dissenting.
The basic question that this case presents is whether Congress intended to prevent a State from computing workers’ compensation benefits on the basis of the entire remuneration of injured employees when a portion of that remuneration is provided by an employee benefit plan. By converting unnecessarily broad dicta interpreting the words “relate to” as used in § 514(a) of the Employee Retirement Income Security Act of 1974 (ERISA), 29 U. S. C. § 1144(a), into a rule of law, and by underestimating the significance of the exemption of workers’ compensation plans from the coverage of the Act, the Court has reached an incorrect conclusion in an unusually important case.
In today’s world the typical employee’s compensation is not just her take-home pay; it often includes fringe benefits such as vacation pay and health insurance. If an employee loses her job, by reason of either a wrongful discharge or a negligently inflicted physical injury, normal contract or tort principles would allow her to recover damages measured by her entire loss of earnings — including the value of fringe benefits such as health insurance. If I understand the Court’s reasoning today, a state statute that merely announced that basic rule of damages law would be pre-empted *134by ERISA if it “specifically refers” to each component of the damages calculation. Ante, at 130.1
Workers’ compensation laws provide a substitute for tort actions by employees against their employers. They typically base the amount of the compensation award on the level of the employee’s earnings at the time of the injury. In the District of Columbia’s workers’ compensation law, for example, an employee’s “average weekly wages” provide the basic standard for computing the award regardless of the nature of the injury. D. C. Code Ann. § 36-308 (1988 and Supp. 1992). Because an employee who receives health insurance benefits typically has a correspondingly reduced average weekly wage, the District decided to supplement the standard level of workers’ compensation with a component reflecting any health insurance benefits the worker receives. The Court seems to be holding today that such a supplement may never be measured by the level of the employee’s health insurance coverage — at least if the state statutes or regulations specifically refer to that component of the calculation.
It is true, as the Court points out, that in Shaw v. Delta Air Lines, Inc., 463 U. S. 85, 96-97 (1983), we stated that a law “related to” an employee benefit plan, “in the normal sense of the phrase, if it has a connection with or reference to such a plan.” It is also true that we have repeatedly quoted that language in later opinions.2 Indeed, it has been *135reiterated so often that petitioners did not challenge the proposition that the statute at issue in this case “related to” respondent’s ERISA plan. It nevertheless is equally true that until today that broad reading of the phrase has not been necessary to support any of this Court’s actual holdings.
Given the open-ended implications of today’s holding and the burgeoning volume of litigation involving ERISA preemption claims,31 think it is time to take a fresh look at the intended scope of the pre-emption provision that Congress enacted. Let me begin by repeating the qualifying language in the Shaw opinion itself and by emphasizing one word in the statutory text that is often overlooked.
After explaining why the two New York statutes at issue related to benefit plans, we noted:
“Some state actions may affect employee benefit plans in too tenuous, remote, or peripheral a manner to warrant a finding that the law ‘relates to’ the plan. Cf. American Telephone and Telegraph Co. v. Merry, 592 F. 2d 118, 121 (CA2 1979) (state garnishment of a spouse’s pension income to enforce alimony and support orders is not pre-empted). The present litigation plainly does not present a borderline question, and we express no views about where it would be appropriate to draw the line.” Id., at 100, n. 21.
*136In deciding where that line should be drawn, I would begin by emphasizing the fact that the so-called “pre-emption” provision in ERISA does not use the word “pre-empt.” It provides that the provisions of the federal statute shall “supersede any and all State laws insofar as they may now or hereafter relate to any employee benefit plan described in section 1003(a) of this title and not exempt under section 1003(b) of this title.” 29 U. S. C. § 1144(a) (emphasis added). Thus the federal statute displaces state regulation in the field that is regulated by ERISA; it expressly disavows an intent to supersede state regulation of exempt plans; and its text is silent about possible pre-emption of state regulation of subjects not regulated by the federal statute. Thus, if we were to decide this case on the basis of nothing more than the text of the statute itself, we would find no pre-emption (more precisely, no “supersession”) of the District’s regulation of health benefits for employees receiving workers’ compensation because that subject is entirely unregulated by ERISA.4
I would not decide this case on that narrow ground, however, because both the legislative history of ERISA and *137prior holdings by this Court have given the supersession provision a broader reading. Thus, for example, in Shaw itself we held that the New York Human Rights Law, which prohibited employers from structuring their employee benefit plans in a manner that discriminated on the basis of pregnancy, was pre-empted even though ERISA did not contain any superseding regulatory provisions. 463 U. S., at 98. State laws that directly regulate ERISA plans, or that make it necessary for plan administrators to operate such plans differently, “relate to” such plans in the sense intended by Congress. In my opinion, a state law’s mere reference to an ERISA plan is an insufficient reason for concluding that it is pre-empted — particularly when the state law itself is related almost solely to plans that Congress expressly excluded from the coverage of ERISA. It is anomalous to conclude that ERISA has superseded state regulation in an area that is expressly excluded from the coverage of ERISA.
The statute at issue in this case does not regulate any ERISA plan or require any ERISA plan administrator to make any changes in the administration of such a plan. Although the statute may grant injured employees who receive health insurance a better compensation package than those who are not so insured, it does so only to prevent a converse windfall going to injured employees who receive high weekly wages and little or no health insurance coverage.5 Even if the District’s statute did encourage an employer to pay higher wages instead of providing better fringe benefits, that would surely be no reason to infer a congressional intent to supersede state regulation of a category of compensation programs that it exempted from federal coverage. Moreover, by requiring an injured worker’s compensation to reflect his entire pay package, the statute attempts to replace fully the lost earning power of every injured employee. Noth*138ing in ERISA suggests an intent to supersede the State’s efforts to enact fair and complete remedies for work-related injuries; it is difficult to imagine how a State could measure an injured worker’s health benefits without referring to the specific health benefits that worker receives. Any State that wishes to effect the equitable goal of the District’s statute will be forced by the Court’s opinion to require a predetermined rate of health insurance coverage that bears no relation to the compensation package of each injured worker. The Court thereby requires workers’ compensation laws to shed their most characteristic element: postinjury compensation based on each individual worker’s preinjury level of compensation.
Instead of mechanically repeating earlier dictionary definitions of the word “relate” as its only guide to decision in an important and difficult area of statutory construction, the Court should pause to consider, first, the wisdom of the basic rule disfavoring federal pre-emption of state laws, and second, the specific concerns identified in the legislative history as the basis for federal pre-emption. The most expansive statement of that purpose was quoted in our opinion in Shaw. As explained by Congressman Dent, the “crowning achievement” of the legislation was the “ ‘reservation to Federal authority [of] the sole power to regulate the field of employee benefit plans. With the preemption of the field, we round out the protection afforded participants by eliminating the threat of conflicting and inconsistent State and local regulation.’” Id., at 99 (quoting 120 Cong. Rec. 29197 (1974)).
The statute at issue in this case does not regulate even one inch of the pre-empted field, and poses no threat whatsoever of conflicting and inconsistent state regulation. By its holding today the Court enters uncharted territory. Where that holding will ultimately lead, I do not venture to predict. I am persuaded, however, that the Court has already taken a step that Congress neither intended nor foresaw.
Accordingly, I respectfully dissent.
Similar arguments have been considered and rejected in several cases. See Martori Bros. Distributors v. James-Massengale, 781 F. 2d 1349, 1358-1359 (CA9), modified, 791 F. 2d 799, cert. denied, 479 U. S. 949 (1986); Teper v. Park West Galleries, Inc., 431 Mich. 202, 216, 427 N. W. 2d 535, 541 (1988); Schultz v. National Coalition of Hispanic Mental Health and Human Services Organizations, 678 F. Supp. 936, 938 (DC 1988); Jaskilka v. Carpenter Technology Corp., 757 F. Supp. 175, 178 (Conn. 1991).
See Ingersoll-Rand Co. v. McClendon, 498 U. S. 133, 138-139 (1990); FMC Corp. v. Holliday, 498 U. S. 52, 58-59 (1990); Mackey v. Lanier Collection Agency & Service, Inc., 486 U. S. 825, 829 (1988); Fort Halifax Packing Co. v. Coyne, 482 U. S. 1, 11 (1987); Pilot Life Ins. Co. v. Dedeaux, *135481 U. S. 41, 47-48 (1987); Metropolitan Life Ins. Co. v. Massachusetts, 471 U. S. 724, 739 (1985).
Several years ago a District Judge who had read “nearly 100 cases about the reach of the ERISA preemption clause” concluded that “common sense should not be left at the courthouse door.” See Schultz v. National Coalition of Hispanic Mental Health and Human Services Organizations, 678 F. Supp., at 938. A recent LEXIS search indicates that there are now over 2,800 judicial opinions addressing ERISA pre-emption. This growth may be a consequence of the growing emphasis on the meaning of the words “relate to,” thus pre-empting reliance on what the District Judge referred to as “common sense.”
See, e. g., Cipollone v. Liggett Group, Inc., 505 U. S. 504, 516 (1992): “Consideration of issues arising under the Supremacy Clause ‘start[s] with the assumption that the historic police powers of the States [are] not to be superseded by . . . Federal Act unless that [is] the clear and manifest purpose of Congress.’ Rice v. Santa Fe Elevator Corp., 331 U. S. 218, 230 (1947). Accordingly, ‘“[t]he purpose of Congress is the ultimate touchstone’” of pre-emption analysis. Malone v. White Motor Corp., 436 U. S. 497, 504 (1978) (quoting Retail Clerks v. Schermerhorn, 376 U. S. 96, 103 (1963)).
. . . “In the absence of an express congressional command, state law is pre-empted if that law actually conflicts with federal law, see Pacific Gas & Elec. Co. v. Energy Resources Conservation and Development Comm’n, 461 U. S. 190, 204 (1983), or if federal law so thoroughly occupies a legislative field ‘ “as to make reasonable the inference that Congress left no room for the States to supplement it.”’ Fidelity Fed. Sav. & Loan Assn. v. De la Cuesta, 468 U. S. 141, 153 (1982) (quoting Rice v. Sante Fe Elevator Corp., 331 U. S., at 230).”
One of the statute’s stated goals was “to promote a fairer system of compensation.” Preamble to District of Columbia’s Workers’ Compensation Equity Amendment Act of 1990, 37 D. C. Register 6890 (Nov. 1990).