dissenting.
The Court’s decision substantially alters, in the State of New York, the balance of advantage between management and labor prescribed by the National Labor Relations Act (NLRA). It sustains a New York law that requires the employer, after a specified time, to pay striking employees as much as 50% of their normal wages. In so holding, the Court substantially rewrites the principles of pre-emption that have been developed to protect the free collective bargaining which is the essence of federal labor law.
I
The Policy of Free Collective Bargaining
Free collective bargaining is the cornerstone of the structure of labor-management relations carefully designed by Congress when it enacted the NLRA. Of the numerous actions that labor or management may take during collective bargaining *552to bring economic pressure to bear in support of their respective demands, the NLRA protects or prohibits only some. The availability and usefulness of many others depend entirely upon the relative economic strength of the parties.1
What Congress left unregulated is as important as the regulations that it imposed. It sought to leave labor and management essentially free to bargain for an agreement to govern their relationship.2 Congress also intended, by its limited regulation, to establish a fair balance of bargaining power. That balance, once established, obviates the need for substantive regulation of the fairness of collective-bargaining agreements: whatever agreement emerges from bargaining between fairly matched parties is acceptable.3 Thus, the NLRA’s regulations not only are limited in scope but also must be viewed as carefully chosen to create the congres-sionally desired balance in the bargaining relationship. As the Court observed in Motor Coach Employees v. Lockridge, 403 U. S. 274, 286 (1971), the primary impetus for enactment of “a comprehensive national labor law” was the need to stabilize labor relations by "equitably and delicately structuring the balance of power among competing forces so as to further the common good.” 4
*553Because the NLRA’s limits represent a clear congressional choice with respect to the freedom and fairness of the bargaining process, the Court has been alert to prevent interference with collective bargaining that is unwarranted by the NLRA. For example, in NLRB v. Insurance Agents, 361 U. S. 477 (1960), the Court rejected the conclusion of the National Labor Relations Board (Board) that certain on-the-job conduct undertaken by employees to support their bargaining demands was inconsistent with the union’s duty to bargain in good faith. The Court, noting that the NLRA did not prohibit such actions, id., at 498, concluded that allowing the Board to regulate the availability of such economic weapons would intrude on the area deliberately left unregulated by Congress.5
The Court employed the same analysis in reversing the Board’s determination that the NLRA was violated by a lockout conducted to bring economic pressure to bear in support of the employer’s bargaining position. American Ship Building Co. v. NLRB, 380 U. S. 300, 308 (1965). It rejected the Board’s suggestion that, in enforcing the employer’s duty to bargain in good faith, the Board could deny to the employer the use of certain economic weapons not otherwise proscribed by § 8.
“While a primary purpose of the National Labor Relations Act was to redress the perceived imbalance of economic power between labor and management, it sought *554to accomplish that result by conferring certain affirmative rights on employees and by placing certain enumerated restrictions on the activities of employers. . . . Having protected employee organization in countervailance to the employers’ bargaining power, and having established a system of collective bargaining whereby the newly coequal adversaries might resolve their disputes, the Act also contemplated resort to economic weapons should more peaceful measures not avail. [The NLRA does] not give the Board a general authority to assess the relative economic power of the adversaries in the bargaining process and to deny weapons to one party or the other because of its assessment of that party’s bargaining power.” 380 U. S., at 316-317.
The States have no more authority than the Board to upset the balance that Congress has struck between labor and management in the collective-bargaining relationship. “For a state to impinge on the area of labor combat designed to be free is quite as much an obstruction of federal policy as if the state were to declare picketing free for purposes or by methods which the federal Act prohibits.” Garner v. Teamsters, 346 U. S. 486, 500 (1953). In Teamsters v. Morton, 377 U. S. 252, 259-260 (1964), the Court held that a state law allowing damages for peaceful secondary picketing was pre-empted because “the inevitable result [of its application] would be to frustrate the congressional determination to leave this weapon of self-help available, and to upset the balance of power between labor and management expressed in our national labor policy.” Id., at 259-260. The Court followed the same approach in Machinists v. Wisconsin Employment Relations Comm’n, 427 U. S. 132 (1976), where it held pre-empted a state law under which the union had been enjoined from a concerted refusal to work overtime. Its prior decisions, the Court concluded, indicated that such activities, “whether of employer or employees, were not to be regulable by States any *555more than by the NLRB, for neither States nor the Board is 'afforded flexibility in picking and choosing which economic devices of labor and management shall be branded as unlawful.’ ” Id., at 149, quoting NLRB v. Insurance Agents, supra, at 498.
II
Free Collective Bargaining and the New York Statute
The plurality’s opinion, after acknowledging that the payment of benefits financed ultimately by the employer was “a substantial factor” in the employees’ decision to strike and remain on strike, ante, at 525, further concedes — as it must— that the New York law “has altered the economic balance” between management and labor. Ante, at 532. During the strike out of which the present controversy arose, the petitioners’ employees collected more than $49 million in unemployment compensation. All but a small fraction of these benefits were paid from the petitioners’ accounts in the New York unemployment insurance fund; because of these payments, the petitioners’ tax rates were increased in subsequent periods.6 The challenged provisions of the New York statute thus had a “twofold impact” on the bargaining process (ante, *556at 526 n. 5, 531-532): they substantially cushioned the economic impact of the lengthy strike on the striking employees, and also made the strike more expensive for the employers.7
Nothing in the NLRA or its legislative history indicates that Congress intended unemployment compensation for strikers, let alone employer financing of such compensation, to be part of the legal structure of collective bargaining.8 The New York law therefore alters significantly the bargaining balance prescribed by Congress in that law. The decision upholding it cannot be squared with Morton and Machinists, *557where far less intrusive state statutes were invalidated because they “upset the balance of power between labor and management expressed in our national labor policy.” Morton, 377 U. S., at 260.9
The plurality’s opinion seeks to avoid this conclusion by ignoring the fact that the petitioners are not challenging the entire New York unemployment compensation law but only that portion of it that provides for benefits for striking employees. Although the plurality characterizes the State’s unemployment compensation law as “a law of general applicability” that “implement [s] a broad state policy that does not primarily concern labor-management relations,” ante, at 533, 534, this description bears no relation to reality when applied to the challenged provisions of the law. Those provisions are “of general applicability” only if that term means — contrary to what the plurality itself says — generally applicable only to labor-management relations. It would be difficult to think of a law more specifically focused on labor-management relations than one that compels an employer to finance a strike against itself.10
Even if the challenged portion of the New York statute properly could be viewed as part of a law of “general applica*558bility,” this generality of the law would have little or nothing to do with whether it is pre-empted by the NLRA. A state law with purposes and applications beyond the area of industrial relations nonetheless may impinge upon congressional policy when it is applied to the collective-bargaining relationship.11 The Court has recognized accordingly that pre-emption must turn not on the generality of purpose or applicability of a state law but on the effect of that law when applied in the context of labor-management relations. The “crucial inquiry regarding pre-emption” is whether the application of the state law in question “ 'would frustrate effective implementation of the [NLRA’s] processes.’ ” Machinists, 427 U. S., at 147-148, quoting Railroad Trainmen v. Jacksonville Terminal Co., 394 U. S. 369, 380 (1969). As the Court stated in Farmer v. Carpenters, 430 U. S. 290, 300 (1977):
“[I]t is well settled that the general applicability of a state cause of action is not sufficient to exempt it from pre-emption. '[I]t [has not] mattered whether the States have acted through laws of broad general application rather than laws specifically directed towards the governance of industrial relations.’ Garmon, 359 U. S., at 244. Instead, the cases reflect a balanced inquiry into such factors as the nature of the federal and state interests in regulation and the potential for interference with federal regulation.” (Footnote omitted.)
Accord, Sears, Roebuck & Co. v. Carpenters, 436 U. S. 180, 193, and n. 22 (1978). It is self-evident that the “potential [of the New York law] for interference” (Morton, supra, at *559260) with the federally protected economic balance between management and labor is direct and substantial.12
The Court has identified several categories of state laws whose application is unlikely to interfere with federal regulatory policy under the NLRA. Farmer v. Carpenters, supra, at 296-297. Mr. Justice Frankfurter described one of these categories in broad terms in San Diego Building Trades Council v. Garmon, 359 U. S. 236, 243-244 (1959):
“[States retain authority to regulate] where the regulated conduct touche[s] interests so deeply rooted in local feeling and responsibility that, in the absence of compelling congressional direction, we could not infer that Congress had deprived the States of the power to act.”
The plurality, attempting to draw support from the foregoing generalization, mistakenly treats New York’s requirement that employers pay benefits to striking employees as state action “deeply rooted in local feeling and responsibility.” 13 But *560the broad language from Garmon has been applied only to a narrow class of cases. In Garmon, Mr. Justice Frankfurter identified, as typical of the kind of state law that would not be pre-empted, “the traditional law of torts.” Id., at 247; cf. id., at 244 n. 2. The Court has adhered to this understanding of the “local feeling and responsibility” exception formulated in Garmon. See Machinists, 427 U. S., at 136, and n. 2 (“Policing of actual or threatened violence to persons or destruction of property has been held most clearly a matter for the States”); id., at 151 n. 13 ; Farmer v. Carpenters, supra, at 296-300; cf. Sears, supra, at 194-197. The provisions of the New York law at issue here have nothing in common with the state laws protecting against personal torts or violence to property that have defined the “local feeling and responsibility” exception to pre-emption.
Ill
The Lack of Evidence of Congressional Intent to Alter the Policy of the NLRA
The challenged provisions of the New York law cannot, consistently with prior decisions of this Court, be brought within the “local feeling and responsibility” exception to the pre-emption doctrine. The principles of Morton and Machinists therefore require pre-emption in this case unless in some other law Congress has modified the policy of the NLRA. The plurality, acknowledging the need to look beyond the NLRA to support its conclusion, relies primarily on the Social Security Act. In that Act, adopted only five weeks after the passage of the NLRA, it finds an indication that Congress did intend that the States be free to make unemployment compensation payments part of the collective-bargaining relation*561ship structured by the NLRA. But it is extremely unlikely that little over a month after enacting a detailed and carefully designed statute to structure industrial relations, the Congress would alter so dramatically the balance struck in that law. It would be even more remarkable if such a change were made, as the plurality suggests, without any explicit statutory expression, and indeed absent any congressional discussion whatever of the problem.
The Social Security Act, as the plurality acknowledges, ante, at 540, is silent on the question, neither authorizing the States to provide unemployment compensation for strikers nor prohibiting the States from making such aid available. Congress did explicitly forbid the States to condition unemployment compensation benefits upon acceptance of work as strikebreakers, or membership in a company union, or nonmember-ship in any labor union,14 thereby indicating an intention to prohibit interference with the collective-bargaining balance struck in the NLRA.
Nor does the legislative history of the Social Security Act reflect any congressional intention to allow unemployment compensation for strikers.15 Senator Wagner, a sponsor of *562the prop.osed legislation, made no reference to any such feature of the Social Security Act in his remarks to the Senate Finance Committee. Hearings on S. 1130 before the Senate Committee on Finance, 74th Cong., 1st Sess., 1-30 (1935).16 Although the suggestion that the Act should contain an explicit prohibition of unemployment compensation to strikers was included in several written submissions to the Senate Committee, there is no evidence whatever that the Committee considered the suggestion.17 Indeed, it is clear that the prob*563lem never received congressional attention, for the subject is mentioned nowhere in the Committee Reports or the congressional debates on the Social Security Act. H. R. Rep. No. 615, 74th Cong., 1st Sess. (1935); S. Rep. No. 628, 74th Cong., 1st Sess. (1935); 79 Cong. Rec. 5467-5478, 5528-5563, 5579-5606, 5678-5715, 5768, 5771-5817, 5856-5909, 5948-5994, 6037-6068, 9191, 9267-9273, 9282-9297, 9351-9362, 9366, 9418-9438, 9440, 9510-9543, 9625-9650, 11320-11343 (1935).18
*564Faced with the absence of any specific indications in the Social Security Act or its legislative history that Congress intended for the States to have the authority to upset the NLRA’s collective-bargaining relationship by paying compensation to strikers, the plurality relies on the general policy embodied in the Social Security Act of leaving to the States the determination of eligibility requirements for compensation. Ante, at 537-538, 542, and n. 42.19 That policy sup*565ports the narrow interpretation of the few conditions on eligibility imposed on the States by the Social Security Act itself. Ohio Bureau of Employment Services v. Hodory, 431 U. S. 471, 475 n. 3, 482-489 (1977). But there is no indication in that Act or its legislative history that Congress thought that this general policy relieved the States of constraints imposed by other federal statutes such as the NLRA.20 In particular, it would be difficult indeed to infer from this feature of the Act that Congress intended to leave the States free to require employers to fund unemployment compensation for their striking employees without regard to the effect on the bargaining relationship structured by the NLRA.
The plurality holds, nonetheless, that New York may require employers to pay unemployment compensation to strikers amounting to some 50% of their average wage. Nothing in the plurality’s opinion, moreover, limits such compensation to 50% of average wages, for the plurality indicates that the Social Security Act gives the States complete control over this aspect of their unemployment compensation programs. Accordingly, New York and other States are free not only to increase compensation to 100% but also to eliminate the waiting period now imposed on striking employees.21 The plurality’s *566sweeping view of the Act thus lays open the way for any State to undermine completely the collective-bargaining process within its borders.
A much more cautious approach to implied amendments of the NLRA is required if the Court is to give proper effect to the legislative judgments of the Congress. Having once resolved the balance to be struck in the collective-bargaining relationship, and having embodied that balance in the NLRA, Congress should not be expected by the Court to reaffirm the balance explicitly each time it later enacts legislation that may touch in some way on the collective-bargaining relationship. Absent explicit modification of the NLRA, or clear inconsistency between the terms of the NLRA and a subsequent statute, the Court should assume that Congress intended to leave the NLRA unaltered.22 This assumption is especially *567appropriate in considering the intent of Congress when it enacted the Social Security Act just five weeks after completing its deliberations on the NLRA.
IY
The effect of the New York statute is to require an employer to pay a substantial portion of the wages of employees who are performing no services in return because they have voluntarily gone on strike. This distorts the core policy of the NLRA — the protection of free collective bargaining. Whether that national policy should be subject to such substantial alteration by any state legislature is a decision that the Congress should make after the plenary consideration and public debate that customarily accompany major legislation. The financing of striking employees by employers under unemployment compensation systems such as that of New York has never received any such consideration by Congress. The Court today, finding nothing in any statute, congressional committee report, or debate that indicates any intention to allow States to alter the balance of collective bargaining in this major way, rests its decision on inferences drawn from only the most fragmentary evidence.
I would hold, as it seems to me our prior decisions compel, that the New York statute contravenes federal law. It would then be open to the elected representatives of the people in Congress to address this issue in the way that our system contemplates.
See Machinists v. Wisconsin Employment Relations Comm’n, 427 U. S. 132, 134-135, 140-148 (1976).
The tension between the value of freedom of contract and the legal ordering of the collective-bargaining relationship is discussed in H. Wellington, Labor and the Legal Process, ch. 2 (1968).
See NLRA §8 (d), 29 U. S. C. § 158 (d); Porter Co. v. NLRB, 397 U. S. 99, 102-104 (1970); Teamsters v. Oliver, 358 U. S. 283, 295-296 (1959); NLRB v. Jones & Laughlin Steel Corp., 301 U. S. 1, 45 (1937).
“An appreciation of the true character of the national labor policy-expressed in the [NLRA] indicates that in providing a legal framework for union organization, collective bargaining, and the conduct of labor disputes, Congress struck a balance of protection, prohibition, and laissez faire in respect to union organization, collective bargaining, and labor disputes that would be upset if a state could also enforce statutes or rules of decision resting upon its views concerning accommodation of the same *553interests.” Cox, Labor Law Preemption Revisited, 85 Harv. L. Rev. 1337, 1352 (1972).
The Court stated:
“[I]f the Board could regulate the choice of economic weapons that may be used as part of collective bargaining, it would be in a position to exercise considerable influence upon the substantive terms on which the parties contract. . . . Our labor policy is not presently erected on a foundation of government control of the results of negotiations. . . . Nor does it contain a charter for the [Board] to act at large in equalizing disparities of bargaining power between employer and union.” 361 U. S., at 490.
Petitioner TELCO’s employees collected $43 million in compensation. Of this amount, approximately $40 million was paid from TELGO’s account in the unemployment insurance fund. 566 F. 2d 388, 390 (CA2 1977); 434 F. Supp. 810, 812-813 (SDNY 1977). The proportion of the $6 million in compensation paid to employees of the other petitioners from the accounts of their employers does not appear in the record. But the overall element of nonemployer financing of compensation is so small that the Court of Appeals simply stated that “New York’s unemployment insurance system is financed entirely by employer contributions, so the cost of making these payments was borne by the struck employers.” 566 F. 2d, at 391.
The petitioners’ own tax rates are tied directly to the payments made to their employees by the so-called “experience rating system.” Under that system, an employer’s rate in any given period varies from the standard of 2.7% primarily according to the amount of benefits paid to its employees during prior periods. N. Y. Lab. Law § 581 (McKinney 1977 and Supp. 1978-1979).
The impact of unemployment compensation for strikers on the collective-bargaining process could be reduced significantly if such payments were funded from general tax revenues. The disruptive effect also would be lessened, though not as markedly, if such payments were funded by the unemployment compensation tax but were not taken into account in calculating experience ratings of individual employers. New York has eschewed both of these middle paths, however, in favor of a system in which such payments are financed directly by the struck employer.
New York is not alone in the course it has chosen. Although New York and Rhode Island are the only States that provide unemployment compensation for all covered employees idled by a strike, a number of other States pay unemployment compensation to strikers under varying conditions. See Grinnell Corp. v. Hackett, 475 F. 2d 449, 457, and n. 7 (CA1), cert. denied, 414 U. S. 858 (1973); Albuquerque-Phoenix Exp., Inc. v. Employment Security Comm’n, 88 N. M. 596, 600-601, 544 P. 2d 1161, 1165-1166 (1975), appeal dismissed sub nom. Kimbell, Inc. v. Employment Security Comm’n, 429 U. S. 804 (1976); U. S. Dept. of Labor, Comparison of State Unemployment Insurance Laws 4^41 (1972). All of those States appear to fund such payments from the unemployment compensation taxes paid by employers and calculated under an experience rating system. Staff Study of House Committee on Ways and Means, Information Relating to Federal-State Unemployment Compensation Laws 2-3 (1974).
At the time that Congress enacted the NLRA, unemployment compensation laws had been enacted in only five States, and only in Wisconsin had the State’s program gone into operation, a year earlier. S. Rep. No. 628, 74th Cong., 1st Sess., 11 (1935). Wisconsin and three of the other States denied unemployment compensation to strikers. The New York law, with its limited provision for compensation to striking employees, would not pay any benefits for another two years. It is not at all remark*557able, therefore, that Congress overlooked the subject of unemployment compensation for strikers under these novel state programs during its consideration of the NLRA. Nor did Congress discuss the subject during its deliberations on the Social Security Act, which deals directly with state unemployment compensation programs. See Part III, infra.
The State’s adjustment of the relative economic strength of the parties to the collective-bargaining relationship is equally effective, and equally disruptive of the balance established by the NLRA, whether it takes the form of restricting or supporting a party’s activities in furtherance of its bargaining demands.
This assessment and readjustment of the collective-bargaining relationship by the state legislature is especially obvious in the challenged New York statute, which contains a special eligibility rule requiring strikers to wait seven weeks longer than other unemployed workers before collecting compensation. See ante, at 523 n. 3.
In reviewing the history of the analogous decisions on the pre-emption of state-court jurisdiction, the Court has observed that “some early cases suggested the true distinction lay between judicial application of general common law, which was permissible, as opposed to state rules specifically designed to regulate labor relations, which were pre-empted,” but that this approach had been unsatisfactory. Motor Coach Employees v. Lockridge, 403 U. S. 274, 290-291 (1971).
The District Court found that the availability of unemployment compensation had a significant effect on the willingness of the petitioners' employees to remain on strike.
“Notwithstanding the State’s adamant position to the contrary, I regard it as a fundamental truism that the availability to, or expectation or receipt of a substantial weekly tax-free payment of money by, a striker is a substantial factor affecting his willingness to go on strike or, once on strike, to remain on strike, in the pursuit of desired goals. This being a truism, one therefore would expect to find confirmation of it everywhere. One does.” 434 F. Supp., at 813-814.
The Court of Appeals accepted this finding by the District Court. 566 F. 2d, at 390. The plurality’s opinion, as already noted, supra, at 555-556, also accepts without question the District Court’s findings on this point.
The plurality supports this approach to the New York law by reference to the Social Security Act, which commits to the States broad control over eligibility requirements for unemployment compensation. This aspect of the Social Security Act, the plurality concludes, makes it
“appropriate to treat New York’s statute with the same deference that we have afforded analogous state laws of general applicability that protect interests ‘deeply rooted in local feeling and responsibility.’ With respect *560to such, laws, we have stated 'that, in the absence of compelling congressional direction, we could not infer that Congress had deprived the States of the power to act,’ San Diego Building Trades Council v. Garmon, 359 U. S. 236, 244.” Ante, at 539-540.
To qualify under federal law, a State’s unemployment compensation program must, among other things, provide that:
“(5) compensation shall not be denied in such State to any otherwise eligible individual for refusing to accept new work under any of the following conditions:
“(A) if the position offered is vacant due directly to a strike, lockout, or other labor dispute;
“(C) if as a condition of being employed the individual would be required to join a company union or to resign from or refrain from joining any bona fide labor organization.” Social Security Act § 903 (a)(5), 49 Stat. 640, 26 U. S. C. § 3304 (a) (5).
The Court of Appeals for the First Circuit, after reviewing the legislative history, also concluded that “unambiguous Congressional intent is lacking” regarding the authorization of state unemployment compensation for striking employees. Grinnell Corp. v. Hackett, 475 F. 2d, at 457. As *562one commentator has concluded, “the absence of legislation and the absence of any discussion in the committee reports relating to this legislation are indicative [that] Congress did not anticipate in detail the problems which would arise when workers claimed benefits when their own unemployment was related either directly or indirectly to a labor dispute.” Haggart, Unemployment Compensation During Labor Disputes, 37 Neb. L. Rev. 668, 674 (1958).
The plurality also finds support for its holding by noting that Senator Wagner, a principal sponsor of both the NLRA and the Social Security Act, was familiar with New York’s unemployment compensation law, and that the Senate Report on the Social Security bill — in the portion thereof discussing the States’ freedom of choice with respect to such laws— expressly mentioned the New York statute as an example. The plurality’s opinion then reasons:
“Even though that reference [in the Senate Report] did not mention the subject of benefits for strikers, it is difficult to believe that Senator Wagner and his colleagues were unaware of such a controversial provision . . . .” Ante, at 541-542.
I agree with the plurality that any provision for unemployment compensation for strikers would have been controversial. Indeed, it strains credulity to think that the entire Congress and the scores of witnesses who testified with respect to this legislation ignored so controversial an issue. On a question of this importance, especially in its relation to the NLRA, there would have been hearings, testimony, lobbying, and debate. I am unwilling to assume that Senator Wagner was “aware of [this] controversial provision” and elected to avoid, by remaining silent, the normal democratic processes of legislation. In any event, the unexpressed awareness of Senator Wagner hardly can be imputed to other Members of the Congress.
Contrary to the implication in the plurality’s opinion, ante, at 543 n. 41, Mr. Witte, the Executive Director of the President’s Committee on *563Economic Security, did not recommend withholding benefits from strikers during a strike. The issue of unemployment compensation for strikers never arose during Mr. Witte’s testimony. The plurality’s reference is to a Report of the Advisory Council to the Committee on Economic Security, a group of 23 “laymen” assembled to “give practical advice to the committee [on Economic Security].” Hearings on S. 1130, at 225. See H. R. Rep. No. 615, 74th Cong., 1st Sess., App. (1935). Mr. Witte did not appear before the Senate Committee to support the report of the Advisory Council, and placed it in the record only at the request of the Senate Committee. The Report of the Committee on Economic Security did not refer to or comment on the subject of compensation for strikers, except perhaps indirectly in its statement that “[t]o serve its purposes, unemployment compensation must be paid only to workers involuntarily unemployed.” Report of the President’s Committee on Economic Security 21 (1935).
Similarly, the question of compensation for striking workers did not arise during the examination of the other two witnesses whose written submissions included suggestions that the Social Security Act should contain an explicit disqualification of strikers. See Hearings on S. 1130, supra, at 458-478, 919-959. The Court should be “extremely hesitant to presume general congressional awareness” of the issue of unemployment compensation for strikers “based only upon a few isolated statements in the thousands of pages of legislative documents.” SEC v. Sloan, 436 U. S. 103, 121 (1978).
Subsequent congressional inaction does not demonstrate an understanding that the Social Security Act modified the NLRA to allow payment of unemployment compensation to strikers. See ante, at 544-545, and n. 44. As the plurality acknowledges, ibid., the 1947 Conference Committee gave no reason for its rejection of an amendment to the NLRA that would have excluded strikers from the statute’s coverage if they collected unemployment compensation. The Committee may have decided that the amend*564ment was redundant, and so not worth the controversy it might provoke if included in the final bill sent to Congress: the House Report approving the amendment had stated that it was recommended to halt the “perversion” of the purposes of social security legislation. H. R. Rep. No. 245, 80th Cong., 1st Sess., 12 (1947). The comments in 1969 of a single Congressman, delivered long after the original passage of the Social Security Act, are of no aid in determining congressional intent on this matter.
The plurality also cites the Railroad Unemployment Insurance Act (RUIA) and the Food Stamp Act, as evidence that Congress intended to allow the States to require employers to finance unemployment compensation to their striking employees. See ante, at 544-545, n. 44. These statutes are simply irrelevant to the question raised by this case. The RUIA, together with the Railway Labor Act, is part of a special system of labor-management relations separate and distinct from the general structure established in the NLRA. The availability of unemployment compensation for strikers within the jurisdiction of the RUIA is conditioned upon their compliance with restrictions on the right to strike that are much more onerous than those imposed by the NLRA. See Detroit & Toledo Shore Line R. Co. v. Transportation Union, 396 U. S. 142, 148-153 (1969); Railway & Steamship Clerks v. Railroad Retirement Bd., 99 U. S. App. D. C. 217, 222-223, 239 F. 2d 37, 42-43 (1956).
Unlike unemployment compensation, which is linked only to an interruption in the employee’s income, food stamps and other general welfare programs are available only when income and assets have become insufficient to supply necessities. See, e. g., 7 U. S. C. § 2014 (a) (1976 ed., Supp. Ill) (“Participation in the food stamp program shall be limited to those households whose incomes and other financial resources . . . are determined to be a substantial limiting factor in permitting them to obtain a more nutritious diet”). Such welfare programs are funded out of general revenues rather than by taxes levied on the employers of those using the stamps. Moreover, when 7 U. S. C. § 2014 (c) was amended in 1977, the Congress *565deleted the proviso that “[r]efusal to work at a plant or site subject to a strike or a lockout for the duration of such strike or lockout shall not be deemed to be a refusal to accept employment.” See 7 U. S. C. § 2014 (c) (1976 ed., Supp. III).
Cf. Nash v. Florida Industrial Comm’n, 389 U. S. 235, 239 (1967) (eligibility requirement in the State’s unemployment compensation law, interfering with NLRA’s policy of protection for employees filing unfair labor practice charges with the Board, held pre-empted).
The Solicitor General would escape this implication of the plurality’s construction of the Social Security Act by concluding that at some point between 50% and 100% of weekly wages, or between an 8-week waiting period and none at all, the policy of the Social Security Act would give way to that of the NLRA.
“It is unnecessary to determine in this case the ultimate scope of the states’ freedom to make payments to strikers that may intrude on or dis*566rupt the collective bargaining process. . . . For example, a statute requiring an employer to pay its employees — through the state unemployment compensation system — 100 percent of wages from the beginning of a strike to- the end would appear to be so far beyond the focus of the Social Security Act and so destructive of the principles of the NLRA as to be beyond the contemplation of Congress in permitting some freedom of choice to the states.” Brief for United States as Amicus Curiae 25 n. 25.
But the Solicitor General is no more successful in identifying the source of this limitation on the modification of the NLRA by the Social Security Act than is the plurality in identifying the source of the modification itself. The plurality refrains from compounding insupportable inferences, apparently accepting instead the open-ended implications of its conclusion that New York is free to pay such unemployment benefits to strikers as it desires.
See Malone v. White Motor Corp., 435 U. S. 497, 515-516 (1978) (Stewart, J., dissenting) (“I do not believe, however, that inferences drawn largely from what Congress did not do in enacting the Disclosure Act are sufficient to override the fundamental policy of the national labor laws to leave undisturbed 'the parties’ solution of a problem which Congress has required them to negotiate in good faith toward solving . . . .’ Teamsters v. Oliver, 358 U. S. 283, 296”). This Court has often stated that implied repeals and modifications of statutes by subsequent congres*567sional enactments are justified only when the two statutes are otherwise irreconcilable. Morton v. Mancari, 417 U. S. 535, 550 (1974); United States v. Welden, 377 U. S. 95, 103 n. 12 (1964); United States v. Borden Co., 308 U. S. 188, 198-199 (1939); cf. Bulova Watch Co. v. United States, 365 U. S. 753, 758 (1961) (a specific statute controls over a general one without regard to priority of enactment).