dissenting.

With great reluctance, I have concluded that I am unable to concur in the majority *1452 opinion. Like the majority, I believe that homosexuals have been unfairly treated both historically and in the United States today. Were I free to apply my own view of the meaning of the Constitution and in that light to pass upon the validity of the Army’s regulations, I too would conclude that the Army may not refuse to enlist homosexuals. I am bound, however, as a circuit judge to apply the Constitution as it has been interpreted by the Supreme Court and our own circuit, whether or not I agree with those interpretations. Because of this requirement, I am sometimes compelled to reach a result I believe to be contrary to the proper interpretation of constitutional principles. This is, regrettably, one of those times.

I.

In this case we consider the constitutionality of a regulation which bars homosexuals from enlisting in the Army. Sergeant Perry Watkins challenges that regulation under the Equal Protection Clause. The majority holds that homosexuals are a suspect class, and that the regulation cannot survive strict scrutiny. Because I am compelled by recent Supreme Court and Ninth Circuit precedent to conclude first, that homosexuals are not a suspect class and second, that the regulation survives both rational and intermediate level scrutiny, I must dissent.

Bowers v. Hardwick, 478 U.S. 186 , 106 S.Ct. 2841 , 92 L.Ed.2d 140 (1986), is the landmark case involving homosexual conduct. In Hardwick , the Supreme Court decided that homosexual sodomy is not pro: tected by the right to privacy, and thus that the states are free to criminalize that conduct. Because Hardwick did not challenge the Georgia sodomy statute under the Equal Protection Clause, and neither party presented that issue in its briefs or at oral argument, the Court limited its holding to due process and properly refrained from reaching any direct conclusion regarding an equal protection challenge to the statute. 1 See id. at 2846 n. 8. However, the fact that Hardwick does not address the equal protection question directly does not mean that the case is not of substantial significance to such an inquiry.

An important part of the function of circuit court judges is to interpret the Supreme Court’s opinions. In doing so, we must attempt to understand the principles underlying those opinions, so that we may determine how past decisions affect subsequent cases. With respect to Hardwick , the májority balks at performing this task. Instead, it states: “the Hardwick Court simply did not address either the question whether heterosexual sodomy also falls outside the scope of the right to privacy or the separate question whether homosexual but not heterosexual sodomy may be criminalized without violating the equal protection clause.” Maj. op. at 1439. The duty to interpret Supreme Court precedent cannot be so easily avoided. Logic and reason are among the tools available to judges who wish to determine the meaning of cases.

The answer to the meaning of Hardwick is not difficult to find. There are only two choices: either Hardwick is about “sodomy”, and heterosexual sodomy is as constitutionally unprotected as homosexual sodomy, or it is about “homosexuality”, and there are some acts which are protected if done by heterosexuals but not if done by homosexuals. In applying the opinion to future cases our first effort must be to decide which of the two propositions Hard-wick stands for. 2 Although the majority *1453 refuses to acknowledge that it is making a choice, there can be no doubt that it does so. The sentence after the text quoted above reads: “We cannot read Hardwick as standing for the proposition that government may outlaw sodomy only when committed by a disfavored class of persons.” 3 Maj. op. at 1439. By expressly rejecting the “homosexuality” option, the majority implicitly but necessarily selects the “sodomy” alternative. I do not believe that Hardwick can reasonably be so construed.

In my opinion, Hardwick must be read as standing precisely for the proposition the majority rejects. To put it simply, I believe that after Hardwick the government may outlaw homosexual sodomy even though it fails to regulate the private sexual conduct of heterosexuals. In Hardwick the Court took great care to make clear that it was saying only that homosexual sodomy is not constitutionally protected, and not that all sexual acts — both heterosexual and homosexual — that fall within the definition of sodomy can be prohibited.

The Georgia statute at issue in Hard-wick on its face barred all acts of sodomy. The Court could simply have upheld the statute without even mentioning the word “homosexual”. Instead it carefully crafted its opinion to proscribe and condemn only homosexual sodomy. While it can be argued that the Court was faced with only a homosexual sodomy case, under the majority’s theory the fact that the particular act of sodomy was homosexual in nature is of no significance. According to the majority, the race and sexual preference of the defendant are equally irrelevant. The majority says: “Surely, for example, Hardwick cannot be read as a license to outlaw sodomy only when committed by blacks.” Maj. op. at 1439. Surely not. And surely, had Hardwick been black rather than a homosexual, the Court would not, throughout its opinion, have written about “black sodomy” or black sodomists. It would simply have written about sodomy. Here, however, from the Court’s standpoint the crucial fact was that Hardwick was a homosexual. For that reason, throughout its opinion the Court wrote about “homosexual sodomy”.

It is significant that whatever one may think of the soundness of Hardwick’s assumptions or conclusions, the decision came as no surprise to those familiar with the rulings of the lower federal courts on the subject of homosexual rights. Well before Hardwick , this court, along with most other federal courts, had concluded that the Supreme Court had determined that the right to privacy was inapplicable to homosexual conduct. As we said in Better v. Middendorf, 632 F.2d 788 , 809-10 (9th Cir.1980), “Most federal courts ... have understood the holding [in Doe v. Commonwealth’s Attorney, 425 U.S. 901 , 96 S.Ct. 1489 , 47 L.Ed.2d 751 (1976)] to be that homosexual conduct does not enjoy special constitutional protection under the due process clause.”

The anti-homosexual thrust of Hardwick , and the Court’s willingness to condone anti-homosexual animus in the actions of the government, are clear. A prominent constitutional scholar makes this point succinctly. Professor Laurence Tribe, after strongly criticizing the Court’s holding and reasoning in Hardwick , states that the “ ‘good news’ about the Court’s decision” is that it was so clearly based on prejudice against homosexuals that it “may therefore pose less of a threat to other privacy precedents than would otherwise be the case.” Constitutional Law, supra, § 15-21, at 1430. Justice Blackmun characterized the decision as being “obsessively focus[ed] on homosexual activity”, and “proceed[ing] on the assumption that homosexuals are so different from other citizens that their lives may be controlled in a way that would not be tolerated if it limited the choices of ... other citizens.” Hardwick, 106 S.Ct. at 2849 (Blackmun, J., dissenting). Indeed, it is hard to find any basis in the Court’s *1454 opinion for interpreting it the way the majority chooses: the Court says explicitly that the statute is justified by “majority sentiments about homosexuality”, 106 S.Ct. at 2846 , not by “majority sentiments about sodomy”.

• My colleagues’ interpretation of Hard-wick is not only unsound, it also unnecessarily and incorrectly increases — exponentially — the damage to the right to privacy caused by Hardwick . While in Hardwick the Court made it clear that homosexual conduct is not protected by the right to privacy, the Court has never held that the government has the authority to regulate the private heterosexual acts of consenting adults. See Griswold v. Connecticut, 381 U.S. 479 , 85 S.Ct. 1678 , 14 L.Ed.2d 510 (1965); Eisenstadt v. Baird, 405 U.S. 438 , 92 S.Ct. 1029 , 31 L.Ed.2d 349 (1972); Carey v. Population Services International, 431 U.S. 678 , 97 S.Ct. 2010 , 52 L.Ed.2d 675 (1977); see also Hardwick, 106 S.Ct. at 2857-58 (Stevens, J., dissenting). To the contrary, it has expressly stated that “intimate relationships” (though apparently only of the heterosexual variety) are constitutionally protected. See Board of Directors of Rotary International v. Rotary Club, — U.S.-, 107 S.Ct. 1940, 1945 , 95 L.Ed.2d 474 (1987). Reading Hardwick as implicitly permitting the regulation of heterosexual conduct, as the majority’s analysis forces it to do, constitutes a serious retreat in the privacy area. 4 If the majority’s interpretation of Hardwick were correct, states could, for example, criminalize the act of oral sex when engaged in by heterosexuals, including married couples, and in fact would be required to do so if they wished to criminalize homosexual sodomy. Moreover, states would be required to enforce these statutes equally, against heterosexuals and homosexuals alike, 5 a practice not heretofore common in our society. I view the Constitution differently than the majority apparently does: I believe the Constitution protects most, if not all, private heterosexual acts between consenting adults.

The majority opinion undermines the right to privacy in another way. In its eagerness to promote its equal protection analysis and to bolster its characterization of Hardwick as an anti-privacy decision, it terms equal protection more objective and more democratic than substantive due process, which it describes as “value-based line-drawing” arising not from the Constitution itself but from “judicial fiat”. Maj. op. at 1440. It is not necessary to denigrate the right to privacy in order to appreciate the importance of the equal protection clause. The majority’s attack on substantive due process is unreasoned and unjustified. 6 As the Supreme Court has made clear on numerous occasions, the right to privacy is a fundamental part of our constitutional protections, originating in the First, Third, Fourth, Fifth, and Ninth Amendments, and of course the due pro *1455 cess clause of the Fourteenth Amendment. See, e.g., Griswold v. Connecticut, 381 U.S. at 484-85 , 85 S.Ct. at 1681-82 . The protections guaranteed by the right to privacy are no less central to the Constitution than those guaranteed by the equal protection clause. See generally Note, “Process, Privacy, and the Supreme Court”, 28 B.C.L.Rev. 691 (1987). Also, notwithstanding the views of Dean Ely, see maj. op. at 1440 ns. 21, 22, most commentators agree that equal protection analysis is no more objective and no less difficult to apply than substantive due process analysis. See, e.g., Tribe, “The Puzzling Persistence of Process-Based Constitutional Theory”, 89 Yale L.J. 1063 (1980); Westen, “The Empty Idea of Equality”, 95 Harv.L.Rev. 537 (1982). Unlike the majority, I believe we should afford both these fundamental constitutional protections full and equal dignity.

II.

The majority opinion concludes that under the criteria established by equal protection case law, homosexuals must be treated as a suspect class. Maj. op. at 1444-48. Were it not for Hardwick (and other cases discussed infra), I would agree, for in my opinion the group meets all the applicable criteria. See, e.g., Note, “The Constitutional Status of Sexual Orientation: Homosexuality as a Suspect Classification”, 98 Harv.L.Rev. 1285 (1985). However, after Hardwick , we are no longer free to reach that conclusion. 7

The majority opinion treats as a suspect class a group of persons whose defining characteristic is their desire, predisposition, or propensity to engage in conduct that the Supreme Court has held to be constitutionally unprotected, an act that the states can — and approximately half the states have 8 — criminalized. 9 Homosexuals are different from groups previously afforded protection under the equal protection clause in that homosexuals are defined by their conduct — or, at the least, by their desire to engage in certain conduct. With other groups, such as blacks or women, there is no connection between particular conduct and the definition of the group. When conduct that plays a central role in defining a group may be prohibited by the state, it cannot be asserted with any legitimacy that the group is specially protected by the Constitution. 10

Sodomy is an act basic to homosexuality. In the relevant state statutes, sodomy is usually defined broadly to include “any sexual act involving the sex organs of one person and the mouth or anus of another.” See, e.g., Hardwick, 106 S.Ct. at 2842 n. 1. The practices covered by this definition are, not surprisingly, the most common sexual practices of homosexuals. Specifically, oral sex is the primary form of homosexual activity. See A. Bell & M. Weinberg, Homosexualities 106-11, 327-30 (1978). When the Supreme Court declares that an *1456 act that is done by a vast majority of a group’s members and is fundamental to their very nature can be criminalized and further states that the basis for such criminalization is “the presumed belief of a majority of the electorate ... that [the practice] is immoral and unacceptable” 11 , I do not think that we are free, whatever our personal views, to describe discriminatory treatment of the group as based on “unreasoning prejudice”. See maj. op. at 1444-45. Rather we are obligated to accept the Supreme Court’s conclusion that what the majority of this panel calls “unreasoning prejudice” is instead a permissible societal moral judgment.

I have already explained the principal reasons why the majority’s interpretation of Hardwick as covering heterosexual sodomy is not only incorrect but also damaging to constitutional principles. I must now add that the majority errs for another important reason. The majority states that the equal protection clause requires the government (if it wishes to criminalize homosexual sodomy) to prohibit all persons from engaging in “the proscribed sexual acts”. Maj. op. at 1439. This analysis affords equal treatment only in the most superficial meaning of the term. Government actions, neutral on their face, can sometimes have distinctly unequal effects, and carry implicit statements of inequality. See L. Tribe, Constitutional Choices 238-45 (1985). Laws against sodomy do not affect homosexuals and heterosexuals equally. Homosexuals are more heavily burdened by such legislation, even if we ignore the governmental tendency to prosecute general sodomy statutes selectively against them. See Hardwick, 106 S.Ct. 2850 n. 2 (Blackmun, J., dissenting). Oral sex, a form of sodomy, is the primary form of sexual activity among homosexuals; however, sexual intercourse is the primary form of sexual activity among heterosexuals. 12 If homosexuals were in fact a suspect class, a statute criminalizing both heterosexual sodomy and homosexual sodomy would still not survive equal protection analysis. For the prohibition to be equal, the government would have to prohibit sexual intercourse — conduct as basic to heterosexuals as sodomy is to homosexuals. 13 This, obviously, the government would not and could not do. Therefore, if equal protection rules apply (i.e. if homosexuals are a suspect class), a ban on homosexual sodomy could not stand no matter how the statute was drawn. Hardwick makes it plain that the contrary is true.

Finally, the “protection” of homosexual rights provided by the majority opinion is hollow indeed. The majority unwittingly denigrates the equal protection clause as well as the right to privacy. Until now, a “suspect class” has been a group whose members were afforded special solicitude. That is patently not the case with respect to homosexuals. Many states deny that group the right to engage in their most fundamental form of sexual activity. A “life without any physical intimacy”, Hardwick, 106 S.Ct. at 2850 n. 2 (Blackmun, J., dissenting), is hardly the life contemplated for our citizens by the Declaration of Independence (“the pursuit of happiness”) or, one would have thought, by the Constitution. While Hardwick may not wholly preclude the possibility of lawful physical intimacy for homosexuals, it drastically limits that right. To proclaim that under these circumstances homosexuals are afforded special protection by the Constitution would be hypocritical at best.

*1457 Before concluding my discussion of Hardwick, I wish to record my own view of the opinion. I have delayed doing so until I have applied the case as I believe we have a duty to apply it. Now, I must add that as I understand our Constitution, a state simply has no business treating any group of persons as the State of Georgia and other states with sodomy statutes treat homosexuals. In my opinion, invidious discrimination against a group of persons with immutable characteristics can never be justified on the grounds of society’s moral disapproval. No lesson regarding the meaning of our Constitution could be more important for us as a nation to learn. I believe that the Supreme Court egregiously misinterpreted the Constitution in Hard-wick. In my view, Hardwick improperly condones official bias and prejudice against homosexuals, and authorizes the criminalization of conduct that is an essential part of the intimate sexual life of our many homosexual citizens, a group that has historically been the victim of unfair and irrational treatment. I believe that history will view Hardwick much as it views Plessy v. Ferguson, 163 U.S. 537 , 16 S.Ct. 1138 , 41 L.Ed. 256 (1896). And I am confident that, in the long run, Hardwick, like Plessy , will be overruled by a wiser and more enlightened Court. See Hardwick, 106 S.Ct. at 2856 (Blackmun, J., dissenting).

The decision in Hardwick has not affected my firm belief that the Constitution, properly interpreted, does afford homosexuals the same protections it affords other groups that are historic victims of invidious discrimination. Nevertheless, for the reasons I have already stated, it is my obligation to follow Hardwick as long as it has precedential force — and for now it does.

III.

Even if the majority’s analysis could survive Hardwick, we would be precluded by our own circuit precedent from concluding that homosexuals are a suspect class. In Hatheway v. Secretary of Army, 641 F.2d 1376 (9th Cir.1981), we considered a challenge brought by an army officer convicted of sodomy by a general court-martial. We rejected Lieutenant Hatheway’s claim that the practice of prosecuting homosexuals but not heterosexuals under a general sodomy statute was unconstitutional. We stated: “We understand Hatheway’s claim (that the commission of a homosexual act is an impermissible basis for prosecution) to be an equal protection argument.” Id. at 1382 . We then applied intermediate level scrutiny and concluded that the government could single out those who engage in “homosexual acts”. Id. Our determination that strict scrutiny did not apply to Lt. Hatheway’s claim, id., must necessarily be interpreted as meaning that we concluded that homosexuals are not a suspect class.

The majority argues that because our analysis in Hatheway was apparently based on the “fundamental rights branch” of equal protection analysis rather than on the “suspect class branch”, Hatheway does not preclude the holding that homosexuals are a suspect class. Maj. op. at 1441-42. I disagree. The majority’s position is based on too narrow a view of how courts decide constitutional questions and too narrow a view of the extent to which we are bound by constitutional holdings. Had we thought in Hatheway that strict scrutiny was required by a “different branch” of the equal protection clause, it would have been our obligation to apply the higher test. The equal protection issue was squarely presented by Lt. Hatheway. We could not have ruled against him, as we did, and failed to apply a standard under which he might have prevailed unless we believed the higher standard was inapplicable. Nowhere does the opinion state that Lt. Hatheway relied on one particular branch of the doctrine to the exclusion of the other, and we may not fairly make that assumption. Nor are we free to refuse to apply our own precedent simply because the reasoning may be unpersuasive or the explanation less than complete. The holding in Hatheway is clear: intermediate level scrutiny, rather than strict scrutiny, applies to an equal protection claim based on discrimination against homosexuals. Because in Hatheway we recognized the equal protection claim, acknowledged the “three-tier approach”, applied the interme *1458 diate level of scrutiny, and ruled against the plaintiff, I do not believe we can blithely ignore its holding. 14

IV.

Because we are not free to hold that homosexuals are a suspect class, we can not apply strict scrutiny to the Army’s regulations. At the most the regulations must pass intermediate scrutiny — and in Hatheway we decided that the military’s singling out of homosexual conduct for special adverse treatment survives that level of review: applying intermediate level scrutiny we concluded that prosecutions by the military on the basis of sexual preference bear “a substantial relationship to an important government interest.” Id. at 1382 . We then upheld the Army’s discriminatory treatment of Hatheway. We are bound by Hatheway to conclude that military “[classifications which are based solely on sexual preference” survive an intermediate level of review. 15 Id.

Courts must give special deference when adjudicating matters involving the military. Goldman v. Weinberger, 475 U.S. 503 , 106 S.Ct. 1310 , 89 L.Ed.2d 478 (1986); Rostker v. Goldberg, 453 U.S. 57 , 101 S.Ct. 2646 , 69 L.Ed.2d 478 (1981). In the context of a first amendment challenge, the Supreme Court has recently stated: “Our review of military regulations ... is far more deferential than constitutional review of similar laws or regulations designed for civilian society.” Goldman v. Weinberger, 106 S.Ct. at 1313. In Better v. Middendorf, Judge, now Justice, Kennedy writing for our court -said: “constitutional rights must be viewed in light of the special circumstances of the armed forces”. 632 F.2d at 810-11.

In rejecting the Army’s justifications for the regulation, the majority fails to give proper deference to the Army’s determinations. Its failure to do so may result in part from its unwillingness to recognize the moral judgments regarding homosexuality approved in Hardwick, 106 S.Ct. at 2846 , and deemed permissible, at least for the purpose of military regulations, by Judge Kennedy in Better, 632 F.2d at 811-12. Although I see no merit in the Army’s ideas about homosexuals, its beliefs about the consequences of allowing homosexuals to serve in the Army, and its pandering to negative stereotypes of homosexuals, see maj. op. at 1448-50, we are not permitted to substitute our views for the Army’s “considered professional judgment” as to what kind of persons should be barred from enlisting in order to ensure a disciplined fighting force. 16 Goldman v. Weinberger, 106 S.Ct. at 1313.

After analyzing the various explanations offered by the Army, the majority dismisses the purposes of the regulations as illegitimate or irrational. Maj. op. at 1448-1451. Again, the majority takes a position that is not open to us. For not only have our cases told us we must defer to the military judgment in matters of this kind, *1459 they have upheld the very reasoning the majority now rejects. The justifications advanced by the Army involving negative views about homosexuals and homosexuality have been accepted by earlier decisions of this court as both legitimate and important. Better, 632 F.2d at 811-12; see Hatheway, 641 F.2d at 1381-82 . We are not free to reconsider those prior conclusions unless or until our court as a whole agrees to do so en bane.

It is true that, as the majority says on several occasions, maj. op. at 1439, 1449-50, the Army could not treat blacks as it treats homosexuals and could not base its regulations on negative judgments regarding blacks. No matter how appealing the analogy may be, we are not free to draw it here. Better and Hatheway both approve discriminatory treatment against homosexuals, by the military, based on moral judgments regarding homosexuality. See Belter, 632 F.2d at 811-12; Hatheway, 641 F.2d at 1381-82 ; see also Hardwick, 106 S.Ct. at 2846 . As the majority points out, similar biases against blacks could not form the basis for state action against that group. Palmore v. Sidoti, 466 U.S. 429 , 104 S.Ct. 1879 , 80 L.Ed.2d 421 (1984). Thus, cases regarding blacks are simply irrelevant. 17

Y.

The majority attempts to overcome the problems posed by Hardwick (and, to some extent, by Hatheway) by distinguishing between the class of persons who engage in homosexual acts and the slightly broader class of persons who have a homosexual orientation. Relying on this distinction, the majority also argues that the Army regulation is about status, not conduct. It is unclear whether the majority is arguing that homosexuals are a suspect class simply because they are a class defined by status rather than conduct, or whether it is arguing that the equal protection question is unaffected by Hardwick because that case involves conduct rather than orientation. In either event, I do not believe we can escape the conclusion that “homosexuals”, however defined, cannot qualify as a suspect class.

Even if we define the class as those who have a “homosexual orientation”, its members will consist principally of active, practicing homosexuals. 18 That the class may also include a small number of persons who are or wish to be celibate is irrelevant for purposes of determining whether the group as a whole constitutes a suspect class. I simply see no way to say that homosexuals defined broadly (by status) are a suspect class, but that the same group, if more narrowly defined (by conduct) is not. Whether the group is defined by status or by conduct, its composition is essentially the same. In short, “homosexuals” are either a suspect class or they aren’t. The answer cannot depend on the niceties of the class definition. 19

*1460 What the majority may be arguing is that a regulation targeted at “orientation” is too broad to survive rationality review. However, if the majority is making this argument, there are a number of difficult questions it must answer first. For example, under the majority’s status/conduct distinction, Watkins could be excluded from the Army based on regulations slightly more narrowly drawn so as to target only the class of persons who have engaged in homosexual conduct. If Watkins’ actions fall within that narrower category (and they do), and Watkins is therefore a member of a class of persons that is not constitutionally protected, does he have standing to challenge the constitutionality of these regulations? 20 If he does, would the correct remedy be simply to strike the few words that make the regulations too broad, rather than invalidating all of the regulations? 21 See Brockett v. Spokane Arcades, Inc., 472 U.S. 491 , 105 S.Ct. 2794 , 86 L.Ed.2d 394 (1985). The majority simply does not discuss these and other similarly troublesome questions.

Moreover, I disagree with the majority’s status/conduct distinction, as applied to this case, for another reason. I view the case before us as a conduct case. In my opinion, the facts regarding Watkins clearly demonstrate disqualifying acts and the regulations before us may properly be viewed as conduct regulations. First, Watkins has admitted to engaging in homosexual conduct with other servicemen while in the Army. Those admissions form an integral part of the reasons for the Army’s refusal to permit him to reenlist. Second, the regulations must be construed in light of the Army’s stated policy regarding homosexuality:

Army Regulation 635-200 1115-2. 22 Read in this light, the regulations constitute an attempt to exclude those who engage in or will engage in homosexual acts. The majority makes much of the fact that the regulation allows a soldier an opportunity to prove that a homosexual act he has engaged in was aberrational. Maj. op. at 1437. Contrary to the majority, I do not think that this proves that the regulation is about orientation rather than conduct. The regulation’s exception relates to conduct: it allows the Army to distinguish between soldiers who are likely to engage in homosexual conduct in the future (practicing homosexuals) and those who are not likely to do so (heterosexuals who engage in an isolated homosexual act due to intoxication or some similar reason). I consider the inclusion of this exception in the regulation to constitute a rational exercise of discretion — a legitimate attempt to predict future conduct on the basis of past conduct. However, I see little purpose in analyzing the Army’s regulations in detail here. Suffice it to say that I disagree with the majority’s characterization of them. In my opinion, the regulations are targeted at conduct —past, present, and future, but conduct nonetheless.

In the end the majority’s status/conduct distinction does not advance its cause. 23 *1461 With or without that part of its analysis, the majority’s effort ultimately comes a cropper on Hardwick, Hatheway, Beller, Goldman and Rostker, 24

CONCLUSION

As the majority points out, Sgt. Watkins has every reason to feel aggrieved. His homosexuality has been well known for many years. During that entire period, his army service has been exemplary. Those who have worked with him, including his supervisors, are anxious to see him continue with his military career. Yet, under the Supreme Court’s (and our own circuit’s) interpretation of the Constitution, the Army is free to terminate that career solely because he is a homosexual. There are only three entities which have the authority to afford Sgt. Watkins the relief which I, like the majority, believe a proper interpretation of the Constitution would require. First, the Supreme Court could undo the damage to the Constitution wrought by Hardwick; it could overrule that precedent directly or implicitly. Second, the Army could voluntarily abandon its unfair and discriminatory regulation (or, I would assume, the Department of Defense could direct it to do so). Third, the Congress could enact appropriate legislation prohibiting the armed services from excluding homosexuals. I recognize that from a practical standpoint the existence of these forums may offer Sgt. Watkins little solace. Nevertheless, I do not believe that a panel of the Ninth Circuit may, consistent with its duty to apply precedent properly, afford him the relief he seeks.

For the above reasons, I must reluctantly dissent.

. Cf. L. Tribe, American Constitutional Law § 15-21, at 1431 n. 7 (2nd ed. 1988) [hereinafter, Constitutional Law ]:

In his treatise, Professor Tribe notes, in the interest of full disclosure, that he served as Hardwick's counsel before the Supreme Court.

. Ultimately, as far as the case before us is concerned, the result is the same regardless of which choice we make. See note 10, infra. *1453 However, the analysis by which we reach that result differs substantially.

. But see Baker v. Wade, supra note 1, where the Fifth Circuit, sitting en banc, considered and rejected the claims that homosexuals are a suspect class and that a statute that criminalized homosexual sodomy but not heterosexual sodomy violates equal protection.

.The majority's disingenuous statement that it reads Hardwick neither one way nor the other on this point, maj. op. at 1439 n. 20, is simply neither logical nor credible. Only if heterosexual sodomy is not protected by the right to privacy could the majority’s equal protection argument conceivably have any validity. If the right to privacy does apply in the case of heterosexuals who, for example, engage in oral sex, then the equal protection clause obviously does not require equal treatment of homosexual and heterosexual sexual conduct. The clause is procedural in nature and cannot afford substantive rights to a particular group when the Constitution does not otherwise provide them. To put it differently, if one group's sexual conduct is protected by the right to privacy and the other’s is not, it is the Constitution itself that distinguishes between the treatment the two groups constitutionally receive. We cannot, then, use the equal protection clause to say that the two groups must be treated identically with respect to that conduct. Since the Court has already held that homosexuals are not protected by the privacy provisions of the Constitution when they commit sodomy, the right to equal treatment exists only if heterosexuals are similarly unprotected. Thus the majority’s argument that equal protection applies is necessarily premised on the view that heterosexual sodomy, including oral sex between married couples, is not protected by the right to privacy and may be criminalized.

. See Yick Wo v. Hopkins, 118 U.S. 356 , 6 S.Ct. 1064 , 30 L.Ed. 220 (1886).

. That Dean John Hart Ely has “severely criticized” Roe v. Wade, maj. op. at 1440 n. 21, makes the majority’s position no more persuasive and Roe v. Wade no less binding or important a constitutional decision.

. See Constitutional Law, supra, § 16-33, at 1616 n. 47:

I note that Professor Tribe’s pessimistic forecast relates to intermediate scrutiny and not the even stricter standard that the majority today holds applicable.

. See Hardwick, 106 S.Ct. at 2845 .

. Cf. Plyler v. Doe, 457 U.S. 202 , 219 n. 19, 102 S.Ct. 2382 , 2396 n. 19, 72 L.Ed.2d 786 (1982) (rejecting the claim that illegal aliens are a suspect class, in part because the defining element of the class is a criminal act).

. Thus, it is not even necessary to decide whether the majority’s view of Hardwick — that it is based on a condemnation of sodomy rather than of homosexuality — is correct. Whatever the explanation for the Court’s willingness to allow sodomy to be criminalized — whether its decision is based on its views as to the morality of homosexuality or on its disapproval of sodomy, including the heterosexual variety — that willingness is inconsistent with affording special constitutional protection to homosexuals — a group whose primary form of sexual activity, the Court tells us, may be declared criminal.

. Hardwick, 106 S.Ct. at 2846 .

. Oral sex, though practiced by a substantial majority of heterosexuals, is not the primary sexual activity for that group. See W. Masters, V. Johnson & R. Kolodny, Human Sexuality 388-92, 418-22 (2nd ed. 1985).

. Statutes which are neutral on their face survive equal protection scrutiny unless they are the product of discriminatory intent. However, as the Supreme Court has noted, "when a neutral law has a disparate impact upon a group that has historically been the victim of discrimination, an unconstitutional purpose may still be at work. ... Certainly, when the adverse consequences of a law upon an identifiable group are ... inevitable ... a strong inference that the adverse effects were desired can reasonably be drawn.” Personnel Administrator of Massachusetts v. Feeney, 442 U.S. 256, 273 , 279 n. 25, 99 S.Ct. 2282, 2293 , 2296 n. 25, 60 L.Ed.2d 870 (1979).

. Moreover, the majority’s assumption that Hatheway relied solely on the "fundamental rights branch" may be based on an overly-simplified view of equal protection doctrine. While many equal protection cases clearly fall exclusively within one branch or the other, some, like Hatheway , do not. The facts in Hatheway involved both an activity to which special protection may be applicable — private sexual activity —and a group for which special protection may be required — homosexuals. In addition, the group is defined by its connection to the activity. Thus, Hatheway presents the paradigmatic circumstances that call for a mixture or merger of the two branches. It appears that we may, intentionally or inadvertently, have used that approach in that case. See Hatheway, 641 F.2d at 1382 .

. The only real question regarding Hatheway is whether our holding that mid-level scrutiny is applicable survives Hardwick. Hatheway established a ceiling for the appropriate level of scrutiny; Hardwick makes it clear, at a minimum, that the ceiling can be no higher.

.To the claim in Goldman v. Weinberger that the military's regulation as applied was irrational and without empirical support, the Court stated:

106 S.Ct. at 1314.

. I am not suggesting, by any means, that all discriminatory statutes affecting homosexuals are valid. We are here dealing only with military regulations. Other governmental action, including state statutes, would still be subject to examination under a number of constitutional principles, including the equal protection clause. As far as that clause is concerned, for purposes of this case it is necessary for me to conclude only that strict scrutiny is not the proper standard. See note 15, supra.

. The majority appears to be unwilling to acknowledge this point. See maj. op. at 1438 n. 14. However, the fact that homosexuals (or persons of "homosexual orientation”) engage in or seek to engage in homosexual conduct is as unremarkable as the fact that "heterosexuals" (or persons of "heterosexual orientation") engage in or seek to engage in heterosexual conduct. To pretend that homosexuality or heterosexuality is unrelated to sexual conduct borders on the absurd. What distinguishes the class of homosexuals from the class of heterosexuals is not some vague "range of emotions", but the nature of the member’s sexual proclivities or interests.

.Nowhere in equal protection jurisprudence can there be found a protected class that is merely a slightly broader form of an unprotected class. In the end, the majority’s distinction between status and conduct comes to nought. For if homosexuals were truly a suspect class, an Army regulation based on conduct would be as unconstitutional as one based on status. See maj. op. at 1439 (“We cannot read Hardwick as standing for the proposition that government may outlaw sodomy only when committed by a disfavored class of persons.”).

. See, e.g., Hatheway, 641 F.2d at 1382-83 (person accused of homosexual sodomy lacks standing to challenge the constitutionality of applying a sodomy law to heterosexuals).

. For example, the underlined words could be removed from the regulation's definition of "homosexual”: "Homosexual means a person, regardless of sex, who engages in, desires to engage in, or intends to engage in homosexual acts." Army Regulation 635-200 If 15-2(a) (emphasis added).

. The statement of policy is from the section' "Separation for Homosexuality". However, there is no reason to believe that it would not be equally applicable to the section which states that homosexuality is a nonwaivable disqualification for reenlistment. Both sections use the same definition of "homosexual”.

.I do not reach the question whether that distinction is relevant for purposes other than criminal law. Cf. Robinson v. California, 370 U.S. 660 , 82 S.Ct. 1417 , 8 L.Ed.2d 758 (1962) (criminal penalties must be based on some act; they may not be inflicted merely on the basis of a person’s condition).

. As the majority acknowledges, its conclusions are also contrary to those of several other circuits. In particular, see Padula v. Webster, 822 F.2d 97 (D.C.Cir.1987), and maj. op. at 1443 & n. 25.