Senior District Judge, concurring in part and dissenting in part:.
I concur in the result reached in this case by the majority, however I have no desire to join in the majority’s issuance of an absolute prohibition against Rule 23(b)(2) defendant class actions. There are respected authorities on both sides of this difficult issue and both schools of thought, in my opinion, do a sufficient job of advocating their beliefs. I believe 23(b)(2) defendant class actions should be reserved for rare sets of circumstances. Yet they are viable for stubborn or neglectful groups of defendants who refuse to make honest attempts to comply with settled substantive law. Let me say at the outset that the Supreme Court had its chance to condemn and prohibit 23(b)(2) defendant class actions on the appeal of Marcera v. Chinlund, 595 F.2d 1231 (2d Cir.1979).1 It didn’t. While I realize a lack of comment by the Court on an issue does not necessarily mean approval, it seems to me inescapable that the practical effect of the Court’s inaction here was to allow such a class action under such circumstances; that is, when attempting to effectuate general compliance with statutory law or uniform, well-recognized administrative policies. I believe under circumstances constituting an “enforcement” setting, certifying defendant classes under 23(b)(2) can, on occasion, pose remarkably few constitutional problems (which could probably be avoided subsequently in the litigation through the wise use of Rule 23(c)(4) and (d)). In their best light, 23(b)(2) defendant class actions can eliminate the perpetuation of substantial injustices.
Allow me to elaborate upon the Marcera case, which serves as an example of how 23(b)(2) defendant class actions can be a liberating force for the federal bench, ridding it of mischievous defendants and festering problems — and actually doing justice for obviously deprived plaintiffs along the way. In Marcera, Chief Judge Irving Kaufman of the Second Circuit astutely— and neatly — put the sheriffs’ departments of New York’s 62 counties in their place. The issue was what kind of “contact visits” should be allowed pretrial detainees in New York state jails. Five years earlier, in 1974, in Rhem v. Malcolm, 371 F.Supp. 594 (S.D.N.Y.), aff'd 507 F.2d 333 (2d Cir.1974), the Second Circuit held that due process mandated that pretrial detainees have the right “to shake hands with a friend, to kiss *418a wife, or to fondle a child.” 371 F.Supp. at 626. The rationale was that pretrial detainees should be considered innocent until proven guilty and that a kiss, handshake or fondle was not too much to ask.
Two years later, in 1976, the issue was still being litigated one penal institution (or county) at a time, each case culminating with the issuance of a decree mandating to that particular county or sheriff what was already known — contact visits must be made available. Also in 1976, the New York State Corrections Commission took the pains to promulgate regulations requiring each corrections facility to implement a contact visitation program. The sheriffs still refused to cooperate, claiming the Commission had infringed on their turf as administrators of the jails.
In November of 1976, two state prison inmates, Joseph Marcera and John Dillon, brought an action in federal district court to enforce their right to contact visitations. Yet this time, on behalf of themselves and a plaintiff class of all pretrial detainees in the state prison system, they sought injunctive relief from a defendant class of 47 sheriffs who all still denied such contact visits in their jails. Sheriff Lombard of Monroe County, where Marcera and Dillon were being held, was designated as representative of the defendant class. At first the district court denied plaintiffs’ requested injunctive relief, stating it believed plaintiffs’ failed to prove a likelihood of success on the merits or irreparable harm. It also declined to certify either a plaintiff or defendant class.
The Second Circuit reversed and instructed that on remand district court should certify the proposed defendant class under 23(b)(2). Chief Judge Kaufman explained the group of 47 sheriffs was large enough to meet the numerosity requirement of Rule 23(a)(1). Questions of law common to the class were also presented to satisfy 23(b)(2). Chief Judge Kaufman emphasized Sheriff Lombard’s defenses were typical of those of the class. Typical defenses were that contact visits would threaten security and increase contraband smuggling within the system. Lombard could not cite a case where the defenses were different. While Lombard contended the court’s expectation of similar defenses was unrealistic since variations in architecture, staffing and inmate population at different detention facilities would lead to different defenses, Chief Judge Kaufman found this reasoning unacceptable. He rebutted that these considerations would be relevant in fashioning various decrees, yet it was settled law that the prospect of minor variations of decrees covering these areas did not justify a blanket refusal of contact visitations. He reasoned individualized relief should not necessarily preclude a court from certifying a defendant class. A subsequent restructuring of the litigation could always be done under 23(c)(4) and (d). As for whether Lombard could adequately represent and protect the interests of the absentee sheriffs, Chief Judge Kaufman noted sheriff Lombard had all-too-effectively vigorously defended his position which was why, two years after Marcera had instituted the suit and five years after Rhem, Lombard was still employing whatever visitation procedure he chose, regardless of federal district court and state correctional commission enunciations on the matter.
Marcera is not a case without its administrative complexities and constitutional soft spots. Chief among them, as Judge Van Graafeiland stated in his thoughtful dissent, is what are the parameters of the term “contact visit” and how can such a visit be effectuated at different jails under different architectural and managerial circumstances? The cost could be high and federal courts could be caught in the “minutiae of jail administration.” 595 F.2d at 1246. Yet as Judge Kaufman implicitly recognized, before' Marcera, by having to handle pretrial detainee suits on a case by case basis, the federal courts were already caught in the “minutiae of jail administration.” Is a protracted battle more desirable? Certainly the interests of justice point toward the 23(b)(2) defendant class certification in this situation.2 Hence when *419a court finds itself in a dilemma as to how to enforce compliance with settled substantive law or promulgated regulations against a group of similarly situated defendants (who may even be acting in concert to flout the law), I believe 23(b)(2) defendant class actions are useful and can present no overriding constitutional problems.
Such a conclusion is not inconsistent with the latest enunciations from the Fourth Circuit on this issue. In Bazemore v. Friday, 751 F.2d 662 (4th Cir.1984) the Fourth Circuit stated:
In this connection, two cases are persuasive, although decided under FRCP 23(b)(2). In Paxman v. Campbell, 612 F.2d 848 (en banc) (4th Cir.1980), we held that a class of defendants of 130 Virginia school boards was inappropriate when each was free to adopt maternity leave policies of entirely unknown differences or similarities, it being uncontradicted there was no statewide policy in force, centrally directed or otherwise. The Sixth Circuit followed Paxman in Thompson v. Board of Education, etc., 709 F.2d 1200 (6th Cir.1983), upon the same fact situation obtaining in Paxman. While those two cases were decided under a different part of the class action rule, the principle remains the same, that to have a proper class of defendants in a case such as this there must be either a statewide rule or practice so that relief is available if the rule or practice is invalid, or the adjudication with respect to a member of a defendant class must as a practical matter be dispositive of the interests of the other members of the class as provided in FRCP 23(b)(1)(B).
Id. at 669-70. (Emphasis added).
Similarly, in Thompson, supra, when the Sixth Circuit held the district court erred in certifying a defendant class under 23(b)(2), it still noted that:
The present case involves approximately 500 separate school districts, each of which was free to adopt its own maternity leave policies. No state statute or uniform administrative policies are involved in the case. The district court acknowledged that “[t]his is not a case where all the defendants are acting under one directive from some superior authority or where there is an obvious binding link among all defendants on this particular policy issue. ”
709 F.2d at 1204. (Emphasis added).
When to certify a defendant class under 23(b)(2) is admittedly a very difficult question to answer. Maybe when in doubt, a court shouldn’t. However, I do not believe this problem, as perplexing as it can be, should lead to an absolute prohibition of 23(b)(2) class actions. A systematic analysis employing rigid criteria would obviously not be desirable. Inherent in the process is a flexible approach focusing on various factors including the degree of complexity (constitutionally, administratively and financially) that would be created by a 23(b)(2) defendant class certification. In Certification of Defendant Classes Under Rule 23(b)(2), 84 Colum.L.Rev. 1371 (1984), a helpful article mentioned by the majority, several guideposts are listed. First, the less heterogeneity in a defendant class, the better. Also, it is preferable that the defendants be “judicially linked” — the members of the proposed class should share some type of a relationship which predates the litigation. In Marcera, the defendants were responsible for complying with or enforcing a particular law long before the litigation and, quite arguably, they were acting in concert to avoid the implementation of the law. This enabled the Marcera court to more comfortably conclude that their defenses would be similar. Sheer numbers of defendants should also be considered. The majority raises a good point when it asks what would stop the instant suit from becoming a nationwide class action on behalf of all welfare applicants in the United States (although there is the personal jurisdiction issue). Further, why *420should one medium sized law firm in Illinois have to defend the interests of hundreds of others? And should the hundreds of others have to depend on that Illinois law firm? These are good points, yet as a longtime district judge I’m more willing to trust the district courts and give them the responsibility and discretion to create and certify defendant classes that will not become constitutional or administrative monsters. A district judge can also use Rule 23(d) to alleviate headaches as the litigation progresses.
In the instant case, I concur in the result because I view the suit as too unmanageable and the remedies too unfocused. It is true that, in some respects, the difference between Marcera and the case at bar is one of degree. But matters of degree are often at the heart of rulings addressing whether to certify a class. In the action at bar there are 770 defendants. All 770 defendants presumably would have to formulate notice and hearing procedures subject to court approval. Various defendants undoubtedly represent populations of various sizes in locales with varied resources. In Marcera the court was concerned with physical visits in rooms. In the case at bar, the concern is the creation of a great number of administrative or quasi-judicial approval boards — arguably 770 bureaucratic systems. Again, with 23(b)(2) defendant class actions, it’s best that when in doubt, don’t. In the instant case there is no inference of a willful defiance of well-recognized law or statewide administrative policy as witnessed in Marcera. I agree with the majority that defendant class certification here is too problematical.
A constant undercurrent throughout the majority’s argument is that allowing 23(b)(2) defendant class actions would result in large numbers of uncontrollable lawsuits with managerial burdens galore. The majority hesitates “to embark on these uncharted and ... perilous seas.” Yet these uncharted seas have been visible since 1966 and there have been few casualties since. Indeed, there is a bit of a contradiction in the majority’s thought process here. On the one hand, it is worried 23(b)(2) defendant class actions would create unwieldy scenarios and great constitutional wrongs, yet on the other hand it admits there have been few 23(b)(2) defendant class actions (45 or so) since 1966 and it uses this low number to conclude their precedential value is insignificant. Hence the majority uses the small number of past 23(b)(2) defendant class actions to doubt their legitimacy, while at the same time saying that if they were condoned the floodgates to constitutional deprivation and administrative chaos would be opened. The reality appears to be that district courts have used the 23(b)(2) defendant class action tool scarcely and responsibly since the 23(b)(2) defendant class action outlet has been available to them. Looking at the problem from this angle, I must conclude that allowing such defendant class actions would not necessarily subject the federal bench to constitutional chaos or judicial disorder.
I see no need to indulge in a detailed word by word analysis of Rule 23(b)(2). As stated earlier, there are respected schools of thought on both sides of this issue and both sides have repeatedly done a remarkable job of proving why the Advisory Committee meant whatever each side wants it to say. There is also the double-edged argument as to why the Advisory Committee hasn’t more definitely stated what it means. The majority would have it that if the Advisory Committee really approved of 23(b)(2) defendant class actions it would have amended the rule by now. I say if the Advisory Committee saw such class actions as a threat, it (or the Supreme Court) could have just as easily ended the debate by now, some 20 years after the creation of Rule 23(b)(2). The arguments in this area on both sides are, at this point, stale and inconclusive.
In summation, I see little reason to dig a. hole and bury a potentially helpful tool for litigants as well as the federal bench until directed to do so by higher authorities. I believe the approach taken by the majority today borders on the simplistic in view of past cases like Marcera. Further, the majority's worries about constitutional (mainly due process) and managerial concerns are *421less based on history and more on an unsubstantiated view of the future. The majority is short on constitutional horror stories from cases where courts certified 23(b)(2) defendant class actions. The majority refuses to reflect upon future possibilities, in view of cases like Marcera, to envision cases where 23(b)(2) defendant class actions are constitutionally and managerially plausible and just. Because I prefer to look at this question in a different light, I am forced to separate myself from the broad scope of the opinion. Therefore, I respectfully dissent from the part of this opinion that expands itself from the denying of the 23(b)(2) defendant class certification as it relates to this case.
. Appealed as Lombard v. Marcera, 442 U.S. 915, 99 S.Ct. 2833, 61 L.Ed.2d 281 (1979).
. Interestingly, while Judge Van Graafeiland expressed concern about the costs and unruliness *419of the result reached by the court, he did not condemn or denounce the viability of 23(b)(2) defendant class actions in general.