Cheeseman v. Carey

OPINION SOFAER, District Judge: This is an action by employee members of the Security Services Unit of New York state to enjoin the state from deducting from their wages the penalty for illegal strikes provided by Section 210 of the New *206 York Civil Service Law, commonly known as the Taylor Law. 1 A motion for a temporary restraining order and a preliminary injunction was granted on September 8, 1979, to the limited extent described in the order filed. Thereafter additional argument was heard and evidence presented. The parties have agreed that their submissions to date should be treated as complete, and that a final opinion and judgment is now appropriate. The named plaintiffs represent some 6,500 employees who work in security positions at New York State’s prisons, mental institutions, universities, and other places requiring security personnel.- The New York State Inspection, Security and Law Enforcement Employees, District Council 82, AFSCME, AFL-CIO (“Council 82”) is their collective bargaining agent. Also included in the plaintiff class are additional employees who joined in the same strike, and to whom Section 210 is equally applicable. 2 The class has been duly certified. On April 18,1979, employees in the plaintiff class engaged in a work stoppage, determined by the Director of the Office of Employee Relations (“Director”) to be an illegal strike. The stoppage continued for sixteen days, until May 4, 1979. Defendants allege without contravention that considerable disruption ensued. During the stoppage, prisoners were kept in their cells for much longer periods than usual, the National Guard was called up to serve on an emergency basis, and visitors were advised to stay home. Plaintiffs contend that the strike resulted from unwarranted provocation — in particular, from an announcement by state negotiators that a new contract had been reached when in fact no such agreement had been made. Although there appears to be some truth to this allegation, 3 it is clearly insufficient under state law to justify a strike. Consequently, soon after the strike, the Director commenced proceedings under Section 210 for imposing penalties. Striking state employees automatically lose their pay for the days they are out. Plaintiffs do not seek to be paid, or to recoup, any wages for days on which they did not perform normal services. But under the Taylor Law, striking employees may be fined an additional day’s pay for every day they are on strike. Plaintiffs assert that the procedure by which this two-for-one penalty is imposed is constitutionally deficient in that it deprives them of their wages without due process. The procedure followed by the Director, and mandated by statute, is described in detail in Sanford v. Rockefeller, 35 N.Y.2d 547, 364 N.Y.S.2d 450, 324 N.E.2d 113 (1974), appeal dismissed, 421 U.S. 973, 95 S.Ct. 1972, 44 L.Ed.2d 465 (1975). In that case, the Court of Appeals upheld the two-for-one penalty when challenged by school teachers on the same grounds raised by plaintiffs in this case. In a careful and persuasive opinion, the court held that due process did not require a hearing before the penalty was exacted, because the procedure for imposing the penalty was reasonably reliable and because adequate procedures existed after imposition of the penalty to review challenges. The court dealt at length with the Supreme Court’s decisions in Arnett v. Kennedy, 416 U.S. 134, 94 S.Ct. 1633, 40 L.Ed.2d 15 (1974), and related cases, and concluded that a loss of wages under the circumstances was not so substantial a deprivation as to mandate a hearing before the deduction of wages. The Supreme Court dismissed the appeal in Sanford for want of a substantial federal question. I. VENUE Defendants claim this suit should be dismissed or transferred due to improper or *207 inconvenient venue. This claim raises a number of interesting and difficult issues. The principal governing statute, 28 U.S.C. § 1391(b), is among the most restrictive venue provisions in the Code: A civil action wherein jurisdiction is not founded solely on diversity of citizenship may be brought only in the judicial district where all defendants reside, or in which the claim arose. . The statute has been periodically liberalized since its adoption in 1877, but it continues to create substantial obstacles for certain classes of plaintiffs. In particular, plaintiffs suing state officers on federal constitutional grounds must often overcome venue limitations not otherwise confronted by federal litigants. Unlike 28 U.S.C. § 1391(a), applicable in diversity cases, Section 1391(b) fails to authorize venue on the basis of plaintiffs’ residence. And, it is far more restrictive in permitting venue than 28 U.S.C. § 1391(e), which provides several bases of venue in suits against federal officers, including the residence of any plaintiff or defendant. A. Venue Based on Defendants’ Residence Under 28 U.S.C. § 1391(b) Plaintiffs claim they meet the requirements of Section 1391(b) in that all the defendants “reside” in the Southern District of New York. The difficulty plaintiffs face is that several of the defendants, and in particular those actually responsible for the allegedly illegal acts, maintain their offices and perform their duties in Albany, in the Northern District of New York. The “residence” of a public officer under Section 1391 means his “official” not “actual” residence. Butterworth v. Hill, 114 U.S. 128, 5 S.Ct. 796, 29 L.Ed. 119 (1885). See 1 Moore, Modern Federal Practice ¶ 0.142[5.-l-2], at 1396 (1978). Moreover, in cases involving federal officers, the courts have found only one official residence for venue purposes. See, e. g., Reuben H. Donnelley Corp. v. F. T. C., 580 F.2d 264, 266-67 (7th Cir. 1978). Whatever the rationale for this rule, a different result with respect to state officers seems warranted. As Judge Frankel reasoned in Buffalo Teachers’ Federation v. Helsby, 426 F.Supp. 828 (S.D.N.Y.1976), the potential inconvenience to federal officials resulting from a rule recognizing multiple official residences is far greater than for state officials, who can never be deemed to reside in another state. Moreover., the fact that a plaintiff can generally sue federal officials in the district in which the plaintiff resides, 28 U.S.C. § 1391(e), makes a multiple-residence rule far less necessary in cases against federal officials. Therefore, in suits against state officials, courts should be willing to consider whether any defendant has more than one official residence for purposes of the particular litigation. But see Procario v. Ambach, 466 F.Supp. 452 (S.D.N.Y.1979). Whether a state defendant has a second official residence would seem to turn on three factors: (1) the defendant’s presence in the district in which the plaintiff has sued; (2) the extent of defendant’s official activities in the district; and (3) the relationship of the defendant’s activities within the district to the cause of action asserted. 4 The complaint here names several defendants including the Governor, the Office of Employee Relations and its Director, the State Comptroller, and officials in charge of agencies or institutions employing security officers. Many of the defendants have offices in the Southern District of New York. But the only defendants neces *208 sary for plaintiffs to gain relief — and the only ones whose conduct is challenged — appear to be the Office of Employee Relations (by its Director) and the Comptroller. These are the defendants who determined that plaintiffs were on strike in violation of Section 210 and who are withholding penalties from plaintiffs’ wages. The Office of Employee Relations has no office in the Southern District. Plaintiffs argue, however, that the Office is part of the Executive Branch, which maintains a substantial office in this district, and that the Director has acted on behalf of “the chief fiscal officer of the government involved.” Civil Service Law § 210(3)(h). Such contacts are insufficient to constitute a presence in this district for purposes of venue. They have nothing to do with the state official actually being sued. Although the State Comptroller (Department of Audit and Control) maintains an office in this district, his involvement with this litigation extends only to the issuance of plaintiffs’ pay checks. No part of the payroll function is performed here, and all the deductions from plaintiffs’ checks take place in Albany. Under the circumstances, venue cannot be found under § 1391(b). B. Venue Based on Defendants’ Residence Under 28 U.S.C. § 1392(a) Although the present record precludes a finding that all defendants reside in this district, 28 U.S.C. § 1392(a) offers an alternative means of basing venue on the residences of the defendants. That section provides: Any civil action, not of a local nature, against defendants residing in different districts in the same State, may be brought in any of such districts. In this case, all the defendants reside in New York. Furthermore, some of the defendants do seem to have official residences in this district; the complaint names as defendants, among others, the Governor of New York, who conducts significant amounts of business from offices in the Southern District; the Acting Director of the Department of Correctional Services, which maintains offices and operates many state prisons in this district; the Commissioner of the Office of Mental Hygiene, which operates many facilities in this district; and the Chancellor of the State University of New York, which operates several educational institutions in this district. Plaintiffs allege that over half the plaintiff class are in fact employed in facilities run by these named defendants. In Kirkland v. New York State Department of Correctional Services, 358 F.Supp. 1349 (S.D.N.Y. 1973), a case involving similar circumstances, Judge Lasker found venue to lie under Section 1392(c). The state seeks, in effect, to distinguish Kirkland by arguing that the defendants in this case who can reasonably be found to have official residences in the Southern District are not necessary parties to this action. Plaintiffs should be able to obtain all the relief they seek, the state contends, without the participation of any defendant who resides in this district. By contrast, in Kirkland, where the requested relief apparently pertained to prison practices or conditions, the Department of Correctional Services was a necessary party. Defendants’ argument rests on the unarticulated premise that Section 1392(a) becomes a device to circumvent Section 1391(b) if a plaintiff can obtain venue in non-diversity cases by simply joining as a defendant a non-essential party who resides in the district of plaintiff’s choice. Only a handful of dated cases have examined this question. Most of these cases rest on a distinction between “necessary parties” — defendants whose joinder is needed for plaintiff to obtain relief — and “nominal” or “formal parties” — defendants who, although proper parties, are not essential for plaintiff to obtain relief. Relying on predecessor statutes to Section 1392(a), the courts in those cases held that venue could lie only if “necessary” parties resided in the forum district. United Office & Prof. Workers of America v. Smiley, 75 F.Supp. 695 (E.D.Pa.1946); Nakken Patents Corp. v. Westinghouse Elec. & Mfg. Co., 21 F.Supp. 336 (E.D.Pa.1937); Baltimore & O. R. Co. v. *209 Board of Public Works, 17 F.Supp. 170 (N.D.W.Ya.1936). 5 In Gunn v. Mathis, 157 F.Supp. 169 (W.D.Ark.1957 & 1958), seemingly the most recent case on point, the court adopted a significantly more flexible approach. It held that an unnecessary party’s residence in the district could establish venue if that party was nevertheless “proper” and had a “substantial connection” to the lawsuit in question. 6 The test suggested by Gunn is preferable to the approach of the earlier cases. Nothing in the history or language of Section 1392(a) suggests an ungenerous application. 7 Indeed, to limit its scope to “necessary” parties, would cut against the legislative policy that favors permissive joinder of non-essential but proper parties. Rule 20(a) of the Federal Rules of Civil Procedure permits the joinder “as defendants” of “all persons” if “there is asserted against them jointly, severally, or in the alternative, any right to relief in respect of or arising out of the same transaction, occurrence . and if any question of law or fact common to all defendants will arise in the action.” This provision represents a considered determination, reached over years of experi *210 ence and after several revisions, that the convenience of parties and of the courts is most effectively advanced by permitting the joinder of parties even if their presence is neither indispensable nor necessary. See generally, Moore, supra 120. Venue rules should be applied consistently with this strongly expressed congressional policy. Furthermore, the danger posed by possible circumvention of Section 1391(b) can easily be exaggerated. Section 1392(a) applies only when multiple defendants are involved, all of whom reside in a single state. It applies, moreover, only to transitory actions, not local actions. It is, as Wright, Miller & Cooper have called it, “a limited statutory escape” from the normal rule. Federal Practice & Procedure § 3807, p. 38 (1976). Even where Section 1392(a) would apply, if construed as in Gunn, no drastic consequences would result. At worst, defendants residing in a single state would have to answer in a district in which they do not reside. And, even then, discretion would remain under 28 U.S.C. § 1404(a) to transfer the case to other proper districts in the interests of justice. For these reasons, any defendant who is not merely a formal or nominal party, and who can properly be joined under the Federal Rules of Civil Procedure, should be considered a defendant upon whose residence venue may be based under Section 1392(a). See Moore, supra 119.02. 8 In the present case, the wardens and other institutional directors are not necessary parties, since plaintiffs can obtain all the relief they seek without joining these defendants. Nevertheless, these heads of institutions are more than merely formal or nominal parties. They have a substantial relationship to plaintiffs, and are officers who enforce the state’s policy against strikes by public employees. They may have considerable information to contribute to the proper resolution of the litigation, and their presence facilitates the implementation of discovery, trial, and the preliminary and final determinations of the court. They are, moreover, parties with a significant interest in the outcome of the litigation, since they will be responsible for operating the institutions in which plaintiffs work. They are, in short, proper parties, since their interests stem from the same transactions and events as those of the other defendants, and since the same questions of fact and law must be resolved as to all defendants. Given these circumstances, the fact that plaintiffs may have joined some defendants for the purpose of obtaining venue is not controlling. Their substantial interest in and relationship to the litigation make them proper defendants for purposes of applying Section 1392(a). C. Venue Based on Where the Claim Arose Under 28 U.S.C. § 1391(b) Plaintiffs argue that, even if venue were not found under Section 1392(a), venue is proper under Section 1391(b) because this is “the district ... in which the claim arose . . . .” Section 1391(b) creates problems even more difficult than those engendered by Section 1392(a). The greatest difficulty is that posed by the language of Section 1391(b), also found in Section 1391(a). The statute specifically states that venue may be established “only” in “the” district where “the” claim arose. Professor Moore surely understates when he writes that this language lends “some inference that the draftsmen assumed that every claim arises in some one district.” Moore, supra ¶ 0.142[5.-2] at p. 1431. The language is as clear as Congress could have made it, there *211 by suggesting that a claim can arise only in a single district. Moore’s suggestion that Congress gave no thought to what it had done, see Moore, supra, 10.142[5.-2], p. 1426, is speculative, and seems dubious in light of the precedents and discussion of this question prior to 1966. 9 The only legislative history available supports the view that Congress intended to ensure that plaintiffs in all federal suits would be able to find a district in which venue is proper when no other basis for venue exists. 10 This limited objective supports a literal reading of Sections 1391(a) and (b), and implicitly refutes the view that the statutory language can be ignored as mere legislative blunder. Despite this background, many cases have avoided the statute’s clear language, holding without explanation that the section allows suit in multiple districts. E. g., Miller v. Cousins Properties, Inc., 378 F.Supp. 711 (D.Vt.1974); and cases cited in Wood, supra, at 410-11. Other decisions at least discuss the policy considerations supporting the results reached. See generally, Wood, supra, at 399-400. But these cases — many based on the notion that suit may be brought in any district in which substantial contacts relating to the claim occurred — fail to explain adequately the statutory language. E. g., Ghazoul v. International Management Services, Inc., 398 F.Supp. 307, 315 (S.D.N.Y.1975); United States ex rel. Flemings v. Chafee, 330 F.Supp. 193, 194 (E.D.N.Y.1971), aff’d, 458 F.2d 544 (2d Cir. 1972), rev’d on other grounds, 413 U.S. 665, 93 S.Ct. 2926, 37 L.Ed.2d 873 (1973); Iranian Shipping Lines v. Moraites, 377 F.Supp. 644, 648 (S.D.N.Y.1974) (quoting from a case arising under Section 1391(c)). But see Honda Associates, Inc. v. Nozawa Trading, Inc., 374 F.Supp. 886, 889-90 (S.D. N.Y.1974) (explaining why venue does not automatically follow contacts justifying personal jurisdiction). To the extent that substantial contacts can exist simultaneously in more than one forum, these decisions are inconsistent with the plain meaning of the statute. 11 The Supreme Court has recently refused to rule on whether Section 1391(b) permits suit in more than one district. Leroy v. Great Western United Corp., 443 U.S. 173, 99 S.Ct. 2710, 2717, 61 L.Ed.2d 464 (1979). If the Supreme Court were to read Section 1391(b) to require the designation of a single district as the one in which “the claim arose,” presumably on the basis of its being the place having the most substantial con *212 tacts with the claim, 12 venue would not lie in this district. However substantial the contacts or relationship of the overall claim to the Southern District, no court could reasonably hold that the case can be brought here if the ruling meant that the case could not as a result be brought in the Northern District. See Weil v. New York State Dep’t of Transportation, 400 F.Supp. 1364 (S.D.N.Y.1975). But the Court in Leroy left room for a construction of Section 1391(b) that would permit courts to find a claim arose in more than one district. The Court found “arguably acceptable” the following interpretation (99 S.Ct. at 2718): [I]n the unusual case in which it is not clear that the claim arose in only one specific district, a plaintiff may choose between those two (or conceivably even more) districts that with approximately equal plausibility — in terms of the availability of witnesses, the accessibility of other relevant evidence, and the convenience of the defendant (but not the plaintiff) — may be assigned as the locus of the claim. Whether this is the “unusual case” in which it is unclear that the claim arose in only one specific district is difficult to know. Many courts have attempted to determine the district in which a claim arose, and . the varying tests used and results reached do not indicate that claims arise in multiple districts only in “unusual” cases. Courts have frequently found that claims arose in multiple districts because of substantial contacts with more than one forum, or because of a significant relationship with more than one district, or because of a substantial impact in more than one district. 13 The Court’s analysis would seem clearly to preclude at least some prior approaches. At a minimum, the Court has indicated that no approach is permissible (or “arguable” as the Court put it) which would lead to multi-district venue in all but a few cases. Multidistrict venue would appear permissible only where the claim had very substantial contacts with, or a close relationship to, each of the districts claimed as potentially appropriate. In this case, all the officials who determined that plaintiffs’ pay should be reduced reside in Albany and made their decisions there. The effects of their decisions were profoundly felt in the Southern District, however. Many prisons and other institutions in which approximately half the members of the plaintiff class work are located here. Furthermore, the Albany decision makers were not merely initiating a policy which had an impact in this district. They were reacting to an apparent strike, which had its roots in the institutions and the conditions of employment that prevailed in those institutions. This district would appear, therefore, to have a substantial relationship to the claim. Assuming the Southern District is a potentially proper forum under Leroy, the *213 plaintiff must also show that this district may “with approximately equal plausibility [as the Northern District] ... be assigned as the locus of the claim.” And the plausibility of such an assertion must turn on “the availability of witnesses, the accessibility of other relevant evidence, and the convenience of the defendant (but not the plaintiff) . . . ” 99 S.Ct. at 2718. In this respect, the argument for venue in this district is far stronger than plaintiff’s argument for venue in Leroy. A finding of venue in the Texas district in which plaintiff resided in Leroy, would have subjected the Idaho defendants to suit in virtually every district in the nation. All the important witnesses and evidence were located in Idaho, moreover, and the federal judges in Idaho were presumed more able to handle the many state law issues that were involved. By contrast, defendants in this case will be subjected to potential suit only in districts within New York, perhaps only in the Southern District in light of the relatively large number of guarded institutions here. This district is not an inconvenient forum for defendants, since it is close to Albany, and since most state agencies have facilities and offices here. Furthermore, location of the necessary evidence is insignificant in this case, since no issues of fact exist; the entire matter was presented by affidavits. Finally, the federal judges of this district would appear presumptively as able to construe and apply New York law as federal judges in any other district within this state. This analysis suggests that, even under the relatively restrictive — but “arguably acceptable” — test adopted in Leroy, an alternative basis for venue in this district exists under Section 1391(b), and that basis is hereby adopted. Some doubt must persist, however, as to the Supreme Court’s willingness to accept any interpretation of the claim-arising language of Section 1391(b) that does more than fill a “venue gap” in particular cases. The majority in Leroy expressly adopted the suggestion in Brunette Machine Works, Ltd. v. Kockum Industries, Inc., 406 U.S. 706, 710 n. 8, 92 S.Ct. 1936, 1939, 32 L.Ed.2d 428 (1972), that the “claim arose” language was added to the venue statutes in 1966 because Congress wanted to fill the venue gap that occasionally confronted plaintiffs prior to that time by assuring them at least one proper district. If venue gap filling was Congress’ only purpose in amending Sections 1391(a) and (b), it would appear that any given cause of action should be deemed to arise in only one district, since at that point the venue gap would be filled. And the most suitable analytical tool for determining in which district venue should lie would seem to be the “most significant contacts” or “weight of the contacts” approach. See Philadelphia Housing Authority v. American Radiator & Standard Sanitary Corp., 291 F.Supp. 252 (E.D.Pa.1968). In this case, that test would place venue in the Northern District of New York. The difficulties posed by Leroy seem certain to make venue issues an even greater problem for the lower federal courts than they were before the Supreme Court spoke. Pressures for increased venue flexibility have caused successive statutory liberalizations since the Judiciary Act of 1887, 14 and the liberal rules developed over the years by numerous judges in literally hundreds of cases reflect their desire to prevent the needless expenditure of judicial and litiga-tive resources on technical and abstruse constructions of venue provisions, the true purposes of which can never be known with confidence. 15 The Court in Leroy referred *214 to virtually none of the many lower federal court decisions construing the claim-arising language, relying instead on the premise that the language was adopted only to fill venue gaps and not for the convenience of plaintiffs. However indisputable this proposition now becomes, by reason of the Court’s having stated it, the significance of the claim-arising language under Section 1391(e) remains unexplained, although it would appear to have been the most relevant analogy in construing Sections 1391(a) and (b), 16 and the possibility of protecting defendants while at the same time accommodating plaintiffs remains unexplored. D. Request for Transfer Under 28 U.S.G. § 1404(a) Defendants’ request for a transfer under Section 1404(a) is the final venue issue presented. The Northern District of New York, defendants claim, is more convenient for the parties and witnesses in this case than this district. Moreover, they argue that a transfer is “in the interest of justice” because plaintiffs engaged in forum shopping by bringing suit far from the place where all the important witnesses and records are located, and far from where plaintiffs initiated their state court actions. In fact, the principal witnesses in this case will be state officials, most of whom have offices in Albany. Little need exists for the testimony of members of the plaintiff class who reside here. Even the lawyers in this case are from Albany. However welcome such able counsel may be in this district, one cannot help but be suspicious of a party’s motivations when his attorney has to travel several hours to attend a session in his chosen forum. Plaintiffs’ counsel concedes that he consciously chose the Southern District despite its distance from his office. He argues, however, that his motivation was not to inconvenience defendants, but to avoid anticipated delays in an overburdened forum. Section 1404(a) poses little difficulty for courts resolved either to grant or deny motions for transfer. If a court is inclined to grant a transfer, it has available the established doctrine that the section was intended to prevent forum shopping. If a court is inclined to deny a transfer, it usually suffices to invoke the equally well established *215 doctrines that a plaintiff’s choice of forum is entitled to respect, and that a moving defendant has the burden of proving a case for transfer. See generally, Moore, supra ¶ 0.145[5]. A court seeking reasons for choosing one or the other of these conflicting doctrines is inevitably led to Mr. Justice Jackson’s analysis of the doctrine of forum non conveniens in Gulf Oil Corp. v. Gilbert, 330 U.S. 501, 67 S.Ct. 839, 91 L.Ed. 1055 (1947). Forum shopping is no more an evil than any other tactical determination a party makes in its behalf. Any competent lawyer chooses a forum with his or her client’s interests in mind. This is undoubtedly why Justice Jackson made clear that a party had to do more than forum shop to justify invoking forum non conveniens; the choice of forum must have been made “with some harassment” in mind to have justified dismissal of the action. Id. at 507, 67 S.Ct. 839. Among the factors identified in Gulf Oil as warranting consideration in determining whether to invoke forum non conveniens was the fact, alleged in this case, that court congestion in a forum may cause administrative difficulties. Furthermore, that the defendants here have offices in the Southern District, and are represented by the State Attorney General whose principal office is in this district, are also facts to which Justice Jackson would have given considerable weight. In addition, Gulf Oil would require that consideration be given to the fact that plaintiffs’ counsel has a fairly substantial practice in this district, despite its distance from his office, and that many of the plaintiffs’ attendance records are located here. Finally, Justice Jackson’s analysis would incorporate the fundamental fact that this case does not require a trial in which the Albany based witnesses would need to appear. It is true that Section 1404 permits “courts to grant transfers upon a lesser showing of inconvenience” than had been required to dismiss an action under the doctrine of forum non conveniens. Norwood v. Kirkpatrick, 349 U.S. 29, 32, 75 S.Ct. 544, 546, 99 L.Ed. 789 (1955). But nothing written since the 1948 passage of Section 1404 negates Justice Jackson’s reasoning, particularly with respect to the significance of forum shopping. Courts continue to recognize that the relative condition of dockets in alternative districts is a proper consideration in deciding motions to transfer. E. g., Solomon v. Continental American Life Insurance Co., 472 F.2d 1043, 1047 (3d Cir. 1973); Buffalo Teachers’ Federation, supra, 426 F.Supp. at 830; Farben-fabriken Bayer A. G. v. National Distillers & Chemical Corp., 324 F.Supp. 156 (S.D.N. Y.1971). In this case, speed of disposition was a particularly important objective, since plaintiffs were seeking to stop a program of payroll deductions that, absent judicial intervention, would have been completed in less than two months. Had plaintiffs selected this forum to avoid specific precedents in the Northern District, the case for a transfer would have been far stronger. Just as this district’s local rules are designed to prevent shopping for individual judges — and thereby for the application of some anticipated view of the law — efforts to select one district to avoid or to obtain specific rulings of another district court should be disfavored and discouraged. No such argument can be made here, however. Defendants concede that no ruling has been rendered by any Northern District judge that would predictably control the decision of this case. Under the circumstances, plaintiffs appear to be proceeding in this district for legitimate tactical considerations, and not for the purpose of harassing defendants. No substantial injustice is caused by plaintiffs’ choice. II. ABSTENTION Defendants also seek dismissal of plaintiffs’ complaint on the theory that this court should abstain. Defendants note that plaintiffs have brought two separate cases in state court challenging the Director’s determination, but have carefully refrained from raising or pursuing their constitutional claims in the state actions. They argue that, since plaintiffs had the opportunity to *216 present their federal claims in state court, plaintiffs should be precluded from doing so here. “Abstention from the exercise of federal jurisdiction is the exception, not the rule.” Colorado River Water Conservation District v. United States, 424 U.S. 800, 813, 96 S.Ct. 1236, 1244, 47 L.Ed.2d 483 (1976). At the time this action was brought, it appeared to fall into none of the categories in which the Supreme Court has indicated that federal court abstention is appropriate. Id. at 813-20, 96 S.Ct. 1236. See generally, 17 Wright, Miller & Cooper, Federal Practice and Procedure § 4241 (1978). Defendants initially relied on the principles of federal judicial restraint established by Younger v. Harris, 401 U.S. 37, 91 S.Ct. 746, 27 L.Ed.2d 669 (1941), and its progeny involving civil litigation. See, e. g., Moore v. Sims, 442 U.S. 415, 99 S.Ct. 2371, 60 L.Ed.2d 994 (1978); Juidice v. Vail, 430 U.S. 327, 97 S.Ct. 1211, 51 L.Ed.2d 376 (1977); Trainor v. Hernandez, 431 U.S. 434, 97 S.Ct. 1911, 52 L.Ed.2d 486 (1977). But these principles are inapplicable here. Younger problems arise when the federal plaintiff is a defendant in a state judicial proceeding in which the state or a private litigant has sued to enforce an important state interest. The purpose of the federal suit is to vindicate a constitutional right by either enjoining the state judicial proceeding or by obtaining declaratory relief which effectively moots that proceeding. In such cases, principles of equity, comity and federalism have been held by the Supreme Court to require that federal courts not interfere with the state judicial proceeding, absent a showing of bad faith, if the federal constitutional right asserted by the federal plaintiff can be vindicated in the state judicial proceeding. Plaintiffs here are not defendants in state court and do not seek to enjoin or curtail any judicial proceeding brought to enforce a state interest. The decision to implement the Taylor Law was administrative, and no effort has been made, or is presently needed under the statute, to enforce that decision in the state courts. The injunction sought is against state officers who, through executive action, will allegedly violate plaintiffs’ constitutional rights. Exercise of federal jurisdiction in such a case is singularly appropriate. As the Supreme Court has noted, “there is, of course, no doctrine requiring abstention merely because resolution of a federal question may result in the overturning of a state policy.” Zablocki v. Redhail, 434 U.S. 374, 379 n. 5, 98 S.Ct. 673, 678, 54 L.Ed.2d 618 (1978). Defendants also suggested abstention because plaintiffs could have raised their federal claims in the state court actions they brought. The Supreme Court ' has recognized that, under “exceptional” circumstances, “considerations of ‘wise judicial administration, giving regard to conservation of judicial resources and comprehensive disposition of litigation,’ ” may justify dismissal of a federal action in light of a concurrent state proceeding. Colorado River Water Conservation District v. United States, supra, 424 U.S. at 817-18, 96 S.Ct. at 1246. But absent “exceptional” circumstances, the general rule “as between state and federal courts ... is that ‘the pendency of an action in the state court is no bar to proceedings concerning the same matter in the Federal court having jurisdiction . . . Id. at 817, 96 S.Ct. at 1246. “Only the clearest of justifications will warrant dismissal.” Id. at 819, 96 S.Ct. at 1247. No “exceptional” circumstances are present here. This forum is clearly convenient. Further, the fact that the state actions were brought first is not in itself determinative. Although it is generally desirable to avoid piecemeal litigation, the concurrent pendency of these suits may in fact enhance the efficient administration of justice. The legal issues raised in each action are addressed to the expertise of the respective fora; the state courts are considering statutory questions while this court is asked to decide a federal issue. Moreover, those legal issues can be determined directly and expeditiously by the respective courts, since no material facts are in dispute. Indeed, the absence of disputed facts *217 in itself eliminates the greatest potential problems of concurrent adjudications. Finally, this case originally came before the court on an order to show cause and a motion for a temporary restraining order. To have abstained at that time might have mooted plaintiffs’ argument that the extent to which their salaries were being reduced without prior hearing violated due process. At the outset of this litigation, defendants also mentioned, but did not press, the possibility of abstention under Railroad Commission of Texas v. Pullman Co., 312 U.S. 496, 61 S.Ct. 643, 85 L.Ed. 971 (1941). That case requires federal courts to abstain from deciding federal constitutional questions when a state court decision on state law grounds might obviate the need to reach the federal constitutional question. Pullman initially seemed inapposite because plaintiffs brought this suit as a broad challenge to the Taylor Law’s penalty provision, arguing that the deduction of an amount as great as that imposed after their strike violated due process. Defendants properly argued that the broad challenge had been rejected in the state courts, and in the Supreme Court as well. Further, plaintiffs’ efforts to distinguish Sanford as involving a de minimus penalty, 17 and as involving a determination challenged by far fewer employees, 18 were not persuasive. After plaintiffs’ broad contentions were rejected on the motion for preliminary relief, an additional argument was advanced based on the speed at which the penalties were being taken from plaintiffs’ salaries. 19 The state assumed in its presentation that the statute left no discretion to either the Director or to the state courts to construe or to apply the Taylor Law so as to enable the Director to extend the period for imposing the penalty beyond ninety (90) days of his determination. 20 The fact that both *218 parties shared this assumption, together with the absence of any ambiguity on the face of the statute, indicated that Pullman abstention was unwarranted. An order was therefore entered on federal constitutional grounds, preliminarily enjoining defendants to exact the statutory penalty over a longer period of time than the Taylor Law seemed to require. The preliminary injunction rested on the proposition that, although the absolute amount of the penalties contemplated in this case may be justifiable under due process, without a prior hearing, the speed with which the penalty was being exacted constituted a matter of independent concern. Because the state claimed it was bound by Section 210 to deduct the full penalty no more than ninety days after the Director’s determination, it proposed a schedule whereby approximately 33% of gross and over 50% of take-home salary would be deducted from the pay of all plaintiffs for four consecutive two-week periods. 21 To have left employees with so small a portion of their wages, for a two-month period, would have created substantial hardship for all and extreme hardship for many; 22 the state’s effort to controvert this hardship was not persuasive. 23 Furthermore, the arguments pressed by the state to justify this stringent exaction were unconvincing; plaintiffs clearly have a property interest in their salaries, which the state has not (and probably could not) deny them. 24 Arnett v. Kennedy, 416 U.S. 134, *219 94 S.Ct. 1633, 40 L.Ed.2d 15 (1974) is distiguishable; in the context of this case a suspension without pay would have less severe economic consequences than the deductions contemplated, at least in the short run. 25 In light of widely adopted policies limiting wage garnishment to at most 25% of gross income, 26 which policies are fully pertinent here, 27 the length of time that can be expected to pass before plaintiffs would be afforded hearings, and the fact that a large proportion of plaintiffs seemed likely to secure some degree of relief as a result of the hearings, 28 it appeared that plaintiffs *220 were likely to succeed on the merits and would suffer irreparable harm if an injunction did not issue. Neither party appealed from the preliminary injunction requiring a longer payout period but denying relief on plaintiffs’ more sweeping claims. Both parties recently agreed, moreover, to an entry of final judgment without further proceedings. Nevertheless, a recent state court decision brings this case squarely within the scope of Pullman and requires abstention on the merits. During November 1979, a suit was filed in this District which challenged the application of the Taylor Law penalty provision to workers at New York’s Off-Track Betting Corporation. Plaintiffs claimed, as in this case, that the state was constitutionally required to limit withholding of their pay to sums approximating those allowed by the preliminary injunction in this case. Judge Duffy abstained in an oral opinion, dismissing the complaint on the ground that “the question of the constitutionality and legality of the application of that state law should be determined primarily by a state court.” Local 2021 of Dist. Council 37 v. Off-Track Betting Corp., 79 Civ. 6005, Transcript, p. 3 (S.D.N.Y. November, 1979). This ruling, though doctrinally difficult to rationalize, proved sound in retrospect. Plaintiffs proceeded to state court and promptly obtained from New York County Supreme Court an injunction mandating a longer period of withholding than the statute appears to authorize. Characterizing the state’s statutory argument as “a harsh, strict interpretation of the law,” Justice Okin held that principles of equity, and the unlikelihood that the legislature intended to reduce civil servants to paupers, justified reading the statute to enable alleged strikers “to provide for their families basic needs and monthly necessities.” Taking notice of the impending holiday season, and of the fact that two of the three deductions had already been effected, the Court ordered that no further deduction take place until after December 25. Local 2021 of Dist. Council 37 v. N.Y.C. Off-Track Betting Corp., Index No. 6065/79, Memo. Decision, pp. 2-3 (Sup.Ct. N.Y.Co. Nov. 30, 1979). The state court’s decision in Local 2021 makes clear that both the parties in this case may have been mistaken in assuming that no state law question was lurking here, the decision of which might eliminate the need to decide the constitutional claim. The state courts may well, by construing the Taylor Law, place restraints upon the state’s power to enforce the Act that would sufficiently ameliorate the plight of employees in plaintiffs’ position to eliminate what otherwise seems a substantial constitutional issue. Granted, abstention at this stage is inefficient because much of the work of deciding this litigation has already been done. But Pullman abstention has always been inefficient. See generally Field, The Abstention Doctrine Today, 125 U.Pa.L.Rev. 590 (1977). It is mandated by the Supreme Court as a means for avoiding constitutional questions, and in that sense it applies even at this stage of the case, since the only decision reached on the merits in this case — that a preliminary injunction should issue — was not a final adjudication of the constitutional question. One modification of ordinary Pullman abstention practice does seem appropriate, however. While Pullman abstention normally requires the federal court to retain jurisdiction in case the need to decide the federal question is not obviated by state court action on state law grounds, a dismissal here seems proper. The deduction schedule established by the preliminary injunction should by now be complete. Nothing therefore remains for federal courts to correct in this controversy, except perhaps *221 on appeal from this ruling, which a dismissal should facilitate. See Field, supra, 125 U.Pa.L.Rev. at 592-601. The complaint is therefore dismissed on the ground that it would be improper to decide the constitutional question presented. SO ORDERED. Footnotes [1] Jurisdiction is based on 28 U.S.C. §§ 1331 & 1343. The substantive claims are made under 42 U.S.C. §§ 1983 § 1988, and declaratory relief is sought under 28 U.S.C. §§ 2201 & 2202. [2] See September 5, 1979 letter from plaintiffs’ counsel, Richard R. Rowley. [3] See New York State Inspection, Security and Law Enforcement Employees, District Council 82, AFSCME v. Carey, 99 Misc.2d 565, 416 N.Y.S.2d 697 (N.Y.Sup.Ct. Albany Co., 1979). [4] In Buffalo Teachers’ Federation v. Helsby, supra, the defendants were members of the New York Public Employment Relations Board, which maintained an office in New York City and conducted, from that office, “a significant portion of its business.” 426 F.Supp. at 830. It is unclear whether defendants’ activities were related to the cause of action asserted. The defendants in Procario v. Ambach, supra, by contrast, were members of New York’s Board of Regents, which conducted “only a very limited amount of business from its New York City branch office,” none of which was related to the issues in that case. 466 F.Supp. at 453. [5] In Baltimore & O. R. Co. v. Board of Public Works, plaintiff sought to enjoin the collection of a state tax. Although all the defendants resided in West Virginia, the only defendants residing in the forum district were local sheriffs who would enforce one of the challenged laws if it were not enjoined. The court, construing a predecessor to Section 1392(a), determined that the sheriffs would be powerless to do anything “which would affect the plaintiff’ if an injunction were granted against the State Auditor. Thus, according to the court, these defendants were “mere nominal” or “merely formal” parties, not “necessary” parties, and venue could not be based on their residences. In Nakken Patents Corp. v. Westinghouse Elec. & Mfg. Co., plaintiff sued to establish its right to a patent. The defendant who had assigned the patent to Westinghouse was a resident of the forum district. The court, construing the predecessor to Section 1392(a), concluded that although the assignor was a proper party, his presence was purely formal and unnecessary and therefore his residence was insufficient to establish venue. Finally, in United Office & Prof. Workers of America v. Smiley, the court followed Baltimore & O. R. Co. and Nakken, and refused to base venue on the presence of a defendant employee whose participation was deemed unnecessary. Although other factors were present, the court relied on the loose distinction between “necessary” and “nominal or formal” parties. [6] The plaintiff in Gunn v. Mathis sought to enjoin representatives of the Internal Revenue Service from collecting taxes under a certain assessment and lien. The only defendant joined who resided in the district where suit was brought was a Revenue Officer named Krone; the District Director and Chief Collection Officer both resided in another district. In an initial decision, the court, relying on Baltimore & O. R. Co. v. Board of Public Works, supra, Nakken Patents Corp. v. Westinghouse, supra, and United Office & Prof. Workers of America v. Smiley, supra, found that Krone’s presence did not establish venue. While the District Director was an indispensible party, Krone was merely subject to his orders; conversely, an injunction against Krone would not obtain the relief plaintiff sought — expunging of the challenged tax lien. On rehearing, the court reversed its ruling. The plaintiffs, the court noted, sought not only an expunging of the lien but also an order against any levy or seizure pursuant to the lien. This made Krone “more than a nominal party,” particularly since Krone had previously discussed enforcing the lien with plaintiff and his attorney. This “substantial connection with the collection of the taxes” made Krone a “proper party” and thus established venue in the district of his residence. Id. at 176-77. [7] Until 1887, defendants were usually subject to suit anywhere they could be found. The notion that defendants should be forced to answer only in the district of their residence appeared for the first time in that year. Moore, supra ¶ 0.141, p. 1337. Nevertheless, since relatively few districts existed within the states, the 1887 legislative change did not have a significant effect. Indeed, as the number of districts within individual states increased, exceptions were made to lessen the impact of the 1887 change. Thus, venue in Baltimore & O. R. Co. v. Board of Public Works, supra note 6, would probably now lie on the ground that the cause of action arose in the district in which the challenged tax was to be collected. 28 U.S.C. § 1391(b). Since 1948, suits against corporations such as Nakken Patents Corp. v. Westinghouse Elec. & Mfg. Co., supra, can be brought in any district in which the company is “doing business,” in addition to the districts in which it was incorporated or licensed to do business. 28 U.S.C. § 1391(c). Venue in suits against federal officers, such as Gunn v. Mathis, supra, can be based on four separate grounds. 28 U.S.C. § 1391(e). In short, Congress has itself minimized the impact of the 1887 venue statute, reflecting no strong legislative interest in a restrictive interpretation. [8] In general, this rule would exclude, as defendants under 1392(a) parties such as guardians, beneficiaries (when a trustee is sued), or other mere figureheads. There are several decisions — especially in jurisdictional cases — that equate “merely formal” with “unnecessary” parties. E. g., Nunn v. Feltinton, 294 F.2d 450, 455 (5th Cir. 1961), cert. denied, 369 U.S. 817, 82 S.Ct. 829, 7 L.Ed.2d 784 (1962). Although these decisions may be appropriate in their special contexts, such as in ascertaining if diversity of citizenship is present, they should not be applied in determining venue under Section 1392(a) given the relatively inconsequential interests at stake, and given the fact that 28 U.S.C. § 1404(a) has been available since 1948 to avoid such injustices as may result from this policy. [9] Scholars had suggested alternative formulations to correct the 1887 Act’s restrictiveness. See, e. g., Stevens, Venue Statutes: Diagnosis and Proposed Cure, 49 Mich.L.Rev. 307, 310 (1951) (discussion of state statutes); Blume, Place of Trial of Civil Cases, 48 Mich.L.Rev. 1, 39 (1949). Even if it is safe to assume that Congress ignored these scholars, other suggestions for a more expansive version emanated from sources less readily dismissed. In 1946, the American Law Institute’s Tentative Draft No. 2 on the Division of Jurisdiction Between the State and Federal Courts, proposed that Congress make venue proper in any district where a “substantial part of the events or omissions giving rise to the claim occurred.” The Judicial Conference of the United States suggested a more limited statute. It recommended in 1959 that venue in tort cases be permitted in the district in which the tortious act took place, and in 1962 and 1963 the Conference repeated and expanded this recommendation to make it applicable in all civil actions. 1959 Jud.Conf.Rep. 316; 1962 id. at 11; 1963 id. at 17. Congress in fact adopted the Conference’s suggestion regarding tort actions, permitting suit in many such cases where the act or omission complained of occurred. Act of Dec. 23, 1963, Pub.Law 88-234, 77 Stat. 473. In 1965, the Judicial Conference again pressed for the relatively limited legislation it preferred, and it was this proposal that Congress adopted in 1966. Wright, Miller & Cooper, supra, § 3806, pp. 26-27. [10] See generally Moore, supra, ¶ 0.142[5.-2], p. 1426; Wood, Federal Venue: Locating the Place Where the Claim Arose, 54 Texas L.Rev. 392, 398 (1976). Ms. Wood’s article is particularly helpful, though it advocates a view that ignores Congress’ language. [11] Judge Friendly’s comment in Liberation News Service v. Eastland, 426 F.2d 1379, 1382 n. 4 (2d Cir. 1970), that venue might have been available under Section 1391(b), “perhaps” in multiple districts, was at most a suggested possibility, unaccompanied by the reasoning and analysis that so often make his considered pronouncements authoritative. [12] The substantial contacts or weight of the contacts approach has been particularly popular in multi-defendant antitrust and security cases. See, e. g., Philadelphia Housing Authority v. American Radiator & Standard Sanitary Corp., 291 F.Supp. 252 (E.D.Pa.1968); and cases cited in Wood, supra, at 403-07 where the author argues that the decisions are unclear as to whether venue can lie in only one district. Other courts have construed the statute to permit suit in “the” district in which the defendant’s actions had their greatest impact upon any individual plaintiff. See, e.g., Iranian Shipping Lines v. Moraites, 377 F.Supp. 644 (S.D.N. Y.1974) (Gurfein, J.), discussed with approval in Brient & Scheindlin, Venue in the Second Circuit: A Topic Whose Time Has Come, 43 Bklyn.L.Rev. 841, 843 (1977); Jimenez v. Pierce, 315 F.Supp. 365 (S.D.N.Y.1970) (transferring to N.D.N.Y. as locus of impact despite presence of defendants’ headquarters in S.D.N. Y.). [13] Some courts require only minimal contacts to justify finding venue, a test very similar to that for acquiring jurisdiction by service of process. Put another way, these courts require a showing only that any part of the claim arose in the chosen district. E. g., Arnold v. Smith Motor Co., 389 F.Supp. 1020, 1023-24 (N.D. Iowa 1974). Others require at least that a substantial part of the claim—the weight of the contacts—arose in the'chosen district. E. g., Honda Associates, Inc. v. Nozawa Trading, Inc., 374 F.Supp. 886, 891-92 (S.D.N.Y.1974); Mendelson v. Fleischmann, 386 F.Supp. 436 (S.D.N. Y. 1973). [14] See generally 1 Moore, supra ¶ 0.141, pp. 1337-39. The 1887 Act’s contraction of venue options resulted from the expanding number of federal judicial districts within individual states, in an age when even intrastate travel frequently posed substantial problems for litigants. Times have changed. Travel within most states presents little if any problem for most parties. Furthermore, 28 U.S.C. § 1404 enables district courts to transfer any case for the convenience of parties or witnesses, in the interests of justice, to other districts in which the action could have been brought. [15] See the discussion in Brient & Scheindlin, supra, 43 Bklyn.L.Rev. at 842-45. Bastille Properties Inc. v. Hometels of America, Inc., 476 F.Supp. 175, 178-82 (S.D.N.Y.1979), is a *214 recent decision that might well reflect how able District Judges will construe and apply Leroy. [16] The “cause of action” language of Section 1391(e) was not necessary to fill a venue gap. The Leroy Court distinguished the statute, however, with a “But cf.” citation and no discussion. 99 S.Ct. at 2717. Yet, the language in Sections 1391(a) and (b) (“in which the claim arose”) is functionally identical to that used in Section 1391(e) (“in which . . . the cause of action arose”). The latter section, adopted in 1962, only four years before Sections 1391(a) and (b), provides the best available analogy. See Moore, supra, 1] 0.142 [5.-2], p. 1427. Section 1391(e) was designed to implement 28 U.S.C. § 1361, which confers jurisdiction on the district courts to compel federal officials to perform duties owed to plaintiffs. Section 1391(e) provided several alternative bases for venue in such suits, so that citizens could bring actions against federal officials in districts other than Washington, D. C. (as prior law required), even though the policy authorizing the challenged action often emanated from the nation’s capital. See generally, Natural Resources Defense Council, Inc. v. Tennessee Valley Authority, 459 F.2d 255 (2d Cir. 1972) (Friendly, C. J.). Courts have not had to consider what Congress meant when it allowed suit under Section 1391(e) where the “cause of action arose,” because other provisions — particularly that allowing suit in the district where any plaintiff resides — eliminated the need to rely on the cause-arising clause. But the fact that these clauses of Section 1391(e) overlap suggests the purpose of both provisions was essentially the same — to provide plaintiffs with the most convenient forum under the circumstances. Given this purpose, it is a fair inference that the district “in which . the cause of action arose” should be read to mean the district in which the impact of a decision or policy on a particular plaintiff is most acutely felt. Usually, the district of greatest impact would be the plaintiff’s residence. But where it is not, a greatest-impact reading would serve to maximize the plaintiff’s convenience, as the statute intended. If the cause-arising language of Section 1391(e) is read to mean the place of greatest impact, a statutory basis exists for a similar reading of the claim-arising language of Sections 1391(a) and (b). Such a reading would satisfy the gap-filling purpose of the 1966 amendment, and at the same time would also give significance to the similarity of language and purpose of Section 1391(e). [17] Sanford involved only a fine of two days pay. Here, by contrast, the individual penalties will average about $1,000, and the total collected from all strikers will exceed $6,000,000. Although the total amounts of the penalties are substantial, plaintiffs knew, or should have known, in advance that penalties would be imposed and would increase each day as the strike continued. Furthermore, nothing in Sanford supports the view that the Court of Appeals regarded the limited amount of the penalty — two days pay — as significant; rather the Court was moved by the fact that the two-for-one penalty was a far less serious sanction than a discharge from employment, as in Arnett v. Kennedy, 416 U.S. 134, 94 S.Ct. 1633, 40 L.Ed.2d 15 (1974). The legislative scheme seems generally reasonable: the longer the strike and the resulting social disruption, the greater the need to penalize the striking employees. In sum, the total amount of the penalty should not, in itself, be regarded as significant. [18] In the present case, some 6,500 employees have filed affidavits challenging the Director’s determination that they were on strike. Similarly, in Sanford, many of the employees (approximately 3,400) filed affidavits protesting the Director’s determination. [19] After oral argument, the court received from plaintiffs’ counsel a telegram, dated September 6, 1979, which reads: PAYCHECKS DELIVERED TO MEMBERS OF THE SECURITY SERVICES BARGAINING UNIT ON SEPTEMBER 5, CONTAINED NO EXPLANATION OF THE COMPUTATION OF PENAL SEIZURE OF WAGES. EXAMPLES OF THE EXTREME HARDSHIP BEING IMPOSED ARE SHOWN BY THE FOLLOWING INDIVIDUAL PAYCHECKS WHICH MUST COVER A TWO-WEEK PERIOD. CORRECTION OFFICER WILLIAM TRUMBULL, 4 DEPENDENTS, EMPLOYED AT CAMP ADIRONDACK, PAYCHECK $3.30. CORRECTION OFFICER DANIEL DE-CHICK, 1 DEPENDENT, EMPLOYED AT GREAT MEADOWS CORRECTIONAL FACILITY, PAYCHECK $20.50. CORRECTION OFFICER ROBERT DIAN-NING, 5 DEPENDENTS, EMPLOYED AT DOWNSTATE CORRECTIONAL FACILITY, PAYCHECK $11.50. CORRECTION OFFICER G. DILLARD, 4 DEPENDENTS, EMPLOYED AT DOWNSTATE CORRECTIONAL FACILITY, PAYCHECK $35.57. WE RESPECTFULLY URGE THAT THE COURT IMMEDIATELY RESTRAIN THE STATE FROM WITHHOLDING MORE THAN ONE DAY’S PAY FROM THE WAGES FOR EACH TWO-WEEK PERIOD. [20] Section 210(2)(g) of the Civil Service Law provides: Payroll deductions. Not earlier than thirty nor later than ninety days following the date of such determination, the chief fiscal officer of the government involved shall deduct from the compensation of such public employee an amount equal to twice his daily rate of pay *218 for each day or part thereof that it was determined that he had violated this subdivision; such rate of pay to be computed as of the time of such violation. In computing such deduction, credit shall be allowed for amounts already withheld from such employee’s compensation on account of his absence from work or other withholding services on such day or days. In computing the aforesaid thirty or ninety day period of time following the determination of a violation pursuant to subdivision (d) of paragraph two of this section and where the employee’s annual compensation is paid over a period of time which is less than fifty-two weeks, that period of time between the last day of the last payroll period of the employment term in which the violation occurred and the first day of the first payroll period of the next succeeding employment term shall be disregarded and not counted. [21] The parties agree that the total penalty to be imposed on the average plaintiff in this case is roughly thirteen days pay (sixteen less three regular days off, which the state no longer intends to count). If that amount were taken proportionately from four consecutive pay checks, somewhat more than three days pay would be taken from the paychecks of most of the plaintiffs. Each pay period includes ten-days pay, so a reduction in each pay period of over three days means a reduction of over thirty percent before taxes. Since the actual reduction is based on net, after-tax income, the penalty the state intended to deduct each pay period amounts to over 50% of take-home pay. [22] The affidavits on file indicate that most of the employees involved have families to support. Some have support and alimony obligations. Many must make monthly mortgage or other payments. Others have pre-existing debts which result in garnishment of part of their wages. These situations, on reflection, are hardly surprising. In a large group of salaried employees one would expect a sizeable number who have substantial financial obligations that necessarily exacerbate the difficulties a fifty percent reduction in wages could be expected to cause. The record already contains evidence that for the first pay period in which deductions occurred, some individuals received paychecks of as little as $3.30 and $11.50. [23] The state argues that the financial hardship of the deductions it intended to implement was ameliorated by a $600 lump-sum payment made to all the plaintiffs in August. This hardly seems a factor upon which a court could rely to cushion the impact of financial deductions commencing in September. About half of the lump-sum payment represented the amount owed on a 7% retroactive pay increase. The other half was a special training stipend. Both payments were negotiated in April 1979, and had been expected by the employees as of May 1979. No doubt many employees, anticipating these payments, planned their expenditures accordingly; which is to say that since guards are humans, many of them can be counted on to have spent the money before getting it. Thus, while the $600 payment (reduced by taxes) should not be disregarded in evaluating the impact of the state’s planned schedule of deductions, the schedule remains onerous, and the extreme hardship that has resulted for many is not surprising. [24] Illegal strikers have no “legitimate claim of entitlement” to salary under Board of Regents v. Roth, 408 U.S. 564, 577, 92 S.Ct. 2701, 33 L.Ed.2d 548 (1972), the state contends, because *219 they accepted the Taylor Law as part of the terms of their employment. But the doctrine that states have power to establish property interests in more recently developed forms of entitlement does not permit states to withdraw from any of its citizens property interests firmly established at the time the Constitution was adopted. A person’s earned salary has always been among the property interests protected by the Constitution. See e. g. Sniadach v. Family Finance Corp., 395 U.S. 337, 89 S.Ct. 1820, 23 L.Ed.2d 349 (1969). In any event, to say that no individual plaintiff has a property claim in wages forfeited by an illegal strike begs the real issue. Sanford establishes that salary can be forfeited, and even that it can be taken without a prior hearing. The issue here is whether the salary can be taken so quickly as to impose separate and severe hardship; whether the deductions are so painful in the circumstances as to constitute or approximate a penal measure, for which a prior hearing would be necessary. If the state’s analysis is accepted, it could deduct 100% of wages for any violation of a contract term, without a prior hearing. Furthermore, Sanford is based on the premise of a hearing after-the-fact. The state in the Taylor Law has not — even if it could have — taken away plaintiffs’ property interest in their wages. [25] If an employee can be suspended without a hearing and without any pay, the state argues, it follows that part of the employee’s pay can be withheld pending a hearing at a later date. Here, again, the issue is not whether some pay can be withheld without a prior hearing, but how much. The analogy is strong, and justifies deference to the extent of allowing some reasonable degree of hardship. But suspension without pay is not an entirely apposite comparison in this context. When suspended, an employee may become eligible for union benefits, welfare, or unemployment compensation; and a suspended employee (even if unprepared to resign) may find other work to provide earnings until the promised hearing is held. Plaintiffs here are ineligible for any benefits, and can supplement their reduced income only by obtaining employment beyond their full-time (and rather demanding) jobs. Some weight must therefore be given to the fact that deductions from the salaries of employees whose services are retained may be so substantial that they can reasonably be expected to suffer greater hardship then employees who are suspended without pay. Due process might well require additional protection to ensure that the facts justify the additional hardship imposed, [26] New York, for example, provides that no more than 10% of gross income regularly received may be garnished by any creditors at any time. See 7B N.Y.C.P.L.R. § 5231. The Federal Consumer Credit Protection Act limits garnishment in most cases to 25% of a person’s disposable earnings. See 15 U.S.C. § 1673. See generally, Note, Consumer Protection — Creditors Rights — Restrictions on Garnishment, 23 Wayne L.Rev. 225, 228 (1976). Indeed, no statute that the parties have been able to find permits the garnishment of over 25% of an individual’s gross income; many limit garnishment to substantially less. See Appendix. This analogy is reinforced by the fact that, to proceed under the garnishment laws, a party must usually have a final judgment against the wage earner. Here, by contrast, the state is proceeding without having granted plaintiffs a hearing on the penalty, and is seeking to extract over 50% of take-home pay. A practice so out of line with the national consensus is suspect, and bears a substantial burden of justification. The state has advanced no basis whatever for the proposition that so drastic a remedy is necessary or any more useful than some more humane but nevertheless exacting schedule for deterring strikes. [27] The state argues that the garnishment analogy should not be applied when the party deducting wages is the employer, since the statutes are designed in part to help employers. It refers to no authority to support this argument, however. The garnishment statutes are as applicable to garnishments by employers as by any other creditors, and the interests in enforcing the statute in the employer-employee context are substantially the same as in other contexts. [28] Prior practice strongly suggests that hearings for plaintiffs will not commence until about six months after the Director’s determination, which was made on or about July 20, 1979. Since over 6,500 employees have requested hearings, it will be a long time after hearings commence before all employees will have had a chance to secure repayment of salary deducted because of the strike. Furthermore, prior experience, including Sanford, indi *220 cates that many employees will succeed in recovering part or all the salary deducted. In Sanford, for example, some 500 teachers proved they were improperly determined to have been on strike. They recovered all the pay they had lost. Many more recovered part of the pay deducted. Though these corrections reflect well on the adequacy of the state’s hearing process, they show that the Director’s determinations as to which individuals were on strike and for how long are not so reliable as to justify the extreme hardship now threatened.