OPINION

MOTZ, Chief Judge.

Presently pending before me is plaintiffs’ motion for class certification. The class representatives seek certification of all liability and damages issues raised by all of plaintiffs’ claims, both state and federal. The motion will be granted in part and denied in part. A class will be certified as to the liability issues presented by plaintiffs’ RICO claims, specifically (1) whether the Honda defendants and Lyon & Lyon are hable under the substantive RICO claims that have survived motions to dismiss, and (2) whether there was a single nationwide RICO conspiracy with a common purpose and, if so, whether the defendants and various other dealers were members of that conspiracy.

Those issues will be tried to a jury. If the jury answers them (or any of them) in favor of the plaintiff class, the trial will proceed to a final judgment as to the damages claims asserted by the representative plaintiffs only. 1 Since this limited issues class certification will not result in the entry of a final judgment in favor of any member of the class other than the representative plaintiffs, the reasons for tolling the statute of limitations as to any class member who has not yet filed suit no longer exist. Therefore, I am further ruling that the tolling period will end ten days after notice is given to the class of the limited issues class certification.

As to the limited liability issues I am certifying for class treatment, the numerosity, the common question of law or fact, and the adequacy of representation requirements of Rule 23(a)(1), (2), and (4) are clearly met and require no discussion. The typicality requirement of Rule 26(a)(3) is likewise *367 beyond dispute 2 Defendants do contend that the predominance of common questions of law or fact requirement of Rule 28(b)(3) is not met. I disagree. Quantitatively, almost by definition there will always be more individual damages issues than common liability issues where the potential class includes, as here, upwards of 800 plaintiffs. Qualitatively, however, this is a case in which the liability issues—concerning alleged wrongdoing over a substantial period of time by many executives of the Honda defendants, numerous Honda dealers, and a law firm said to be involved in an ongoing coverup of the illegal scheme—far exceed in complexity the more mundane individual damages issues. 3

II.

The issue remaining to be considered is whether, as also required by Rule 23(b)(3), “a class action is superior to other available methods for the fair and efficient adjudication of the controversy.” Almost countless pages and hours of argument have been devoted to consideration of this question. In the interest of the timely issuance of this opinion (which is necessary for keeping this litigation on schedule), I will not repeat what has been said on the record as counsel and I have explored the issues but will simply state the reasons I conclude that class action treatment is superior to any available alternative.

*368 A.

I use as my starting point the “test case” approach suggested by defendants. 4 Under that approach the five plaintiffs in the Borman case (the class representatives) would try all of their claims to a final judgment in the first trial. Within months thereafter, assuming that at least some of the defendants lost on at least some of the claims, there would be a series of consolidated trials to determine causation, individual injury, and damages for all remaining plaintiffs in the MDL proceeding. In the second round of trials, the findings made in the Borman trial would have such collateral estoppel/res judicata effect as found appropriate by the judge(s) presiding over those trials.

This test case approach has several advantages. It would result in a final judgment on all of plaintiffs’ claims, not simply their RICO claims. It would involve a traditional lawsuit familiar to our professional experience and comfortable to our legal training. It would allow the dispositive issues to mature through pretrial proceedings and the trial itself rather than having to be formulated when notice of the class certification is given to the potential members of the class. Finally, defendants’ approach would end the tolling of the statute of limitations and require potential plaintiffs to file lawsuits if they want to pursue their claims, thereby helping to quantify plaintiffs’ potential recovery and defendants’ potential financial exposure—a step necessary to the progress of just and effective settlement negotiations.

There are, however, three severe disadvantages to defendants’ proposed test case approach. First, the collateral estoppel/res judicata effect of any findings made against defendants in the first trial cannot now be predicted with any accuracy. A judgment adverse to defendants in the Borman trial might simply lead to prolonged debate about what issues were actually resolved by that judgment and the binding nature of specific findings made by the jury. Second, a judgment in favor of defendants in the Borman trial would have no preclusive effect upon other plaintiffs under the test ease approach. Although defendants have, at least implicitly, indicated a willingness to accept this consequence, I will not assume a plaintiffs’ victory in the first trial. Obviously, it would be a waste of judicial resources to have the test case end up having no collaterally preclusive effect at all.

Third, pursuit of the test case approach might well create chaos among plaintiffs’ counsel. If Borman becomes no more than a test case and if all other cases (including those subsequently filed) are of equal weight and stature, at least some class counsel may withdraw their appearances. Even if they do not, their leadership and authority will almost certainly be undermined, at least for some period of time. This is a material concern, not only to plaintiffs but also to the orderly and efficient resolution of this litigation. Lawyers of the highest quality have assumed the role of lead counsel on both sides of this litigation. Battles are fought hard, but they are fought responsibly, “meet and confer” procedures result in the resolution of many disputes and the succinct presentation of those that cannot be resolved, schedules are adhered to, and complex issues are briefed and argued with remarkable skill. The satisfactory progress of the proceedings is attributable to the efforts of lead counsel for all parties. It would be against the public interest for the leadership structure that has developed to be unnecessarily disrupted.

The limited issues class approach does not suffer from these disadvantages. There will be no question in any subsequent proceedings about the binding effect of rulings and findings made in the class action proceeding upon all members of the plaintiff class and all of the defendants. The preclusive effect will *369 be mutual. Further, the limited issues class approach will serve to keep the leadership structure among plaintiffs’ counsel in place. 5

Moreover, the corresponding disadvantages from which the limited issues class approach suffers are less critical. Arguably, from an academic standpoint the most serious of these disadvantages is the fact that the class approach, as I am structuring it, will result in resolution only of plaintiffs’ RICO claims and not of any common law claims. However, as a practical matter, if plaintiffs prevail on their RICO claims, there will be no reason for them later to pursue their common law claims. Further, if the test ease approach were adopted, the significance for the litigation as a whole of findings in favor of the Borman plaintiffs on their common law claims would probably not be substantial. Given the factual particularities surrounding an individual plaintiffs claims and the legal variations depending upon which state’s law is to be applied, there might be virtually no collateral estoppel or res judicata effects from the jury’s findings.

As for the issue of attorney/judge “comfort level,” the trial under the limited class certification approach will not be far different than it would be under the test case approach. Of course, many complex issues will have to be decided by the jury but this would be the same under either approach. Whether the plaintiffs are proceeding individually or as class representatives, the same liability questions must be formulated and resolved, and the damages phase of the trial under the limited issues class approach will be limited to the Borman plaintiffs’ individual claims. Similarly, although I would prefer to defer formulation of the issues until a later stage of the proceedings, I have confidence that the issues can now be defined with sufficient generality to give fair notice to the class of what is to be decided in the class action phase of the case while preserving sufficient flexibility under Rule 28(d) to formulate the issues more narrowly as the day of final reckoning approaches. After all, the distinction between what is to be decided on a classwide basis and what is not is really quite clear to anyone who does not have an inclination to obfuscate. The liability, classwide phase will involve the questions of what was done by whom, whether the conduct was unlawful, and, if so, whether any of the defendants are legally responsible for it. The second, non-classwide phase (assuming it is reached) will involve the questions of whether the unlawful conduct caused damages to particular dealers and, if so, how much.

III.

The remaining disadvantage from which the limited class certification suffers would, unless it is overcome, be of immense practical importance. Until and unless a reasonable estimate can be made of plaintiffs’ potential recovery and defendants’ potential exposure, settlement negotiations are unlikely to be fruitful. 6 Defendants assert that no such estimate can be made until dealers who have not yet filed suit are placed under a deadline to do so since, according to defendants, the preliminary economic model employed by those plaintiffs seeking class certification on injury-in-fact and damages issues is far removed from reality. Defendants contend that the class representatives are asserting claims on behalf of what defendants characterize as “phantom plaintiffs.”

The class representatives disagree with defendants’ assessment of their economic model, and the merits of the dispute will have to be litigated in due course. I am fully persuaded, however, that the existence of the *370 alleged phantom plaintiff problem is a genuine obstacle to settlement negotiations. It became clear during the course of oral argument on the class certification motion that several of the major direct action plaintiffs who have already filed suit would opt out of a class certified for damages as well as liability issues. The reason is fundamental. The preliminary economic model used by the class representatives is based upon the contention that the financial effects of defendants’ alleged unlawful conduct were national in their scope and that dealers, wherever they are located, suffered relatively similar damages (depending upon the size of their dealership). The direct action plaintiffs who have indicated that they would opt out, on the other hand, are of the view that although there was (as all plaintiffs allege) a single national conspiracy with a common purpose, the effects of that conspiracy varied regionally and that dealers in some regions suffered substantially greater relative damages than did dealers in other regions.

Obviously, if an important subgroup of plaintiffs were to opt out of a class certified for damages as well as liability issues, maintenance of a class action would not be an effective management tool. Moreover, to permit one group of plaintiffs to proceed with a class action premised upon a “national” damages theory while permitting other plaintiffs to proceed with individual claims premised upon a “regional” damages theory might well undermine one of the fundamental purposes of Rule 23—to protect against the risk of inconsistent adjudications.

This is not to say that if defendants (or some of them) ultimately lose on the liability issues, a mechanism for grouping plaintiffs in the damages phase of the litigation will not be devised. However, because elasswide certification is now clearly inappropriate and because of the vast uncertainty of existing damages estimates, whatever catalysts can be added to the mix to enable the parties to assess their respective risks and recoveries will be of benefit to all of them. Reasonable settlement negotiations depend upon what is perceived as reasonably reliable data from reasonably reliable sources.

The absent members of the plaintiff class constitute one such source. Honda dealers themselves are in a singularly good position to know if defendants’ alleged misconduct (which has received widespread publicity) caused them damages. The success of their business turns on the allocation of automobiles they receive and their ability to earn profits by the sale and service of those vehicles. There is no reason to hypothesize or speculate about whether they were injured by what is alleged to have occurred. If they believe they were, they need only file suit and ask that the harm done to them be remedied. Although one of the purposes of class actions is to protect potential plaintiffs with small claims, here the members of the plaintiff class are knowledgeable businesspeople who have been sufficiently successful to obtain what the record suggests are highly prized Honda franchises. It is not asking too much of them to decide—before the liability issues have been resolved—if they believe they have viable claims and, if so, to file a lawsuit.

I recognize that this course requires dealers who have not yet filed claims to consult with their lawyers. However, the barrier to entry into the MDL and limited issues class proceedings is extremely low; new plaintiffs can borrow heavily from complaints that have already withstood the test of motions to dismiss. Even as to this limited work, the legal fees incurred by new plaintiffs will be shifted to any defendant against whom they ultimately prevail on their RICO claims. Once they have filed suit, they can choose not to opt out and be represented by class counsel through the liability phase of the proceedings. In the meantime their presence (or absence) may prove to be an important factor during the settlement negotiations between the parties. 7

*371 rv.

I recognize that my ruling on the limitations issue may be said to run afoul of the Supreme Court decisions in Crown, Cork & Seal Co. v. Parker, 462 U.S. 345, 103 S.Ct. 2392, 76 L.Ed.2d 628 (1983), and American Pipe & Construction Co. v. Utah, 414 U.S. 538, 94 S.Ct. 756, 38 L.Ed.2d 713 (1974). Those cases have come to stand for the black letter rule that the pendency of a class action suspends the tolling of limitations until the termination of the class action proceedings.

What makes the present case different from Crown, Cork & Seal and American Pipe is that in both of the latter cases a class action had been pending during the tolling period as to all aspects of the absent plaintiffs’ claims. In contrast, here I am ruling that the tolling period ceases only after I have rendered my decision to certify the class solely as to liability issues and not to certify it as to damages issues. In my view this difference is critical. As the Supreme Court recognized in Crown, Cork & Seal, the rule adopted in American Pipe served “[t]he principal purposes of the class action procedure—promotion of efficiency and economy of litigation.” 462 U.S. at 349, 103 S.Ct. at 2395. If tolling were not permitted “[t]he result would be a needless multiplicity of actions—precisely the situation that Federal Rule of Civil Procedure 23 ... [was] designed to avoid.” Id. at 351, 103 S.Ct. at 2396.

This reasoning is absolutely correct in the context of a proceeding in which all dispositive issues have been certified for class treatment. But it fails where only limited, not fully dispositive, issues have been certified under Rule 23(c)(4). Indeed, rote application of the Crown, Cork & Seal and American Pipe rule in the context of such a ease may well frustrate efficiency and economy of litigation by postponing the filing of multiple actions which, instead of being “needless,” will be necessary to the ultimate resolution of the MDL proceedings.

That certainly is the case here. By virtue of my ruling today, if the plaintiff class succeeds in the liability phase of the litigation, the institution and prosecution of individual lawsuits by the class members (other than the representative plaintiffs, whose damages claims will be submitted to the jury in phase one in order to accomplish an appealable final judgment on all issues resolved in that phase) will be required. 8 If limitations have been tolled in the interim so that these suits have not yet been filed at the completion of the liability phase, valuable time will be unnecessarily lost while the parties and the court wait to see how many new plaintiffs come forward. The pace and momentum of the proceedings will be dissipated, perhaps never to be recovered. Clearly, this manner of proceeding would promote neither the efficiency nor the economy of the litigation.

I will enter an order effecting the rulings made in this opinion after conferring with counsel.

Notes

. When I refer to "damages issues” in this opinion, I intend to include issues relating to causation, injuiy-in-fact, and extent of damages.

. Defendants contend that plaintiffs have failed in their threshold obligation to define the class adequately. Plaintiffs initially defined the class as "Honda dealers in the United States who owned Honda dealerships at any time between 1975 and 1992. Excluded from the Class are Defendants and Co-Conspirators." This definition is too open-ended. The determination of all the persons who plaintiffs (or defendants) allege to be co-conspirators cannot simply be left to be resolved at trial. See Harris v. General Development Corp., 127 F.R.D. 655, 658 (N.D.Ill.1989) (noting requirement that class members be ascertainable before action may proceed). However, this problem can be avoided by redefining the class to make it administratively feasible for any dealer’s class membership to be resolved prior to trial. For the reasons that follow, the class will be defined to exclude defendants, various dealers (identified by name) who plaintiffs have a good faith basis for alleging were co-conspirators, and, subject to further exploration of the issue with counsel, various dealers (again identified by name) who defendants allege were co-conspirators, provided that the alleged conspiracy existed at all.

Plaintiffs have now provided defendants a list of 158 dealers, in addition to the dealer defendants, who plaintiffs allege, based upon what they have learned during the discovery period, were co-conspirators. I will require plaintiffs to provide defendants (if they have not already done so) with the evidence upon which they base their allegation as to each of these 158 dealers. If defendants contend that there is not a good faith basis for plaintiffs to make their allegation as to a particular dealer, they may challenge in pretrial proceedings the non-inclusion of that dealer in the plaintiff class. Conversely, if plaintiffs obtain evidence between now and the time of the submission of the pretrial order that there are other dealers as to whom there is a good faith basis for alleging that they were co-conspirators, they should move to have those dealers excluded from the plaintiff class. I will resolve any dispute concerning whether there is a good faith basis for excluding any such dealers prior to trial. Whether any of these dealers were in fact co-conspirators will be determined by the jury at trial, not for the purpose of determining whether they are class members (since they will have been excluded from the class prior to trial) but for the purpose of making findings that will be relevant at the subsequent damages phase of the proceedings when the number of dealers found to have participated in the conspiracy may impact the number of cars found to have been misallocated.

According to defendants, there is another group of dealers who must also be excluded from the class. This group consists of dealers (including one of the present direct action plaintiffs) who the class action plaintiffs apparently have not alleged were co-conspirators but who defendants do allege paid bribes to American Honda executives. This group can be handled in exactly the same manner as that outlined above. If there is a good faith basis for defendants’ allegations, any such dealer can be excluded from the class and his status determined at trial. However, if any such dealers choose to opt out of the class, it may be in the interest of the defendants (and the court system) to have them participate in the class phase of the trial as parties with the jury determining whether or not they were co-conspirators. I will confer with counsel about this issue.

. As I find that the liability issues in the case as a whole predominate over the individual issues as required by Rule 23(b)(3), I need not decide whether the predominancy inquiry is altered for a limited issues class under Rule 23(c)(4)(A) Compare Castano v. American Tobacco Co., 84 F.3d 734, 745 n. 21 (5th Cir.1996) (requiring that case as a whole satisfy predominancy requirement) with In re Tetracycline Cases, 107 F.R.D. 719, 727 (W.D.Mo.1985) (relaxing predominancy standard for certification of individual issues under Rule 23(c)(4)(A)).

. Of course, the burden is upon plaintiffs to demonstrate that class action treatment is superi- or, not upon defendants to devise a more suitable litigation approach. However, defendants have acted quite responsibly in suggesting something positive rather than simply poking holes in what plaintiffs propose. I might add that their responsibility is also good litigation strategy. Their proposal highlighted questions that helped lead me to conclude that a class certification more limited than that originally proposed by plaintiffs is appropriate.

. It has been suggested from plaintiffs’ side that my failure to certify the class for damages purposes and my ruling that the tolling of limitations will cease when notice is given of my decision to certify a limited issues class only may itself have an adverse effect upon the relationship among plaintiffs’ counsel because these rulings may reduce plaintiffs’ potential claims. The short answer to this concern is that if the underlying economics of the case do not support the claims being made on behalf of the class, that is something which should be known earlier rather than later in the proceedings so that the parties can properly assess their positions.

. Settlement, of course, is not a desirable end in and of itself. Some wars must be waged until there is a clear victor, regardless of cost and collateral injury. However, if settlement will result in a just, cost-effective, and constructive resolution of the matters in dispute, its social and economic worth is self-evident.

. Plaintiffs have suggested that limitations will not expire as to some dealers until after January, 1999, when the trial on the RICO liability issues is scheduled to begin. That may or may not be so. Further, I recognize that if plaintiffs are successful in that trial, there may be some dealers who may thereafter seek to file suits challenging my ruling that the tolling of limitations ceases within ten days after notice of the ruling is given to potential class members. However, if *371 dealers believe that they have viable claims, it makes no business sense for them to defer filing suit, thereby running the risk that their claims ultimately will be found untimely because of their delay. Therefore, cessation of the tolling of limitations is likely to put to an effective test defendants’ theory that there are a large number of “phantom claims" among the absent members of the plaintiff class.

. I recognize that reasonable people might disagree with my judgment that a damages class should not he certified. However, that judgment stands separate and apart from my ruling that limitations should not be tolled, and criticism of the former does not detract from the logic of the latter. In that regard, I note that defendants contend that issues relating to RICO’s injury-in-fact requirement cannot be decided on a class-wide basis. Cf. Windham v. American Brands, Inc., 565 F.2d 59, 71 (4th Cir.1977) (en banc) (finding classwide determination of injury inappropriate in Sherman Act case because injury and damages inquiries are intertwined). I am not now so holding. However, if that contention is correct, it underscores the reason why the Crown, Cork & Seal and American Pipe rule should not apply in the context of limited issues class certification. It clearly would be senseless to continue to toll limitations and delay the filing of suits in a situation where class certification of liability issues is appropriate but where class treatment of damages issues is improper as a matter of law. Waiting until the liability issues have been decided before stopping the tolling of limitations would serve no useful purpose since by definition individual damages suits would ultimately have to be filed.