CIRCUIT CITY STORES, INC. v. ADAMS

Justice Kennedy, delivered the opinion of the Court.

Section 1 of the Federal Arbitration Act (FAA or Act) excludes from the Act's coverage "contracts of employment of seamen, railroad employees, or any other class of workers engaged in foreign or interstate commerce." 9 U. S. C. § 1. All but one of the Courts of Appeals which have addressed the issue interpret this provision as exempting contracts of employment of transportation workers, but not other employment contracts, from the FAA's coverage. A different interpretation has been adopted by the Court of Appeals for the Ninth Circuit, which construes the exemption so that all contracts of employment are beyond the FAA's reach, whether or not the worker is engaged in transportation. It applied that rule to the instant case. We now decide that the better interpretation is to construe the statute, as most of the Courts of Appeals have done, to confine the exemption to transportation workers.

I

In October 1995, respondent Saint Clair Adams applied for a job at petitioner Circuit City Stores, Inc., a national retailer of consumer electronics. Adams signed an employment application which included the following provision:

Adams was hired as a sales counselor in Circuit City's store in Santa Rosa, California.

Two years later, Adams filed an employment discrimination lawsuit against Circuit City in state court, asserting claims under California's Fair Employment and Housing Act, Cal. Govt. Code Ann. § 12900 et seq. (West 1992 and Supp. 1997), and other claims based on general tort theories under California law. Circuit City filed suit in the United States District Court for the Northern District of California, seeking to enjoin the state-court action and to compel arbitration of respondent's claims pursuant to the FAA, 9 U. S. C. §§ 1— 16. The District Court entered the requested order. Respondent, the court concluded, was obligated by the arbitration agreement to submit his claims against the employer to binding arbitration. An appeal followed.

While respondent's appeal was pending in the Court of Appeals for the Ninth Circuit, the court ruled on the key issue in an unrelated case. The court held the FAA does not apply to contracts of employment. See Craft v. Campbell Soup Co., 177 F.3d 1083 (1999). In the instant case, following the rule announced in Craft, the Court of Appeals held the arbitration agreement between Adams and Circuit City was contained in a "contract of employment," and so was not subject to the FAA. 194 F.3d 1070 (1999). Circuit City petitioned this Court, noting that the Ninth Circuit's conclusion that all employment contracts are excluded from the FAA conflicts with every other Court of Appeals to have addressed the question. See, e. g., McWilliams v. Logicon, Inc., 143 F.3d 573, 575-576 (CA10 1998); O'Neil v. Hilton Head Hospital, 115 F.3d 272, 274 (CA4 1997); Pryner v. Tractor Supply Co., 109 F.3d 354, 358 (CA7 1997); Cole v. Burns Int'l Security Servs., 105 F.3d 1465, 1470-1472 (CADC 1997); Rojas v. TK Communications, Inc., 87 F.3d 745, 747-748 (CA5 1996); Asplundh Tree Co. v. Bates, 71 F.3d 592, 596-601 (CA6 1995); Erving v. Virginia Squires Basketball Club, 468 F.2d 1064, 1069 (CA2 1972); Dickstein v. DuPont, 443 F.2d 783, 785 (CA1 1971); Tenney Engineering, Inc. v. United Elec. & Machine Workers of Am., 207 F.2d 450 (CA3 1953). We granted certiorari to resolve the issue. 529 U.S. 1129 (2000).

II

A

Congress enacted the FAA in 1925. As the Court has explained, the FAA was a response to hostility of American courts to the enforcement of arbitration agreements, a judicial disposition inherited from then-longstanding English practice. See, e. g., Allied-Bruce Terminix Cos. v. Dobson, 513 U.S. 265, 270-271 (1995); Gilmer v. Interstate/Johnson Lane Corp., 500 U.S. 20, 24 (1991). To give effect to this purpose, the FAA compels judicial enforcement of a wide range of written arbitration agreements. The FAA's coverage provision, § 2, provides that

We had occasion in Allied-Bruce, supra, at 273-277, to consider the significance of Congress' use of the words "involving commerce" in § 2. The analysis began with a reaffirmation of earlier decisions concluding that the FAA was enacted pursuant to Congress' substantive power to regulate interstate commerce and admiralty, see Prima Paint Corp. v. Flood & Conklin Mfg. Co., 388 U.S. 395, 405 (1967), and that the Act was applicable in state courts and pre-emptive of state laws hostile to arbitration, see Southland Corp. v. Keating, 465 U.S. 1 (1984). Relying upon these background principles and upon the evident reach of the words "involving commerce," the Court interpreted § 2 as implementing Congress' intent "to exercise [its] commerce power to the full." Allied-Bruce, supra, at 277.

The instant case, of course, involves not the basic coverage authorization under § 2 of the Act, but the exemption from coverage under § 1. The exemption clause provides the Act shall not apply "to contracts of employment of seamen, railroad employees, or any other class of workers engaged in foreign or interstate commerce." 9 U. S. C. § 1. Most Courts of Appeals conclude the exclusion provision is limited to transportation workers, defined, for instance, as those workers "`actually engaged in the movement of goods in interstate commerce.' " Cole, supra, at 1471. As we stated at the outset, the Court of Appeals for the Ninth Circuit takes a different view and interprets the § 1 exception to exclude all contracts of employment from the reach of the FAA. This comprehensive exemption had been advocated by amici curiae in Gilmer, where we addressed the question whether a registered securities representative's employment discrimination claim under the Age Discrimination in Employment Act of 1967, 81 Stat. 602, as amended, 29 U. S. C. § 621 et seq., could be submitted to arbitration pursuant to an agreement in his securities registration application. Concluding that the application was not a "contract of employment" at all, we found it unnecessary to reach the meaning of § 1. See Gilmer, supra, at 25, n. 2. There is no such dispute in this case; while Circuit City argued in its petition for certiorari that the employment application signed by Adams was not a "contract of employment," we declined to grant certiorari on this point. So the issue reserved in Gilmer is presented here.

B

Respondent, at the outset, contends that we need not address the meaning of the § 1 exclusion provision to decide the case in his favor. In his view, an employment contract is not a "contract evidencing a transaction involving interstate commerce" at all, since the word "transaction" in § 2 extends only to commercial contracts. See Craft, 177 F. 3d, at 1085 (concluding that § 2 covers only "commercial deal[s] or merchant's sale[s]"). This line of reasoning proves too much, for it would make the § 1 exclusion provision superfluous. If all contracts of employment are beyond the scope of the Act under the § 2 coverage provision, the separate exemption for "contracts of employment of seamen, railroad employees, or any other class of workers engaged in . . . interstate commerce" would be pointless. See, e. g., Pennsylvania Dept. of Public Welfare v. Davenport, 495 U.S. 552, 562 (1990) ("Our cases express a deep reluctance to interpret a statutory provision so as to render superfluous other provisions in the same enactment"). The proffered interpretation of "evidencing a transaction involving commerce," furthermore, would be inconsistent with Gilmer v. Interstate/Johnson Lane Corp., 500 U.S. 20 (1991), where we held that § 2 required the arbitration of an age discrimination claim based on an agreement in a securities registration application, a dispute that did not arise from a "commercial deal or merchant's sale." Nor could respondent's construction of § 2 be reconciled with the expansive reading of those words adopted in Allied-Bruce, 513 U. S., at 277, 279-280. If, then, there is an argument to be made that arbitration agreements in employment contracts are not covered by the Act, it must be premised on the language of the § 1 exclusion provision itself.

Respondent, endorsing the reasoning of the Court of Appeals for the Ninth Circuit that the provision excludes all employment contracts, relies on the asserted breadth of the words "contracts of employment of . . . any other class of workers engaged in . . . commerce." Referring to our construction of § 2's coverage provision in Allied-Bruce —concluding that the words "involving commerce" evidence the congressional intent to regulate to the full extent of its commerce power—respondent contends § 1's interpretation should have a like reach, thus exempting all employment contracts. The two provisions, it is argued, are coterminous; under this view the "involving commerce" provision brings within the FAA's scope all contracts within the Congress' commerce power, and the "engaged in . . . commerce" language in § 1 in turn exempts from the FAA all employment contracts falling within that authority.

This reading of § 1, however, runs into an immediate and, in our view, insurmountable textual obstacle. Unlike the "involving commerce" language in § 2, the words "any other class of workers engaged in . . . commerce" constitute a residual phrase, following, in the same sentence, explicit reference to "seamen" and "railroad employees." Construing the residual phrase to exclude all employment contracts fails to give independent effect to the statute's enumeration of the specific categories of workers which precedes it; there would be no need for Congress to use the phrases "seamen" and "railroad employees" if those same classes of workers were subsumed within the meaning of the "engaged in . . . commerce" residual clause. The wording of § 1 calls for the application of the maxim ejusdem generis, the statutory canon that "[w]here general words follow specific words in a statutory enumeration, the general words are construed to embrace only objects similar in nature to those objects enumerated by the preceding specific words." 2A N. Singer, Sutherland on Statutes and Statutory Construction § 47.17 (1991); see also Norfolk & Western R. Co. v. Train Dispatchers, 499 U.S. 117, 129 (1991). Under this rule of construction the residual clause should be read to give effect to the terms "seamen" and "railroad employees," and should itself be controlled and defined by reference to the enumerated categories of workers which are recited just before it; the interpretation of the clause pressed by respondent fails to produce these results.

Canons of construction need not be conclusive and are often countered, of course, by some maxim pointing in a different direction. The application of the rule ejusdem generis in this case, however, is in full accord with other sound considerations bearing upon the proper interpretation of the clause. For even if the term "engaged in commerce" stood alone in § 1, we would not construe the provision to exclude all contracts of employment from the FAA. Congress uses different modifiers to the word "commerce" in the design and enactment of its statutes. The phrase "affecting commerce" indicates Congress' intent to regulate to the outer limits of its authority under the Commerce Clause. See, e. g., Allied-Bruce, 513 U. S., at 277. The "involving commerce" phrase, the operative words for the reach of the basic coverage provision in § 2, was at issue in Allied-Bruce. That particular phrase had not been interpreted before by this Court. Considering the usual meaning of the word "involving," and the pro-arbitration purposes of the FAA, Allied-Bruce held the "word `involving,' like `affecting,' signals an intent to exercise Congress' commerce power to the full." Ibid. Unlike those phrases, however, the general words "in commerce" and the specific phrase "engaged in commerce" are understood to have a more limited reach. In Allied-Bruce itself the Court said the words "in commerce" are "oftenfound words of art" that we have not read as expressing congressional intent to regulate to the outer limits of authority under the Commerce Clause. Id., at 273; see also United States v. American Building Maintenance Industries, 422 U.S. 271, 279-280 (1975) (phrase "engaged in commerce" is "a term of art, indicating a limited assertion of federal jurisdiction"); Jones v. United States, 529 U.S. 848, 855 (2000) (phrase "used in commerce" "is most sensibly read to mean active employment for commercial purposes, and not merely a passive, passing, or past connection to commerce").

It is argued that we should assess the meaning of the phrase "engaged in commerce" in a different manner here, because the FAA was enacted when congressional authority to regulate under the commerce power was to a large extent confined by our decisions. See United States v. Lopez, 514 U.S. 549, 556 (1995) (noting that Supreme Court decisions beginning in 1937 "ushered in an era of Commerce Clause jurisprudence that greatly expanded the previously defined authority of Congress under that Clause"). When the FAA was enacted in 1925, respondent reasons, the phrase "engaged in commerce" was not a term of art indicating a limited assertion of congressional jurisdiction; to the contrary, it is said, the formulation came close to expressing the outer limits of Congress' power as then understood. See, e. g., The Employers' Liability Cases, 207 U.S. 463, 498 (1908) (holding unconstitutional jurisdictional provision in Federal Employers Liability Act (FELA) covering the employees of "every common carrier engaged in trade or commerce"); Second Employers' Liability Cases, 223 U.S. 1, 48-49 (1912); but cf. Illinois Central R. Co. v. Behrens, 233 U.S. 473 (1914) (noting in dicta that the amended FELA's application to common carriers "while engaging in commerce" did not reach all employment relationships within Congress' commerce power). Were this mode of interpretation to prevail, we would take into account the scope of the Commerce Clause, as then elaborated by the Court, at the date of the FAA's enactment in order to interpret what the statute means now.

A variable standard for interpreting common, jurisdictional phrases would contradict our earlier cases and bring instability to statutory interpretation. The Court has declined in past cases to afford significance, in construing the meaning of the statutory jurisdictional provisions "in commerce" and "engaged in commerce," to the circumstance that the statute predated shifts in the Court's Commerce Clause cases. In FTC v. Bunte Brothers, Inc., 312 U.S. 349 (1941), the Court rejected the contention that the phrase "in commerce" in § 5 of the Federal Trade Commission Act, 38 Stat. 719, 15 U. S. C. § 45, a provision enacted by Congress in 1914, should be read in as expansive a manner as "affecting commerce." See Bunte Bros., supra, at 350-351. We entertained a similar argument in a pair of cases decided in the 1974 Term concerning the meaning of the phrase "engaged in commerce" in § 7 of the Clayton Act, 38 Stat. 731, 15 U. S. C. § 18, another 1914 congressional enactment. See American Building Maintenance, supra, at 277-283; Gulf Oil Corp. v. Copp Paving Co., 419 U.S. 186, 199-202 (1974). We held that the phrase "engaged in commerce" in § 7 "means engaged in the flow of interstate commerce, and was not intended to reach all corporations engaged in activities subject to the federal commerce power." American Building Maintenance, supra, at 283; cf.Gulf Oil, supra, at 202 (expressing doubt as to whether an "argument from the history and practical purposes of the Clayton Act" could justify "radical expansion of the Clayton Act's scope beyond that which the statutory language defines").

The Court's reluctance to accept contentions that Congress used the words "in commerce" or "engaged in commerce" to regulate to the full extent of its commerce power rests on sound foundation, as it affords objective and consistent significance to the meaning of the words Congress uses when it defines the reach of a statute. To say that the statutory words "engaged in commerce" are subject to variable interpretations depending upon the date of adoption, even a date before the phrase became a term of art, ignores the reason why the formulation became a term of art in the first place: The plain meaning of the words "engaged in commerce" is narrower than the more open-ended formulations "affecting commerce" and "involving commerce." See, e. g., Gulf Oil, supra, at 195 (phrase "engaged in commerce" "appears to denote only persons or activities within the flow of interstate commerce"). It would be unwieldy for Congress, for the Court, and for litigants to be required to deconstruct statutory Commerce Clause phrases depending upon the year of a particular statutory enactment.

In rejecting the contention that the meaning of the phrase "engaged in commerce" in § 1 of the FAA should be given a broader construction than justified by its evident language simply because it was enacted in 1925 rather than 1938, we do not mean to suggest that statutory jurisdictional formulations "necessarily have a uniform meaning whenever used by Congress." American Building Maintenance Industries, supra, at 277. As the Court has noted: "The judicial task in marking out the extent to which Congress has exercised its constitutional power over commerce is not that of devising an abstract formula." A. B. Kirschbaum Co. v. Walling, 316 U.S. 517, 520 (1942). We must, of course, construe the "engaged in commerce" language in the FAA with reference to the statutory context in which it is found and in a manner consistent with the FAA's purpose. These considerations, however, further compel that the § 1 exclusion provision be afforded a narrow construction. As discussed above, the location of the phrase "any other class of workers engaged in . . . commerce" in a residual provision, after specific categories of workers have been enumerated, undermines any attempt to give the provision a sweeping, openended construction. And the fact that the provision is contained in a statute that "seeks broadly to overcome judicial hostility to arbitration agreements," Allied-Bruce, 513 U. S., at 272-273, which the Court concluded in Allied-Bruce counseled in favor of an expansive reading of § 2, gives no reason to abandon the precise reading of a provision that exempts contracts from the FAA's coverage.

In sum, the text of the FAA forecloses the construction of § 1 followed by the Court of Appeals in the case under review, a construction which would exclude all employment contracts from the FAA. While the historical arguments respecting Congress' understanding of its power in 1925 are not insubstantial, this fact alone does not give us basis to adopt, "by judicial decision rather than amendatory legislation," Gulf Oil, supra, at 202, an expansive construction of the FAA's exclusion provision that goes beyond the meaning of the words Congress used. While it is of course possible to speculate that Congress might have chosen a different jurisdictional formulation had it known that the Court would soon embrace a less restrictive reading of the Commerce Clause, the text of § 1 precludes interpreting the exclusion provision to defeat the language of § 2 as to all employment contracts. Section 1 exempts from the FAA only contracts of employment of transportation workers.

C

As the conclusion we reach today is directed by the text of § 1, we need not assess the legislative history of the exclusion provision. See Ratzlaf v. United States, 510 U.S. 135, 147— 148 (1994) ("[W]e do not resort to legislative history to cloud a statutory text that is clear"). We do note, however, that the legislative record on the § 1 exemption is quite sparse. Respondent points to no language in either Committee Report addressing the meaning of the provision, nor to any mention of the § 1 exclusion during debate on the FAA on the floor of the House or Senate. Instead, respondent places greatest reliance upon testimony before a Senate subcommittee hearing suggesting that the exception may have been added in response to the objections of the president of the International Seamen's Union of America. See Hearing on S. 4213 and S. 4214 before a Subcommittee of the Senate Committee on the Judiciary, 67th Cong., 4th Sess., 9 (1923). Legislative history is problematic even when the attempt is to draw inferences from the intent of duly appointed committees of the Congress. It becomes far more so when we consult sources still more steps removed from the full Congress and speculate upon the significance of the fact that a certain interest group sponsored or opposed particular legislation. Cf. Kelly v. Robinson, 479 U.S. 36, 51, n. 13 (1986) ("[N]one of those statements was made by a Member of Congress, nor were they included in the official Senate and House Reports. We decline to accord any significance to these statements"). We ought not attribute to Congress an official purpose based on the motives of a particular group that lobbied for or against a certain proposal—even assuming the precise intent of the group can be determined, a point doubtful both as a general rule and in the instant case. It is for the Congress, not the courts, to consult political forces and then decide how best to resolve conflicts in the course of writing the objective embodiments of law we know as statutes.

Nor can we accept respondent's argument that our holding attributes an irrational intent to Congress. "Under petitioner's reading of § 1," he contends, "those employment contracts most involving interstate commerce, and thus most assuredly within the Commerce Clause power in 1925 . . . are excluded from [the] Act's coverage; while those employment contracts having a less direct and less certain connection to interstate commerce . . . would come within the Act's affirmative coverage and would not be excluded." Brief for Respondent 38 (emphases in original).

We see no paradox in the congressional decision to exempt the workers over whom the commerce power was most apparent. To the contrary, it is a permissible inference that the employment contracts of the classes of workers in § 1 were excluded from the FAA precisely because of Congress' undoubted authority to govern the employment relationships at issue by the enactment of statutes specific to them. By the time the FAA was passed, Congress had already enacted federal legislation providing for the arbitration of disputes between seamen and their employers, see Shipping Commissioners Act of 1872, 17 Stat. 262. When the FAA was adopted, moreover, grievance procedures existed for railroad employees under federal law, see Transportation Act of 1920, §§ 300-316, 41 Stat. 456, and the passage of a more comprehensive statute providing for the mediation and arbitration of railroad labor disputes was imminent, see Railway Labor Act of 1926, 44 Stat. 577, 46 U. S. C. § 651 (repealed). It is reasonable to assume that Congress excluded "seamen" and "railroad employees" from the FAA for the simple reason that it did not wish to unsettle established or developing statutory dispute resolution schemes covering specific workers.

As for the residual exclusion of "any other class of workers engaged in foreign or interstate commerce," Congress' demonstrated concern with transportation workers and their necessary role in the free flow of goods explains the linkage to the two specific, enumerated types of workers identified in the preceding portion of the sentence. It would be rational for Congress to ensure that workers in general would be covered by the provisions of the FAA, while reserving for itself more specific legislation for those engaged in transportation. See Pryner v. Tractor Supply Co., 109 F. 3d, at 358 (Posner, C. J.). Indeed, such legislation was soon to follow, with the amendment of the Railway Labor Act in 1936 to include air carriers and their employees, see 49 Stat. 1189, 45 U. S. C. §§ 181-188.

III

Various amici, including the attorneys general of 21 States, object that the reading of the § 1 exclusion provision adopted today intrudes upon the policies of the separate States. They point out that, by requiring arbitration agreements in most employment contracts to be covered by the FAA, the statute in effect pre-empts those state employment laws which restrict or limit the ability of employees and employers to enter into arbitration agreements. It is argued that States should be permitted, pursuant to their traditional role in regulating employment relationships, to prohibit employees like respondent from contracting away their right to pursue state-law discrimination claims in court.

It is not our holding today which is the proper target of this criticism. The line of argument is relevant instead to the Court's decision in Southland Corp. v. Keating, 465 U.S. 1 (1984), holding that Congress intended the FAA to apply in state courts, and to pre-empt state antiarbitration laws to the contrary. See id., at 16.

The question of Southland `s continuing vitality was given explicit consideration in Allied-Bruce, and the Court declined to overrule it. 513 U. S., at 272; see also id., at 282 (O'Connor, J., concurring). The decision, furthermore, is not directly implicated in this case, which concerns the application of the FAA in a federal, rather than in a state, court. The Court should not chip away at Southland by indirection, especially by the adoption of the variable statutory interpretation theory advanced by the respondent in the instant case. Not all of the Justices who join today's holding agreed with Allied-Bruce, see 513 U. S., at 284 (Scalia, J., dissenting); id., at 285 (Thomas, J., dissenting), but it would be incongruous to adopt, as we did in Allied-Bruce, a conventional reading of the FAA's coverage in § 2 in order to implement proarbitration policies and an unconventional reading of the reach of § 1 in order to undo the same coverage. In AlliedBruce the Court noted that Congress had not moved to overturn Southland, see 513 U. S., at 272; and we now note that it has not done so in response to Allied-Bruce itself.

Furthermore, for parties to employment contracts not involving the specific exempted categories set forth in § 1, it is true here, just as it was for the parties to the contract at issue in Allied-Bruce, that there are real benefits to the enforcement of arbitration provisions. We have been clear in rejecting the supposition that the advantages of the arbitration process somehow disappear when transferred to the employment context. See Gilmer, 500 U. S., at 30-32. Arbitration agreements allow parties to avoid the costs of litigation, a benefit that may be of particular importance in employment litigation, which often involves smaller sums of money than disputes concerning commercial contracts. These litigation costs to parties (and the accompanying burden to the courts) would be compounded by the difficult choice-of-law questions that are often presented in disputes arising from the employment relationship, cf. Egelhoff v. Egelhoff, post, at 149 (noting possible "choice-of-law problems" presented by state laws affecting administration of Employee Retirement Income Security Act of 1974 plans), and the necessity of bifurcation of proceedings in those cases where state law precludes arbitration of certain types of employment claims but not others. The considerable complexity and uncertainty that the construction of § 1 urged by respondent would introduce into the enforceability of arbitration agreements in employment contracts would call into doubt the efficacy of alternative dispute resolution procedures adopted by many of the Nation's employers, in the process undermining the FAA's proarbitration purposes and "breeding litigation from a statute that seeks to avoid it." Allied-Bruce, supra, at 275. The Court has been quite specific in holding that arbitration agreements can be enforced under the FAA without contravening the policies of congressional enactments giving employees specific protection against discrimination prohibited by federal law; as we noted in Gilmer, "`[b]y agreeing to arbitrate a statutory claim, a party does not forgo the substantive rights afforded by the statute; it only submits to their resolution in an arbitral, rather than a judicial, forum.' " 500 U. S., at 26 (quoting Mitsubishi Motors Corp. v. Soler Chrysler-Plymouth, Inc., 473 U.S. 614, 628 (1985)). Gilmer, of course, involved a federal statute, while the argument here is that a state statute ought not be denied state judicial enforcement while awaiting the outcome of arbitration. That matter, though, was addressed in Southland and Allied-Bruce, and we do not revisit the question here.

* * *

For the foregoing reasons, the judgment of the Court of Appeals for the Ninth Circuit is reversed, and the case is remanded for further proceedings consistent with this opinion.

It is so ordered.

Justice Stevens, with whom Justice Ginsburg and Justice Breyer join, and with whom Justice Souter joins as to Parts II and III, dissenting.

Justice Souter has cogently explained why the Court's parsimonious construction of § 1 of the Federal Arbitration Act (FAA or Act) is not consistent with its expansive reading of § 2. I join his dissent, but believe that the Court's heavy reliance on the views expressed by the Courts of Appeals during the past decade makes it appropriate to comment on three earlier chapters in the history of this venerable statute.

I

Section 2 of the FAA makes enforceable written agreements to arbitrate "in any maritime transaction or a contract evidencing a transaction involving commerce." 9 U. S. C. § 2. If we were writing on a clean slate, there would be good reason to conclude that neither the phrase "maritime transaction" nor the phrase "contract evidencing a transaction involving commerce" was intended to encompass employment contracts.1

The history of the Act, which is extensive and well documented, makes clear that the FAA was a response to the refusal of courts to enforce commercial arbitration agreements, which were commonly used in the maritime context. The original bill was drafted by the Committee on Commerce, Trade, and Commercial Law of the American Bar Association (ABA) upon consideration of "the further extension of the principle of commercial arbitration." Report of the Forty-third Annual Meeting of the ABA, 45 A. B. A. Rep. 75 (1920) (emphasis added). As drafted, the bill was understood by Members of Congress to "simply provid[e] for one thing, and that is to give an opportunity to enforce an agreement in commercial contracts and admiralty contracts." 65 Cong. Rec. 1931 (1924) (remarks of Rep. Graham) (emphasis added).2 It is no surprise, then, that when the legislation was first introduced in 1922,3 it did not mention employment contracts, but did contain a rather precise definition of the term "maritime transactions" that underscored the commercial character of the proposed bill.4 Indeed, neither the history of the drafting of the original bill by the ABA, nor the records of the deliberations in Congress during the years preceding the ultimate enactment of the Act in 1925, contain any evidence that the proponents of the legislation intended it to apply to agreements affecting employment.

Nevertheless, the original bill was opposed by representatives of organized labor, most notably the president of the International Seamen's Union of America,5 because of their concern that the legislation might authorize federal judicial enforcement of arbitration clauses in employment contracts and collective-bargaining agreements.6 In response to those objections, the chairman of the ABA committee that drafted the legislation emphasized at a Senate Judiciary Subcommittee hearing that "[i]t is not intended that this shall be an act referring to labor disputes, at all," but he also observed that "if your honorable committee should feel that there is any danger of that, they should add to the bill the following language, `but nothing herein contained shall apply to seamen or any class of workers in interstate and foreign commerce.' " Hearing 9. Similarly, another supporter of the bill, then Secretary of Commerce Herbert Hoover, suggested that "[i]f objection appears to the inclusion of workers' contracts in the law's scheme, it might be well amended by stating `but nothing herein contained shall apply to contracts of employment of seamen, railroad employees, or any other class of workers engaged in interstate or foreign commerce.' " Id., at 14. The legislation was reintroduced in the next session of Congress with Secretary Hoover's exclusionary language added to § 1,7 and the amendment eliminated organized labor's opposition to the proposed law.8

That amendment is what the Court construes today. History amply supports the proposition that it was an uncontroversial provision that merely confirmed the fact that no one interested in the enactment of the FAA ever intended or expected that § 2 would apply to employment contracts. It is particularly ironic, therefore, that the amendment has provided the Court with its sole justification for refusing to give the text of § 2 a natural reading. Playing ostrich to the substantial history behind the amendment, see ante, at 119 ("[W]e need not assess the legislative history of the exclusion provision"), the Court reasons in a vacuum that "[i]f all contracts of employment are beyond the scope of the Act under the § 2 coverage provision, the separate exemption" in § 1 "would be pointless," ante, at 113. But contrary to the Court's suggestion, it is not "pointless" to adopt a clarifying amendment in order to eliminate opposition to a bill. Moreover, the majority's reasoning is squarely contradicted by the Court's approach in Bernhardt v. Polygraphic Co. of America, 350 U.S. 198, 200, 201, n. 3 (1956), where the Court concluded that an employment contract did not "evidence `a transaction involving commerce' within the meaning of § 2 of the Act," and therefore did not "reach the further question whether in any event petitioner would be included in `any other class of workers' within the exceptions of § 1 of the Act."

The irony of the Court's reading of § 2 to include contracts of employment is compounded by its cramped interpretation of the exclusion inserted into § 1. As proposed and enacted, the exclusion fully responded to the concerns of the Seamen's Union and other labor organizations that § 2 might encompass employment contracts by expressly exempting the labor agreements not only of "seamen" and "railroad employees," but also of "any other class of workers engaged in foreign or interstate commerce." 9 U. S. C. § 1 (emphasis added). Today, however, the Court fulfills the original—and originally unfounded—fears of organized labor by essentially rewriting the text of § 1 to exclude the employment contracts solely of "seamen, railroad employees, or any other class of [transportation] workers engaged in foreign or interstate commerce." See ante, at 119. In contrast, whether one views the legislation before or after the amendment to § 1, it is clear that it was not intended to apply to employment contracts at all.

II

A quarter century after the FAA was passed, many Courts of Appeals were presented with the question whether collective-bargaining agreements were "contracts of employment" for purposes of § 1's exclusion. The courts split over that question, with at least the Third, Fourth, and Fifth Circuits answering in the affirmative,9 and the First and Sixth Circuits answering in the negative.10 Most of these cases neither involved employees engaged in transportation nor turned on whether the workers were so occupied. Indeed, the general assumption seemed to be, as the Sixth Circuit stated early on, that § 1 "was deliberately worded by the Congress to exclude from the [FAA] all contracts of employment of workers engaged in interstate commerce." Gatliff Coal Co. v. Cox, 142 F.2d 876, 882 (1944).

The contrary view that the Court endorses today—namely, that only employees engaged in interstate transportation are excluded by § 1—was not expressed until 1954, by the Third Circuit in Tenney Engineering, Inc. v. Electrical Workers, 207 F.2d 450, 452 (1953). And that decision, significantly, was rejected shortly thereafter by the Fourth Circuit. See Electrical Workers v. Miller Metal Products, Inc., 215 F.2d 221, 224 (1954). The conflict among the Circuits that persisted in the 1950's thus suggests that it may be inappropriate to attach as much weight to recent Court of Appeals opinions as the Court does in this case. See ante, at 109, 110-111, 112.

Even more important than the 1950's conflict, however, is the way in which this Court tried to resolve the debate. In Textile Workers v. Lincoln Mills of Ala., 353 U.S. 448 (1957), the Court granted certiorari to consider the union's claim that, in a suit brought under § 301 of the Labor Management Relations Act, 1947 (LMRA), a federal court may enforce the arbitration clause in a collective-bargaining agreement. The union argued that such authority was implicitly granted by § 301 and explicitly granted by § 2 of the FAA. In support of the latter argument, the union asked the Court to rule either that a collective-bargaining agreement is not a "contrac[t] of employment" within the meaning of the exclusion in § 1, or that the exclusion is limited to transportation workers.11 The Court did not accept either argument, but held that § 301 itself provided the authority to compel arbitration. The fact that the Court relied on § 301 of the LMRA, a statutory provision that does not mention arbitration, rather than the FAA, a statute that expressly authorizes the enforcement of arbitration agreements, strongly implies that the Court had concluded that the FAA simply did not apply because § 1 exempts labor contracts. That was how Justice Frankfurter, who of course was present during the deliberations on the case, explained the disposition of the FAA issues. See 353 U. S., at 466-468 (dissenting opinion).12

Even if Justice Frankfurter's description of the majority's rejection of the applicability of the FAA does not suffice to establish Textile Workers as precedent for the meaning of § 1, his opinion unquestionably reveals his own interpretation of the Act. Moreover, given that Justice Marshall and I have also subscribed to that reading of § 1,13 and that three more Members of this Court do so in dissenting from today's decision, it follows that more Justices have endorsed that view than the one the Court now adopts. That fact, of course, does not control the disposition of this case, but it does seem to me that it is entitled to at least as much respect as the number of Court of Appeals decisions to which the Court repeatedly refers.

III

Times have changed. Judges in the 19th century disfavored private arbitration. The 1925 Act was intended to overcome that attitude, but a number of this Court's cases decided in the last several decades have pushed the pendulum far beyond a neutral attitude and endorsed a policy that strongly favors private arbitration.14 The strength of that policy preference has been echoed in the recent Court of Appeals opinions on which the Court relies.15 In a sense, therefore, the Court is standing on its own shoulders when it points to those cases as the basis for its narrow construction of the exclusion in § 1. There is little doubt that the Court's interpretation of the Act has given it a scope far beyond the expectations of the Congress that enacted it. See, e. g., Southland Corp. v. Keating, 465 U.S. 1, 17-21 (1984) (Stevens, J., concurring in part and dissenting in part); id., at 21-36 (O'Connor, J., dissenting).

It is not necessarily wrong for the Court to put its own imprint on a statute. But when its refusal to look beyond the raw statutory text enables it to disregard countervailing considerations that were expressed by Members of the enacting Congress and that remain valid today, the Court misuses its authority. As the history of the legislation indicates, the potential disparity in bargaining power between individual employees and large employers was the source of organized labor's opposition to the Act, which it feared would require courts to enforce unfair employment contracts. That same concern, as Justice Souter points out, see post, at 138, n. 2, underlay Congress' exemption of contracts of employment from mandatory arbitration. When the Court simply ignores the interest of the unrepresented employee, it skews its interpretation with its own policy preferences.

This case illustrates the wisdom of an observation made by Justice Aharon Barak of the Supreme Court of Israel. He has perceptively noted that the "minimalist" judge "who holds that the purpose of the statute may be learned only from its language" has more discretion than the judge "who will seek guidance from every reliable source." Judicial Discretion 62 (Y. Kaufmann transl. 1989). A method of statutory interpretation that is deliberately uninformed, and hence unconstrained, may produce a result that is consistent with a court's own views of how things should be, but it may also defeat the very purpose for which a provision was enacted. That is the sad result in this case.

I respectfully dissent.

Justice Souter, with whom Justice Stevens, Justice Ginsburg, and Justice Breyer join, dissenting.

Section 2 of the Federal Arbitration Act (FAA or Act) provides for the enforceability of a written arbitration clause in "any maritime transaction or a contract evidencing a transaction involving commerce," 9 U. S. C. § 2, while § 1 exempts from the Act's coverage "contracts of employment of seamen, railroad employees, or any other class of workers engaged in foreign or interstate commerce." Whatever the understanding of Congress's implied admiralty power may have been when the Act was passed in 1925, the commerce power was then thought to be far narrower than we have subsequently come to see it. As a consequence, there are two quite different ways of reading the scope of the Act's provisions. One way would be to say, for example, that the coverage provision extends only to those contracts "involving commerce" that were understood to be covered in 1925; the other would be to read it as exercising Congress's commerce jurisdiction in its modern conception in the same way it was thought to implement the more limited view of the Commerce Clause in 1925. The first possibility would result in a statutory ambit frozen in time, behooving Congress to amend the statute whenever it desired to expand arbitration clause enforcement beyond its scope in 1925; the second would produce an elastic reach, based on an understanding that Congress used language intended to go as far as Congress could go, whatever that might be over time.

In Allied-Bruce Terminix Cos. v. Dobson, 513 U.S. 265 (1995), we decided that the elastic understanding of § 2 was the more sensible way to give effect to what Congress intended when it legislated to cover contracts "involving commerce," a phrase that we found an apt way of providing that coverage would extend to the outer constitutional limits under the Commerce Clause. The question here is whether a similarly general phrase in the § 1 exemption, referring to contracts of "any . . . class of workers engaged in foreign or interstate commerce," should receive a correspondingly evolutionary reading, so as to expand the exemption for employment contracts to keep pace with the enhanced reach of the general enforceability provision. If it is tempting to answer yes, on the principle that what is sauce for the goose is sauce for the gander, it is sobering to realize that the Courts of Appeals have, albeit with some fits and starts as noted by Justice Stevens, ante, at 129-130 (dissenting opinion),1 overwhelmingly rejected the evolutionary reading of § 1 accepted by the Court of Appeals in this case. See ante, at 110-111 (opinion of the Court) (citing cases). A majority of this Court now puts its imprimatur on the majority view among the Courts of Appeals.

The number of courts arrayed against reading the § 1 exemption in a way that would allow it to grow parallel to the expanding § 2 coverage reflects the fact that this minority view faces two hurdles, each textually based and apparent from the face of the Act. First, the language of coverage (a contract evidencing a transaction "involving commerce") is different from the language of the exemption (a contract of a worker "engaged in . . . commerce"). Second, the "engaged in . . .commerce" catchall phrase in the exemption is placed in the text following more specific exemptions for employment contracts of "seamen" and "railroad employees." The placement possibly indicates that workers who are excused from arbitrating by virtue of the catchall exclusion must resemble seamen and railroad workers, perhaps by being employees who actually handle and move goods as they are shipped interstate or internationally.

Neither hurdle turns out to be a bar, however. The first objection is at best inconclusive and weaker than the grounds to reject it; the second is even more certainly inapposite, for reasons the Court itself has stated but misunderstood.

I

Is Congress further from a plenary exercise of the commerce power when it deals with contracts of workers "engaged in . . . commerce" than with contracts detailing transactions "involving commerce?" The answer is an easy yes, insofar as the former are only the class of labor contracts, while the latter are not so limited. But that is not the point. The question is whether Congress used language indicating that it meant to cover as many contracts as the Commerce Clause allows it to reach within each class of contracts addressed. In Allied-Bruce we examined the 1925 context and held that "involving commerce" showed just such a plenary intention, even though at the time we decided that case we had long understood "affecting commerce" to be the quintessential expression of an intended plenary exercise of commerce power. 513 U. S., at 273-274; see also Wickard v. Filburn, 317 U.S. 111 (1942).

Again looking to the context of the time, I reach the same conclusion about the phrase "engaged in commerce" as a description of employment contracts exempted from the Act. When the Act was passed (and the commerce power was closely confined) our case law indicated that the only employment relationships subject to the commerce power were those in which workers were actually engaged in interstate commerce. Compare The Employers' Liability Cases, 207 U.S. 463, 496, 498 (1908) (suggesting that regulation of the employment relations of railroad employees "actually engaged in an operation of interstate commerce" is permissible under the Commerce Clause but that regulation of a railroad company's clerical force is not), with Hammer v. Dagenhart, 247 U.S. 251, 271-276 (1918) (invalidating statute that had the "necessary effect" of "regulat[ing] the hours of labor of children in factories and mines within the States"). Thus, by using "engaged in" for the exclusion, Congress showed an intent to exclude to the limit of its power to cover employment contracts in the first place, and it did so just as clearly as its use of "involving commerce" showed its intent to legislate to the hilt over commercial contracts at a more general level. That conclusion is in fact borne out by the statement of the then-Secretary of Commerce, Herbert Hoover, who suggested to Congress that the § 1 exclusion language should be adopted "[i]f objection appears to the inclusion of workers' contracts in the law's scheme." Sales and Contracts to Sell in Interstate and Foreign Commerce, and Federal Commercial Arbitration: Hearing on S. 4213 and S. 4214 before a Subcommittee of the Senate Committee on the Judiciary, 67th Cong., 4th Sess., 14 (1923) (hereinafter Hearing on S. 4213 et al.).

The Court cites FTC v. Bunte Brothers, Inc., 312 U.S. 349 (1941), United States v. American Building Maintenance Industries, 422 U.S. 271 (1975), and Gulf Oil Corp. v. Copp Paving Co., 419 U.S. 186 (1974), for the proposition that "engaged in" has acquired a more restricted meaning as a term of art, immune to tampering now. Ante, at 117-118. But none of the cited cases dealt with the question here, whether exemption language is to be read as petrified when coverage language is read to grow. Nor do the cases support the Court's unwillingness to look beyond the four corners of the statute to determine whether the words in question necessarily "`have a uniform meaning whenever used by Congress,' " ante, at 118 (quoting American Building Maintenance, supra, at 277). Compare ante, at 119 ("[W]e need not assess the legislative history of the exclusion provision"), with, e. g., American Building Maintenance, supra, at 279— 283 (examining legislative history and agency enforcement of the Clayton Act before resolving meaning of "engaged in commerce").

The Court has no good reason, therefore, to reject a reading of "engaged in" as an expression of intent to legislate to the full extent of the commerce power over employment contracts. The statute is accordingly entitled to a coherent reading as a whole, see, e. g., King v. St. Vincent's Hospital, 502 U.S. 215, 221 (1991), by treating the exemption for employment contracts as keeping pace with the expanded understanding of the commerce power generally.

II

The second hurdle is cleared more easily still,and the Court has shown how. Like some Courts of Appeals before it, the majority today finds great significance in the fact that the generally phrased exemption for the employment contracts of workers "engaged in commerce" does not stand alone, but occurs at the end of a sequence of more specific exemptions: for "contracts of employment of seamen, railroad employees, or any other class of workers engaged in foreign or interstate commerce." Like those other courts, this Court sees the sequence as an occasion to apply the interpretive maxim of ejusdem generis, that is, when specific terms are followed by a general one, the latter is meant to cover only examples of the same sort as the preceding specifics. Here, the same sort is thought to be contracts of transportation workers, or employees of transporters, the very carriers of commerce. And that, of course, excludes respondent Adams from benefit of the exemption, for he is employed by a retail seller.

Like many interpretive canons, however, ejusdem generis is a fallback, and if there are good reasons not to apply it, it is put aside. E. g., Norfolk & Western R. Co. v. Train Dispatchers, 499 U.S. 117, 129 (1991).2 There are good reasons here. As Adams argued, it is imputing something very odd to the working of the congressional brain to say that Congress took care to bar application of the Act to the class of employment contracts it most obviously had authority to legislate about in 1925, contracts of workers employed by carriers and handlers of commerce, while covering only employees "engaged" in less obvious ways, over whose coverage litigation might be anticipated with uncertain results. It would seem to have made more sense either to cover all coverable employment contracts or to exclude them all. In fact, exclusion might well have been in order based on concern that arbitration could prove expensive or unfavorable to employees, many of whom lack the bargaining power to resist an arbitration clause if their prospective employers insist on one.3 And excluding all employment contracts from the Act's enforcement of mandatory arbitration clauses is consistent with Secretary Hoover's suggestion that the exemption language would respond to any "objection . . . to the inclusion of workers' contracts."

The Court tries to deflect the anomaly of excluding only carrier contracts by suggesting that Congress used the reference to seamen and rail workers to indicate the class of employees whose employment relations it had already legislated about and would be most likely to legislate about in the future. Ante, at 120-121. This explanation, however, does nothing to eliminate the anomaly. On the contrary, the explanation tells us why Congress might have referred specifically to the sea and rail workers; but, if so, it also indicates that Congress almost certainly intended the catchall phrase to be just as broad as its terms, without any interpretive squeeze in the name of ejusdem generis.

The very fact, as the Court points out, that Congress already had spoken on the subjects of sailors and rail workers and had tailored the legislation to the particular circumstances of the sea and rail carriers may well have been reason for mentioning them specifically. But making the specific references was in that case an act of special care to make sure that the FAA not be construed to modify the existing legislation so exactly aimed; that was no reason at all to limit the general FAA exclusion from applying to employment contracts that had not been targeted with special legislation. Congress did not need to worry especially about the FAA's effect on legislation that did not exist and was not contemplated. As to workers uncovered by any specific legislation, Congress could write on a clean slate, and what it wrote was a general exclusion for employment contracts within Congress's power to regulate. The Court has understood this point before, holding that the existence of a special reason for emphasizing specific examples of a statutory class can negate any inference that an otherwise unqualified general phrase was meant to apply only to matters ejusdem generis.4 On the Court's own reading of the history, then, the explanation for the catchall is not ejusdem generis; instead, the explanation for the specifics is ex abundanti cautela, abundance of caution, see Fort Stewart Schools v. FLRA, 495 U.S. 641, 646 (1990).

Nothing stands in the way of construing the coverage and exclusion clauses together, consistently and coherently. I respectfully dissent.