OPINION
McCUNE, District Judge.
This case, before this court on remand from the Court of Appeals,1 presents at this juncture the limited question of whether class certification pursuant to Rule 23 of the Federal Rules of Civil Procedure is appropriate, given the nature of the claims asserted. In our Memorandum and Order of December 17, 1975, we dismissed this action on grounds not here relevant, and stated that even if the action could be brought, class certification would not be proper, due to the unmanageability of the class asserted. Our dismissal was reversed by the Court of Appeals and remanded with directions for reconsideration of the problems of manageability.
This memorandum constitutes the reconsideration directed by the Court of Appeals, with the perspective gained from a hearing on the question of the data available to [517]*517determine the membership of the asserted class. The evidence so adduced and further analysis of the problems of discovery require certification of only part of the asserted class. This class will be composed of claimants in this judicial district only. We remain convinced that any larger class would present problems of management too significant to ignore.
I
The scope of our reconsideration is governed by the remand of the Court of Appeals. That remand freely allows reconsideration of the problems of management, with the caveat that any denial of certification not be based upon speculation. In our December 17,1975 Memorandum, we denied certification because, inter alia, the Federal Set-Off Statute, 31 U.S.C. § 227, would require the government to bring all claims available against plaintiffs as a set-off to any recovery. This requirement, we held, in all probability would cause a class action to become unwieldy, due to the profusion of individual set-off claims which would result. The Court of Appeals rejected this rationale because it was speculative, justifying only decertification after the claims were brought. In effect, it was improper to take judicial notice of the probability that individual set-off claims would be brought. Manageability problems, it was therefore held, must be factual, unavoidable obstacles to the litigation before they can be considered obstacles to class certification.
In addition to rejecting any speculative rationale, the Court of Appeals became aware at oral argument of the possibility that an incomplete examination of the data available concerning the identity of the membership of the class had been made. The parties were directed to the Administrative Office of the United States Courts and to other agencies for potentially computerized information concerning class members. The court was concerned with the possibility that manageability decisions would be based on surmise, and not on “hard data.” As a result, a hearing on the problems of manageability was necessary, and the case was remanded.
II
A hearing on the availability of data was held on December 6,1977. At that time the defendant submitted the results of inquiries made to the Administrative Office of the United States Courts, the Internal Revenue Service, the various Clerks of the District Courts of the United States, and the Legal Management Systems Service Office of Management and Finance of the Department of Justice. The results of these inquiries were largely uniform. No agency could produce the information concerning the identity of class members from any computerized source. All information would, of necessity, be available only as a result of a manual search of records. The only records which would provide the necessary data are the criminal dockets of the various district courts. Because of the age of much of the information, the records are not necessarily complete, and many of the addresses of class members are either missing or no longer correct.
Three conclusions can be made on the basis of the evidence produced at the hearing. First, the discovery of the identity of class members will be arduous and expensive. Second, this discovery will require an extensive search of criminal dockets in each judicial district in the United States. Third, this search will be extensive enough to disrupt the operations of the various district courts if the Clerks are required to make the search and there are no funds available to pay for additional clerks.
These conclusions support a finding that a national class is unmanageable. They also define a manageable class. Since each judicial district controls the evidence which is necessary for a complete determination of this action, each district properly should control the action as it relates to that district. A class action is appropriate in this district only when composed of claimants from this district.
Two additional considerations, though not the basis of this finding, support and underlie it. First, much of the discovery for this [518]*518district has already been completed by plaintiff. Second, although certification cannot be denied on the basis of a possible profusion of set-off claims, those claims may well be brought. Whether they will or will not be brought cannot be known until the membership of the class is determined. Since there is a possibility that decertification may become necessary when these claims are brought, it is best to limit expenditures by the parties at this stage of the litigation. This can only be accomplished by limiting the scope of the class to a manageable number. The claimants in this district present such a number.
Ill
The propriety of the certification of this limited class can be seen through consideration of precedent construing Rule 23. Because this precedent is complex .and at times contradictory, it is necessary to fully review the problem of manageability and its treatment.
Manageability is properly not a separate area of inquiry in consideration of the certification of a Rule 23(b)(3) class, but rather it is one of several problems to be considered. The larger questions are, of course, whether the “questions of law or fact common to the members of the class predominate over any questions affecting only individual members, and that a class action is superior to other available methods for the fair and efficient adjudication of the controversy,” Rule 23(b)(3). Manageability is listed as only one element of the determination of these two somewhat overlapping questions. It is thus important to keep in mind the secondary role manageability plays as a consideration necessary to class certification. Any problems of management must be weighed against the dual considerations of the predomination of common questions (predomination analysis) and the superiority of proceeding as a class action (superiority analysis).
Problems of management have not always been considered in relation to predomination and superiority questions. To some extent manageability has developed as a miscellaneous category, comprised of those problems of administration which arise out of the novelty of large class actions. See, e. g., Stern v. Lucy Webb Hayes National Training School, 367 F.Supp. 536, 540—41 (D.D.C.1973).
Free access — add to your briefcase to read the full text and ask questions with AI
OPINION
McCUNE, District Judge.
This case, before this court on remand from the Court of Appeals,1 presents at this juncture the limited question of whether class certification pursuant to Rule 23 of the Federal Rules of Civil Procedure is appropriate, given the nature of the claims asserted. In our Memorandum and Order of December 17, 1975, we dismissed this action on grounds not here relevant, and stated that even if the action could be brought, class certification would not be proper, due to the unmanageability of the class asserted. Our dismissal was reversed by the Court of Appeals and remanded with directions for reconsideration of the problems of manageability.
This memorandum constitutes the reconsideration directed by the Court of Appeals, with the perspective gained from a hearing on the question of the data available to [517]*517determine the membership of the asserted class. The evidence so adduced and further analysis of the problems of discovery require certification of only part of the asserted class. This class will be composed of claimants in this judicial district only. We remain convinced that any larger class would present problems of management too significant to ignore.
I
The scope of our reconsideration is governed by the remand of the Court of Appeals. That remand freely allows reconsideration of the problems of management, with the caveat that any denial of certification not be based upon speculation. In our December 17,1975 Memorandum, we denied certification because, inter alia, the Federal Set-Off Statute, 31 U.S.C. § 227, would require the government to bring all claims available against plaintiffs as a set-off to any recovery. This requirement, we held, in all probability would cause a class action to become unwieldy, due to the profusion of individual set-off claims which would result. The Court of Appeals rejected this rationale because it was speculative, justifying only decertification after the claims were brought. In effect, it was improper to take judicial notice of the probability that individual set-off claims would be brought. Manageability problems, it was therefore held, must be factual, unavoidable obstacles to the litigation before they can be considered obstacles to class certification.
In addition to rejecting any speculative rationale, the Court of Appeals became aware at oral argument of the possibility that an incomplete examination of the data available concerning the identity of the membership of the class had been made. The parties were directed to the Administrative Office of the United States Courts and to other agencies for potentially computerized information concerning class members. The court was concerned with the possibility that manageability decisions would be based on surmise, and not on “hard data.” As a result, a hearing on the problems of manageability was necessary, and the case was remanded.
II
A hearing on the availability of data was held on December 6,1977. At that time the defendant submitted the results of inquiries made to the Administrative Office of the United States Courts, the Internal Revenue Service, the various Clerks of the District Courts of the United States, and the Legal Management Systems Service Office of Management and Finance of the Department of Justice. The results of these inquiries were largely uniform. No agency could produce the information concerning the identity of class members from any computerized source. All information would, of necessity, be available only as a result of a manual search of records. The only records which would provide the necessary data are the criminal dockets of the various district courts. Because of the age of much of the information, the records are not necessarily complete, and many of the addresses of class members are either missing or no longer correct.
Three conclusions can be made on the basis of the evidence produced at the hearing. First, the discovery of the identity of class members will be arduous and expensive. Second, this discovery will require an extensive search of criminal dockets in each judicial district in the United States. Third, this search will be extensive enough to disrupt the operations of the various district courts if the Clerks are required to make the search and there are no funds available to pay for additional clerks.
These conclusions support a finding that a national class is unmanageable. They also define a manageable class. Since each judicial district controls the evidence which is necessary for a complete determination of this action, each district properly should control the action as it relates to that district. A class action is appropriate in this district only when composed of claimants from this district.
Two additional considerations, though not the basis of this finding, support and underlie it. First, much of the discovery for this [518]*518district has already been completed by plaintiff. Second, although certification cannot be denied on the basis of a possible profusion of set-off claims, those claims may well be brought. Whether they will or will not be brought cannot be known until the membership of the class is determined. Since there is a possibility that decertification may become necessary when these claims are brought, it is best to limit expenditures by the parties at this stage of the litigation. This can only be accomplished by limiting the scope of the class to a manageable number. The claimants in this district present such a number.
Ill
The propriety of the certification of this limited class can be seen through consideration of precedent construing Rule 23. Because this precedent is complex .and at times contradictory, it is necessary to fully review the problem of manageability and its treatment.
Manageability is properly not a separate area of inquiry in consideration of the certification of a Rule 23(b)(3) class, but rather it is one of several problems to be considered. The larger questions are, of course, whether the “questions of law or fact common to the members of the class predominate over any questions affecting only individual members, and that a class action is superior to other available methods for the fair and efficient adjudication of the controversy,” Rule 23(b)(3). Manageability is listed as only one element of the determination of these two somewhat overlapping questions. It is thus important to keep in mind the secondary role manageability plays as a consideration necessary to class certification. Any problems of management must be weighed against the dual considerations of the predomination of common questions (predomination analysis) and the superiority of proceeding as a class action (superiority analysis).
Problems of management have not always been considered in relation to predomination and superiority questions. To some extent manageability has developed as a miscellaneous category, comprised of those problems of administration which arise out of the novelty of large class actions. See, e. g., Stern v. Lucy Webb Hayes National Training School, 367 F.Supp. 536, 540—41 (D.D.C.1973). Because each large class action is commenced very much as a maiden voyage, each presents new procedural difficulties. These difficulties are generally unique, and offer little guidance to subsequent courts. Manageability is frequently an ad hoc problem. Indeed, in response to this, the Court of Appeals has recognized the need to accord substantial discretion to the district courts in certification matters, effectively consigning the development of manageability rules to the district courts. Link v. Mercedes Benz, 550 F.2d 860 (3d Cir. 1977).
Rules, as normally conceived, have not been quick to develop. Generally, the problems of management are merely enumerated. No controlling questions are considered because none have been adequately formulated. Nonetheless, although no rules are as yet developed, problems of management can be classified in a manner corresponding to the two pertinent inquiries under Rule 23, predomination analysis and superiority analysis. When so classified, these problems can be better assessed in light of the policies of Rule 23 and the gravamen of the substantive claims being asserted.
Predomination analysis generally involves questions arising out of the need to individually treat class members’ claims. At issue is whether the problems of individual members significantly divert the focus of the action from the questions of law or fact common to the class. How significant the diversion must be is unclear, however. Certainly, when individual questions are “overwhelming”, the action cannot be maintained as a class action. See, e. g., In re Hotel Telephone Charges, 500 F.2d 86 (9th Cir. 1974) and Cotchett v. Avis Rent-A-Car Systems, Inc., 56 F.R.D. 549 (S.D.N.Y.1972). Additionally, the extent to which individuality will impose overwhelming obstacles to a class certification varies directly with the size of the class. As the size of the class [519]*519increases, less time can be devoted to individual problems. At some point, excessive size alone precludes certification. See, In re Hotel Telephone Charges, supra, and Boshes v. General Motors Corp., 59 F.R.D. 589, 599 (N.D.Ill.1973). Beyond these general conclusions, however, there is little guidance for a marginal case. In an action in which predomination is a close question, it is proper to consider problems of policy presented by the action itself. See 89 H.L.R. 1309 (1976).
The present action may be a marginal case, however such problems need not be considered now.2 At this juncture, there are no significant individual claims. Although each claimant will present a claim which must be proven, proof, given the assumption that no set-off claim is raised, will not be protracted. In addition, this district contains only 500 odd claimants. If the proof of individual damages amounts only to determining the amount of fine paid, the class should be manageable. Even this limited proof could prove unwieldly, however, if a national class potentially 100 times as large were certified. In such an instance, 50,000 individual damage determinations would present administrative burdens approaching those considered in In re Hotel Telephone Charges, supra. See also, Ralston v. Volkswagenwerk, A.G., 61 F.R.D. 427 (W.D.Mo.1973). Given a class limited to this district, however, the predomination requirements of Rule 23 are met.
Superiority analysis, on the other hand, is somewhat more problematic than predomination analysis, both in general terms and in relation to this action. Even the proper focus of inquiry is unclear. The notes to the Proposed Amendments to Rules of Civil Procedure, 39 F.R.D. 73 (1966), contemplate the focus of superiority analysis to be on the question of whether “another method of handling the litigious situation may be available which has greater practical advantages.” This question has not, however, developed as the primary focus of superiority analysis. The notes suggest that a finding of superiority supporting a class action be based on analysis of other potential forms for the action, yet few courts have so treated the question. Because superiority brings into question the administrative problems attendant to a class proceeding, most courts have considered the specific problems presented by the action, basing a finding of superiority on the magnitude of these problems. As a result, superiority analysis has become not an analysis of potential procedural forms of an action, but rather an analysis of the specific problems of a class proceeding.
The most difficult of these superiority problems arise when it becomes necessary to reconcile class action procedure to the requirements of constitutional due process. It is necessary to notify all known class members individually, if at all possible. See Eisen v. Carlisle & Jacquelin, 417 U.S. 156, 94 S.Ct. 2140, 40 L.Ed.2d 732 (1974). In a class action involving a large class, the class representative (or the defendant in appropriate circumstances) may not be able to afford this notice. If substantial, this expense may at times make the question of who shall bear the cost of notice determinative. See Eisen v. Carlisle & Jacquelin, 479 F.2d 1005, 1016-17 (2d Cir. 1973), vacated, 417 U.S. 156, 94 S.Ct. 2140, 40 L.Ed.2d 732 (1974); Appleton Electric Co. v. Advance United Expressways, 494 F.2d 126, 136 (7th Cir. 1974). Even if this question is not determinative, substantial doubt as to the possibility of identification may arise after certification and substantial expenditure. In such an instance, decertification may be necessary, placing an extreme burden on the class representative. Cf. Souder v. Brennan, 367 F.Supp. 808, 814 n. 29 (D.D.C.1973). These results are to be avoided, if at all possible.
This action does not present substantial notice problems. The actual cost of notice will not be overwhelming. What is substantial, however, is the cost of determining the class membership. This discovery requires a manual search of numerous rec[520]*520ords. Defendant urges, on the basis of representations by the Clerks of the various United States Courts, that this search would require well in excess of 20,000 man hours. Who should bear this expense is just as much a problem here as it is in the notice context. Indeed, since decertification may very possibly become necessary, the expenditure of funds for this discovery could become as speculative here as in the Brennan context. The allocation of this expense is thus a very substantial problem.
Plaintiff has completed a search of the records in this district. The expense of this search has certainly been much less than that required for a national search. Since plaintiff has undertaken this work, the expense of discovery need not be otherwise allocated. For this district, therefore, allocation is not a problem. In addition, since the cost of discovery is not yet substantial, it is proper to proceed in the face of possible decertification to consider the class issues in this district. If a larger class were considered, the speculative discovery investment would be substantial, and proceeding as a class action could be unjustified. Discovery and notice cost problems are therefore solved by allowing class certification only in this district.
Notice problems are not the only difficult superiority problems. Other significant problems arise out of the occasional incompatibility of class action proceedings and a judicial system based primarily on a two party, largely local, practice. One such problem is presented here. Information regarding the class members is under the control of the Clerks of the various district courts. These Clerks are not parties and most are outside the subpoena power of this court. Potentially, information necessary to this action could become unobtainable. See Windham v. American Brands, 68 F.R.D. 641 (D.S.C.1975). Even if the Clerks were within the power of the co.urt, it is questionable whether any order affecting them would be proper, as it would necessarily entail an incursion into the affairs of other district courts. This incursion, we believe, is best avoided. Therefore, to avoid unnecessary involvement in the affairs of other courts, it is proper to certify a class from this district only.
IV
In addition to the class certification question, a motion to intervene on the part of fifteen additional plaintiffs is presently before the court. Given our decision on the question of class certification, we see no reason to deny this motion. All intervenors appear to have been convicted in this district.
We cannot comment concerning the tolling of this statute of limitations by this action for parties concerning whom the class certification was denied. Because these parties are not now before this court it would be improper to rule on this issue. The issue can be raised in other districts when and if other claims are brought.
Finally, defendant asserts that the class certified should not include persons who have plea bargained for conviction under one of the tax statutes which are the subject of this action. This argument is based on the reasoning that such convictions rest not only on the tax violation but also on other constitutional charges dismissed pursuant to the bargain. These charges, the defendant maintains, validate the convictions which, on their face, are violative of the Constitution. This question, too, is not presently before the court. This argument is a defense which should be raised when it is found that there are, in fact, plea bargained convictions included in the class. At that time, the defense can be raised, and the certification issue may, if necessary, be reconsidered.