OPINION OF THE COURT

Bailey v. DeQuevedo, 375 F.2d 72, 74 (3d Cir.), cert. denied, 389 U.S. 923, 88 S.Ct. 247, 19 L.Ed.2d 274 (1967), held that “an enlisted man in the armed services of the United States cannot maintain an action against an Army medical surgeon for negligence in an operation performed at an Army hospital in line of duty.” The question presented here is whether that bar can be extended to an action brought by the representative of a retired enlisted man against two Army surgeons. The district court held that the surgeons possessed immunity and dismissed the complaint with prejudice. We affirm.

Plaintiff’s decedent, a retired Army sergeant and a civilian employee at Fort Dix, New Jersey, died on January 23, 1975, shortly after a gall bladder operation performed on him by defendants, two Army surgeons. Plaintiff initiated survival (N.J. S.A. 2A:15-3) and wrongful death (N.J.S.A. 2A:31-1) claims in the Superior Court of New Jersey contending that defendants’ negligence caused the death. The action was properly removed to the district court pursuant to 28 U.S.C. § 1442(a) where it was dismissed with prejudice. The district court relied on the immunity doctrine enunciated in the defamation case of Barr v. Matteo, 360 U.S. 564, 79 S.Ct. 1335, 3 L.Ed.2d 1434 (1959) and later applied in this circuit in Keiser v. Hartman, 339 F.2d 597 (3d Cir. 1964), cert. denied, 381 U.S. 934, 85 S.Ct. 1764, 14 L.Ed.2d 699 (1965). Plaintiff appealed the order of dismissal.

*767Our starting point is the Bailey v. DeQuevedo rationale,1 anchored not on Barr v. Matteo but on Feres v. United States, 340 U.S. 135, 141, 71 S.Ct. 153, 157, 95 L.Ed. 152 (1950): “We know of no American law which ever has permitted a soldier to recover for negligence, against either his superi- or officers or the Government he is serving.” In denying relief to the plaintiff in Bailey our justification was two-fold: “the plaintiff was on ‘active duty’ and ‘subject to military discipline’, and the defendant’s alleged negligent action was ‘committed in the course of military duty.’ ” 375 F.2d at 74 (emphasis added). Here we must decide if a different rule should apply because: (a) the survival claim is asserted on behalf of a retired serviceman, entitled to military medical care but no longer on active duty at the time of the alleged negligent conduct or (b) the wrongful death claim, in legal theory, belongs to the decedent’s family, not to the decedent.

We concede that the distinction in the status of plaintiff and her decedent deprives these defendants of the benefit of the rationale that “one soldier may [not] sue another for negligent acts performed in the line of duty.” Bailey v. Van Buskirk, 345 F.2d 298 (9th Cir. 1965), cert. denied, 383 U.S. 948, 86 S.Ct. 1205, 16 L.Ed.2d 210 (1966), quoted in Bailey v. DeQuevedo, supra, 375 F.2d at 74. But we do not believe that the applicability of immunity doctrines ought to turn on such distinctions. We perceive more meaningful policy considerations at work — considerations looking not to the particular status of the plaintiff, but to the governmental interests inhering in the duties of the defendants. The Supreme Court instructs:

[ T]he Court has not fashioned a fixed, invariable rule of immunity but has advised a discerning inquiry into whether the contributions of immunity to effective government in particular contexts outweigh the perhaps recurring harm to individual citizens .

Doe v. McMillan, 412 U.S. 306, 320, 93 S.Ct. 2018, 2028, 36 L.Ed.2d 912 (1973) (emphasis added). In particular, two considerations are to be counterbalanced:

[0]n the one hand, the protection of the individual citizen against pecuniary damage caused by oppressive or malicious action on the part of officials of the Federal Government; and on the other, the protection of the public interest by shielding responsible governmental officers against the harassment and inevitable hazards of vindictive or ill-founded damage suits brought on account of action taken in the exercise of their official responsibilities.

Barr v. Matteo, supra, 360 U.S. at 565, 79 S.Ct. at 1336; see Doe v. McMillan, supra, 412 U.S. at 319, 93 S.Ct. at 2018.

Applying these broad directives from the Supreme Court, we perceive significant public policy considerations at work here which tip the balance in favor of immunity.

First, as in Bailey, the defendants here were Army physicians on active duty, performing “in the course of military duty.” They could not pick and choose their patients. The objectives of certainty and uniformity in the law would be ill-served if immunity were wrapped around a surgeon for a nine o’clock operation in an Army hospital on an active-duty soldier who survived the operation, but were removed for a ten o’clock operation in the same operating room on a retired soldier who did not survive the operation. The law should not require the military surgeon, in exercising his informed medical judgment, to concern himself with his patient’s military status or with the technicalities of tort law. The same standard of professional care obviously ought to apply in both situations, and the same legal rules should obtain.

Second, if viewed from the perspective of avoiding “perhaps recurring harm to individual citizens”, the plaintiff’s position in this case is equally untenable. Immunizing these defendants from personal liability does not deprive the plaintiff of a remedy: she may seek relief under the Federal Tort *768Claims Act, 28 U.S.C. § 1346(b), see United States v. Brown, 348 U.S. 110, 75 S.Ct. 141, 99 L.Ed. 139 (1954). Indeed, we were advised at oral argument that a timely Federal Tort Claims Act claim has been filed to protect the plaintiff. In Bailey we held the defendants were immune from personal liability even though the plaintiff was precluded from suing under the Federal Tort Claims Act. See Feres v. United States, supra. There was no alternative remedy in Bailey; there is in the case at bar. Insofar as this factor should affect the outcome, our decision today would seem to follow a fortiori from Bailey.

Third, concerning the contribution of an immunity rule to “effective government”, we note that this claim is not one where liability would be borne ultimately by the United States or by an insurance carrier. Absent a Federal Tort Claims Act suit naming the United States as a defendant, there is no provision for joinder of or indemnity from the United States; and malpractice insurance is not involved here.2 This case raises the spectre of personal liability on the part of military doctors.

We willingly concede that concern for the savings accounts of military doctors should not be controlling in and of itself. But viewed with regard for its possible effect on the delivery of health care to the armed services, the prospect of personal liability takes on added significance. Surely, there is a strong governmental interest in assuring that the military services be able to recruit and retain competent medical personnel. Given today’s litigious medical malpractice environment,3 we believe that such persons would be substantially discouraged from pursuing military medical careers by the prospect of personal liability in malpractice actions. Civilian physicians are having difficulty meeting the spiralling cost of malpractice insurance premiums.4 Some have felt compelled to leave previously lucrative practices. It could hardly be suggested that the purchase of costly insurance would provide a solution for the more modestly compensated military doctor. Therefore, if there is a governmental interest in the delivery of health care to the armed services — and we strongly believe there is — a result imposing personal liability on military medical personnel would be antithetical to that interest.

Finally, to argue, as does the plaintiff, that the protection of military medical officers is a matter for Congress,5 and not for *769the courts, is to ignore the reality that it has been the Supreme Court, not Congress, that has developed the principles of absolute immunity,6 as well as the principles of qualified immunity.7 Moreover, such an argument ignores the fact that it was this court, not Congress, that afforded protection to military physicians from claims by persons on active military duty. Bailey v. DeQuevedo, supra.

We see the issues reduced to this: being limited to the salary of a military officer without the financial emoluments of private civilian practice, and being under military obligation to accept patients sent to him, should a military physician be confronted with potential personal liability for a claim of medical malpractice? Would such liability be consonant with sound public policy? We think not. We think that the contribution of immunity to effective government, relating here to the armed services’ ability to recruit and retain competent medical personnel, outweighs the perhaps recurring harm to individual citizens who, in fact, have recourse under the Federal Tort Claims Act.

The judgment of the district court will be affirmed.

(concurring in the result).

I concur in the judgment of the majority, but since I take a different route to arrive at the result, I write separately to state my views.

A.

The facts presented in Bailey v. DeQuevedo 1 may appear, at first blush, to resemble those of the present case rather closely, since both concern medical malpractice actions brought against physicians employed on a full-time basis by the United States Army. A distinction between the cases exists, though, by virtue of the different relationships the two injured parties bore to the military. Bailey was on active duty in the armed forces, and as a practical matter had no choice as to which doctor to utilize. Martinez, on the other hand, was retired, *770and had the option of going to an Army doctor or to a civilian one.

The distinction is an important one, since the rationale utilized by the Court in Bailey was that of Feres v. United States,2 a Supreme Court decision that also involved tort suits by persons who had been on active military duty at the time of their injuries. The suits in Feres were not brought against individual defendants, as in Bailey, but against the government, under the Federal Tort Claims Act. Focusing upon the unique “relationship of military personnel to the Government,”3 the Supreme Court held the sovereign immune from suits brought under the Act by servicemen “where the injuries arise out of or are in the course of activity incident to service.”4

When, several years later, the Supreme Court was faced in United States v. Brown5 with a suit identical to that in Feres, except that the plaintiff was a discharged veteran — precisely the difference that exists between Bailey and Martinez — it expressly distinguished Feres on that basis. Because Brown was not on active duty when he was injured, the proceedings were allowed to go forward. The Supreme Court observed that the rationales underlying the result in Feres • — “[t]he peculiar and special relationship of the soldier to his superiors, the effects of the maintenance of such suits on discipline, and the extreme results that might obtain if suits under the Tort Claims Act were allowed for negligent orders given or negligent acts committed in the course of military duty”6 — did not apply to actions brought by military personnel not on active duty. In so doing, the Court specifically “adhere[d] ... to the line drawn in the Feres case”7 between injuries that arose in the course of active military duty and those that did not.

The Bailey Court recognized that Feres was not precisely on point, since the defendants in Bailey were the allegedly negligent physicians rather than the government, but it nonetheless adopted the Feres reasoning.8 Significantly, the Court then proceeded to point out the difference between Feres and Brown, in order to demonstrate that the former rather than the latter was the correct analogue to Bailey9 The case before us today is closer to Brown than to Feres, however, since Martinez was not on active duty at the time of the operation in question, and of course his executor was not on duty at the time of suit. As such, the Brown holding would appear to be more persuasive than that of Feres, which applied a rule of absolute immunity.

I thus conclude that Bailey does not require that we automatically immunize the defendants here. Consequently, I cannot agree with the determination of the majority that Bailey should be the starting point of the analysis or with the implication that Bailey is binding.

B.

Even if Bailey could stand for the proposition that Army physicians are absolutely immune from a suit brought by retired military personnel, it ought not govern the result here. The decision in Bailey came at a time when government officials with discretionary duties “enjoyed an absolute immunity from damage suits”10 if they were acting “within the outside perimeter of [their] line of duty . . . .”11 It is not *771disputed that the physicians here were within such perimeter in performing the operation in question.

In the nine years since Bailey, however, the law of immunity has undergone a marked change. One commentator has observed that “[t]he common law doctrines of personal immunity for official acts were revitalized”12 by the decision of the Supreme. Court in Pierson v. Ray,13 which came down eight days after Bailey. Further evolution has occurred since then.14

The rule of absolute immunity that prevailed at the time of Bailey has been modified by the Supreme Court into a standard of qualified immunity which, as stated in Scheuer v. Rhodes,15 is dependent upon “the scope of discretion and responsibilities of the office and all the circumstances as they reasonably appeared at the time of the action on which liability is sought to be based.” In view of the all-the-circumstances test that is to be applied, the Supreme Court, as pointed out by the majority, “has not fashioned a fixed, invariable rule of immunity . . ..”16

As noted in Chief Judge Seitz’s dissenting opinion, a number of the more recent Supreme Court decisions have involved the immunity of policymaking executive officials, and thus are factually distinguishable from the case now before us. The principles and the approach in the extant immunity decisions would appear to be useful in the present context, however, even though the Supreme Court has not addressed such a situation.

Thus, in Doe v. McMillan,17 a suit against the Public Printer, the Superintendent of Documents, and others for invasion of privacy, the Supreme Court stated that the doctrine of official immunity insulates “[government officials of suitable rank for the reason that ‘officials of government should be free to exercise their duties unembarrassed by the fear of damage suits in respect of acts done in the course of those duties — suits which would consume time and energies which would otherwise be devoted to governmental service and the threat of which might appreciably inhibit the fearless, vigorous, and effective administration of policies of government.”18 Although the last three words of the quotation do not specifically apply to army physicians, the two conflicting considerations underlying the doctrine of immunity are as applicable to military doctors as to policy-making officials. They are:

On the one hand, the protection of the individual citizen against pecuniary damage caused by oppressive or malicious ae*772tion on the part of officials of the Federal Government; and on the other, the protection of the public interest by shielding responsible governmental officers against the harassment and inevitable hazards of vindictive or ill-founded damage suits brought on account of action taken in the exercise of their official responsibilities.19

The defendant doctors in this case are skilled professionals employed by the government. They must exercise, in the course of their daily employment, a careful discretion, — albeit not precisely of the same type adverted to in Doe and Barr — that has a direct impact on their patients. It would disserve the public interest if these governmental physicians had to concern themselves with the possibility of a monetary judgment each time they were faced with a discretionary decision. Thus, it would not seem realistic to suggest, as the reasoning employed by the dissent might, that a government agent in charge of a HUD office in a small town, for example, would be clothed with immunity when determining whether the government should insure a mortgage agreement, but that the chief surgeon at a military hospital in a major metropolis, deciding whether to perform difficult surgery, would not be.

Nor can the absence of congressional legislation respecting this situation be deemed controlling on the question whether the defendants should be immunized. It is true that Congress has specifically given immunity to medical personnel in the Veterans’ Administration20 and the Public Health Service,21 but has not yet approved the legislation that has been proposed to immunize military medical personnel. But as is often the case, the meaning of the failure to enact legislation is inscrutable: Has Congress chosen to limit immunity to physicians in the Veterans’ Administration and Public Health Service? Or has it decided that the scope of immunity already afforded military doctors through judicial decisions is sufficient, so that legislation is unnecessary?22 It may be that the proponents of the legislation dealing with army physicians wished all such physicians to be immune, regardless of a balancing approach, so that even if no tort action against the government were available or if the harm resulted from a nondiscretionary act — such as failure to remove a sponge— the physician would nonetheless be protected. The legislative pattern cannot be determinative, for “[t]he official immunity doctrine . . . ‘has in large part been of judicial making . . .’”23 In the absence of guidance from the Congress, the courts face no statutory or constitutional obstruction to the interpretation or fashioning of the doctrine.24

Since qualified immunity would thus appear to be applicable, it is necessary to follow the all-the-circumstances approach mandated by the Supreme Court in such cases. After such evaluation, I conclude that the defendants here should be held immune from a suit for damages.

Many of the important factors bearing on this case are set forth rather fully in the majority opinion. In my view, the most determinative ones, in capsule form, are: (1) the defendants were Army physicians on active duty, and as such were required to perform the operation upon Martinez; (2) the plaintiffs have an alternative remedy *773against the United States under the Federal Tort Claims Act; (3) a holding that the defendants are subject to suit may be unduly harsh, particularly when there is an indication that they do not carry malpractice insurance; and (4) the risk of liability may deter competent physicians from embarking upon a career in the military, a development that would be counter to the national interest.25

Were one or more of the relevant factors missing — for example, if no alternative remedy existed — the balance could well be different. But an approach that regards Bailey as the controlling precedent and thus supplies absolute immunity in all military situations, or which provides for no immunity in any situation involving army physicians, would not permit the flexibility that would allow us to reweigh the totality of the considerations presented by cases that may arise in the future.

(dissenting).

It was long ago said that hard cases make bad law. I think that aphorism is particularly applicable to the result reached by the majority here. Reduced to its essence the majority holds that for purposes of the immunity doctrine a claim by a retired serviceman against military personnel is to be equated with the claim of an individual in active service against military personnel. In this way, the majority attempts to avoid the applicability of Barr v. Matteo, 360 U.S. 564, 79 S.Ct. 1335, 3 L.Ed.2d 1434 (1959). Instead they rely on Feres v. United States, 340 U.S. 135, 71 S.Ct. 153, 95 L.Ed. 152 (1950) which, as the concurring opinion notes, is not applicable to a claim of a retired serviceman arising after his retirement. Compare United States v. Brown, 348 U.S. 110, 75 S.Ct. 141, 99 L.Ed. 139 (1954).

No amount of legal legerdemain can obscure the fact that this is a state law claim by a civilian against military personnel. As the Supreme Court held in Howard v. Lyons, 360 U.S. 593, 79 S.Ct. 1331, 3 L.Ed. 1454 (1959), the immunity claim in such cases is a matter of federal law to be formulated by the courts in the absence of Congressional action. In addition, Howard teaches that the controlling federal principles governing the existence and scope of such immunity, in the absence of Congressional action, are to be found in Barr v. Matteo, supra.

In granting summary judgment in favor of defendants, the district court relied on the doctrine of official immunity enunciated in Barr v. Matteo, supra, and adopted by this Court in Keiser v. Hartman, 339 F.2d 597 (3d Cir. 1964). The majority, as noted, rejects this approach, and, in my view, commits error.

Barr was a libel action against the Acting Director of the Office of Rent Stabilization. There the Supreme Court extended the doctrine of absolute immunity, previously provided as a shield to a cabinet-level official in Spalding v. Vilas, 161 U.S. 483, 16 S.Ct. 631, 40 L.Ed. 780 (1896), to one whose position, although somewhat less exalted, involved the operation of a federal agency. The test which emerges from Barr is twofold. For absolute immunity to apply the government officer must (1) be a “policy-making executive official” and (2) be acting “within the outer perimeter of [his] line of duty.”

I am convinced that the defendant-doctors are not officials entitled to absolute immunity under Barr or subsequent cases.1 Defendants are clearly not policy making executive officials like the agency head in Barr, and thus are not shielded by an absolute immunity. I therefore conclude that the district court erred in holding that de*774fendants were entitled to absolute immunity-

Lesser officials are entitled only to a qualified immunity, the scope of which is related to the nature and extent of their governmental duties, see Doe v. McMillan, 412 U.S. 306, 319-20, 93 S.Ct. 2018, 36 L.Ed.2d 912 (1972).2 There is no rigid formula as to what duties or functions of a public servant carry with them an entitlement to the protection of a qualified immunity and which do not. Each case must be scrutinized to determine the nature of a public servant’s acts. Some of the considerations which may weigh in favor of a finding that certain actions give rise to an official immunity are whether the particular public servant is acting to formulate public policy rather than merely to administer policy decisions, or whether the actions undertaken are broad in scope and prospective in effect, pertain to a large number of persons or to the public in general, and involve the exercise of official discretion. Such actions are considered governmental in nature and, as such, give rise to a qualified immunity. By way of contrast, if the actions in question involve carefully circumscribed or routine application of a policy to particular individuals or situations, they are considered ministerial and do not give rise to governmental immunity.

In this case the acts complained of clearly were not related to the formulation of policy nor to broad decision making with an impact on a wide number of people. Although defendants’ acts were discretionary in nature, it is clear that they were imbued, in the words of the court in Henderson v. Bluemink, 167 U.S.App.D.C. 161, 511 F.2d 399 (1974), with medical and not governmental discretion. I therefore believe that defendants’ acts were not governmental in nature and that defendants are not shielded from any liability which may arise out of the performance of those acts under the doctrine of qualified immunity.

The most casual examination of the policy factors relied on in the majority and concurring opinions shows that they are not relevant under controlling principles. Rather than focusing upon the scope of the defendants’ authority and the nature of their discretionary duties, both the majority and the concurrence seem to rely instead on irrelevant factors reflecting the “equities” of the case. Certainly, the fact that the defendants were required to perform this operation, that they lacked malpractice insurance, and that plaintiff may have an alternate remedy does not further the inquiry into whether defendants are deserving of the qualified immunity reserved to lower public officials. Although these considerations may ultimately sway Congress to pass the bill presently before it, it is not for this court to “force” a result which Congress, though afforded the opportunity, has not yet seen fit to enact into law.

I would reverse the judgment of the district court and remand for further proceedings.

(dissenting).

I join in Chief Judge Seitz’s dissenting opinion, but I believe that one consideration requires clarification and emphasis.

In this case we have a state law which provides a rule of decision in the absence of a federal immunity which would supplant it.1 There is no doubt, of course, that federal law could, subject to due process limitations which need not be explored at this point, immunize all federal employees from state law causes of action for all negligent acts committed in the scope of their federal employment.2 Heretofore neither Congress *775nor this Circuit3 has recognized the absolute immunity claimed by defendants in this case. But Congress, which has previously immunized the medical personnel of the Veterans Administration4 and the Public Health Service5 from the very kind of suit that is the subject of this case, now has under active consideration a bill to extend the same immunity to active-duty military medical personnel.6 Thus, the issue confronting this court is which law-pronouncing branch of the federal government — the federal courts or Congress — should make the immunity determination.

In Doe v. McMillan, 412 U.S. 306, 93 S.Ct. 2018, 36 L.Ed.2d 912 (1972), the Supreme Court, confronted with a virtually identical choice, declined, in the absence of Congressional action, to supplant a cause of action under local law by absolutely immunizing the Public Printer and the Superintendent of Documents. Indeed, Doe v. McMillan is an a fortiori case since the Speech or Debate Clause of the United States Constitution 7 arguably afforded a constitutional basis for extending immunity to agents of the Congress. In this case there is no such constitutional provision. Moreover, in Crilly v. Southeastern Pa. Transp. Authority, 529 F.2d 1355 (3d Cir. 1976), we were asked to supplant local law when the reasons for applying a uniform law of labor-management relations appeared far more persuasive than those given for the extension of absolute immunity to cover these defendants. But in Crilly we deferred, and properly so I submit, to Congress which, just as in this case, was simultaneously considering proposed legislation covering the same subject matter.8

Instead of deferring to future Congressional judgment on the creation of absolute immunity in this kind of suit, the majority has chosen to weigh the competing policy considerations and to make an essentially legislative judgment. It has done so, however, without the benefit of the interplay of the various competing interests which, by design, appropriately occurs within the legislative arena. Indeed, I strongly suspect that not all of the relevant factors have even been identified, since the record before us does not even disclose, for example, the extent to which army surgeons are paid proficiency or incentive allowances over and above the military pay authorized for their rank.9 What principled reasons, I ask, make this case such a strong one for judicial legislation when compared to others in which we have refused to create federal common law? Unless those reasons are at least identified, the plaintiff will be justified in suspecting that they are entirely subjective.