society and our judicial system. The Court quite recently stated, with respect to the rule against assertion of third-party rights as well as certain other doctrines of judicial self-restraint, that “[t]hese principles rest on more than the fussiness of judges. They reflect the conviction that under our constitutional system courts are not roving commissions assigned to pass judgment on the validity of the Nation’s laws. . . . Constitutional judgments . . . are justified only out of the necessity of adjudicating rights in particular cases between the litigants brought before the Court.” Broadrick v. Oklahoma, 413 U. S., at 610-611 (citation omitted). Today’s holding threatens to make just such “roving commissions” of the federal courts.
As the Court notes, ante, at 109-110, Missouri has structured its Medicaid system so that payments for medical services are made directly to the physician rather than to the patient.
The inquiry also has been framed, in appropriate cases, as whether a person with Art. Ill standing is asserting an interest arguably within the zone of interests intended to be protected by the constitutional or statutory provision on which he relies, see, e. g., Data Processing Service v. Camp, 397 U. S. 150, 153-156 (1970), or whether a person should be allowed to attack a statute, not on the ground that it is unconstitutional as applied to him, but that it would be unconstitutional as applied to third parties, see, e. g., United States v. Raines, 362 U. S. 17 (1960); Dombrowski v. Pfister, 380 U. S. 479, 486-488 (1965); Broadrick v. Oklahoma, *124413 U. S. 601, 611-618 (1973). Cf. generally United States v. Richardson, 418 U. S. 166, 196 n. 18 (1974) (Powell, J., concurring).
1 agree with the plurality, ante, at 113-114, that a fundamental policy behind the general rule is a salutary desire to avoid unnecessary constitutional adjudication. See Ashwander v. TV A, 297 U. S., at 346-348 (Brandeis, J., concurring). The plurality perceives a second basis for the rule in the courts' need for effective advocacy. While this concern is relevant, it should receive no more emphasis in this context .than in the context of Art. Ill standing requirements. There the need for effective advocacy or a factual sharpening of issues long was the touchstone of discussion. See Baker v. Carr, 369 U. S. 186, 204 (1962); Flast v. Cohen, 392 U. S. 83, 99 (1968). Perhaps a more accurate formulation of the Art. Ill limitation — one consistent with the concerns underlying the constitutional provision — is that the plaintiff’s stake in a controversy must insure that exercise of the court’s remedial powers is both necessary and sufficient to give him relief. See Warth v. Seldin, 422 U. S. 490, 498-499, 508 (1975); Simon v. Eastern Ky. Welfare Rights Org., 426 U. S. 26, 38, and n. 16 (1976). The Court today uses this formulation. Ante, at 112-113. A similar focus upon the proper judicial role, rather than quality of advocacy, is preferable in the area of prudential limitations upon judicial power. See Warth v. Seldin, supra, at 498; cf. Schlesinger v. Reservists to Stop the War, 418 U. S. 208, 225-226 (1974).
Congress by statute may foreclose any inquiry into competing policy considerations and give a party with Art. Ill standing the *125right to assert the interests of third parties or even the public interest. See Warth v. Seldin, supra, at 500-501.
The plurality retrospectively analyzes the facts in Banows, NAACP, and Eisenstadt in an effort to show that litigation by the rightholders was possible in each case. Ante, at 116 n. 6. While this technically may be true, it also is true that the Court in Barrows and NAACP expressly emphasized the extreme difficulty of such litigation. Moreover, the plurality underestimates the difficulty confronting a would-be Negro vendee in Barrows who attempted to prove that race alone blocked his deal with a covenantor. And the plurality denigrates the difficulty of the NAA.CP members’ assertion of their own right to anonymity when in the text on the same page it quotes, approvingly, the very language in the NAACP case expressing the difficulty of such litigation. As for Eisenstadt, allowing the assertion of third-party rights there was justified not only because of the difficulty of rightholders’ litigation, but also because the State directly interdicted a course of conduct that allegedly enjoyed constitutional protection. As explained infra, Part II-B, the Court rightly shows special solicitude in that situation.
In any event, as argued above in the text, my basic disagreement with the plurality rests on the facts of this case, and the application of the plurality’s own test — “some genuine obstacle” to the right-holder’s assertion of her own rights. There simply is no such obstacle here.
The plurality’s primary emphasis upon this relationship is in *128marked contrast to the Court’s previous position that the relationship between litigant and rightholder was subordinate in importance to “the impact of the litigation on the third-party interests.” Eisenstadt v. Baird, 405 U. S. 438, 445 (1972). I suspect the plurality’s inversion of the previous order results from the weakness of the argument that this litigation is necessary to protect third-party interests. I would keep the emphasis where it has been before, and would consider the closeness of any “relationship” only as a factor imparting confidence that third-party interests will be represented adequately in a case in which allowing their assertion is justified on other grounds. Cf. n. 2, supra.
The plurality contends that assertion of third-party rights has been allowed where “the interference was no more direct than it is here,” ante, at 118 n. 7, and cites Pierce v. Society of Sisters, 268 U. S. 510 (1925). Pierce is of little or no precedential value since the Court did not address — or even mention — the issue of third-party rights in that case. More importantly, however, the interference with the normal functioning of the private school-parent relationship was as complete as if it had been proscribed: as the statute required that children be sent “ 'to a public school for the period of time a public school shall be held during the current year,’ ” id., at 530, there was no practical way for parents to send their children to private schools. As the Court- noted, “[t]he inevitable practical result of enforcing the Act . . . would be destruction of appellees' primary schools,’ and perhaps all other private primary schools for normal children within the State of Oregon.” Id., at 534.
The plurality says it is proceeding “by assessing relevant factors in individual cases . . . , rather than by adopting a set of per se rules,” and implies that I am advocating the latter course. Ante, at 119 n. 7. The fact is that I have not proposed any such set of rules. Rather, my dissent is grounded in the decisions of the Court from which I believe today’s holding departs.
By divining from previous cases two factors, and two factors alone, whose application to the facts of this case “quickly yields its proper result,” ante, at 117, the plurality appears to have articulated a new rule of third-party standing that leaves little room for flexibility. The ease with which the plurality would allow assertion of such standing in this case — based on nothing more substantial than a professional (or perhaps only an abortion-clinic) relationship and dimly perceived “obstacles” to the rightholder’s own litigation — suggests that “the proper result” usually will be third-party standing.
The plurality’s attempt to distinguish this case from the next one involving another provider of services is not reassuring. Three distinguishing factors are suggested. The first one, a “confidential” relationship, is analytically empty (especially when one recognizes that, realistically, the “confidential” relationship in a case of this kind often is set in an assembly-line type abortion clinic). Moreover, it is unsupported by nearly half of the’cases the plurality relies upon in finding “relationship” one of the two elements yielding third-party standing: there was no “confidential” relationship in Barrows or Eisenstadt — or, so far as the opinion shows, with respect to one of the defendants in Griswold. The second suggested distinction is that the woman’s right in this case “is one that may be impaired by its assertion.” I do not understand how a woman’s litigation over her right to malee an abortion decision impairs her ability to make that decision. Finally, the plurality falls back on the contention that the woman’s claim here is “imminently moot,” a point which the plurality’s own citation to Roe proves to be irrelevant. As these three “distinctions” seem insubstantial, I repeat: Today’s holding will be difficult to cabin.