concurring in part and dissenting in part.
I agree with the majority that the plaintiffs in this case may not assert a cause of action implied directly from the Constitution, although I disagree with the majority’s disposition of the state law claims. The reasoning that leads me to agree on the federal question, however, varies from that of the majority in that it focuses on a different element of the test for inference of a constitutional damages remedy. It is for this reason that I write separately to explain my views.
I.
When the Supreme Court, in Bivens v. Six Unknown Named Agents of the Federal Bureau of Narcotics, 403 U.S. 388, 91 S.Ct. 1999, 29 L.Ed.2d 619 (1971), first recognized the entitlement, in the absence of statute, to recover damages for violations of constitutional rights committed by federal officials, it acknowledged that not all constitutional infractions could be redressed in this manner. Specifically, the Court suggested that a damage action would fail if either of two conditions obtained. First, there might exist “special factors counselling hesitation in the absence of affirmative action by Congress.” Id. at 396, 91 S.Ct. at 2005. Second, Congress might have declared that victims who suffer a particular type of constitutional injury at the hands of federal officials have no right to recover money damages, “but must instead be remitted to another remedy, equally effective in the view of Congress.” Id. at 397, 91 S.Ct. at 2005. Subsequent eases have reaffirmed these limitations on the Bivens-type remedy. See Davis v. Passman, 442 U.S. 228, 245, 99 S.Ct. 2264, 2276, 60 L.Ed.2d 846 (1979) ; Carlson v. Green, 446 U.S. 14, 18-19, 100 S.Ct. 1468, 1471-1472, 64 L.Ed.2d 15 (1980) ; Purtill v. Harris, 658 F.2d 134 at 136-139 (3d Cir. 1981).
The majority today concludes that this case presents “special factors counselling hesitation” in the implication of a cause of action under the Constitution. It does not decide that defendants in this case are immune from suit by analogy to Feres v. United States, 340 U.S. 135, 71 S.Ct. 153, 95 L.Ed. 152 (1950) (general waiver of sovereign immunity inherent in Federal Tort Claims Act not applicable to negligence action brought by soldier for service connected injury). Rather, it suggests that the considerations underlying Feres immunity help inform a decision on whether “special factors” foreclose Jaf fee’s lawsuit. The majority concludes that the policies that favored sovereign immunity from negligence liability in Feres apply as well in reaching a conclusion that a soldier may not proceed with a suit against individual defendants for redress of intentional wrongs.
My uneasiness with this analysis stems from Davis v. Passman. There the Supreme Court, in deciding that a Congressional staff member intentionally discharged on the basis of her sex could bring against her former employer a suit premised on the Due Process Clause of the Fifth Amendment, considered whether a suit against a Congressman raises “special factors counselling hesitation.” The Court acknowledged that such concerns exist, but held that they are coextensive with the protections which the Speech or Debate Clause affords. 442 U.S. at 246, 99 S.Ct. at 2277. If that Clause does not shield a Congressman’s actions, the Court proclaimed, the legislator ought to be bound by the law as are other citizens. Id. (citing Gravel v. United States, 408 U.S. 606, 615, 92 S.Ct. 2614, 2622, 33 L.Ed.2d 583 (1972)). In effect, the Court’s analysis appeared to render disposition of the “special factors” issue *1242inseparable from resolution of the immunity question. Analytically, the Court might have remanded for a determination of immunity under the Speech or Debate Clause, and postponed decision on the existence of a constitutional cause of action until a lower court had ascertained whether immunity, and thus a “special factor,” shielded the defendant from suit. Instead, the Court in essence eliminated consideration of “special factors,” held that a cause of action under the Fifth Amendment was available, and remanded for consideration of whether a constitutional immunity blocked prosecution of the cause of action.
For purposes of the “special factors” inquiry, it is difficult to distinguish the case at bar from Davis v. Passman. The conventional defense of military officials who are sued (and the defense on which defendants in this case have relied) is immunity — not a constitutional immunity like that asserted in Davis v. Passman, but a common law military immunity ultimately derived from Feres. Moreover, it would appear equally true of military officials as of legislators that, absent immunity, they should be held accountable for violations of legal norms. In light of Davis v. Passman, I cannot avoid regarding the “special factors counselling hesitation” as congruent with the scope of military immunity. If the “special factors” question constituted the sole inquiry in a case placing in issue the inference of a direct, constitutional cause of action, I would think that the Supreme Court’s disposition of Davis might well control the present case. We initially would hold that Jaffee could assert a cause of action under the Constitution, and then would face the question whether the immunity established in Feres — immunity of the Government from negligence suits arising out of service-connected injuries — should be extended to shield individual military and civilian defendants from suits based on intentional malfeasance.
II.
Absent the element of “special factors,” the Supreme Court has declared, the implication of a constitutional cause of action may be precluded when Congress has provided an adequate alternate remedy. Recently, the Supreme Court addressed this component of the Bivens standard in Carlson v. Green. In upholding a cause of action based on the Eighth Amendment, the Court stated that a defendant can defeat implication of a cause of action if he “show[s] that Congress has provided an alternative remedy which it explicitly declared to be a substitute for recovery directly under the Constitution and viewed as equally effective.” Id. 446 U.S. at 18-19, 100 S.Ct. at 1471-1472. Defendants in Carlson had identified the Federal Tort Claims Act as an adequate alternative to a damages remedy. The Court, however, found no indication that Congress intended the FTCA to preempt a Bivens -type remedy; to the contrary, it asserted that recent amendments to the FTCA made it “crystal clear,” id. at 20, 100 S.Ct. at 1472, that Congress regarded the FTCA and Bivens remedies as parallel and complementary. More generally, the Court cautioned that no “magic words,” id. at 19 n.5, 100 S.Ct. at 1472 n.5, need be discovered in the legislative history; instead, “our inquiry ... is whether Congress has indicated that it intends the statutory remedy to replace, rather than to complement, the Bivens remedy. When Congress decides to enact a statutory remedy which it views as fully adequate only in combination with the Bivens remedy, . . . that congressional decision should be given effect by the courts.” Id.1
*1243In the case at hand, I believe that Congress viewed the compensation system available to injured veterans under the Veterans’ Benefits Act, 38 U.S.C. § 301 et seq., as an adequate remedy within the reasoning of Carlson v. Green. Before applying the Carlson standard, two distinctions between that case and the present controversy should be noted. First, Congress adopted the current Veterans’ Benefits Act in 1958,2 thirteen years before the Supreme Court’s decision in Bivens. Because the Veterans’ Benefits Act predated the Court’s anticipated approval of a damage action based directly on the Constitution, it comes as no surprise that its legislative history fails to disclose a Congressional declaration that benefits under the Act are to substitute for the constitutional damages action. While the same is true of the FTCA, the Supreme Court in Carlson had the benefit of a postBivens amendment to the Federal Tort Claims Act which explicitly addressed the coexistence of statutory and constitutional causes of action. The post-Bivens amendments to the Veterans’ Benefits Act do not manifest a similar Congressional intent— again not surprisingly, since at the time of the amendments to the Veterans’ Benefits Act no court had recognized a cause of action based directly on the Constitution and intended to redress service related injuries. Under the circumstances, I do not understand Carlson to require an explicit Congressional declaration of the exclusivity of veterans’ compensation. In a case like the present one, where the relevant legislative history antedates Bivens, I would look for a Congressional declaration concerning the adequacy, rather than the exclusivity, of the statutory remedy.3 When scrutinizing the activities of a pre-Bivens Congress, I would hold that discovery of a Congressional intention to establish a comprehensive compensation scheme for injuries sustained in the military service will foreclose a private cause of action premised on the Constitution.
A second difference between Carlson and the case at issue here derives from the nature of the alternative remedy. In Carlson, which involved a claim by a mother on behalf of her deceased son against federal prison officials who were alleged to have intentionally deprived him of necessary medical care, plaintiff’s choice was between two lawsuits, one based on the Constitution and one on the FTCA. As between these two, the Supreme Court found the Bivens action more effective, adducing four reasons for this conclusion: constitutional actions have greater deterrent force, allow for the recovery of punitive damages, may be tried to a jury, and are governed by a uniform body of federal, rather than a divergent body of state, law. The Court concluded that the possibility of suit under the FTCA did not indicate a Congressional resolve to preclude a more effective suit directly under the Constitution.
In contrast to Carlson, where a comparison of two damages actions could illumine Congressional intent, the alternative to a lawsuit based on the Constitution is, in the present case, a statutory benefits scheme rather than a judicial action for damages.4 *1244Because of the different character of the two possible routes to recovery, I doubt whether an analysis which tracks the four factors that distinguished the Bivens action from the FTCA action in Carlson would support an inference concerning Congressional views on the adequacy of its compensation plan.5 To be sure, several of the reasons set forth in Carlson for preferring the Bivens remedy likewise distinguish that alternative in the situation before us: trial of a cause of action based on the Constitution may result in an award of punitive damages and might have greater deterrent value than an award under the veterans’ benefits program. Moreover, a plaintiff remitted to the statutory benefits scheme concededly might receive an award significantly less than his possible damages in a successful lawsuit. On the other hand, the barriers to recovery are likewise lower under the Veterans Benefits Act, and the benefits, if modest, are reasonably assured. See, e. g., 38 U.S.C. §§ 313, 333 (presumptions in favor of service-relatedness of disabilities). In light of the inherent differences between civil suits for damages and claims for statutory benefits, I believe that the superiority in some respects of the Bivens remedy has a limited bearing on the determination whether Congress intended its veterans’ benefits system to provide adequate compensation.
The Eighth Circuit, acting in the wake of Carlson v. Green, also has interdicted a cause of action based directly on the Constitution when an adequate administrative remedy exists, notwithstanding the failure of Congress to denominate the alternative as an exclusive remedy. In Bishop v. Tice, 622 F.2d 349 (8th Cir. 1980), plaintiff sought damages under the Fifth Amendment for his allegedly wrongful discharge from federal employment. The court of appeals, however, noted that a federal employee is entitled to an administrative challenge to his dismissal on the ground that there was no sufficient cause for the discharge. See 5 U.S.C. § 7701. If successful, the employee is entitled to reinstatement and back pay. This administrative process, the Eighth Circuit held, would adequately redress any violations of plaintiff’s substantive rights under the Due Process Clause, thereby obviating the need for a Bivens remedy. 622 F.2d at 356-57.
The Fifth Circuit has followed the rationale of Bishop v. Tice in Bush v. Lucas, 647 F.2d 573 (1981). Lucas involved a law suit directly under the Constitution by a Government employee seeking damages from his supervisor for an alleged retaliatory demotion. In denying the employee a right of action under the Constitution, the court disavowed reliance on the “alternative remedy” prong of the Carlson test. Its analysis of the “special factors” prong, however, turned largely on the availability of a comprehensive remedial scheme reflecting an effort by Congress to “achieve a proper balance between promoting governmental efficiency and protecting the rights of em*1245ployees aggrieved by improper personnel action.” The court concluded, “The very comprehensiveness of the legislative and administrative scheme evinces Congress’ awareness of the special relationship and of the government’s responsibilities toward its civil service employees.” Id. at 576-77. Although the court ultimately incorporated this analysis under a “special factors” heading, it appears that Lucas would have been governed by Davis v. Passman but for the presence of an alternative remedial scheme. Thus, the Fifth Circuit’s emphasis on the Congressionally-created alternative supports the analysis presented here and in Bishop v. Tice as to the sufficiency of a comprehensive alternative remedy to indicate a Congressional intent to preclude Bivens actions.
As I read Bivens, Passman, Carlson, Tice, and Lucas, the dispositive question here is whether Congress has created a comprehensive benefits scheme which it regards as fully adequate for the compensation of injured servicemen.6 If this question is answered in the affirmative, a private cause of action for damages, based directly on the Constitution, is precluded, even though such a remedy might, if allowed, result in a larger recovery for the injured veteran.
Although I recognize that there is room for disagreement, two considerations persuade me that the required legislative indications exist. First, the Senate Report to the 1976 amendments to the Veterans’ Benefits Act states: “The [Congress] periodically reviews the service-connected disability compensation program to ensure that authorized benefits provide reasonable and adequate compensation for disabled veterans.” S.Rep.No.94-1226, 94th Cong., 2d Sess., at 10 (1976) (emphasis added), reprinted at [1976] U.S.Code Cong. & Ad.News 2537, 2541. See also H.R.Rep.No.463, 78th Cong., 1st Sess. 1 (1943), reprinted at [1943] U.S. Code Cong.Serv. 2-159 (purpose of 1943 veterans’ benefits legislation was “to provide more adequate . . . veterans’ laws pertaining to compensation, pension, and retirement pay”). I interpret this statement to indicate that Congress has assumed responsibility for ensuring adequate compensation for service-connected injuries. And under Carlson v. Green, it would appear that no greater showing of legislative purpose is required.
Statements in a recent Supreme Court case buttress this inference. In Hatzlachh Supply Co. v. United States, 444 U.S. 460, 100 S.Ct. 647, 62 L.Ed.2d 614 (1980) (per curiam), an importer sued to recover the value of goods lost by the customs service after they had been seized for an alleged customs violation. In deciding in favor of the claimant, the Court distinguished Stencel Aero Engineering Corp. v. United States, 431 U.S. 666, 97 S.Ct. 2054, 52 L.Ed.2d 665 (1977), which denied Government liability to indemnify a third party for damages paid to a member of the armed forces injured in military training.7 The *1246Court in Hatzlachh interpreted its earlier decision in Stencel to be based on the “recogni[tion] that the Veterans’ Benefits Act provided compensation to injured servicemen, which we understood Congress intended to be the sole remedy for service connected injuries.” 444 U.S. at 464, 100 S.Ct. at 650 (emphasis added). The Court continued: “In Stencel, Congress had provided a remedy, which we thought to be exclusive.” Id.
The Supreme Court’s statements in Hatzlachh concerning the exclusivity of veterans’ benefits are admittedly dictum. Nonetheless, in my view they reflect an understanding on the part of the Court that Congress has shouldered the weighty responsibility of assuring an effective and reparative compensation arrangement for injured military personnel. Standing alone, the Court’s remarks might not be persuasive of Congressional intent. But when those remarks are joined with the statement in the Senate Report, I believe it is reasonable for us to infer that Congress has undertaken to guarantee adequate compensation for service-connected injuries, thereby satisfying the Carlson v. Green standard for preclusion of a cause of action implied directly under the Constitution.
III.
Several additional observations are in order. First, I wish to emphasize the limited applicability of my view of Mr. Jaffee’s case. I would hold only that a suit for damages asserted directly under the Constitution is barred when Congress has devised a compensation scheme intended to provide adequate benefits for injured parties. This rule would not preclude a party from receiving benefits and simultaneously pursuing a judicial remedy created, for example, by a non-constitutional provision of federal law. Nor would it compel the conclusion that a federal compensation scheme automatically forecloses any remedy under state law. Thus I do not challenge the continued vitality of Wilkes v. Dinsman, 48 U.S. (7 How.) 89, 122, 12 L.Ed. 618 (1849); Dinsman v. Wilkes, 53 U.S. (12 How.) 390, 401, 13 L.Ed. 1036 (1851), where the Supreme Court allowed a marine to prosecute an action for trespass against his commanding officer, based on the latter’s infliction of imprisonment and corporal punishment after the marine’s term of enlistment had allegedly expired.8 There is, so far as I know, no Supreme Court precedent for extinguishing a traditional common law cause of action whenever Congress has provided an alternative compensation system. But in my view Bivens and its progeny establish that Congressional creation of an adequate alternative will defeat what, at least until recently, has been regarded as an extraordinary remedy — the implication of a cause of action for damages directly under the Constitution.
The district court dismissed the Jaffees’ state law claims as barred- by the Feres doctrine of absolute immunity. Although the majority does not- decide the issue of absolute immunity, it disposes of the state claims by concluding that they are preempted by federal law. In resting its holding on preemption grounds, the majority is reaching out and deciding an issue that was not considered by the district court and was not briefed or argued by the litigants before this Court. Such an approach does not offer the parties a fair chance to be heard, and thus would not appear to be prudent. Because I believe, as the dissent has argued, that absolute immunity does not govern this case, proper resolution of the state claims requires careful consideration of several issues. For that reason, I would return this case to the district court. A remand *1247for this limited purpose would leave a number of questions for the district court to consider: whether to exercise its pendent jurisdiction so as to permit these claims to proceed in federal court; whether the qualified immunity doctrine of Butz v. Economou, 438 U.S. 478, 98 S.Ct. 2894, 57 L.Ed.2d 895 (1978), protects the defendants; whether plaintiffs assert a valid cause of action under state law; and whether such a cause of action is nevertheless preempted by federal law.
Allegations in Jaffee’s complaint — which we must accept as pleaded — maintain that the defendants, through their wilful actions, exposed him and other servicemen to lethal dosages of radiation. If it should be established that the Camp Desert Rock exercise was in fact conducted despite knowledge by the defendants of the gross dangers it presented, then the harm that ensued might well merit special Congressional attention. Indeed, even in the absence of such a showing Congress may wish to take some action. We note that Congress has recently enacted a bill to expand the availability of medical coverage under the Veterans’ Benefits Act both for Vietnam War veterans exposed to the herbicide Agent Orange and for veterans suffering from injuries that appear to have resulted from nuclear weapons tests. H.R. 3499, 97th Cong., 1st Sess., 127 Cong. Rec. 11572 (Oct. 16, 1981). Although presidential action on the bill is still pending, this legislative attention to those in Jaffee’s situation, and Congress’s use of the Veterans’ Benefits Act as the appropriate vehicle for redress, provide dramatic illustration of the extent to which Congress has taken on the responsibility for dealing with grievances arising out of military service, such as those advanced by Jaffee here.
IV.
It is perhaps understandable that cases like the present one can evoke strong emotional tugs regarding the alleged plight of the plaintiff or the problems besetting the defendants. But the role of the court is not to take sides: rather, it is expected to analyze, carefully and quietly, the precedents as well as the legislative efforts in order to arrive at a fair and just result. It is important, too, for the judge to realize that there are other bodies in our society that deal with these problems — the Executive, the Congress and the public itself. The use of epithets and slogans does not solve thorny legal problems. Instead, it tends to obfuscate the issues, issues that are difficult enough even under the best of circumstances. As Justice Holmes taught us, we must use care lest we interject the judiciary into heated controversies that are best settled by the operation of political forces.9
For all these reasons, I believe that the appropriate course — considering the prior decisions of the Supreme Court and paying heed to the words as well as to the deeds of the Congress — is to remand this matter for the limited purpose of having the district court address the questions relating to the claim of a state law cause of action. Aside from that, at least as I see it, any action accruing under the Constitution has been foreclosed, under the tests in Bivens and Carlson and by the actions of Congress.
ROSENN, Circuit Judge, joins in this opinion.. This court has recently emphasized that Carlson calls for an inquiry into the implicit intent of Congress to exclude Bivens-type remedies, and not merely for a mechanical search for an expressed intent. In Purtill v. Harris, at 136-139 (3d Cir. 1981), the court, in an opinion by Judge Aldisert, examined both the comprehensiveness of remedies under Section 15 of the Age Discrimination in Employment Act, 29 U.S.C. § 633a, and the legislative histories of ADEA and of Title VII of the Civil Rights Act. The court was able to conclude, without citing any explicit Congressional statement in support, that Congress intended the ADEA to preclude Bivens remedies for age discrimination claims by federal employees. Id. at 138.
. The 1958 Act was “essentially a consolidation of all of the existing laws administered by the Veterans’ Administration.” S.Rep.No.2259, 85th Cong., 2d Sess. 1 (1958), reprinted at 38 U.S.C.A.App. at 111. The Act, which made only relatively minor changes in the law, was for the most part an “accurate restatement” of existing rules. Id. Veterans’ benefits have a history virtually as old as the Republic. See Act of March 23, 1792, ch. XI, § 2, 1 Stat. 244 (servicemen injured during American Revolution entitled to pension); Act of Apr. 23, 1800, ch. XXXIII, § 8, 2 Stat. 53 (seamen disabled in line of duty entitled to pension not exceeding one-half monthly pay).
. Although the focus on adequacy is not explicit in Carlson, I believe this approach to be consistent with that decision. The Court in Carlson stated that when a contemporary enactment suggests that a statutory remedy is “adequate” only when supplemented by a Bivens remedy, a constitutional claim for damages is not precluded. 446 U.S. at 19 n.5, 100 S.Ct. at 1472 n.5. The converse of this proposition would appear to be that when Congress states that its remedy is adequate, without reference to a Bivens cause of action, a constitutional claim for damages might be precluded.
. Like Carlson, the other Supreme Court cases on implication of a constitutional cause of ac*1244tion did not arise in the context of a statutory compensation scheme. In Bivens, the plaintiff faced the choice of “damages or nothing,” 403 U.S. at 410, 91 S.Ct. at 2012 (Harlan, J., concurring); the asserted violation of plaintiffs Fourth Amendment rights had already occurred, rendering injunctive relief unsatisfactory, and the Court gave no indication that an assured compensation scheme would provide monetary redress for the wrong. Similarly, in Davis v. Passman the deprivation of rights would remain uncompensated if the constitutional lawsuit could not proceed. Even in Carlson, the plaintiff faced a “damages or nothing” choice — the only question was which of two possible damages remedies he could pursue. By contrast, the plaintiff here has an alternative to a suit based on the Constitution, and that remedy does not require judicial resolution of a claim as a prerequisite to an award of damages. Rather, a claimant under the Veterans’ Benefits Act need only show the existence of service-related damages.
. It should be noted that the Supreme Court in Carlson compared the Bivens and FTCA remedies in order to help ascertain whether Congress intended to limit the plaintiff to an FTCA action. See 446 U.S. at 20-21, 100 S.Ct. at 1472-1473. The comparison thus was ancillary to the central issue of Congressional intent; the Court did not announce a general rule that a constitutional cause of action must be implied whenever it would be superior in some respects to an alternative remedy. Instead, the so-called superiority factor was just one element in ascertaining Congressional intent.
. This past term, the Supreme Court applied a similar analysis to the availability of a cause of action under 42 U.S.C. § 1983. In Middlesex County Sewage Authority v. National Sea Clammers Association,-U.S.-, 101 S.Ct. 2615, 69 L.Ed.2d 435 (1981) the Court considered, inter alia, the availability of such a private right of action for damages resulting from discharges and ocean dumping of sewage. It noted that Congress had created “elaborate enforcement provisions” under the Federal Water Pollution Control Act, 33 U.S.C. §§ 1251 et seq., and the Marine Protection, Research, and Sanctuaries Act of 1972, 33 U.S.C. §§ 1401 et seq., which covered the activities in question, - U.S. at -, 101 S.Ct. at 2622, 69 L.Ed.2d at 446. Citing Carlson as analogous authority, the Court rejected any Section 1983 right of action, reasoning that “When the remedial devices provided in a particular act are sufficiently comprehensive, they may suffice to demonstrate Congressional intent to preclude the remedy of suits under § 1983.” Id.-U.S. at-, 101 S.Ct. at 2626, 69 L.Ed.2d at 450. By parity of reasoning, a preclusive intent similarly may be inferred from a comprehensive scheme such as the Veterans’ Benefits Act, which provides federal benefits to compensate for a wide range of injuries.
. The injured serviceman, Captain John Don-ham, was awarded a lifetime pension under the Veterans’ Benefits Act, but nonetheless brought a negligence action against Stencel, which had manufactured the product responsible for causing his injury. See 431 U.S. at 667-68, 97 S.Ct. at 2055-56. Nothing in this opinion questions the right of a serviceman simultaneously to receive veterans’ benefits *1246and to initiate a legal action based on a common law tort theory. See Part III infra. Thus the adequacy or exclusivity of veterans’ compensation would not bar a non-constitutional action like Donham’s.
. Whatever the merits of Dinsman, decided more than a century ago, it remains the law, and indeed has been cited several times in the recent past by the Supreme Court. E. g., Butz v. Economou, 438 U.S. 478, 492-94, 98 S.Ct. 2894, 2903-04, 57 L.Ed.2d 895 (1978). If it is to be overruled, the Supreme Court should do it. An intermediate appellate court obviously does not possess such authority.
. The following statement by Judge Learned Hand might also be an appropriate guide in a situation such as this:
But the judge must always remember that he should go no further than he is sure the government would have gone, had it been faced with the case before him. If he is in doubt, he must stop, for he cannot tell that the conflicting interests in the society for which he speaks would have come to a just result, even though he is sure that he knows what the just result should be. He is not to substitute even his juster will for theirs; otherwise it would not be the common will which prevails, and to that extent the people would not govern.
Hand, the Spirit of Liberty 109 (3d ed. Employed 1960).