FRIENDLY, Circuit Judge:
We have here appeals from and petitions for mandamus to vacate two orders of Chief Judge Edelstein in the Government’s civil antitrust action in the District Court for the Southern District of New York against International Business Machines Corporation (IBM). These orders denied motions to quash subpoenas directing Felix Kaufman, a partner in the international accounting firm of Coopers & Lybrand, and Frederic G. Withington, a senior staff member of the well-known management consulting firm of Arthur D. Little, Inc., to appear and testify on behalf of the Government. The issue is entirely between the United States on the one hand and Dr. Kaufman and Mr. Withington on the other; IBM has taken no position either in the District Court or here.
As the cases now stand, the sole issue sought to be raised on the merits is whether Dr. Kaufman and Mr. Withington are entitled to be excused from responding to the subpoenas because the Government is seeking to interrogate them on the basis of their expert knowledge of the computer industry. In affidavits opposing the motions to quash the subpoenas, the Government stated its desire as being to develop testimony “concerning the nature and structure of the general purpose electronic digital computer systems market, and of the electronic data processing industry in general” — a subject on which petitioners are highly qualified.1 So far as concerned opinion testimony, the Government made clear that it would not require appellants to state their present opinions but rather “prior opinions expressed during the period from 1960 through 1972.” In its brief here the Government has somewhat amplified the range of the proposed testimony; it intends [813]*813also to ask the witnesses “to explain the nature of their duties as computer systems consultants, and especially to recount the advice which they gave to various users and potential users of computer systems.” It states, however, that the witnesses “will not be asked at trial for their expert evaluation of the government’s evidence” and that “their testimony will be confined to events which occurred between 1960-1972.” It is not seeking to have either witness or his firm conduct any examinations or undertake any special studies in preparation for trial.
The Government also states in brief that it “will pay both witnesses as experts for their services” in amounts to be negotiated between them and Government counsel. According to the Government’s affidavits, although contested in part by petitioners, both witnesses had initially agreed to testify but later declined at the direction of their firms. Although some claims on the score of conflict of interest were raised in the district court, these have not been pressed on appeal.2 The issue on the merits is posed most bluntly in the concluding paragraph of Mr. Withington’s affidavit— “the government is seeking the very core of my expertise which I do not wish to provide and which I consider to be a proprietary asset available solely to my employer or to those for whom I wish to work.”
Chief Judge Edelstein, in opinions, denied the motions to quash. His rulings were based largely on statements in this court’s opinion in Carter-Wallace, Inc. v. Otte, 474 F.2d 529, 536 (2 Cir. 1972), cert. denied, 412 U.S. 929, 93 S.Ct. 2753, 37 L.Ed.2d 156 (1973), which are discussed below. He added “that an even stronger basis may be said to exist in the case at bar for recognizing the court’s power to compel expert testimony than existed in Carter-Wall ace," since “[t]his is not an attempt by a litigant in a private controversy to support its case through the assistance of an unwilling expert” but “an attempt by the United States to summon a member of the public to testify in a major government antitrust case, a case which, by definition, greatly affects the commonweal,” citing Pennsylvania Co. for Insurances On Lives and Granting Annuities v. City of Philadelphia, 262 Pa. 439, 441, 105 A. 630 (1918). Dr. Kaufman and Mr. Withington have appealed and also have petitioned for writs of mandamus to require Chief Judge Edelstein to grant their motions. The United States has moved to dismiss the appeals for want of jurisdiction and asks us to deny the petitions both on that ground and on the basis that this is not a proper case for mandamus. If we should reach the merits, it contends that the district court’s decision was right.
Appealability
Our jurisdiction of appeals in government civil antitrust cases is governed by the 1974 amendment to the Expediting Act, 15 U.S.C. § 29. This, so far as here relevant, provides:
(a) Except as otherwise expressly provided by this section, in every civil action brought in any district court of the United States under the Act entitled “An Act to protect trade and commerce against unlawful restraints and monopolies”, approved July 2, 1890, or any other Acts having like purpose that have been or hereafter may be enacted, in which the United States is the complainant and equitable relief is sought, any appeal from a final judgment entered in any such action shall be taken to the court of appeals pursuant to sections 1291 and 2107 of Title 28. Any appeal from an interlocutory order entered in any such action shall be taken to the court of appeals pursuant to sections 1292(a)(1) and 2107 of Title 28 but not otherwise.
One would have supposed it to be beyond argument that, despite Cohen v. Beneficial Industrial Loan Corp., 337 U.S. 541, 546 — 47, 69 S.Ct. 1221, 1225-1226, 93 L.Ed. 1528, 1536-1537 (1949), “An order compelling tes[814]*814timony . . . in an ordinary civil or criminal action is neither a final order [under § 1291] nor an interlocutory order granting an injunction' [under § 1292(a)(1)] and it is not appealable. This is the oft-cited rule of Alexander v. United States [201 U.S. 117, 26 S.Ct. 356, 50 L.Ed. 686 (1906)].” 9 Moore, Federal Practice H 110.13[2] at 153-54 (Ward ed. 1975). The remedy of the party witness wishing to appeal is to refuse to answer and subject himself to criminal contempt; that of the non-party witness is to refuse to answer and subject himself to civil or criminal contempt. Id. at H 110.-13[4] at 165-66. We have applied this rule in many cases of non-party witnesses, in one of which, United States v. Fried, 386 F.2d 691, 694 (2 Cir. 1967), we specifically rejected a contrary view expressed in Covey Oil Co. v. Continental Oil Co., 340 F.2d 993, 996-97 (10 Cir.), cert. denied, 380 U.S. 964, 85 S.Ct. 1110, 14 L.Ed.2d 155 (1965) — as four other circuits have done, Ryan v. CIR, 517 F.2d 13, 18-20 (7 Cir.), cert. denied, 423 U.S. 892, 96 S.Ct. 190, 46 L.Ed.2d 124 (1975); Gialde v. Time, Inc., 480 F.2d 1295, 1300-01 (8 Cir. 1973); United States v. Anderson, 150 U.S.App.D.C. 336, 464 F.2d 1390 (D.C. Cir. 1972); Borden Co. v. Sylk, 410 F.2d 843, 846 (3 Cir. 1969). The doctrine has since been reaffirmed by a unanimous Court in United States v. Ryan, 402 U.S. 530, 91 S.Ct. 1580, 29 L.Ed.2d 85 (1971), and was again recognized in United States v. Nixon, 418 U.S. 683, 690-92, 94 S.Ct. 3090, 3098-99, 41 L.Ed.2d 1039, 1053-54 (1974), although the Court, for reasons stated in its opinion which are wholly inapplicable here, made an exception where the subpoena was directed to the President of the United States.
Apart from a suggestion that the rule against appeals from orders refusing to quash subpoenas is limited to appeals from such orders in criminal cases, an assertion contradicted by Alexander itself which concerned a subpoena issued in a government civil antitrust case, as well as by Fried, appellants’ sole reliance for our departing from the established rule is the discussion in the majority opinion in International Business Machines Corp. v. United States, 471 F.2d 507, 511-16 (2 Cir. 1972), a case which concerned an order requiring the production of documents by a party, and which, on the merits, turned on issues of privilege and waiver thereof. The majority held that the order was appealable under Cohen’s gloss on § 1291, relying in part on the discussion in Covey Oil. At best this was an alternative holding, since the majority also found that it had power to issue mandamus, under 28 U.S.C. § 1651, 471 F.2d at 516-17. More to the point, the opinion failed to state the basis upon which Fried and Ryan were thought to be inapplicable, although these cases had been correctly cited by the Government as directly rejecting Covey Oil. These cases were again pressed in the Government’s Petition for Rehearing, although they were not in terms discussed in the subsequent en banc opinion, 480 F.2d 293 (1973), cert. denied, 416 U.S. 979, 980, 94 S.Ct. 2413, 40 L.Ed.2d 776 (1974), in which the court reversed the panel primarily on the basis that the Expediting Act in its then form, as recently explicated in Tidewater Oil Co. v. United States, 409 U.S. 151, 93 S.Ct. 408, 34 L.Ed.2d 375 (1972), wholly deprived the courts of appeals of both appellate and mandamus jurisdiction in government civil antitrust suits. However, the court added, 480 F.2d at 298-99, that, even if this were not true with respect to orders not going to the merits, Cohen would be inapplicable. We therefore must decline to recognize the opinion of the majority in 471 F.2d 507 as altering the rule established by many decisions of the Supreme Court and this court (including the later decision of this court en banc) that an order in a civil or a criminal case denying a motion by a non-party witness to quash a subpoena is not appealable until the witness has subjected himself to contempt.3
[815]*815
Availability of Mandamus
At the outset we must consider a possible barrier to the issuance of mandamus arising from the language of 15 U.S.C. § 29(a) which we have quoted above. As our discussion has shown, the orders refusing to quash the subpoenas were not final judgments under 28 U.S.C. § 1291, but were interlocutory orders. However, they were not interlocutory orders granting or denying a temporary injunction under § 1292(a)(1), and 15 U.S.C. § 29(a) directs that appeals from interlocutory orders shall be taken to a court of appeals “pursuant to sections 1292(a)(1) and 2107 of Title 28 but not otherwise.”4
Since the enactment of this statute, this court has granted a writ of mandamus relating to other orders in this same case, International Business Machines Corp. v. Edelstein, 526 F.2d 37 (2 Cir. 1975). IBM argued there that the revisions to the Expediting Act removed the previous impediments to court of appeals’ consideration of a mandamus petition; the Government responded that the terms of the new statute limited review of interlocutory orders to those involving injunctive relief. While the court must have been of the view that mandamus could properly issue, the impact of the cited language in the revised Expediting Act in this respect was not discussed, see 526 F.2d at 40-41. In the present case the Government has renewed its argument, although with the caveat that mandamus might issue “under exceptional circumstances”; petitioners respond that “[a]s the government undoubtedly knows, this argument was recently laid to rest by this Court.” Given the silence of the preceding opinion on this point, we think it useful to record the considerations that impel us to agree that the current Expediting Act does not preclude our issuing mandamus in an appropriate case.
We begin by comparing the language of the Expediting Act quoted above with that of 28 U.S.C. § 1447(d), “An order remanding a case to the State court from which it was removed is not reviewable on appeal or otherwise. . . .” This provision, or rather is predecessors, has long been construed to prohibit review not only by appeal but by mandamus or in any other way. See In re Pennsylvania Co., 137 U.S. 451, 11 S.Ct. 141, 34 L.Ed. 738 (1890); Ex parte Matthew Addy S.S. Co., 256 U.S. 417, 41 S.Ct. 508, 65 L.Ed. 1027 (1921); Employers Reinsurance Corp. v. Bryant, 299 U.S. 374, 57 S.Ct. 273, 81 L.Ed. 289 (1937). But see Thermtron Products, Inc. v. Hermansdorfer, 423 U.S. 336, 96 S.Ct. 584, 46 L.Ed.2d 542 (1976), 44 U.S.L.W. 4085 (restriction against mandamus with respect to remand order inapplicable where it was conceded that removal had been neither improvident nor without jurisdiction under 28 U.S.C. § 1447(c) and district court remanded solely because of congestion of its own docket).
However, the language of the revised Expediting Act does not evince the clear purpose to outlaw mandamus that has been found in the removal statute. The Act states that “Any appeal from an interlocutory order” may be taken in the specified manner “but not otherwise;” syntactically these words as used in the new Expediting Act can be read about equally well to mean that they encompass all forms of review other than by appeals under § 1292(a)(1), or that they prohibit only other forms of “appeal.” We think the legislative history supports the second of these readings and thus leaves the potential mandamus power of this court unimpaired.
Although the final Committee Reports on the bill that was enacted, Sen.Rep. 93-298, 93d Cong., 1st Sess. (1973), and H.R. Rep.93-1463, 93d Cong., 2d Sess. (1974), shed no light on the purpose of the “but not otherwise” language, illumination is cast by [816]*816earlier reports on the legislation, H.R.Rep. 91-1129, 91st Cong., 2d Sess. (1970), and S.Rep.91-1214, 91st Cong., 2d Sess. (1970). These contain a letter from the then Attorney General in which, after explaining the need for court of appeals review of orders granting or denying temporary injunctions, especially in merger or other acquisition cases,5 he added that no such need existed in respect of interlocutory appeals under 28 U.S.C. § 1292(b):
One reason against applicability of section 1292(b) is the desire to avoid undue delay and disruption. ... A second reason is the inappropriateness of review of controlling questions of law by a court which later may never get review of the final judgment. The theory of 1292(b) is that the appellate court should have an opportunity to rule early, before getting the final judgment, on questions that may be decisive. It would be anomalous for the courts of appeals to undertake interlocutory resolution of such issues when, at the end of trial, if a certificate is filed, the final judgment would go directly to the Supreme Court.
While one may debate the weight of these arguments, which would apply with almost equal force to orders granting or denying temporary injunctions, they evidently were what Congress had in mind. In contrast there is no indication that Congress gave any thought to the question of mandamus. Since the courts of appeals are now vested with jurisdiction over final judgments in government civil antitrust cases, save in the special circumstances provided in 15 U.S.C. § 29(b), they have the same power to issue the prerogative writs as the Supreme Court did before the amendment.6
To put the matter in another way, prior to the amendment the law was that prerogative writs could be issued in government civil antitrust cases, but only by the Supreme Court, since it had sole appellate jurisdiction; by contrast, appeals under § 1292(b) simply could not be taken, either to the courts of appeals or to the Supreme Court, Tidewater Oil Co. v. United States, 409 U.S. 151, 93 S.Ct. 408, 34 L.Ed.2d 415 (1972). The legislative history supports the view that the “but not otherwise” phrase in the amendment was intended, despite the general routing of appeals to the courts of appeals, to keep the latter rule in force,7 but gives no reason for belief that the power to issue prerogative writs was still to be confined to the Supreme Court.
The Propriety of Issuing Mandamus
In availing itself of its jurisdiction to issue prerogative writs in two government civil antitrust cases, United States Alkali Export Ass'n v. United States, 325 U.S. 196, 65 S.Ct. 1120, 89 L.Ed. 1554 (1945), and DeBeers Consolidated Mines, Ltd. v. United States, 325 U.S. 212, 65 S.Ct. 1130, 89 L.Ed. 1566 (1945), the Court cautioned against liberality in their use.8 Although “hardship is imposed on parties who are compelled to await the correction of an alleged error at an interlocutory stage by an appeal from a final judgment . . . such hardship does not necessarily justify resort to certiorari or other of the extraordinary writs as a means of review,” United States Alkali, supra, 325 U.S. at 202, 65 S.Ct. at 1125, 89 [817]*817L.Ed. at 1560. Again, “The writs may not be used as a substitute for an authorized appeal; and where, as here, the statutory scheme permits appellate review of interlocutory orders only on appeal from the final judgment, review by certiorari or other extraordinary writ is not permissible in the face of the plain indication of the legislative purpose to avoid piecemeal reviews.” Id. at 203, 65 S.Ct. at 1125, 89 L.Ed. at 1561. “When Congress withholds interlocutory review [§ 202] [the all writs statute] can, of course, not be availed of to correct a mere error in the exercise of conceded judicial power,” DeBeers, supra, 325 U.S. at 217, 65 S.Ct. at 1133, 89 L.Ed. at 1572. If such statements continue to be the law with respect to government civil antitrust suits, issuance of mandamus here would clearly be precluded; the only hardship to petitioners from a refusal to issue the writ is the relatively mild one of declining to answer the questions propounded to them and subjecting themselves to a sentence of contempt — in all likelihood of civil contempt.
The case against this is that these statements reflected the narrow view with respect to the scope of the prerogative writs which the Supreme Court then entertained but from which it has since departed, notably in LaBuy v. Howes Leather Co., 352 U.S. 249, 77 S.Ct. 309, 1 L.Ed.2d 290 (1957), and Schlagenhauf v. Holder, 379 U.S. 104, 85 S.Ct. 234, 13 L.Ed.2d 152 (1964). See also the dicta in Will v. United States, 389 U.S. 90, 88 S.Ct. 269,19 L.Ed.2d 305. (1967), and Note, Supervisor and Advisory Mandamus under the All Writs Act, 86 Harv.L. Rev. 595 (1973). That it has departed in some degree is clear enough; what is unclear is how far. Probably this cannot be completely determined until the Court speaks again; there is an obvious tension between the final judgment rule and increased use of the prerogative writs to review orders that are interlocutory even in light of Cohen, which must be resolved on a more principled basis than how much a court of appeals disapproves the district court’s ruling in the particular case. As we wrote in United States v. DiStefano, 464 F.2d 845, 850 (2 Cir. 1972), even if “the judge was wrong, indeed very wrong . that is not enough.”
In considering the propriety of the issuance of mandamus, we must begin with the basis upon which the decision below was rendered. Chief Judge Edelstein considered that his power to compel the testimony of petitioners here sought by the Government was settled by this court’s decision in Carter-Wallace, Inc. v. Otte, supra, 474 F.2d at 535-38. The question arose there in an unusual manner. Otte was the Chapter XI trustee in Bankruptcy of Davis-Edwards Pharmacal Corporation, which Carter-Wallaee, Inc. had sued for infringement of its patent on meprobamate, widely marketed under the tradenames of Miltown and Equanil. Carter-Wallaee had previously sued in the Court of Claims for infringement by the United States. In that action, which had not yet been decided, the United States had presented several expert witnesses whom Carter-Wallaee had thoroughly cross-examined. Otte, who had insufficient funds to engage experts, was allowed to introduce the Court of Claims testimony of the Government’s experts on the basis that the experts were unavailable since none of them were within the reach of the court’s subpoena power under F.R.Civ.P. 45(e).
One of Carter-Wallace’s arguments against this procedure was that the principle allowing the use of prior testimony of an unavailable witness did not apply “to expert witnesses whose opinions could under no circumstances be obtained through court process.” The court, in a unanimous opinion by the writer, met the point head-on. We said:
Carter-Wallace’s reliance on the court’s alleged lack of power to compel expert testimony is misplaced. The weight of authority holds that, although it is not the usual practice, a court does have the power to subpoena an expert witness and, though it cannot require him to conduct any examinations or experiments to prepare himself for trial, it can require him to state whatever opinions he may have previously formed. Boynton v. R. J. [818]*818Reynolds Tobacco Co., 36 F.Supp. 593 (D.Mass.1941); United States v. 284,392 Square Feet of Floor Space, 203 F.Supp. 75 (E.D.N.Y.1962) (dictum); see 4 Moore, Federal Practice U 26.66[1], at 26-469 (1972); 8 Wigmore, supra, § 2203(2)(c). Since the witnesses involved here had previously testified as to their opinions, it would seem that they could have been subpoenaed to repeat their testimony here.
Petitioners say that this was only dictum because we added the following:
In any event, even if Carter-Wallace were correct that a court cannot require expert witnesses to appear, this argument would seem to lead to the conclusion that an expert witness is “unavailable” even if he is within the 100 mile radius of the court’s usual subpoena powers, rather than to the result Carter-Wallace seeks.
We would regard this statement as an alternative ground of decision, and one on which less reliance was placed, rather than as one that rendered the previous statement dictum. Carter-Wallace’s argument was that it was anomalous to allow the use of testimony of an expert not called by the proponent simply because the expert was outside the reach of process when the proponent could not have required the testimony if the witness had been within its reach. Our main answer was that, so far as concerned previously formed opinions, the latter assumption was wrong. Following this lead, the trial court here ruled that “[t]he law to which this court must adhere does not support [petitioners’] central contention that an unwilling expert may not be compelled to testify.”
We can think of only one basis which, if well-founded, would have justified the district court in disregarding our decision in Carter-Wallace9 Unlike Carter-Wallace, this case is governed by the Federal Rules of Evidence. The trial court considered whether the Rules ran counter to our decision in Carter-Wallace and concluded that they did not. If anything, they tend to confirm it in a case like this which is governed by the “General Rule” of Rule 501 rather than the exception for claims or defenses in civil actions as to which State law provides the rule of decision. Article VII of the Rules deals with “Opinions and Expert Evidence” and contains six rules, 701-706. Although the framers of the rules must have been well aware of the frequently made contention that experts enjoy some kind of privilege, neither this chapter nor the proposed rules on privilege which the framers proposed but Congress rejected contain any suggestion that an expert enjoys either an absolute or a qualified privilege against being called by a party against his will. The only reference to the need of consent by an expert is in Rule 706(a), dealing with court appointed experts. This provides in part that “An expert witness shall not be appointed by the court unless he consents to act” — language taken verbatim from former F.R.Cr.P. 28. The situation of the court appointed expert who is expected to delve deeply into the problem and arrive at an informed and unbiased opinion differs utterly from that of an expert called by a party to state what facts he may know and what opinion he may have formed without being asked to make any further investigation. If any inference is to be drawn from the Federal Rules of Evidence, it is thus against the claim of privilege by an expert, not for it.
It is in light of this careful consideration given by the trial judge to the law as announced by this court and the effect of the Federal Rules of Evidence that we must consider whether mandamus lies. Clearly we are a great distance from the facts of LaBuy, supra, which concerned a district court’s practice of referring cases to masters which the Supreme Court characterized as “little less than an abdication of the [819]*819judicial function,” and not much nearer the facts of Schlagenhauf, supra, which primarily concerned “the basic, undecided question of whether a district court could order the mental or physical examination of a defendant.” Looking beyond the facts of these leading cases, we also think this case falls outside their rationale.
This court’s major effort to outline the criteria for this broadened scope of mandamus, at least as regards situations similar to this one, is Judge Feinberg’s opinion for a divided panel in American Express Warehousing, Ltd. v. Transamerica Ins. Co., 380 F.2d 277, 282-83 (2 Cir. 1967), a discovery case. The majority thought the “touchstones” for the issuance of mandamus now were “usurpation of power, clear abuse of discretion, and the presence of an issue of first impression.” The majority declined to issue mandamus in American Express since none of these factors was present; Judge Lumbard, who has consistently taken an expansive view of the use of mandamus,10 dissented. In this case the judge’s denial of a motion to quash subpoenas clearly was no “usurpation of power.” F.R.Civ.P. 45 implicitly empowers a district judge to rule on such motions as it does explicitly in the case of motions to quash or modify subpoenas for the production of documentary evidence. No one could reasonably suppose that the Court has now subscribed, despite the contrary pronouncements in such cases as Bankers Life & Casualty Co. v. Holland, 346 U.S. 379, 382, 74 S.Ct. 145, 147, 98 L.Ed. 106, 141 (1953), and Parr v. United States, 351 U.S. 513 520,76 S.Ct. 912, 917, 100 L.Ed. 1377, 1384 (1956), to a doctrine that any non-frivolous claim of error in a decision is a claim of “usurpation of power” on the theory that courts are bound to decide all issues correctly. “Jurisdiction is authority to decide the case either way,” The Fair v. Kohler Die Co., 228 U.S. 22, 25, 33 S.Ct. 410, 411, 57 L.Ed. 716, 717 (1913).
Whatever the usefulness of the “clear abuse of discretion” test for the issuance of mandamus may or may not be in other contexts, see A. Olinick & Sons v. Dempster Bros., 365 F.2d 439 (2 Cir. 1966), it would not justify this in a case like the present where the judge reasonably considered that, in requiring petitioners to testify to facts within their ken and opinions “previously formed,” he was acting in conformity with the law as recently declared by this court and consistently with the Federal Rules of Evidence, and that no sufficient grounds for a dispensation had been shown. Petitioners’ argument thus must come down to a contention that, absent any other basis for its issuance, mandamus may be used to review an interlocutory order whenever a case presents an issue claimed to be novel and important.
If we were convinced that an expert witness is privileged against being called to testify against his will either under all circumstances or in the absence of a preliminary showing of “necessity” or on the basis of any other rule under which petitioners could bring themselves, we might be able to find a basis for overruling the pertinent portions of Carter-Wallace without invoking an en banc court. The decision could have been — although it was not — placed on the ground that Carter-Wallace could not in effect invoke the alleged privilege of the expert witnesses; it was wholly speculative wh^her they would have claimed it if they had been subject to the process of the district court. Moreover, as the writer freely concedes, the facts that the claim was made by a party rather than the witness and was one of many issues in a complicated pharmaceutical patent case led the court to place greater reliance on the briefs and to engage in less independent research than if a recalcitrant witness had been before us.
We think it plain, however, that, quite apart from the inference to be drawn from the provision with respect to court-appointed experts, Rule 706(a), “the principles of the common law as they may be interpreted by the courts of the United States in [820]*820the light of reason and experience,” Federal Rule 501 of Evidence, do not recognize any general privilege for experts.
In a notable case the Supreme Court has recently quoted its statement in Branzburg v. Hayes, 408 U.S. 665, 688, 92 S.Ct. 2646, 2660, 33 L.Ed.2d 626, 643 (1972),“that ‘the public . . . has a right to every man’s evidence,’ except for those persons protected by a constitutional, common-law, or statutory privilege,” United States v. Nixon, 418 U.S. 683, 709, 94 S.Ct. 3090, 3108, 41 L.Ed.2d 1039, 1064 (1974).11 Admittedly there is no constitutional or statutory privilege against the compulsion of expert testimony, and we perceive no sufficient basis in principle or precedent for holding that the common law recognizes any general privilege to withhold his expert knowledge.
The truth of the proposition that the high degree of a person’s knowledge of a subject excuses him from giving testimony about it is not self-evident, to say the least. As Wigmore says, 8 Evidence § 2192 at 72 (McNaughton rev. 1961), the giving of such testimony
may be a sacrifice of time and labor, and thus of ease, of profits, of livelihood. This contribution is not to be regarded as a gratuity, or a courtesy, or an ill-required favor. It is a duty not to be grudged or evaded. Whoever is impelled to evade or to resent it should retire from the society of organized and civilized communities, and become a hermit. He who will live by society must let society live by him, when it requires to.
He adds that “all privileges of exemption from this duty [of giving testimony] are exceptional, and are therefore to be discountenanced. There must be good reason, plainly shown, for their existence.” Id. at 73 (emphasis in original). The opinion which, on its facts, most directly supports the view espoused by petitioners, People ex rel. Kraushaar Bros. & Co. v. Thorpe, 296 N.Y. 223, 72 N.E.2d 165 (1947), itself conceded that, as compared with three other states recognizing a privilege for experts,12 there were eleven which compelled experts to testify to opinions which they “are able to give without study of the facts or other preparation.”13 See also Annotation, 77 A.L.R.2d 1182 (1961); Note, Requiring Experts to Testify in Maine, 20 Maine L.Rev. 297 (1968). The reasoning in Kraushaar Bros, for regarding the majority rule as “quite unsatisfactory”14 ignores several important points. One is the enormous range of “expert” knowledge in modern life; if there is a privilege against the compulsion of “expert” testimony, this must extend not only to such learned professions as “medicine, law [and] science” but to the builder, the tug boat captain, the carpenter, the real estate broker, the stock broker, and a legion of other callings. Furthermore it assumes that the only purpose in calling an expert is to ask him to express an opinion about facts [821]*821of which he has no personal knowledge. In such a case we see no reason why an expert cannot decline to testify “without reflection upon his professional ability” if, as may often happen, he believes that a hypothetical question does not give all the facts required for a full and fair answer. Often the inquiry may be about the development of techniques, the stage they had reached, opinions already formulated or expressed, even — as the Government claims to be the situation here — relevant knowledge about a party to the case or its products. To clothe all such expert testimony with privilege solely on the basis that the expert “owns” his knowledge free of any testimonial easement15 would be to seal off too much evidence important to the just determination of disputes.
What gives some pause about compelling expert testimony is not that the expert is called upon to make available to the community relevant knowledge that he has acquired through time and labor, but that since his knowledge,.in contrast to that of the ordinary witness, is relevant to many cases, he may be summoned too often, both because of his eminence and because published views may eliminate the usual practical deterrent against calling an expert with whom counsel has not conferred in advance. See Ex parte Roelker, 20 F.Cas.P. 1092 (No. 11,995) (D.Mass. 1854). A court would indeed have cause for concern if, for example, a world-renowned surgeon were forced to spend a considerable share of his time in the courtroom rather than the operating room;16 the interest in the saving of lives would outweigh á party’s interest in obtaining testimony from an expert having no previous connection with the case. But to say, as we do, that there is no privilege to withhold testimony simply because it represents the result of study or experience does not entail a holding that courts may never extend protection when good cause is shown. It will be time for a federal court to consider what should be done about cases like the one we have framed when and if they should arise in situations governed by the first sentence of Rule 501. We can find no justification for a federal rule that would wholly exempt experts from placing before a tribunal factual knowledge relating to the case in hand,17 opinions already formulated, or, even, in the rare case where a party may seek this and the witness feels able to answer, a freshly formed opinion, simply because they have become expert in a particular calling.
We likewise see no basis for recognizing a narrower principle that an expert is privileged against being called against his will in the absence of a preliminary showing of the unavailability of a voluntary expert equally qualified. The task of determining whether the country contains a voluntary expert who is really as qualified and useful as the witness sought to be compelled would be an almost impossible one. Indeed, the facts of this case vividly illustrate the unworkability of such a “necessity” principle. A substantial 'part of the testimony which the Government here seeks from petitioners is testimony no one else can give. While it would seem that there must be other experts who could describe “the nature and structure of the general purpose electronic digital computer systems market, and of the electronic data processing industry in general,” no one not connected with Coopers & Lybrand or Arthur D. Little can testify what advice these two important consultants gave to the large purchasers of [822]*822computers who were their clients.18 To be sure, there probably are other accounting and management consulting firms with clienteles as large as petitioners’. But if, and it would be a bit naive to assume otherwise, the Government has some reason to think the petitioners will give testimony on this subject considered favorable to its position, we see no reason why it must go down the list of great accounting and consulting firms until it finds one — if it should — that presents the combination of willingness to furnish a qualified expert and the helpfulness of his testimony which the Government evidently thinks it has found in Dr. Kaufman and Mr. Withington.
Since there is thus no federal rule giving expert witnesses any privilege that can be formulated as a principle of law, petitioner's are really asking us to utilize mandamus to set standards that should guide a trial judge in exercising his power to grant dispensation to an expert witness from the general principle of testimonial compulsion and then to hold that the district court violated such standards here. They claim to' find support for this in the portion of Schlagenhauf, supra, where the Supreme Court held that the court of appeals should not merely have decided the novel question of a trial court’s power to order a physical or mental examination of a defendant, but also should have interpreted the terms “in controversy” and “good cause” as used in F.R.CiviP. 35. However, as we read the opinion on this point, 379 U.S. 104 at 111, 85 S.Ct. 234 at 238, 13 L.Ed.2d 152 at 159, it was primarily because the basic issue of power was already presented that the Court urged the court of appeals to reach these subsidiary issues as well. Be that as it may, what we would be doing here is quite different from construing language in a statute or a rule having the effect of one. As indicated, the utmost we could do would be to list a variety of factors appropriate for consideration by a trial judge in determining whether the exercise of compulsion upon an expert would constitute such a misuse of the court’s processes that, despite the lack of any general privilege, the expert should nevertheless be excused from testifying in the particular case. Appropriate factors for consideration — some pointing against a dispensation and some for one — would be the degree to which the expert is being called because of his knowledge of facts relevant to the case rather than in order to give opinion testimony; the difference between testifying to a previously formed or expressed opinion and forming a new one; the possibility that, for other reasons, the witness is a unique expert; the extent to which the calling party is able to show the unlikelihood that any comparable witness will willingly testify; the degree to which the witness is able to show that he has been oppressed by having continually to testify; and, undoubtedly, many others. But, as this discussion shows, the end result is that dispensation would represent an exercise of discretion by the trial judge, and mandamus could not issue unless there had been “clear abuse of discretion,” see American Express Warehousing Ltd. v. Transamerica Ins. Co., supra, 380 F.2d at 282-83- — a showing that would not be remotely approached here. It is not asking too much of an expert who seeks to invoke a discretionary dispensing power of the trial judge that, save in the exceedingly rare case where refusal altogether to quash a subpoena would be a “clear abuse of discretion,” the witness must attend, see just what questions are propounded to him, allow the trial court to rule upon them, and, if he then wishes to pursue the point further, subject himself to contempt.
The appeals are dismissed for want of appellate jurisdiction; the petitions for mandamus are denied.