delivered the opinion of the Court.

Appellants, the Swift and Armour Companies, stuff, freeze, and package turkeys which they ship to retailers throughout the country for ultimate sale to consumers. Each package is labeled with the net weight of the particular bird (including stuffing) in conformity with a governing federal statute, the Poultry Products Inspection Act of 1957, 71 Stat. 441, 21 U. S. C. §§ 451-469 (1964 ed.), and the regulations issued under its authority by the Secretary of Agriculture. 1 Many of these turkeys are *113 sold in New York. Section 193 of New York’s Agriculture and Markets Law 2 has been interpreted through regulations and rulings to require that these packaged turkeys be sold with labels informing the public of the weight of the unstuffed bird as well as of the entire package. Because the amount of stuffing varies with each bird, the State thus seeks to help purchasers ascertain just how much fowl is included in each ready-for-the-oven turkey.

Swift and Armour requested permission of the Poultry Products Section of the Department of Agriculture to change their labels in order to conform with New York’s requirements, but such permission was refused at the initial administrative level and no administrative review of that refusal was sought. Swift and Armour *114 then brought this federal action to enjoin the Commissioner of Agriculture and Markets of New York from enforcing the State’s labeling provisions, asserting that enforcement would violate the Commerce Clause and the Fourteenth Amendment of the Federal Constitution and overriding requirements of the federal poultry enactment.

Pursuant to appellants’ request, a three-judge district court was constituted under 28 U. S. C. § 2281 (1958 ed.), which provides for such a tribunal whenever the enforcement of a state statute is sought to be enjoined “upon the ground of the unconstitutionality of such statute.” The District Court, unsure of its jurisdiction for reasons appearing below, dismissed the suit on the merits 3 acting both in a three-judge and single-judge capacity. 4 Appeals were lodged in the Court of Appeals for the Second Circuit from the single-judge determination, and in this Court from the three-judge decision in accordance with the direct appeal statute, 28 U. S. C. § 1253 (1964 ed.). The threshold question before us, the consideration of which we postponed to the merits (379 U. S. 997), is whether this Court, rather than the Court of Appeals, has jurisdiction to review the District Court determination, and this in turn depends on whether a three-judge court was required. We hold that it was not.

At the outset, we agree with the District Court that the Commerce Clause and Fourteenth Amendment *115 claims alleged in the complaint are too insubstantial to support the jurisdiction of a three-judge court. It has long been held that no such court is called for when the alleged constitutional claim is insubstantial, Ex parte Poresky, 290 U. S. 30; California Water Service Co. v. City of Redding, 304 U. S. 252. Since the only remaining basis put forth for enjoining enforcement of the state enactment was its asserted repugnancy to the federal statute, the District Court was quite right in concluding that the question of a three-judge court turned on the proper application of our 1962 decision in Kesler v. Department of Public Safety, 369 U. S. 153. There we decided that in suits to restrain the enforcement of a state statute allegedly in conflict with or in a field pre-empted by a federal statute, § 2281 comes into play only when the Supremacy Clause of the Federal Constitution is immediately drawn in question, but not when issues of federal or state statutory construction must first be decided even though the Supremacy Clause may ultimately be implicated. Finding itself unable to say with assurance whether its resolution of the merits of this case involved less statutory construction than had taken place in Kesler, the District Court was left with the puzzling question how much more statutory construction than occurred in Kesler is necessary to deprive three judges of their jurisdiction.

It might suffice to dispose of the three-judge court issue for us to hold, in agreement with what the District Court indicated, 230 F. Supp., at 410, that this case involves so much more statutory construction than did Kesler that a three-judge court was inappropriate. (We would indeed find it difficult to say that less or no more statutory construction was involved here than in Kesler and that therefore under that decision a three-judge court was necessary.) We think, however, that such a disposition of this important jurisdictional question would be *116 less than satisfactory, that candor compels us to say that we find the application of the Kesler rule as elusive as did the District Court, and that we would fall short in our responsibilities if we did not accept this opportunity to take a fresh look at the problem. We believe that considerations of stare decisis should not deter us from this course. Unless inexorably commanded by statute, a pro.-cedural principle of this importance should not be kept on the books in the name of stare decisis once it is proved to be unworkable in practice; the mischievous consequences to litigants and courts alike from the perpetuation of an unworkable rule are too great. For reasons given in this opinion, we have concluded that the Kesler doctrine in this area of § 2281 is unsatisfactory, and that Kesler should be pro tanto overruled. The overruling of a six-to-two decision 5 of such recent vintage, which was concurred in by two members of the majority in the present case, 6 and the opinion in support of which was written by an acknowledged expert in the field of federal jurisdiction, demands full explication of our reasons.

I.

The three-judge district court is a unique feature of our jurisprudence, created to alleviate a specific discontent within the federal system. The antecedent of § 2281 was a 1910 Act 7 passed to assuage growing popular displeasure with the frequent grants of injunctions by federal courts against the operation of state legislation regulating railroads and utilities in particular. 8 The *117 federal courts of the early nineteenth century had occasionally issued injunctions at the behest of private litigants against state officials to prevent the enforcement of state statutes, 9 but such cases were rare and generally of a character that did not offend important state policies. The advent of the Granger and labor movements in the late nineteenth century, 10 and the acceleration of state social legislation especially through the creation of regulatory bodies met with opposition in the federal judiciary. In Chicago, M. & St. P. R. Co. v. Minnesota, 134 U. S. 418, this Court held that the setting of rates not permitting a fair return violated the Due Process Clause of the Fourteenth Amendment. Ex parte Young, 209 U. S. 123, established firmly the corollary that inferior federal courts could enjoin state officials from enforcing such unconstitutional state laws.

This confrontation between the uncertain contours of the Due Process Clause and developing state regulatory *118 legislation, arising in district courts that were generally considered unsympathetic to the policies of the States, had severe repercussions. Efforts were made in Congress to limit in various ways the jurisdiction of federal courts in these sensitive areas. 11 State officials spoke out against the obstruction and delay occasioned by these federal injunction suits. 12 The sponsor of the bill establishing the three-judge procedure for these cases, Senator Overman of North Carolina, noted:

“[TJhere are 150 cases of this kind now where one federal judge has tied the hands of the state officers, the governor, and the attorney-general ....

Whenever one judge stands up in a State and enjoins the governor and the attorney-general, the people resent it, and public sentiment is stirred, as it was in my State, when there was almost a rebellion, whereas if three judges declare that a state statute is unconstitutional the people would rest easy under it.” 45 Cong. Rec. 7256. 13

*119 In such an atmosphere was this three-judge court procedure put on the statute books, and although subsequent Congresses have amended the statute 14 its basic structure remains intact.

II.

Section 2281 was designed to provide a more responsible forum for the litigation of suits which, if successful, would render void state statutes embodying important state policies. The statute provides for notification to the State of a pending suit, 28 U. S. C. § 2284 (2) (1964 ed.), thus preventing ex parte injunctions common previously. 15 It provides for three judges, one of whom must be a circuit judge, 28 U. S. C. §2284 (1) (1964 ed.), to allow a more authoritative determination and less opportunity for individual predilection in sensitive and politically emotional areas. It authorizes direct review by this Court, 28 U. S. C. § 1253, as a means of accelerating a final determination on the merits; an important criticism of the pre-1910 procedure was directed at the *120 length of time required to appeal through the circuit courts to the Supreme Court, and the consequent disruption of state tax and regulatory programs caused by the outstanding injunction. 16

That this procedure must be used in any suit for an injunction against state officials on the ground that a state enactment is unconstitutional has been clear from the start. What yet remains unclear, in spite of decisions by this and other courts, is the scope of the phrase “upon the ground of the unconstitutionality of such statute” when the complaint alleges not the traditional Due Process Clause, Equal Protection Clause, Commerce Clause, or Contract Clause arguments, but rather that the state statute or regulation in question is pre-empted by or in conflict with .some federal statute or regulation thereunder. Any such pre-emption or conflict claim is of course grounded in the Supremacy Clause of the Constitution: if a state measure conflicts with a federal requirement, the state provision must give way. Gibbons v. Ogden, 9 Wheat. 1. The basic question involved in these cases, however, is never one of interpretation of the Federal Constitution but inevitably one of comparing two statutes. Whether one district judge or three must carry out this function is the question at hand.

The first decision of this Court casting light on the problem was Ex parte Buder, 271 U. S. 461, in which the question presented was, as here, whether an appeal was properly taken directly from the District Court to the Supreme Court. At issue was whether a Missouri statute authorizing taxation of bank shares remained valid after the enactment of a federal statute which enlarged the scope of the States’ power to tax national banks by permitting taxation of shares, or dividends, or *121 income. Under the federal scheme, States were apparently expected to choose one of the three methods. Although the Missouri law applied the first basis of assessment, the District Court held that because the State did not explicitly choose among the three types of taxation, but instead relied on a prior statute, the assessment was void. Mr. Justice Brandéis, writing for a unanimous Court, held that this was not properly a three-judge court case . . because no state statute was assailed as being repugnant to the Federal Constitution.” 271 U. S., at 465. Although the complaint in Buder did not explicitly invoke the Supremacy Clause, it should be noted that the defendants’ answer asserted that if the federal statute was constitutional under the Tenth Amendment, then it would indeed be the “ 'supreme law of the land’ within the meaning and provisions of Article VI of the Constitution of the United States,” and thus controlling over the particular state statute unless that statute could be construed as consistent with the federal law. The District Court in Buder was thus clearly presented with the Supremacy Clause basis of the statutory conflict.

Ex parte Bransford, 310 U. S. 354, raised a similar problem, also in the context of the validity of a state tax. The Court again held this type of federal-state confrontation outside the purview of the predecessor of § 2281:

“If such assessments are invalid, it is because they levy taxes upon property withdrawn from taxation by federal law or in a manner forbidden by the National Banking Act. The declaration of the supremacy clause gives superiority to valid federal acts over conflicting state statutes but this superiority for present purposes involves merely the construction of an act of Congress, not the constitutionality of the state enactment.” 310 U. S., at 358-359.

*122 In a third case, Case v. Bowles, 327 U. S. 92, the question involved the proposed sale by the State of Washington of timber on state-owned land at a price violating the Federal Emergency Price Control Act of 1942. A federal district court enjoined the sale, and on appeal the State argued that the single judge lacked jurisdiction. This Court held otherwise: “the complaint did not challenge the constitutionality of the state statute but alleged merely that its enforcement would violate the Emergency Price Control Act. Consequently a three-judge court is not required.” 327 U. S., at 97. 17

The upshot of these decisions seems abundantly clear: Supremacy Clause cases are not within the purview of § 2281. 18 This distinction between cases involving claims *123 that state statutes are unconstitutional within the scope of § 2281 and cases involving statutory pre-emption or conflict remained firm until Kesler v. Department of Public Safety, 369 U. S. 153, in which the plaintiff alleged a conflict between the federal bankruptcy laws and a state statute suspending the driving licenses of persons who are judgment debtors as a result of an adverse decision in an action involving the negligent operation of an automobile. It was argued that federal policy underlying the bankruptcy law overrode the State’s otherwise legitimate exercise of its police power. Mr. Justice Frankfurter, for a majority, declared first that § 2281 made no distinction between the Supremacy Clause and other provisions of the Constitution as a ground for denying enforcement of a state statute, and second that Buder, Bransford, and Case could be distinguished on the ground that they presented no claims of unconstitutionality as such: “If in immediate controversy is not the unconstitutionality of a state law but merely the construction of a state law or the federal law, the three-judge requirement does not become operative.” 369 U. S., at 157. In the Kesler case itself, Mr. Justice Frankfurter said, there was no problem of statutory construction but only a “constitutional question” whether the state enactment was pre-empted. After what can only be characterized as extensive statutory analysis (369 U. S., at 158-174) the majority concluded that there had in fact been no pre-emption. 19

*124 III.

In re-examining the Kesler rule the admonition that § 2281 is to be viewed “not as a measure of broad social policy to be construed with great liberality, but as an enactment technical in the strict sense of the term and to be applied as such,” Phillips v. United States, 312 U. S. 246, 251, should be kept in mind. The Kesler opinion itself reflects this admonition, for its rationalization of Buder, Bmnsford, and Case as being consistent with the view that Supremacy Clause cases are not excluded from “the comprehensive language of § 2281,” 369 U. S., at 156, is otherwise most difficult to explain.

As a procedural rule governing the distribution of judicial responsibility the test for applying § 2281 must be clearly formulated. The purpose of the three-judge scheme was in major part to expedite important litigation: it should not be interpreted in such a way that litigation, like the present one, is delayed while the proper composition of the tribunal is litigated. We are now convinced that the Kesler rule, distinguishing between cases in which substantial statutory construction is required and those in which the constitutional issue is “immediately” apparent, is in practice unworkable. Not only has it been uniformly criticized by commentators, 20 but lower courts have quite evidently sought to avoid dealing with its application 21 or have interpreted it with uncertainty. 22 As Judge Friendly’s opinion for the court below demonstrates, in order to ascertain the *125 correct forum, the merits must first be adjudicated in order to discover whether the court has “engaged in so much more construction than in Kesler as to make that ruling inapplicable.” 230 F. Supp., at 410. Such a formulation, whatever its abstract justification, cannot stand as an every-day test for allocating litigation between district courts of one and three judges.

Two possible interpretations of § 2281 would provide a more practicable rule for three-judge court jurisdiction. The first is that Kesler might be extended to hold, as some of its language might be thought to indicate, 23 that all suits to enjoin the enforcement of a state statute, whatever the federal ground, must be channeled through three-judge courts. The second is that no such suits resting solely on “supremacy” grounds fall within the statute.

The first alternative holds some attraction. First, it is relatively straightforward: a court need not distinguish among different constitutional grounds for the requested injunction; it need look only at the relief sought. Moreover, in those cases, as in that before us, in which an injunction is sought on several grounds, the proper forum would not depend on whether certain alleged constitutional grounds turn out to be insubstantial. Second, § 2281 speaks of “unconstitutionality,” and, to be sure, any determination that a state statute is void for obstructing a federal statute does rest on the Supremacy Clause of the Federal Constitution. And, third, there is some policy justification for a wider rule. In a broad sense, what concerned the legislators who passed the progenitor of § 2281 was the voiding of state legislation by inferior federal courts. The sensibilities of the citizens, and *126 perhaps more particularly of the state officials, were less likely to be offended, the Congress thought, by a judgment considered and handed down by three judges rather than by one judge. This rationale can be thought to be as applicable to a suit voiding state legislation on grounds of conflict with a federal statute as it is to an identical suit alleging a conflict with the Federal Constitution directly.

Persuasive as these considerations may be, we believe that the reasons supporting the second interpretation, that is, returning to the traditional Buder-Bransford-Case rule, should carry the day. This restrictive view of the application of § 2281 is more consistent with a discriminating reading of the statute itself than is the first and more embracing interpretation. The statute requires a three-judge court in order to restrain the enforcement of a state statute “upon the ground of the unconstitutionality of such statute.” Since all federal actions to enjoin a state enactment rest ultimately on the Supremacy Clause, 24 the words “upon the ground of the unconstitutionality of such statute” would appear to be superfluous unless they are read to exclude some types of such injunc-tive suits. 25 For a simple provision prohibiting the restraint of the enforcement of any state statute except by a three-judge court would manifestly have sufficed to embrace every such suit whatever its particular constitutional ground. It is thus quite permissible to read *127 the phrase in question as one of limitation, signifying a congressional purpose to confine the three-judge court requirement to injunction suits depending directly upon a substantive provision of the Constitution, leaving cases of conflict with a federal statute (or treaty) to follow their normal course in a single-judge court. We do not suggest that this reading of § 2281 is compelled. We do say, however, that it is an entirely appropriate reading, and one that is supported by all the precedents in this Court until Kesler and by sound policy considerations.

An examination of the origins of the three-judge procedure does not suggest what the legislators would have thought about this particular problem, but it does show quite clearly what sort of cases were of concern to them. Their ire was aroused by the frequent grants of injunctions against the enforcement of progressive state regulatory legislation, usually on substantive due process grounds. (See pp. 116-119, supra.) Requiring the collective judgment of three judges and accelerating appeals to this Court were designed to safeguard important state interests. In contrast, a case involving an alleged incompatibility between state and federal statutes, such as the litigation before us, involves more confining legal analysis and can hardly be thought to raise the worrisome possibilities that economic or political predilections will find their way into a judgment. Moreover, those who enacted the three-judge court statute should not be deemed to have been insensitive to the circumstance that single-judge decisions in conflict and pre-emption cases were always subject to the corrective power of Congress, whereas a “constitutional” decision by such a judge would be beyond that ready means of correction and could be dealt with only by constitutional amendment. The purpose of § 2281 to provide greater restraint and dignity at the district court level cannot well be thought generally applicable to cases that involve con *128 flicts between state and federal statutes, in this instance determining whether the Department of Agriculture’s regulations as applied to the labeling of total net weight on frozen stuffed turkeys necessarily renders invalid a New York statute requiring a supplemental net weight figure which excludes the stuffing.

Our decision that three-judge courts are not required in Supremacy Clause cases involving only federal-state statutory conflicts, in addition to being most consistent with the statute’s structure, with pre-Kesler precedent, and with the section’s historical purpose, is buttressed by important considerations of judicial administration. As Mr. Justice Frankfurter observed in Florida Lime & Avocado Growers, Inc. v. Jacobsen, 362 U. S. 73, 92-93 (dissenting opinion):

“[T]he convening of a three-judge trial court makes for dislocation of the normal structure and functioning of the lower federal courts, particularly in the vast non-metropolitan regions; and direct review of District Court judgments by this Court not only expands this Court’s obligatory jurisdiction but contradicts the dominant principle of having this Court review decisions only after they have gone through two judicial sieves . . . .”

Although the number of three-judge determinations each year should not be exaggerated, 26 this Court’s concern for efficient operation of the lower federal courts persuades us to return to the Buder-Bransford-Case rule, *129 thereby conforming with the constrictive view of the three-judge jurisdiction which this Court has traditionally taken. Ex parte Collins, 277 U. S. 565; Oklahoma Gas & Elec. Co. v. Oklahoma Packing Co., 292 U. S. 386; Rorick v. Board of Commissioners, 307 U. S. 208; Phillips v. United States, 312 U. S. 246.

We hold therefore that this appeal is not properly before us under 28 U. S. C. § 1253 and that appellate review lies in the Court of Appeals, where appellants’ alternative appeal is now pending. The appeal is dismissed for lack of jurisdiction.

It is so ordered.

Less than four years ago, this Court decided that a three-judge district court was required in suits brought under 28 U. S. C. § 2281, even though the alleged “ground of the unconstitutionality” of the challenged statute was based upon a conflict between state and federal statutes. Kesler v. Department of Public Safety, 369 U. S. 153.

A state statute may violate the Equal Protection Clause of the Fourteenth Amendment or the Due Process Clause or some other express provision of the Constitution. If so a three-judge court is plainly required by 28 U. S. C. § 2281. But the issue of the “unconstitutionality” of a state statute can be raised as clearly by a conflict between it and an Act of Congress as by a conflict between it and a provision of the Constitution. The Supremacy Clause, contained in Art. VI, cl. 2, of the Constitution, states as much in clear language:

“This Constitution, and the Laws of the United States which shall be made in Pursuance thereof . . . shall be the supreme Law of the Land; and the *130 Judges in every State shall be bound thereby, any Thing in the Constitution or Laws of any State to the Contrary notwithstanding.”

An issue of the “unconstitutionality” of a state statute is therefore presented whether the conflict is between a provision of the Constitution and a state enactment or between the latter and an Act of Congress. What Senator Overman, author of the three-judge provision, said of it in 1910 is as relevant to enjoining a state law on the ground of federal pre-emption as it is to enjoining it because it violates the Fourteenth Amendment:

“The point is, this amendment is for peace and good order in the State. Whenever one judge stands up in a State and enjoins the governor and the attorney-general, the people resent it, and public sentiment is stirred, as it was in my State, when there was almost a rebellion, whereas if three judges declare that a state statute is unconstitutional the people would rest easy under it. But let one little judge stand up against the whole State, and you find the people of the State rising up in rebellion. The whole purpose of the proposed statute is for peace and good order among the people of the States.” 45 Cong. Rec. 7256.

Some of the most heated controversies between State and Nation which this Court has supervised have involved questions whether there was a conflict between a state statute and a federal one or whether a federal Act was so inclusive as to pre-empt state action in the particular area. One of the earliest and most tumultuous was Cohens v. Virginia, 6 Wheat. 264, 440, where the alleged unconstitutionality of a Virginia law was based on the argument that an Act of Congress, authorizing a lottery in the District of Columbia, barred Virginia from making it a criminal offense to sell lottery *131 tickets within that State. The protest from the States was vociferous 1 even though the Court in the end construed the federal Act to keep it from operating in Virginia. Id., at 447. I therefore see no difference between a charge of “unconstitutionality” of a state statute whether the conflict be between it and the Constitution or between it and a federal law. Neither the language of the Supremacy Clause nor reason nor history makes any difference plain.

Pre-emption or conflict of a state law with a federal one is a recurring theme 2 arising in various contexts. The storm against Cohens v. Virginia was a protest against this Court’s acting as referee in a federal-state contest involving pre-emption or a conflict between the *132 laws of the two regimes. Congress has recently been concerned with the problem in another aspect of the matter, 3 when efforts were made to curb the doctrine of preemption by establishing standards for an interpretation of an Act of Congress. 4 The three-judge court is only another facet of the self-same problem.

The history of 28 U. S. C. § 2281, as related by the Court, speaks of the concern of Congress over the power *133 of one judge to bring a halt to an entire state regulatory-scheme. That can — -and will hereafter — happen in all cases of pre-emption or conflict where the Supremacy Clause is thought to require state policy to give way. A fairly recent example is Cloverleaf Co. v. Patterson, 315 U. S. 148, where a federal court injunction in a preemption case suspended Alabama’s program for control of renovated butter — a demonstrably important health measure. The Court in Florida Lime Growers v. Jacobsen, 362 U. S. 73, where one of the issues was pre-emption or conflict between two statutory systems, emphasized that the interest of the States in being free from such injunctive interference at the instance of a single judge outweighed the additional burdens that such a rule imposed on the federal court system. On reflection I think that result better reflects congressional policy even though, as in Cohens v. Virginia, the end result is only a matter of statutory construction.

On the basis of virtually no experience in applying that interpretation of the statute, a majority has now decided that the rule of Kesler is “unworkable” and, therefore, that our previous interpretation of the statute must have been incorrect. I regret that I am unable to join in that decision. My objection is not that the Court has not given Kesler “a more respectful burial,” Gideon v. Wainwright, 372 U. S. 335, 349 (concurring opinion), but that the Court has engaged in unwarranted infanticide.

Stare decisis is no immutable principle. 5 There are many occasions when this Court has overturned a prior decision, especially in matters involving an interpretation of the Constitution or where the problem of statutory construction had constitutional overtones.

An error in interpreting a federal statute may be easily remedied. If this Court has failed to perceive the inten *134 tion of Congress, or has interpreted a statute in such a manner as to thwart the legislative purpose, Congress may change it. The lessons of experience are not learned by judges alone.

I am unable to find a justification for overturning a decision of this Court interpreting this Act of Congress, announced only on March 26, 1962.

If the Court were able to show that our decision in Kesler had thrown the lower courts into chaos, a fair case for its demise might be made out. The Court calls the rule “unworkable.” But it is not enough to attach that label. The Court broadly asserts that “lower courts have quite evidently sought to avoid dealing with its [Kesler’s] application or have interpreted it with uncertainty.” For this proposition, only three cases (in addition to the instant case) are cited. The Court's failure to provide more compelling documentation for its indictment of Kesler is not the result of less than meticulous scholarship, for so far as I have been able to discover, the truth of the matter is that there are no cases (not even the three cited) even remotely warranting the conclusion that Kesler is “unworkable.”

Kesler was an attempt to harmonize our earlier cases. If the Kesler test is “unworkable” as the Court asserts, we should nonetheless accept its basic premise:

“Neither the language of § 2281 nor the purpose which gave rise to it affords the remotest reason for carving out an unfrivolous claim of unconstitutionality because of the Supremacy Clause from the comprehensive language of § 2281.” 369 U. S., at 156.

If there is overruling to be done, we should overrule Ex parte Buder, 271 U. S. 461, and Ex parte Bransford, 310 U. S. 354.

That the ground of unconstitutionality in many so-called Supremacy Clause cases is found only in the asserted conflict between federal and state statutes is, *135 as I have said, no basis for distinguishing that class of cases from others in which three-judge courts are plainly required. While courts are, strictly speaking, engaging in statutory construction in such cases, the task of adjudication is much the same as in what all would concede to be constitutional adjudication. Though the purpose of Congress is the final touchstone, the interests which must be taken into account in either case are much the same, as Cohens v. Virginia eloquently demonstrates.

The Court has decided, on no more than the gloomy predictions contained in a handful of law review articles, that Kesler would inevitably produce chaos in the federal courts, that the rule announced there is “unworkable.” Those predictions have plainly not been borne out. If difficulties arise, Congress can cure them. Until Congress acts, I would let Kesler stand.

I therefore believe that a three-judge court was properly convened and that we should decide this appeal on the merits.