concurring in part and dissenting in part.
I concur in that part of the majority’s opinion which holds that ERISA precludes the Casino Control Commission from regulating the management of pension and welfare funds.1 And, though I do not agree with the majority’s conclusion that we lack *834jurisdiction over the cross-appeals,21 nevertheless share the majority’s view that the district court was correct in declining to abstain.3
*835I cannot subscribe, however, either to the sweeping preemptive scope that the majority ascribes to the NLRA and the LMRDA or to the holding that section 7 of the NLRA precludes New Jersey from restricting who may hold office in unions representing employees in the casino industry. Relying primarily on Hill v. Florida ex rel. Watson, 325 U.S. 538, 65 S.Ct. 1373, 89 L.Ed. 1782 (1945), and Congress’ subsequent failure to overrule Hill, the majority fashions a preemption doctrine that appears to leave no room for state regulation that affects section 7 rights. As I read the relevant case law, however, preemption doctrine is far from absolute, and the strength and importance of New Jersey’s interest in enacting a comprehensive regulatory scheme is critical to a determination whether the provisions challenged here constitute an impermissible intrusion on federally created and protected rights. After considering two post-Hill manifestations of congressional intent strongly suggesting that state regulation is not necessarily inconsistent with national labor policy, the colossal problems associated with casino gambling, and New Jersey’s interest in attempting to pre*836vent the incidence of such problems and the poisoning of its polity, I conclude that federal labor law does not preempt the Casino Control Act’s restrictions on the right of casino-industry employees to select certain individuals as union officials. Accordingly, I dissent from so much of the majority’s opinion and judgment that reaches the contrary conclusion.
I. Labor Preemption — The Generally Applicable Principles
Appellants’ preemption argument is relatively straightforward: section 7 of the NLRA, 29 U.S.C. § 157 (1976), guarantees employees the right “to bargain collectively through representatives of their own choosing.” Section 93 of the Casino Control Act, however, precludes casino employees from selecting as their bargaining representatives individuals who do not meet the Act’s qualification criteria. If employees ignore the statute’s admonition, the Casino Control Commission may prevent their union from collecting dues from its members, thus vitiating the union’s ability to function. Appellants argue that this restriction imper-missibly intrudes on their unqualified section 7 right to select their collective bargaining representatives. Indeed appellants go so far as to say that section 93’s operation would “totally disrupt and prevent the application to [the casino] industry of federal policies as expressed by the [NLRA].” Br. for Appellants at 13.
The majority’s reaction to this contention is aptly characterized by the following passage:
[T]here are in labor law two separate preemption doctrines. The first, covering protected activity, is absolute. ‘When it is clear or may fairly be assumed that the activities which a State purports to regulate are protected by § 7 of the [NLRA], or constitute an unfair labor practice under § 8, due regard for the federal enactment requires that state jurisdiction must yield. To leave the states free to regulate conduct so plainly within the central aim of federal regulation involves too great a danger of conflict between power asserted by Congress and requirements imposed by state law.’ ... On the other hand where the activity in question is not specifically protected by section 7, but is nevertheless federally regulated, a case by case determination of the interaction between state and federal regulatory schemes is required. The Hill v. Florida rule, and this case, fall in the first Category-
Majority op., at 828 (quoting San Diego Building Trades Council v. Garmon, 359 U.S. 236, 244, 79 S.Ct. 773, 779, 3 L.Ed.2d 775 (1959)). Although this formulation supports the majority’s judgment, it does so only by drawing an inaccurate distinction and by oversimplifying preemption analysis.
There are certain areas in which congressional action precludes the states from regulating at all. But Congress generally does not identify these areas by creating amorphous categories of “protected activity.” Rather, the areas in which Congress has usurped the field are generally defined by explicit statutory language (and, in certain rare situations, legislative history). One example of which the majority itself takes note is ERISA, which states:
Except as provided in subsection (b) of this section, the provisions of this sub-chapter and subchapter III of this chapter 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 (1976); see Majority op., at 830. Similar examples may be found in various individual provisions of the NLRA and the LMRDA. E.g, 29 U.S.C. §§ 164(a), 483 (1976) (discussed infra at pp. 838 n. 5, 841).
Most federal statutes, however, do not wear their preemptive nature on their sleeves; many provisions of federal labor law, including section 7 of the NLRA, fall into this latter category. See New York Telephone Co. v. New York State Department of Labor, 440 U.S. 519, 540, 99 S.Ct. 1328, 1341, 59 L.Ed.2d 553 (1979) (plurality *837op.). Obviously, it would be a mistake to infer from such statutory silence that any and all state regulation is permissible. But it also would be a mistake hastily or uncritically to characterize the particular conduct governed by such federal statutes as “plainly within the central aim of federal regulation” and therefore beyond the reach of the states. As Justice Frankfurter wrote in De Veau v. Braisted, 363 U.S. 144, 153, 80 S.Ct. 1146, 1151, 4 L.Ed.2d 1109 (1960) (plurality op.): “It would misconceive the constitutional doctrine of preemption ... to decide this case mechanically on an absolute concept of free choice of representatives on the part of employees, heedless of the light that Congress has shed for our guidance.”
Our task, therefore, is more complex than the majority suggests: in order for us properly to determine the implicit preemptive scope of a federal statute, we must examine thoroughly the purposes and policies of the federal scheme and the extent to which state regulation would be incompatible.
The fact that there is some restriction due to the operation of state law does not settle the issue of preemption. The doctrine of pre-emption does not present a problem in physics but one of adjustment because of the interdependence of federal and state interests and of the interaction of federal and state powers.
Id. at 152, 80 S.Ct. at 1151.
The Supreme Court recently has restated the direction that our inquiry must take in this context:
First, we determine whether the conduct that the state seeks to regulate or to make the basis of liability is actually or arguably protected or prohibited by the NLRA.... [I]f the conduct at issue is arguably prohibited or protected otherwise applicable state law and procedures are ordinarily preempted.... When, however, the conduct at issue is only a peripheral concern of the Act or touches on interests so deeply rooted in local feeling and responsibility that, in the absence of compelling congressional direction, it could not be inferred that Congress intended to deprive the state of the power to act, we refuse to invalidate state regulation or sanction of the conduct.. .. The question of whether regulation should be allowed because of the deeply-rooted nature of the local interest involves a sensitive balancing of any harm to the regulatory scheme established by Congress, either in terms of negating the [NLRB’s] exclusive jurisdiction or in terms of conflicting substantive rules, and the importance of the asserted cause of action to the state as a protection to its citizens.
Local 926, International Union of Operating Engineers v. Jones, - U.S. -, 103 S.Ct. 1453, 1458-59, 75 L.Ed.2d 368 (1983)4 (citations omitted) (holding state-court action preempted when conduct at issue arguably constituted unfair labor practice, even though NLRB declined to issue complaint). Proceeding within Local 926’s framework, I therefore turn first to the question whether Congress intended section 7 to preempt all attempts by states to impose criteria on whom employees may select as their bargaining representatives. Having determined that all such regulation does not run afoul of congressional intent, I then explain why New Jersey’s interest in prohibiting certain individuals from representing employees in the casino industry is so “deeply rooted in local feeling and responsibility” that it justifies any resulting intrusion on section 7 rights.
II. Congressional Preemptive Intent
Like the majority, I look to that most elusive of guides — congressional intent — to determine whether state-imposed restrictions like those of the Casino Control Act are compatible with federal labor policy. Unlike the majority, however, I conclude from two manifestations of that intent— the enactment of the LMRDA and congressional acquiescence in the state-imposed restrictions upheld in De Veau v. Braisted *838—that national labor policy permits state-imposed restrictions in certain cireumstanc-es. Before turning to those factors, however, I first explain why Hill v. Florida’s explication of congressional intent is not dispositive here.
A. Hill v. Florida
I agree with the majority that the starting point for our analysis must be Hill v. Florida, supra, for Hill represents the seminal explication of section 7’s preemptive scope.5 I disagree, however, with the majority’s conclusion that the Hill Court’s reading of congressional intent remains either definitive or controlling.
At issue in Hill were two provisions of a Florida statute governing all labor unions operating within the state. Section 4 of the *839law required that anyone serving as the “business agent” of a union be licensed by a three-man board composed of the Governor, the Secretary of State, and the Superintendent of Education. The section also provided that licenses would be issued only to individuals who had been citizens of the United States for at least ten years, had not been convicted of a felony, and were of “good moral character.”
The second contested provision, section 6, required every labor union operating within the state to file with the Secretary of State annual written reports containing the name of the union, the location of its offices, and the names and addresses of all union officers. The filing of these reports was a prerequisite to a union’s obtaining a license to function within the state; failure to comply with either section 4 or section 6 was punishable as a misdemeanor.
Hill, a business agent within the meaning of the Florida law, and his union, Local No. 234 of the United Association of Journeymen Plumbers and Steamfitters, failed to register with the state and to obtain the licenses required by sections 4 and 6. Alleging violations of the Act, the Florida Attorney General successfully sued in state court for injunctions against Hill’s further functioning as the union’s business agent and the union’s further functioning at all until each complied with the statutory requirements.
The Florida Supreme Court upheld the injunctions; the United States Supreme Court reversed, finding both injunctions inconsistent with the NLRA. With regard to section 4, the Court stated:
Since the Labor Board has held that an employer must bargain with a properly selected union agent despite his failure to secure a Florida license, it is argued that the state law does not interfere with the collective bargaining process. But here, this agent has been enjoined, and if the Florida law is valid he could be found guilty of a contempt for doing that which the act of Congress permits him to do. Furthermore, he could ... be convicted of a misdemeanor and subjected to fine and imprisonment. The collective bargaining which Congress has authorized contemplates two parties free to bargain, and cannot thus be frustrated by state legislation. We hold that § 4 of the Florida Act is repugnant to the National Labor Relations Act.
325 U.S. at 542, 65 S.Ct. at 1375. The Court reached a similar conclusion regarding section 6. Id. at 543, 65 S.Ct. at 1375.
Relying on this holding, the majority reads Hill as standing for the proposition that section 7 precludes the states from fixing standards or qualifications for labor unions, their officers, or their agents if such standards or qualifications would preclude the unions or their representatives from being chosen and functioning as bargaining agents under section 7 of the NLRA. The majority therefore concludes that section 93 of the Casino Control Act is preempted.
Although section 93 of the Casino Control Act does attempt to define qualifications for union officials in the casino industry, I disagree with the majority’s conclusion that Hill therefore requires us to invalidate it. The central distinction between Hill and this case concerns the challenged laws themselves. The Florida statute at issue in Hill did nothing but establish qualification criteria for the officials of all labor unions in the state. Apparently the state’s sole purpose was the regulation of labor unions. Section 93 of the Casino Control Act, by contrast, does not seek to regulate all unions, but only those with members working in the casino industry. Moreover, section 93 is part of a larger regulatory scheme whose purpose is by no means labor-oriented. Thus Hill’s rejection of Florida’s attempt to establish qualifications for all union officials does not lead inexorably to the conclusion that the legislation here at issue also is inconsistent with congressional intent.6
The majority seeks to buttress its construction of congressional intent by focus*840ing on Congress’ having twice enacted major labor legislation without overruling Hill. In particular, the majority asserts that Taft-Hartley’s “reenactment of the original section 7 ... must be considered to be a Congressional approval of the holding in Hill that state law on eligibility is preempted.” Majority op., 826. The majority reaches a similar conclusion regarding the effect of the LMRDA:
[I]n 1959 Congress, fully aware of the holding in Hill v. Florida that section 7 preempted state disqualification laws, of the rule of statutory interpretation ... that Congressional deference to state law must be specific, and of the holdings ... that even state regulation of activity prohibited by federal law is preempted, chose to legislate on the subject of union officer disqualification, with no deference to state authority, either with respect to parallel disqualification criteria or with respect to conflicting disqualification criteria. Since both disqualification and preemption were carefully considered in the same legislation no intention can be attributed to Congress other than preservation of the Hill v. Florida rule.
Majority op., at 828.
Although appealing, this dwelling on nonevents relies too heavily on the Supreme Court’s approach in Merrill, Lynch, Pierce, Fenner & Smith, Inc. v. Curran, 456 U.S. 353, 102 S.Ct. 1825, 72 L.Ed.2d 182 (1982). In Curran, the Supreme Court, addressing Congress’ failure to eliminate judicially created private causes of action, stated:
[T]he fact that a comprehensive reexamination and significant amendment of the [Commodities Exchange Act] left intact the statutory provisions under which the federal courts had implied a cause of action is itself evidence that Congress affirmatively intended to preserve that remedy.
Id. at 381-82, 102 S.Ct. at 1840-41 (footnote omitted). But the Supreme Court’s approach in Curran cannot be taken too far; nor can it automatically be transferred to other contexts. First, as Curran itself noted, congressional inaction is only evidence of congressional acquiescence; neither in Curran nor elsewhere has the Supreme Court said that such inaction resolves the question of congressional intent. Indeed Curran itself went on to review the particulars of the legislative history in order to confirm the Court’s belief that Congress intended to preserve the private remedy. See id. at 382-88, 102 S.Ct. at 1841-44.
More importantly, Curran’s formulation was adopted in a simpler context: Congress *841either did or did not intend to create a private right of action. By contrast, the question whether Congress intended to preempt state regulation of labor unions cannot so easily be answered in one case for all future cases. Rather, the answer will depend on the type and extent of regulation, the industry involved, the need for such regulation — in short, on the circumstances of each case, see Local 926 v. Jones, supra, 103 S.Ct. at 1458-59. Thus woodenly to transfer the Curran approach to the more fluid context of labor-law preemption at once overestimates the comprehensiveness of congressional decisionmaking and underestimates the complexity of preemption doctrine.
In fact, it would be manifestly unrealistic to expect Congress to have contemplated, addressed, and resolved all (or even most) of the potential state/federal conflicts arising from overlapping regulatory schemes. As Professor Summers has pointed out in discussing the preemptive effect of the LMRDA:
Congress did not, and could not, work out in detail the coordination of federal and state law. To do so would require at the very least a comprehensive knowledge of existing state law and the kinds of cases which came before the courts. Furthermore, it would require a clear understanding by Congress of what it was prescribing as federal law. Congress lacked both of these since such precision of knowledge and understanding is not possible.
Summers, Pre-emption and the Labor Reform Act—Dual Rights and Remedies, 22 Ohio State L.J. 119, 152 (1961); cf. United States v. Little Lake Misere Land Co., 412 U.S. 580, 593, 93 S.Ct. 2389, 2397, 37 L.Ed.2d 187 (1973) (noting the “inevitable incompleteness presented by all legislation”). I do not suggest that Congress was unaware of Hill or of extant preemption jurisprudence. Nor do I suggest that we can or should ignore Congress’ failure to disavow Hill. But I cannot agree with the majority that we must read that failure as a congressional stamp of approval.
In short, I do not view either Hill’s explication of congressional purpose or Congress’ failure to reject Hill's holding as dispositive evidence of congressional intent to preempt all state legislation regulating union officials. Moreover, the majority’s attempt to parlay the 1945 Hill decision into a current and continuing expression of congressional intent is belied by more recent manifestations of congressional intent that suggest the contrary. I turn now to these.
B. The LMRDA
As the majority correctly points out, the NLRA, as originally enacted, placed no restrictions on employees’ choice of representatives. “Their own best judgment, not that of someone else, was to be their guide.” Hill v. Florida, supra, 325 U.S. at 541, 65 S.Ct. at 1374. In 1959, however, Congress enacted the LMRDA, Pub.L. No. 86-257, 73 Stat. 519 (1959), explicitly recognizing the danger of criminal infiltration of the labor movement. See H.R.Rep. 741, 86th Cong., 1st Sess., reprinted in 1959 U.S.Code Cong. & Ad.News 2424, 2431. In particular, section 504(a) provides that individuals convicted of various felonies are prohibited from holding union office for five years.7 29 U.S.C. § 504(a) (1976). Thus, whether or not restrictions on who may hold union office would be inconsistent with the unqualified free choice that the majority would read into the original section 7, Congress itself has established that the choice is no longer unfettered and that national labor policy admits of some such limitations.
The majority holds, however, that although the LMRDA promulgates federal criteria for the holding of union office, it does not permit the states to impose the additional and more stringent criteria at issue here. To support its position, the majority points out that two provisions of the LMRDA, sections 603(a) and 604, contain “savings clauses” that explicitly recognize the right of the states to provide remedies beyond those provided by the statute; Congress did not append such a clause to see*842tion 504, however, and the majority therefore infers a congressional intent to preclude the states from imposing more stringent disqualification criteria.
That conclusion is incorrect for several reasons. First, the majority fails to mention that at least one other section of the LMRDA is expressly preemptive: section 483 states that “[t]he remedy provided by this subchapter for challenging an election already conducted shall be exclusive.” 29 U.S.C. § 483 (1976). Thus it is possible to draw an inference, at least as strong as the majority’s, that Congress intended to preempt only when it did so explicitly.
A more realistic approach is that the LMRDA’s erratic allocation of enforcement jurisdiction between the state and federal governments makes futile any attempt to divine congressional intent from the absence of a savings clause in section 504. As the minority report accompanying the Senate version of the LMRDA protested:
The .. . bill distributes its remedies ... in accordance with no discernible standard or consistent principle. Some remedies are exclusively Federal, some are left to the States and denied to the Federal Government, some are given to the States but only if they apply Federal law, and some are allocated to both the States and Federal Government. In some instances, the majority insisted upon exclusive Federal jurisdiction asserting that absolute uniformity was essential; in others, it was insisted that diversity of treatment under varying State laws was absolutely necessary.
S.Rep. No. 187, 86th Cong., 1st Sess., reprinted in 1959 U.S.Code Cong. & Ad.News 2318, 2407; see also Summers, supra, 22 Ohio St.L.J. at 122 (“Congressional fear of destroying valuable state remedies without providing adequate federal substitutes, and the spectre of creating a new ‘no man’s land,’ routed all objections toward overlapping and possibly conflicting remedies. The statute became studded with assorted provisions cut from different patterns.”).
More important, the majority incorrectly asserts that neither of the savings clauses subsumes the subject matter of section 504. Section 603(a) declares:
Except as explicitly provided to the contrary, nothing in this chapter shall reduce or limit the responsibilities of any labor organization or any officer, agent, shop steward, or other representative of a labor organization, or of any trust in which a labor organization is interested, under any other Federal law or under the laws of any State, and, except as explicitly provided to the contrary, nothing in this chapter shall take away any right or bar any remedy to which members of a labor organization are entitled under such other Federal law or law of any State.
29 U.S.C. § 523(a) (1976).8 The majority contends that the only state-law remedies that this provision preserves are those for breach of fiduciary duties by union officials. See Majority op., at 25. Not only is there no support in the legislative history for such a narrow construction, see Summers, supra, 22 Ohio State L.J. at 122, 140, but the Supreme Court expressly rejected it in De *843Veau v. Braisted, supra: “§ 603(a) is an express disclaimer of pre-emption of state laws regulating the responsibilities of union officials, except where such preemption is expressly provided.... ” 363 U.S. at 157, 80 S.Ct. at 1153.9 Although the language was that of a four-Justice plurality, Justice Brennan, who provided the fifth vote, explained in a concurrence that the LMRDA “explicitly provides that it shall not displace such legislation of the States.”10 Id. at 160-61, 80 S.Ct. at 1154-1155; see also Local 1804, International Longshoremen’s Ass’n v. Waterfront Commission, 85 N.J. 606, 613, 428 A.2d 1283, 1287 (1981) (relying in part on section 603(a)’s savings clause in concluding that section 504 of the LMRDA does not preempt section 8 of the New Jersey Waterfront Commission Act).
Finally, it is important to note that one of the factors that motivated Congress to enact the LMRDA was the inability of the states to deal effectively with criminal infiltration of unions. “The LMRDA was adopted in large part because state and local authorities had failed to adopt ‘effective measures to stamp out crime and corruption [in unions] and guarantee internal union democracy.... ’ Consistent with this legislative purpose, Congress could reasonably allow a state to adopt more restrictive eligibility requirements.... ” International Longshoremen’s Association v. Waterfront Commission, 495 F.Supp. 1101, 1123 (S.D.N.Y.1980), affirmed in part and reversed in part on other grounds, 642 F.2d 666 (2d Cir.), cert. denied, 454 U.S. 966, 102 S.Ct. 509, 70 L.Ed.2d 383 (1981). In short, it is simply too late to assert that the LMRDA supersedes section 93 of the Casino Control Act; the more reasonable interpretation is that Congress viewed such legislation as complementing the federal scheme.
C. De Veau v. Braisted:
Congressional Acquiescence in State Imposed Restrictions
In addition to establishing the non-preemptive effect of the LMRDA, De Veau sheds considerable light on the preemptive scope of the NLRA. De Veau was an officer of the Longshoremen’s Union against whom New York invoked section 8 of its Waterfront Commission Act of 1953, a provision quite similar to the one at issue here. Section 8 provided that no one could collect dues on behalf of a waterfront union if any officer or agent of the union had been convicted of a felony and either had not been pardoned by the governor or had not received a good conduct certificate from the parole board.11 De Veau contended that the law impermissibly restricted the rights guaranteed by sections 1 and 7 of the NLRA.
The Supreme Court upheld the New York statute. Section 8, the Court explained in a plurality opinion by Justice Frankfurter, *844was enacted in conjunction with and in furtherance of an interstate compact between the states of New York and New Jersey. The compact, which established the bi-state Waterfront Commission of New York Harbor, had been approved by Congress pursuant to Article I, section 10 of the United States Constitution. Waterfront Commission Compact, Pub.L. No. 83-252, 67 Stat. 541 (1953). It therefore could not be maintained that section 8 was inconsistent with congressional intent.12
Appellants contend that the absence of a similar compact in this case renders De Vea u inapposite here. This argument, however, ignores three critical aspects of De Veau. First, four Justices explicitly rejected an inference of congressional intent to preempt all state legislation relating to employee choice of collective bargaining representatives; rather, they framed the issue as
whether we may fairly infer a congressional purpose incompatible with the very narrow and historically explained restrictions upon the choice of a bargaining representative embodied in § 8 of the New York Waterfront Commission Act. Would Congress, with a lively regard for its own federal labor policy, find in this state enactment a true, real frustration, however dialectically plausible, of that policy?
363 U.S. at 153, 80 S.Ct. at 1151. Congressional approval of the Waterfront Commission Compact made answering that question easy in De Veau: “In this case we need not imaginatively summon the likely reaction of Congress to the state legislation, as a basis for ascertaining whether due regard for congressional purpose bars the state regulation.” Id. But Justice Frankfurter’s formulation strongly implies that a court without access to similarly conclusive extrinsic evidence nevertheless should attempt to determine whether Congress would have intended to preclude the particular state legislation at issue.
Second, De Veau expressly recognized the importance of considering the particular situation and regulation at issue, rather than adopting a blanket approach to the preemption question.13 Distinguishing Hill v. Florida, supra, Justice Frankfurter wrote:
Nor was it true of Hill v. Florida, as it is here, that the challenged state legislation was part of a program, fully canvassed by Congress through its own investigations, to vindicate a legitimate and compelling state interest, namely, the interest in combatting local crime infesting a particular industry.
363 U.S. at 155, 80 S.Ct. at 1152; see also id. at 154, 80 S.Ct. at 1151 (“It is instructive that this unique provision has occurred in connection with approval of a compact dealing with the prevention of crime where, because of the peculiarly local nature of the problem, the inference is strongest that local policies are not to be thwarted.”).
Third, De Veau is most important for the insight it provides courts trying to “imaginatively summon the likely reaction of Congress” to allegedly preempted legislation. Justice Frankfurter’s opinion notes that *845Congress approved the Waterfront Compact despite its awareness of the restrictions embodied in section 8 and despite the urging of counsel for the International Longshoremen’s Association (the union most likely to be affected by section 8) that the compact’s implementation would conflict with federal labor policy. See id. at 151, 80 S.Ct. at 1150 (citing Hearing on H.R. 6286, H.R. 6321, H.R. 6343, and S. 2383 Before Subcomm. No. 3 of the House Comm, on the Judiciary, 83d Cong., 1st Sess. 136). Congress apparently perceived the problem of labor corruption on the waterfront to be of sufficient severity that state regulation would not be inimical to federal labor policy.
One could argue, of course, that the Waterfront Commission Compact constituted a specifically legislated exception to the NLRA manifesting congressional assent only to the state legislation there at issue. Such a characterization of De Veau and its facts, however, would be inconsistent with the compact’s genesis. The Waterfront Commission Compact was presented to and considered by Congress not because its authors and- sponsors believed it to embody a scheme that contravened the NLRA but because the Compact contemplated the establishment of a bi-state commission to combat evils that could not be eradicated by the two states acting individually. See H.R.Rep. No. 998, 83d Cong., 1st Sess. 3 (1953). In other words, it was the bi-state nature of the compact, and not any potential intrusion on national labor policy, that demanded congressional attention. Nothing in the Senate or House reports accompanying the compact suggests that Congress viewed its implementation as an exception to federal labor policy; rather, the Report of the House Judiciary Committee explicitly noted that “[t]he compact to which the committee here recommends that Congress grant its consent is in no sense antilabor legislation, but rather, antiracke-teering legislation.” Id. at 6.
I do not suggest that De Veau compels the conclusion that the NLRA does not preempt section 93 of the Casino Control Act; certainly the absence of a majority opinion there, and of a congressionally approved compact here, would undercut any such argument. At the same time, De Veau cannot be written off solely because it involved a compact, especially since the most plausible interpretation of its underlying scenario is that Congress did not view all state regulation of the qualifications of union officials as incompatible with the overall federal regulatory scheme.14 The fact that the LMRDA does not preclude the states from enacting more stringent criteria than are embodied in section 504 supports that conclusion.
Because I conclude that section 7 does not necessarily preclude such state regulation, I therefore turn to the question whether New Jersey’s regulation of unions associated with the casino industry is intended to effectuate the type of deeply rooted local interest contemplated by Local 926 v. Jones, supra.
*846III. New Jersey's Interest
It is virtually beyond dispute that comprehensive regulation of casino gambling, including regulation of unions representing industry employees, is “important ... to [New Jersey] as a protection to its citizens.” 15 Id. 103 S.Ct. at 1459. Gambling is an industry that is particularly susceptible to infiltration and control by organized crime and racketeers. See National Institute of Law Enforcement and Criminal Justice, Law Enforcement Assistance Administration, The Development of the Law of Gambling: 1776-1976 (1977). Not only is this vulnerability documented by gambling’s “checkered history in other jurisdictions,” Staff Policy Group on Casino Gambling, Second Interim Report 2 (1977),16 but the FBI has labelled gambling the “lifeblood of organized crime,” Hearings Before the Commission on the Review of the National Policy Toward Gambling (May 10, 1976) (testimony of Frederick Fehl, Acting Assistant Director, FBI); see United States v. Garrison, 348 F.Supp. 1112, 1119 (E.D.La. 1972). As the Superior Court of New Jersey remarked in a similar context, “[t]he undesirability of an association between those previously convicted of a crime or those in affinity with such a person and this sensitive ‘business of racing and the legalized gambling attendant thereupon,’ is too apparent to justify extended discussion.” Niglio v. New Jersey Racing Commission, 158 N.J.Super. 182, 188, 385 A.2d 925, 928 (App.Div.1978).
Legalized casino gambling is particularly attractive to organized crime for two reasons:
First, a casino contains a vast amount of liquid assets in the form of cash and gaming chips which are very attractive and susceptible to misappropriation. Second, these liquid assets remain uncounted and unrecorded as the gaming activity takes place. Casinos are unique because millions of dollars are continually changing hands among thousands of people on the casino floor without any record being made of how much money is exchanged, how many people are involved, or who those individuals are.
Santaniello, Casino Gambling: The Elements of Effective Control, 6 Seton Hall Legis.J. 23, 23 (1982) (footnote omitted); see also New Jersey Commission of Investigation, Report and Recommendations on Casino Gambling, at III (1977) (“[T]he nature of the industry ... makes it a vulnerable target for criminal intrusion.”).
Awareness of the attractiveness of casino gambling to racketeers and concern that the industry therefore could not remain “clean” contributed to the defeat in 1974 of a state-wide referendum to allow casino gambling in Atlantic City; a second referendum passed in 1976 only after proponents assured the public that New Jersey would have the “strongest regulations of casinos in the world,” that organized crime would be prevented from infiltrating the industry, that operating controls would be stringent, and that only those individuals of the highest character, integrity, and competence would be permitted to participate in casino operations. See, e.g., Strongest Law in *847World Offered for Atlantic City Casinos, N.Y. Daily News, Oct. 1, 1976, at 40; Lawmakers Reveal Casino Guidelines, Newark Star Ledger, Oct. 1, 1976, at l.17
Following the voters’ approval of the referendum, the legislature held extensive hearings and, together with the Governor, commissioned reports on how best to insure that Atlantic City’s casinos would remain clean. See Hotel and Restaurant Employees v. Danziger, supra, 536 F.Supp. at 322. These studies concluded that merely keeping organized crime from infiltrating the industry would be an insufficient goal; the success of the venture depend on gaining and retaining public trust in the industry’s integrity. Thus the studies recommended legislation directed toward preventing even the appearance of impropriety. See Report and Recommendations, supra; Second Interim Report, supra. The reports also made clear that the achievement of those goals would require the regulation of every aspect of the industry, and not just the gaming operations.
Profitability is ... not a direct function of the quality of gaming or of the environment in and around the casino. Corporate corruption, cheating, loansharking, overextension of credit, insobriety, prostitution and a honky-tonk atmosphere are not antithetical to a desire for profit, and in the industry are occasionally viewed as legitimate societal overhead so long as they encourage, or at least do not interfere with, the vitality of the gambling market.
The interests of the State in the success of casino gambling are not coterminous with the interests of the entrepreneur. While the latter measures a net return on investment against the degree of risk, the State must measure social and economic benefits against offsetting social, economic and environmental costs. What is acceptable by one measure may be unacceptable by the other. The staff policy group has concluded that the uniqueness of the industry, taken with its potential societal consequences and its checkered history in other jurisdictions, compels a state regulatory interest in virtually every aspect of casinos and related operations.
Staff Policy Group, Second Interim Report, supra, at 1-2 (emphasis added).
The State Commission of Investigation (SCI), in formulating more specific recommendations, operated on the same premises and concluded that “only the most stringent of gambling control laws can thwart the infiltration of casinos and related services and suppliers by organized crime.” Commission of Investigation, Report and Recommendations, supra, at II.18 The SCI then made explicit that it was concerned not only with related services and suppliers but with labor unions as well. Labor unions, *848the SCI asserted, could offer organized crime an entree into the casino industry:
The S.C.I.’s experience and collected intelligence regarding organized crime strongly suggests that there are few better vehicles utilized by organized crime to gain a stranglehold on an entire industry than labor racketeering. Organized crime control of certain unions often requires the legitimate businessmen who employ the services of the union members to pay extra homage to the representatives of the underworld. Moreover the ready source of cash which union coffers provide can be employed as financing of all sorts of legitimate or illicit ventures.
Id. at 1-H.19 The SCI therefore urged that unions representing employees of the easi-nos or their adjoining hotels be required to satisfy the same stringent qualification criteria required of casino operators. Id. at 1-H, 2-H.
Against this background, the New Jersey legislature enacted the Casino Control Act, including the provision challenged here. At the outset, the legislature displayed its awareness of the potential crime problems and their magnitude. The Act expressly declares its goals to include preventing criminal elements from gaining a foothold in the industry and avoiding any public perception that such a foothold even is available.20
The Act seeks to achieve these goals in two ways. First, the legislature created *849two separate agencies — the Casino Control Commission and the Division of Gaming Enforcement — to share responsibility for insuring the integrity of the casino industry; a system of checks and balances exists between the two agencies to keep the regulatory process itself above reproach.21 Second, and more important for our purposes, the legislature accepted the recommendations of the SCI and the Governor’s Staff Policy Group regarding the necessary breadth of control and granted the two agencies authority to regulate all aspects of the casino industry, including labor unions. As the Commission points out in its brief, the statutory and administrative controls over the casino gaming industry are “extraordinary pervasive and intensive.” Knight v. City of Margate, 86 N.J. 374, 381, 431 A.2d 833, 836 (1981).
In sum, New Jersey’s comprehensive regulation of the casino industry is a matter of intense and extraordinary local interest. Such regulation is not only essential to the State’s struggle to maintain the integrity of the industry, but the very prospect of such comprehensive legislation was the basis upon which New Jersey’s citizens consented to casino gambling in the first place. Given the unique nature of the industry — in particular its tremendous, unmonitored cash flow and its consequent attractiveness to racketeers and organized crime — the concerns of the legislature and citizenry cannot be characterized as anything less than “deeply rooted in local feeling and responsibility.” Local 926 v. Jones, supra, 103 S.Ct. at 1459.
IV. Preemption Doctrine Applied
As I noted at the outset, Local 926 v. Jones teaches that the NLRA does not necessarily preempt state legislation when the state’s interest is as strong as New Jersey’s is here. Rather, “[t]he question of whether regulation should be allowed ... involves a sensitive balancing of any harm to the regulatory scheme established by Congress ... and the importance of the asserted cause of action to the state as a protection to its citizens.” Local 926 v. Jones, supra, 103 S.Ct. at 1459. It is indisputable that New Jersey’s interest in comprehensive regulation of casino gambling is “deeply rooted in local feeling and responsibility,” id. at 1458, and that the restrictions on persons who may be associated with the industry — in whatever capacity — are of manifest importance to the State “as a protection to its citizens,” id. at 1459. And while I acknowledge that invocation of the sanctions provided by section 93 of the Casino Control Act may curtail the full exercise of section 7 rights, given the unique dangers posed by criminal infiltration of casino gambling, I cannot say that New Jersey’s interest is outweighed by “any harm to the regulatory scheme established by Congress.” Rather, Congress’ acquiescence in the legislation at issue in De Vea u suggests a congressional perception that, in situations where the threat to the welfare to the state is so *850pervasive and so well-documented that the need for comprehensive state regulation is manifest, such regulation will be permissible even if it includes a component that directly restricts section 7 rights.22 I believe such situations to be rare, but I also believe that one is presented here.
Moreover, as the Court of Appeals for the Second Circuit has noted in discussing section 8 of the New York Waterfront Commission Act, the type of restrictions here at issue are a rather limited intrusion on section 7 rights. “The Act does not prohibit the election of these men to offices in [other] locals nor does it prohibit the collection of dues outside New York State from [other] longshoremen. It simply prevents corrupt union officials from contributing to a longstanding problem of criminal wrong-doing on the New York waterfront.”23 International Longshoremen’s Association v. Waterfront Commission, 642 F.2d 666, 672 (2d Cir.), cert. denied, 454 U.S. 966, 102 S.Ct. 509, 70 L.Ed.2d 383 (1981).
Indeed the parallels between the New York Waterfront Commission Act and the Casino Control Act are striking. Like the New York Waterfront Commission Act, see De Vean, 363 U.S. at 147-51, 80 S.Ct. at 1148-1150, the Casino Control Act was passed only after extensive hearings, reports, and public outcry demanded the institution of stringent regulation to keep the industry clean. In both cases, the intent and effect of the state regulation was not directed particularly toward labor but, rather, toward the establishment of a comprehensive scheme in which union regulation was only one necessary component.24 To be sure, the New York legislature enacted the Waterfront Commission Act only after the situation on the docks had already reached “appalling” proportions, see id. at 147, 80 S.Ct. at 1148, and New York’s action therefore might be said to have been based on “harder” evidence of need than was New Jersey’s. But to write into preemption jurisprudence a distinction between remedial and prophylactic legislation would prevent states from acting until an industry is so rife with corruption that “criminals, racketeers, and hoodlums [have] acquired a stranglehold,” Hazelton v. Murray, 21 N.J. 115, 120, 121 A.2d 1, 4 (1956) (Brennan, J.) (describing condition of New York/New Jersey waterfront prior to compact and sustaining constitutionality of provision of New Jersey law identical to provision sustained in De Veau). The inefficiency of such a distinction is manifest; to say that federal labor policy requires it would offend reason.
The majority rejects New Jersey’s contention that gambling is unique and notes that other industries, such as solid-waste disposal, “have also been identified as susceptible of infiltration by organized crime.” Majority op., at 830 (footnote omitted). Although it is true that other industries *851have been so identified, the majority’s argument ignores the specifics of this situation and this industry. The casino industry is different: the huge number and the staggering aggregate amount of unrecorded cash transactions25 make legalized gambling far more attractive to and ripe for criminal infiltration than are other industries. Moreover, New Jersey has adequately demonstrated through studies, hearings, and the experience of other jurisdictions, that these dangers can be avoided only by means of a comprehensive scheme that includes regulation of all aspects of the industry.26 We are not dealing here with a display of anti-labor animus on the part of the New Jersey legislature; rather, we have before us a conscientious and well-reasoned attempt to erect a breakwater against a tide of vice and corruption that could engulf Atlantic City’s casinos.
Finally, the mere fact that another state may attempt to regulate another industry in a way or to an extent that effectively would undermine the right of employees to choose their bargaining representatives does not mean that section 93 offends federal labor policy. The admitted difficulty of drawing lines does not allow us to abdicate our responsibility to do so. There are distinctions both between the provisions at issue here and the broad regulation of labor unions struck down in Hill, as well as between the pressing need to regulate this industry and the mere desire to do so in other areas. These distinctions are not merely theoretical; they are readily apparent and enormously important.
In sum, given the magnitude of New Jersey’s interest in comprehensively regulating of the casino industry, and my conclusion that such regulation does not “stand[] as an obstacle to the accomplishment and execution of the full purposes and objectives of Congress,” Hines v. Davidowitz, 312 U.S. 52, 67, 61 S.Ct. 399, 404, 85 L.Ed. 581 (1941) (footnote omitted), I would hold that the NLRA does not preempt section 93 of the Casino Control Act.
. By reaching the merits of this appeal, the majority assumes that we are presented with a live controversy. I have evinced my agreement with that conclusion by joining the majority’s holding that ERISA preempts some of the Casino Control Act’s sanctions. I nonetheless find the question sufficiently troublesome that it warrants some discussion.
This is an appeal from a district court order denying appellants’ motion for a preliminary injunction. Unlike the complaint, however, the focus of that motion was the holding of hearings that have now taken place: the motion principally alleged that the “proceedings [i.e., the Casino Control Commission’s hearings and investigations] present an immediate and irreparable injury” to appellants’ section 7 rights. Motion for Preliminary Injunction ¶ 5 (emphasis added). The district court denied the motion on the basis that appellants could not demonstrate that the proceedings would result in the requisite irreparable harm: “The inquiry into the qualifications of the union and its leaders may not be considered to visit irreparable harm on the plaintiffs; something further must befall them before preemption may be found.” Hotel and Restaurant Employees and Bartenders Int’l Union Local 54 v. Danzinger [sic], 536 F.Supp. 317, 330 (D.N.J.1982). As the supplemental record now reveals, and as the majority now acknowledges, the proceedings that the district court refused to enjoin have been completed. It thus might appear that this appeal should be dismissed as moot.
Such a disposition, however, would be severe, as all that stands between appellants and enforcement of the Commission’s order that Local 54 rid itself of certain officers or be barred from collecting dues from its members is the district court’s stay of that order pending *834the disposition of this appeal. Although I do not think that this unfortunate circumstance is itself sufficient to resurrect what otherwise would be a moot appeal, I do believe that the same considerations which persuaded the district court to grant the stay also suggest that this case is one of a rare breed in which it is appropriate for us to reach the merits, notwithstanding the interlocutory posture of the appeal. As Judge Friendly wrote for the Court of Appeals for the Second Circuit in a similar context:
[A]n appellate court has the power, on review of a denial of a temporary injunction, to consider the case on the merits and decide whether the complaint states a claim on which relief can be granted.... Although in most of the cases where this course has been followed the appellate court has ordered the complaint to be dismissed, such a procedure is also appropriate in the converse situation at least where, as here, no material facts are contested, the lower court has considered the merits in detail, and these have also been argued here.... To “save the parties from further litigation” we should therefore “proceed to consider and decide the case upon its merits,” [Mast, Foos & Co. v. Stover Mfg. Co., 177 U.S. 485, 494, 20 S.Ct. 708, 712, 44 L.Ed. 856 (1900)], unless a ruling on the entire complaint, with its requests for declaratory and permanent injunctive ... relief, has also become moot.
Montano v. Lefkowitz, 575 F.2d 378, 382 (2d Cir.1978) (citations omitted). This case, too, is amenable to such treatment, despite the fact that the district court did not reach the viability of the Casino Control Act’s ultimate sanctions. There are no disputed issues of material fact; the parties have extensively and comprehensively briefed and argued the merits both before the district court and this Court; and the preemption issue, essentially a question of law, is one as to which we ultimately would have plenary review. Because appellants’ requests for permanent injunctive and declaratory relief are still very much alive, I therefore agree with the majority’s decision to reach the merits of this appeal.
. Despite its acknowledgment that Kershner v. Mazurkiewicz, 670 F.2d 440 (3d Cir.1982) (in banc), recognizes a limited class of pendent appeals under 28 U.S.C. § 1292(a)(1) (1976), the majority posits three reasons why we do not have jurisdiction over the cross-appeals. None of these considerations, however, poses an insurmountable obstacle to our taking jurisdiction.
First, the majority offers no support for its assertion that the Casino Control Commission and the Division of Gaming Enforcement may not raise their cross-appeals here because they were not aggrieved by the denial of the preliminary injunction. Moreover, that proposition is inconsistent with the fact that courts of appeals reviewing denials of preliminary relief frequently dismiss cases for failure to state a cause of action, thereby acting as if they had granted a cross-motion to dismiss. See Montano v. Lefkowitz, 575 F.2d 378, 382 (2d Cir.1978) (quoted supra note 1).
Second, the majority is too willing to dismiss the cross-appeals on the ground that the district court (apparently inadvertently) neglected to enter an order denying the motions to dismiss or abstain. The majority states: “The trial court’s opinion discussed these motions, but appeals do not ordinarily lie from opinions.” Majority op., at 821 (emphasis added). But this formulation, by its own terms, admits of an exception, and it is one that we previously have recognized. Where, as here, a district court opinion necessarily and definitively resolves a question as a predicate for its holding, the district court’s resolution is reviewable on appeal. See United States ex rel. Fielding v. Degnan, 587 F.2d 619, 621 (3d Cir.1978). Here the district court could not have reached the merits of appellants’ motion for preliminary relief without first having denied appellees’ motion to abstain.
Finally, the majority incorrectly asserts that a cross-appeal from a denial of a motion to abstain does not satisfy Kershner v. Mazurkiewicz’s test for pendent interlocutory cross-appeals. Kershner, which did not permit the cross-appeal of a class certification order, stated that the issue on cross-appeal must be “inextricably bound” with or controlling of the disposition of the preliminary injunction issue. 670 F.2d at 449. Since injunctive relief should not be granted if abstention is required, it seems quite clear that the propriety of abstention is inextricably bound with the review of a decision to grant or to deny preliminary injunc-tive relief.
. Although the Supreme Court initially promulgated the Younger abstention doctrine in order to prevent federal courts from interfering with ongoing state criminal proceedings, see Younger v. Harris, 401 U.S. 37, 91 S.Ct. 746, 27 L.Ed.2d 669 (1971). the Court has since broadened the doctrine’s scope. Federal courts now are admonished to refrain from taking action, absent extraordinary circumstances, when a plaintiff requests relief that would impinge on state proceedings of virtually any variety. “The policies underlying Younger are fully applicable to noncriminal judicial proceedings when important state interests are involved.” Middlesex County Ethics Comm. v. Garden State Bar Ass’n, 457 U.S. 423, 102 S.Ct. 2515, *8352521, 73 L.Ed.2d 116 (1982) (holding that federal court should have abstained from interfering with ongoing state-bar disciplinary proceedings).
Thus, at first blush, this case appears to fall within the class of cases in which district courts should abstain from adjudicating the claims at issue. Proceedings to enforce the Act’s prohibitions are essential to vindicating the state’s policies, see infra Part III; New Jersey’s interest surely is as strong as, if not stronger than, those that have been deemed sufficient in other cases. Cf. Trainor v. Hernandez, 431 U.S. 434, 97 S.Ct. 1911, 52 L.Ed.2d 486 (1977) (barring interference with state proceeding to recover welfare money); Juidice v. Vail, 430 U.S. 327, 97 S.Ct. 1211, 51 L.Ed.2d 376 (1977) (barring injunction against state civil contempt proceedings); Huffman v. Pursue, Ltd., 420 U.S. 592, 95 S.Ct. 1200, 43 L.Ed.2d 482 (1975) (barring injunction against civil enforcement action to close down theater showing obscene movies); Williams v. Red Bank Bd. of Educ., 662 F.2d 1008 (3d Cir.1981) (barring interference with tenure-revocation proceeding).
This conclusion is not altered by the fact that the proceedings, unlike those in Middlesex County, are not conducted under the auspices of the judiciary. In Williams v. Red Bank Bd. of Educ., supra, this Court addressed the question whether the federal courts should entertain a challenge to ongoing tenure-revocation proceedings. Noting that “[a]dministrative regulation often forms a crucial aspect of a state’s implementation of its laws ...we held that “where federal intervention into state administrative proceedings would be substantial and disruptive, and where the state proceedings are adequate to vindicate federal claims and reflect strong and compelling state interests, the district court, pursuant to Younger, should abstain.” 662 F.2d at 1016, 1017.
While the administrative nature of the Casino Control Act’s proceedings therefore do not themselves remove this case from the ambit of Younger abstention, I do not believe that the proceedings afford an adequate opportunity for the vindication of appellants’ federal claims. See Middlesex County, supra, 102 S.Ct. at 2521. To be sure, as in Red Bank, any decision of the Casino Control Commission is appealable to the state courts, and Younger's premise is that state courts are as competent as are federal courts to adjudicate federal claims. Here, however, appellants attack not merely the sanctions for which the Casino Act provides; appellants also assert that their federally created rights are violated by appellants’ being subjected to the administrative proceedings themselves. Thus, regardless of the competence of the state judiciary, all of appellants’ federal claims simply cannot be vindicated in state court.
In such circumstances, abstention is inappropriate because it would deprive a putative plaintiff of the opportunity to vindicate his federal claims. I therefore would hold that, where an individual who is subject to state proceedings to which the federal courts otherwise would defer raises a colorable claim that the proceedings themselves constitute a violation of a constitutional or statutory right, the principles of comity and federalism motivating Younger are superseded. Cf. Abney v. United States, 431 U.S. 651, 97 S.Ct. 2034, 52 L.Ed.2d 651 (1977) (denial of claim of double jeopardy is an appealable collateral order); United States ex rel. Webb v. Court of Common Pleas, 516 F.2d 1034, 1037 (3d Cir.1975) (pre-trial ha-beas relief available where alleged nature of constitutional right asserted, i.e., not to be tried twice, makes post-conviction relief incapable of vindicating the right); Jackson v. Justices, 549 F.2d 215, 220 n. 1 (1st Cir.) (Coffin, C.J., dissenting) (“Whether or not there is to be a general ‘futility’ exception to the Younger doctrine, an exception seems appropriate in the case of double jeopardy claimants who are to be denied the opportunity in the state court system to demonstrate that they have a right not to be retried.”), cert. denied, 430 U.S. 975, 97 S.Ct. 1666, 52 L.Ed.2d 370 (1977).
. Thus the majority is incorrect to the extent that it admonishes against “engaging in weighing or balancing,” Majority op., at 828.
. Ordinarily, of course, the language and legislative history of the statute itself provide the starting point for inquiries into congressional intent. In this case, however, contrary to the intimations of the majority, neither the NLRA. nor the Taft-Hartley Act is of much assistance in that regard.
1. The National Labor Relations Act of 1935 (The Wagner Act).
In the course of congressional debate over the Wagner Act, Pub.L. No. 74-198, 49 Stat. 449 (1935), the only real mention of the relationship between federal and state regulation of the selection of bargaining representatives was an objection that the Act’s regulation of local employment would violate the tenth amendment. See Letter from Counsel for Appellant to the Court (Dec. 3, 1982) (submitted at the request of the Court) (citing Hearings on S. 1958 at 243-44, 840 (Mar. 11-14, 1935) (statement of James A. Emery)). The ultimate rejection of that position, however, in no way implies that Congress intended to preempt the field. Indeed as the majority eloquently demonstrates, subsequent case law implies the contrary. The majority, however, infers a preemptive intent from the fact that Congress did not impose eligibility requirements for service as collective bargaining representatives despite having had drawn to its attention related problems of labor racketeering. Though the omission no doubt represented, as the majority asserts, “a conscious legislative choice,” it would be equally plausible to infer that Congress preferred any such restrictions to be imposed by the states. Thus the omission of restrictions provides little insight into congressional intent.
2. Labor Management Relations Act of 1947 (Tañ-Hartley)
Resort to the legislative history of the Taft-Hartley Act, Pub.L. No. 80-101, 61 Stat. 136 (1947), is equally unilluminating. These debates, too, contain no discussion of the extent to which Congress wished to preclude imposition by the states of qualifications for union representatives. See Garner v. Teamsters, Chauffers, and Helpers Local Union No. 776, 346 U.S. 485, 488, 74 S.Ct. 161, 164, 98 L.Ed. 228 (1953) (“The [LMRA] ... leaves much to the states, though Congress has refrained from telling us how much. We must spell out from conflicting indications of congressional will the area in which state action is still permissible.”). Indeed only a few of Taft-Hartley’s provisions refer to state action, and those provisions can support conflicting inferences. Two provisions arguably support an inference that Congress intended to preempt unless it stated otherwise. First, Congress amended § 10(a), 29 U.S.C. § 160(a) (1976), to eliminate the NLRB’s exclusive jurisdiction, and instead to allow the NLRB, with certain restrictions, to cede jurisdiction to state agencies. Second, § 14(b), id. § 164(b), explicitly permits the states to prohibit union security clauses. The House Committee explained that “[i]t was never the intention of the [NLRA], as is disclosed by the legislative history of that act, to preempt the field in this regard so as to deprive the States of their powers to prevent compulsory unionism.” H.R.Rep. No. 510, 80th Cong., 1st Sess., reprinted in 1947 U.S.Code Cong. & Ad.News 1135, 1166.
But if §§ 10(a) and 14(b) might be read to imply that state action is precluded unless specifically permitted, § 14(a), 29 U.S.C. § 164(a) (1976), leaves the contrary impression. That section states:
Nothing herein shall prohibit any individual employed as a supervisor from becoming or remaining a member of a labor organization, but no employer subject to this subchapter shall be compelled to deem individuals defined herein as supervisors as employees for the purpose of any law, either national or local, relating to collective bargaining.
The fact that Congress felt the need explicitly to preclude states from defining supervisors as employees for purposes of collective bargaining might be read to suggest that the states do retain some freedom to enact other proscriptions not mentioned by the NLRA. The question whether § 14(a) should be so interpreted is not presented here, and I do not necessarily advocate such a reading. I posit the argument only to underscore that we should be wary of inferring preemptive intent from provisions that do not expressly license the states to act.
In short, despite the fact that the Taft-Hart-ley Congress was clearly attuned to the problems associated with dual systems of regulation, there was no discussion of the preemptive effect of § 7. The legislative history of Taft-Hartley is thus of little more help than is that of the Wagner Act.
. There is another reason that Hill does not control all aspects of this case. The Hill Court made relatively clear with respect to section 4, and explicitly so with respect to section 6, that it was the statutes’ sanctions that conflicted *840with congressional purpose. In particular, the Court noted in discussing section 6:
The requirement as to the filing of information and the payment of a $1.00 annual fee does not, in and of itself, conflict with the Federal Act. But, for failure to comply, this union has been enjoined from functioning as a labor union. It could not without violating the injunction and also subjecting itself to the possibility of criminal punishment even attempt to bargain to settle a controversy or a strike. It is the sanction here imposed, and not the duty to report, which brings about a situation inconsistent with the federally protected process of collective bargaining.
325 U.S. at 543, 65 S.Ct. at 1375 (emphasis added). Thus even if Hill were controlling here, this formulation suggests only that a state may not prevent a union collecting dues from its members. But this case involves more than just the Casino Control Commission’s ultimate sanction power. Also at issue are the Casino Act’s provisions for registration and licensing of unions representing casino employees, for investigations by the Division of Gaming Enforcement, and for hearings by the Commission. By failing to distinguish these various elements of the Casino Act’s scheme from the sanctions, the majority intimates that Hill would preclude them as well. Indeed the majority characterizes the hearings as an “illegal proceeding,” Majority op., at 831, without any discussion to that effect. But given the Supreme Court’s holding that it was the “sanction [t]here imposed, and not the duty to report,” that conflicted with the NLRA, Hill simply cannot be construed as affecting the validity of the non-sanction provisions; to the extent that the majority holds the contrary, I dissent from that part of the holding as well. See also Alabama State Fed’n of Labor v. McAdory, 325 U.S. 450, 466, 65 S.Ct. 1384, 1392, 89 L.Ed. 1725 (1945) (“Nor can we say in the absence of any showing to the contrary that the filing of information returns will impose such burdens on any of petitioners as to interfere with the performance of their functions under the [NLRA].... ”).
. The full text of § 504 is set forth at note 8 of the majority’s opinion.
. The other savings clause is contained in § 604, which explicitly preserves the ability of the states to enact “garden variety criminal laws,” see Majority op., at 828, notwithstanding the LMRDA’s “federalization” of conduct that might otherwise come within the scope of state criminal law.
Nothing in this chapter shall be construed to impair or diminish the authority of any State to enact and enforce general criminal laws with respect to robbery, bribery, extortion, embezzlement, grand larceny, burglary, arson, violation of narcotics laws, murder, rape, assault with intent to kill, or assault which inflicts grievous bodily injury, or conspiracy to commit any of such crimes.
29 U.S.C. § 524 (1976). This provision, the majority notes, was enacted in response to dicta in Garner v. Teamsters, Chauffeurs and Helpers Local Union No. 776, 346 U.S. 485, 74 S.Ct. 161, 98 L.Ed. 228 (1953), and San Diego Building Trades Council v. Garmon, 359 U.S. 236, 79 S.Ct. 773, 3 L.Ed.2d 775 (1959), that states might be precluded from regulating conduct proscribed by federal law. Although this savings clause does not authorize the type of restriction here at issue, I do not agree, for the reasons stated in text, that Congress’ decision to include the clause in the LMRDA suggests that other sections without similar clauses must be read as preemptive.
. For the facts of X>e Veau, see infra Part II.C.
. I note in passing that appellants here contend that only the NLRA and ERISA preempt the sanctions provided by the Casino Control Act. They have apparently abandoned the argument they advanced to the district court that the LMRDA also has such a preemptive effect. The district court rejected that argument, noting that the contention was “easily disposed of by reference to DeVeau.” Hotel and Restaurant Employees v. Danziger, supra, 536 F.Supp. at 327; see id. at 326-28.
. Unlike § 504 of the LMRDA, which bars convicted felons from holding union office for five years, § 8 of the New York Waterfront Commission Act imposes a bar that may be permanent:
No person shall solicit, collect or receive any dues, assessments, levies, fines or contribu-lions, or other charges within the state for or on behalf of any labor organization which represents employees registered or licensed pursuant to the provisions of this act ... if any officer, agent or employee of such labor organization ... has been convicted by a court of the United States, or any state or territory thereof, of a felony [or] any misdemeanor involving moral turpitude ... unless he has been subsequently pardoned therefor by the governor or other appropriate authority of the state or jurisdiction in which such conviction was had or has received a certificate of good conduct from the board of parole pursuant to the provisions of the executive law to remove the disability.
N.Y.Unconsol.Laws § 9933 (McKinney 1974). In fact, New York was seeking to bar De Veau from holding union office because of a 36-year old conviction for which De Veau had received a suspended sentence.
. ■ Although section 8 was not itself part of the compact, it had been enacted prior to submission of the compact and was intended to be part of its comprehensive scheme.
In light of the purpose, scope and background of this New York legislation and Congress’ relation to it, ... an inference of incompatibility has no foundation.... Here the States presented their legislative program to cope with an urgent local program to the Congress, and the Congress unambiguously supported what is at the core of this reform. Had § 8 been written into the compact, even the most subtle casuistry could not conjure up a claim of pre-emption.
De Veau, 363 U.S. at 153, 80 S.Ct. at 1151. Section 8 itself did not contemplate or require bi-state enforcement. Id. New Jersey, however, enacted a virtually identical provision. See N.J.Stat.Ann. § 32:23-80 (West 1963).
. See also Local 1804, Internat’l Longshoremen’s Ass’n v. Waterfront Comm’n, 85 N.J. 606, 613, 428 A.2d 1283, 1287 (1981):
The leading case of De Veau v. Braisted ... suggests three criteria for determining whether a federal law preempts section 8. The criteria are: (1) the extent to which the matter regulated involves matters of local responsibility or concern; (2) the degree to which the state law interferes with federal law; and (3) the amount of interference Congress was willing to tolerate.
. The majority limits its discussion of De Veau solely to rebutting the notion that De Veau overruled Hill. The majority concludes quite correctly that it did not; indeed appellees have not argued otherwise, contrary to the majority’s assertion. But having established that Hili has not been overruled, the majority cursorily concludes that “De Veau ... cannot, therefore, be read as impairing the authority of Hill v. Florida.” Majority op., at 829.
This off-hand dismissal is too hasty. First, to the extent that the majority predicates its conclusion on De Veau’s failure to overrule Hill, it is important to keep in mind that the Supreme Court continually modifies prior holdings without overruling them. See, e.g., EEOC v. Wyoming, - U.S. -, 103 S.Ct. 1054, 75 L.Ed.2d 18 (1983) (limiting, without overruling, National League of Cities v. Usery, 426 U.S. 833, 96 S.Ct. 2465, 49 L.Ed.2d 245 (1976)). Indeed Justice Frankfurter’s opinion in De Veau expressly distinguished Hill, and not solely on the basis of the compact: “Nor was it true of Hill v. Florida, as it is here, that the challenged state legislation was part of a program, fully canvassed by Congress through its own investigations, to vindicate a legitimate and compelling state interest, namely, the interest in combat-ting local crime infesting a particular industry.” 363 U.S. at 155, 80 S.Ct. at 1152.
Second, the majority’s analysis ignores De Veau’s refusal blindly to apply Hill’s prohibition against state imposition of dues-collecting sanctions. Indeed the fact that Congress acquiesced in the type of state-imposed sanctions struck down in Hill further undercuts the majority’s reliance on congressional failure explicitly to overrule Hill.
. Casino gambling traditionally has been considered to be within the purview of the states to regulate, and one court even has suggested that “gaming is a matter reserved to the states within the meaning of the Tenth Amendment ... [leaving] no room for federally protected constitutional rights,” State v. Rosenthal, 93 Nev. 36, 44, 559 P.2d 830, 836, appeal dismissed, 434 U.S. 803, 98 S.Ct. 32, 54 L.Ed.2d 61 (1977). Gambling does affect interstate commerce, however, and Congress has enacted a number of laws regulating the interstate aspects of gambling, see, e.g., 15 U.S.C. §§ 1171— 1178 (1976) (barring interstate shipment of gambling devices); 18 U.S.C. §§ 1082-1083 (1976) (prohibiting gambling on certain ships); id. § 1084 (1976) (barring transmission of wagering information over interstate wires); id. §§ 1301-1307 (1976 & Supp. V 1981) (regulating lotteries); id. § 1953 (1976 & Supp. V 1981) (barring interstate shipment of gambling paraphernalia).
. See also Marshall v. Sawyer, 301 F.2d 639, 648-50 (9th Cir.1962) (Pope, J., concurring) (discussing the known hazards of gambling that led to strict controls in Nevada); Nevada Tax Comm’n v. Hicks, 73 Nev. 115, 119, 310 P.2d 852, 854 (1957) (“Throughout this country, ... gambling has necessarily surrounded itself with an aura of crime and corruption.”).
. Cf. Knight v. City of Margate, 86 N.J. 374, 392, 431 A.2d 833, 842 (1981) (“Gambling is an activity rife with evil, so prepotent its mischief in terms of the public welfare and morality that it is governed directly by the [state] Constitution itself.”).
. The SCI went into more detail in formulating its specific recommendations:
Although the State Commission of Investigation and segments of law enforcement have been highly successful in removing many identified members of organized crime from New Jersey society, there remains a well-organized highly functional organized crime network in this state. Additionally, it is important to note that the appearance of organized crime is presently changing. As traditional sources of illegal funds become more limited and more easily detected and prosecuted, elements of organized crime have become significantly more interested in investing funds in legitimate enterprises.
The infiltration of legitimate business by organized crime is a situation which society should be aggressive to prevent. In addition to providing a ready source of funds to capitalize illegal ventures, a convenient cash flow to disguise illegal profits, and an ostensibly legitimate occupation for organized crime members and associates, an incursion into legitimate enterprise is often accompanied by extortion, loansharking, commercial bribery, tax violations and anti-trust law infringements.
Thus, it is the position of the S.C.I. that any casino gambling legislation should not only preserve and protect the integrity of the operation of casinos in Atlantic City, but should also foreclose the possibility of opening up new and fertile areas of legitimate business enterprise to elements of organized crime.
The S.C.I.’s continuing organized crime program discloses that there are several identifiable legitimate enterprises which have *848long been a target of infiltration by organied [sic] crime. Further, the S.C.I.’s monitoring of the workings of organized crime in Atlantic City discloses substantial movement by organied [sic] crime in contemplation of a casino operation and the potential which it represents. The general approach which the S.C.I. deems advisable is the exposure to stringent licensing scrutiny by the gambling regulatory body of all industries of concern and particularly more of an ancillary nature.
The most efficacious legislative scheme must strike the proper balance between the magnitude of the State’s interest in licensing certain service industries and the required administrative resources which will be necessitated. It would be overly burdensome to require licensure for all parties entering into a contractual relationship with the licensed hotel yet, as has been stated, it would be against the interests of the state not to require licensure [in] most other cases. The required scrutiny of the proposed provider should be directly proportional to the risk to the gaming industry and society as a whole. The risk is twofold: (a) that undesirable elements infiltrate the gaming industry through direct involvement with the casino operation, and (b) that undesirable elements either indirectly affect the gaming industry through involvement with the hotel or are financially benefitted through profíts recouped from endeavors relating to casino gaming.
Id. at 1-C to 2-C (emphasis in original).
. New Jersey’s Attorney General recently expressed a similar view when testifying before the Senate Permanent Subcommittee on Investigations:
Organized labor is in a prime position to exert tremendous pressure over the casino industry .... What would a casino owner pay for labor peace? How much is it worth to keep a business that grosses between $500,000 and $1 million a day free of a strike? A corrupt union could extort outright payments or use its power of persuasion to dictate what firms get the lucrative ancillary service contracts within the casino industry. Quoted in Court Delay Seen in Casino Dispute, N.Y. Times, Oct. 10, 1982, at 55, col. 1.
. The Act enumerates seventeen policy declarations and findings of fact, four of which are relevant here:
(6) An integral and essential element of the regulation and control of such casino facilities by the State rests in the public confidence and trust in the credibility and integrity of the regulatory process and of casino operations. To further such public confidence and trust, the regulatory provisions of this act are designed to extend strict State regulation to all persons, locations, practices and associations related to the operation of licensed casino enterprises and all related service industries as herein provided. In addition, licensure of a limited number of casino establishments, with the comprehensive law-enforcement supervision attendant thereto, is further designed to contribute to the public confidence and trust in the efficacy and integrity of the regulatory process.
(7) Legalized casino gaming in New Jersey can attain, maintain and retain integrity, public confidence and trust, and remain compatible with the general public interest only under such a system of control and regulation as insures, so far as practicable, the exclusion from participation therein of persons with known criminal records, habits or associations, and the exclusion or removal from any positions of authority or responsibility *849within casino gaming operations and establishments of any persons known to be so deficient in business probity, ability or experience, either generally or with specific reference to gaming, as to create or enhance the dangers of unsound, unfair or illegal practices, methods and activities in the conduct of gaming or the carrying on of the business and financial arrangements incident thereto.
(9) Since casino operations are especially sensitive and in need of public control and supervision, and since it is vital to the interests of the State to prevent entry, directly or indirectly, into such operations or the ancillary industries regulated by this act of persons who have pursued economic gains in an occupational manner or context which are in violation of the criminal or civil public policies of this State, the regulatory and investigatory powers and duties shall be exercised to the fullest extent consistent with law to avoid entry of such persons into the casino operations or the ancillary industries regulated by this act.
(15) Continuity and stability in casino gaming operations cannot be achieved at the risk of permitting persons with unacceptable backgrounds and records of behavior to control casino gaming operations contrary to the vital law enforcement interest of the State. N.J.Stat.Ann. § 5:12-1 (West Supp.1982).
. For a comprehensive discussion of this regulatory process, including the mechanisms designed to guarantee its integrity, see Cohen, The New Jersey Casino Control Act: Creation of a Regulatory System, 6 Seton Hall Legis. J. 1 (1982).
. I read Local 926 as consistent with this view.
. See also Hazelton v. Murray, 21 N.J. 115, 123, 121 A.2d 1, 5-6 (1956) (Brennan, J.) (sustaining § 8 of New Jersey’s Waterfront Commission Act) (“If the conviction of crime is a proper consideration related to the public interest for the purposes of licensing and registering waterfront workers, much more so is the provision here under review which would wrest from the vicious criminal combine the means through which the corrupt conspiracy was perpetrated.”).
. Like the Casino Control Act, the legislation to clean up the New York Waterfront did not merely regulate union officials:
[The Commission has the] power to license, register and regulate the waterfront employment of pier superintendents, hiring agents, longshoremen and port watchmen, and to license and regulate stevedores. [The compact] entirely prohibits one class of waterfront employment, public loading, found to be unnecessary and particularly infested with corruption.... The issue of licenses to engage in waterfront occupations, or the right to be registered, depends upon findings by the Commission of good character. In particular, past convictions for certain felonies constitute specific disabilities for each occupation, with discretion in the Commission to lift the disability, except in the case of port watchmen, where it constitutes an absolute bar to waterfront employment. A new procedure for the employment of longshoremen is also provided under the supervision of the Commission, replacing the archaic, corrupt ‘shape-up.’
De Veau, 363 U.S. at 149, 80 S.Ct. at 1149.
. The “handle,” i.e., the total amount of money bet, in Atlantic City’s casinos was more than $5 billion in 1982.
. The majority accurately notes that the NLRB has consistently asserted jurisdiction over the casino industry; indeed I do not contend that the NLRB should not do so here. But the majority implies that the exercise of the NLRB’s jurisdiction cannot be reconciled with state regulation such as section 93 and that section 93 must therefore be deemed incompatible with the NLRA’s regulatory framework. This implication is incorrect, for the NLRB continues to exercise jurisdiction over the New York waterfront, despite the vitality of the restriction upheld in De Veau.
In addition, I note that the NLRB, in declining jurisdiction oyer labor disputes in the dog-racing and horse-racing industries, has suggested that state regulation of labor, when it is part of a pervasive industry-wide scheme, is not necessarily inconsistent with national labor policy:
In prior decisions, the Board declined to assert jurisdiction over these industries noting, inter alia, the extensive State control over the industries. It appears that State law sets racing dates of the tracks; State law determines the percentage share of the gross wagers that goes to the State; and State law determines the percentage of gross wagers to be retained by the track. In addition, the State licenses employees, exercises close supervision over the industries through State racing commissions, and in many States retains the right to effect the discharge of employees whose conduct jeopardizes the “integrity” of the industry. As the industries constitute a substantial source of revenue to the States, a unique and special relationship has developed between the States and these industries which is reflected by the States’ continuing interest in and supervision over the industries.
Declination of Assertion of Jurisdiction, 38 Fed. Reg. 9537 (1973) (codified at 29 C.F.R. § 103.3 (1982)). Although a district court held that the rule violates the NLRB’s statutory mandate, see New York Racing Ass’n v. NLRB, 110 L.R. R.M. 3177 (E.D.N.Y.1982) (NLRB may decline jurisdiction only over industries not substantially affecting interstate commerce), that decision was recently vacated, 708 F.2d 46 Nos. 82-6252, 6258 (2d Cir. May 10, 1983) (district court does not have jurisdiction to review such NLRB rules).