« The New Foreword | Main | The Good News About the Oklahoma Anti-Sharia Law »
Friday, November 12, 2010
Design v. Rights
So in my previous post I argued that mass torts are better understood as posing a "property" problem, not, as Richard Nagareda and others have argued, a "governance" property. This is not to say that "property" and "governance" are mutually exclusive. Instead, and as I argue in more detail in my current draft, "Mass Torts and Due Process," the governance problems that arise in mass torts are caused by the plaintiffs' individual ownership of their claims.
Specifically, individual ownership leads plaintiffs to invest much less in common issues relative to the defendant. The defendant, who owns all of the liability associated with a common issue, simply has more at stake. Moreover, strategic behavior prevents plaintiffs from voluntarily aggregating to achieve the same economies of scale as the defendant (although "economies of share" is probably a better term). Some plaintiffs are simply better off going alone. Accordingly, a procedure that collectivizes ownership of the claims -- a class action -- is needed.
While I mildly disagreed with the "governance" story, I think Nagareda is absolutely correct that the solution is a matter of procedural "design." Now what do I mean by that? What, exactly, is the alternative? To answer these questions, I want to start with an anecdote, and then address one problem that may arise in class actions - class attorneys "selling out" the class.
Let me start with the anecdote. I recognize that anecdotal evidence is far from ideal, but telling "stories" is unavoidable, even to discuss collective action problems. When I clerked for my district court judge, the judge scheduled a status conference in an MDL involving the pricing of pharmaceuticals (don't ask, it's complicated). The status conference concerned one of the suits within the MDL, a proposed class action of small out-of-pocket payors with some larger insurers as co-plaintiffs. My judge suggested that the class attorneys coordinate their discovery efforts with the co-plaintiffs, and one of the class attorneys responded: "I have no incentive to do any discovery until you cerfity the class!"What's the point of the story? One concern with class actions is that the class attorneys may enter into "sweetheart" settlements with the defendant. Since the class attorneys are usually paid a percentage of the recovery, they stand a lot to gain from settling the class action for pennies on the dollar, a point recognized by Judge Friendly and the Supremes in Ortiz. In fact, one point of the Class Action Fairness Act was to prohibit "coupon settlements" in which class attorneys were paid in dollars while the class only received worthless coupons.
While plausible, the above account of "sweetheart" settlements does not tell the whole story. Why would a class attorney accept a lowball settlement if the claims are worth much more than that? After all, if the attorney is paid a percentage of the recovery, he or she only stands to gain more by trying to extract more in negotiations. But what leverage does a class attorney have? Well, if class actions are hard to certify, a class attorney has the choice of either (1) accepting a lowball settlement or (2) cobbling together some plaintiffs and pursuing the litigation. In my anecdote, the proposed class suffered damages in the tens or hundreds of dollars, so litigating either alone or in partial groups would not have been feasible.
When the class attorneys' alternatives (no class action, individual or partially aggregated litigation) are acknowledged, it is clear why class actions (may) settle for too little, and why the class attorneys in my anecdote balked at discovery pre-certification. Somewhat counterintuitively, the solution to the problem of "sweetheart" settlements entails mandatory (that is, no opt out) class actions that are easier, not harder, to certify for litigation purposes. That way the class attorneys will be incentivized to seek as much as possible for the entire class, and back any claim for more with a threat of litigation as to the whole pot.
Thus, the solution is a matter of rejiggering the incentives of the class attorneys, or, "design." But what is the alternative? The traditional solution has advocated the opposite approach - more chances to "opt out" of the class action rather than making it easier to bring a mandatory class action. While this may protect an individual plaintiff from a "sell-out" class action, it takes away the leverage both the class attorney and the opt-out plaintiff need to extract as much as possible from the defendant. The opt-out plaintiff will be limited by his or her own potential recovery and the lack of economies of scale that would have come from aggregating. The rump class may blow up altogether.
Notice that the traditional solution can be described as "more rights." Generally, if we are concerned with procedures harming individuals, we give them more "rights" to protect their interests. A day in court, say, or an opportunity to be heard, or notice, or, like in this example, an opportunity to opt out of the class. But here the solution is "less rights." It is instead a restructuring of the incentive structure to make sure that the class attorney maximizes recovery, which results in a severe limit on litigant autonomy. This is nothing to sneeze at, since many mass torts plaintiffs who would want to opt out have very valuable claims.
Now, we cannot easily reject the "more rights" approach, because it not only implicates the optimal "design" of mass tort procedure, but due process itself. I will address the due process ramifications of the "design" approach, as well as the problem of internal conflicts in mass torts, in my next posts.
Posted by Sergio Campos on November 12, 2010 at 04:23 PM | Permalink
TrackBack
TrackBack URL for this entry:
https://www.typepad.com/services/trackback/6a00d8341c6a7953ef0133f5cd1808970b
Listed below are links to weblogs that reference Design v. Rights:
Comments
Now I see what you are saying. Let me back up, and note that I am focusing not just an some investment in common issues, but optimal investment.
In any given case, there are investment opportunities, such as how much legal research to do, how much time to dig through documents, or which expert to hire. As to a common issue, or an issue that is the same for each plaintiff, there are certain economies of scale realized the more a party has at stake with respect to the common issue. If the defendant is facing $100M in liability as to an issue of generic causation, it will probably hire a more expensive expert if the marginal cost of the more expensive expert is outweighed by the marginal benefit in reduced expected liability. The plaintiffs, however, either individually or even partially, have less at stake, so their economies of scale are less. If an individual free rides off of legal research or factual investigation (which many do), then the plaintiffs who, in fact, make investment decisions for these things will not be able to capture the benefits of the freerider to offset the costs. Even if there is no free-riding but simply duplicative work (hiring a less good expert twice rather than hiring a better expert), you still get suboptimal investment relative to the defendant.
Now, the problem of "common benefit work," to use a phrase in a Vanderbilt piece by Charles Silver and Geoffrey Miller discussing MDLs as "quasi-class actions," is well-recognized in the mass tort literature. Even if you get some investment in common issues, at least enough to bring some cases, you still get this asymmetry in investment incentives between the plaintiffs and the defendant. Now, you can argue, like Ryan, that the returns on investment are diminishing to the point of being negative (or so lumpy that the next step up in investment is unattainable by the defendant), so maybe some sufficient form of aggregation is necessary. Why a mandatory class action, which is overkill? Well, I don't know what the production function is between investment in common issues and increases in expected recovery. If I knew that, then I would have some clairvoyant powers and quit my job. But if the possibility is there that the returns on investment in common issues is continuous and sufficiently increasing, which, in my experience, is a good assumption, then why not equalize the scale advantage?
Posted by: Sergio Campos | Nov 16, 2010 11:05:41 AM
Sergio, I feel like I must still be missing something, because your story as I understand it isn't internally consistent. If I'm reading your last comment correctly, the commons problem is underinvestment in the collective good of litigating fully, and individual litigation is "defection." But if the good is investment in litigation, isn't going it alone still a form of cooperation? The only real form of defection is nonlitigation: i.e., free riding off of others' litigation efforts. You say that the class member who files an individual lawsuit is depriving other class members of economies of scale. That explanation only makes sense if each individual lawsuit involves duplicative expenses. If so, the "defector" is the one who bears the principal costs of duplication (all she deprives the class of is her share of the potential attorneys fees to finance the suit), so in fact it's rational for her to stay in the class. You avoid that problem by suggesting that perhaps a plaintiff with a 100% claim doesn't want to be lumped in with the rest of a class which has a meaningfully lower probability of success. But if so, why is class treatment proper in the first place? The 100% plaintiffs have (by hypothesis) sufficient incentives to invest individually, which means they don't fit your underinvestment story in the first place.
Posted by: James Grimmelmann | Nov 16, 2010 7:18:20 AM
Ryan, these are too excellent points. As to declining marginal utility of investment, there is some question whether, in a given case, you may hit a point where additional investment leads to negative returns. Thus, if you permit sufficient aggregation, what's the big deal? The problem is that, like a lot of investment in "research and development" in other contexts, you don't know what you are looking for until you find it. So it is difficult to model the production function of investment in common issues to increases in probability of recovery, at least a priori. In fact, the better model would be something along the lines of a "real options" model which looks at discrete investments and the effects of those investments on your future options to invest. But even if investment returns in an individual case maybe diminishing to the point of being negative, why would anyone use this possibility to force the plaintiffs to litigate with one hand tied behind their backs? Remember that the defendant has a monopoly on its liability no matter what. So simply equalizing this scale advantage of the defendant would ensure parity for those cases in which investments are not sufficiently diminishing.
As to your second point, I agree that something like "statutory damages" can be a substitute, but how do you determine the amount? Wouldn't it be easier to simply equalize the scale of ownership between the plaintiffs and defendants rather than guess as to the appropriate amount to subsidize litigation through statutory damages?
Posted by: Sergio Campos | Nov 15, 2010 11:28:41 PM
Very interesting and I'm looking forward to the article. I share some of Professor Grimmelman's skepticism on the points he addresses. A couple of additional questions. First, how does declining marginal value of litigation investment factor into the property "story" you want to tell? For example, assume a collection of mass tort cases where the total value of all claims exceeds $1 billion. In theory, the defendant should be willing to invest slightly less than $1 billion to defeat the claims. But in the real world, the actual amount spent on the litigation will top out at some point far below that figure because, at some point, throwing more money at the case will not materially improve the odds of success. Isn't it possible that in cases involving large-value claims, coordination among plaintiffs and their attorneys can solve the "collective action" problem at least up to the point where marginal values begin to decline and the defendant's investment of additional resources doesn't materially advance its prospects of success?
Second, if this is truly a property problem, aren't there potential solutions that would avoid the types of governance problems described in your post? For example, a rule of substantive law providing exemplary damages to only the first 20 or so successful plaintiffs who litigate their cases through to a successful judgment at trial? Presumably, by that point, other plaintiffs could "free ride" on the investment of the initial plaintiffs by taking advantage of prior discovery, favorable court rulings, collateral estoppel, etc. I'm not sure mandatory class actions are the only, or even the best way for approaching this problem.
Posted by: Ryan | Nov 15, 2010 11:06:45 PM
James, again thanks. Probably too briefly:
As to your first point, yes. I would be happy to get rid of settlement-only class actions.
As to your second point, I am focusing on (a), not on (b). In particular, I am focusing on investment in common issues. Because the defendant has more at stake, it will have an incentive to invest more. Each plaintiff will not, and because such investments are nonexcludable, others can free ride. A plaintiff may defect if their private interest in going alone would be undermined by collective treatment. You can imagine a situation where a plaintiff has a 100% chance of recovering on a significant claim, when joining with the others might reduce that probability. Hence, the plaintiff would rationally prefer to go alone, but this leads to underinvestment in common issues for the class as a whole relative to the defendant.
As to whether this is a commons problem, a commons problem can manifest itself in a situation where there is "underinvestment in a nonexcludable good." Instead of overgrazing, right, one can see it as a problem of undercultivating the land. As I understand the recent "commons" literature, it does not depend on overuse of underinvestment. Rather, it concerns mismatches in scale. Individual rights to graze coupled with collective ownership of the land leads to a commons tragedy. The same dynamic is present in the mass tort context. Here I am relying on work by Lee Fennell, building of Henry Smith.
I hope this helps. There is the commons issue of "limited funds," where there are not sufficient funds to allow full recovery for each member of the class, but that problem is secondary to my concern with investment in common issues. That is what I care about.
Posted by: Sergio Campos | Nov 15, 2010 11:06:15 PM
I take it, then, that you would allow for easier (and mandatory) class certification, but prohibit the settlement-only class? Your point about optimal incentives requires decoupling the contest for control over the class from the substantive negotiations on settlement terms. Very much in the Macey/Miller tradition, then?
I'm still not convinced by your story about the need to make the class mandatory. I see you raising two points, and tell me if I'm disentangling them correctly. On the one hand, your discussion about the "scale" of rights in the lawsuit leads you to claim that class plaintiffs (a) have insufficient incentives to invest because they don't capture spillovers for other class members, and (b) will wastefully duplicate litigation efforts precisely when their investments don't sufficiently spill over, both of which go away if control over the litigation is concentrated in one set of hands belonging to class counsel. This claim doesn't by itself say anything about mandatory classes; indeed, it would seem to predict that mandatory classes are unnecessary, because defectors from class treatment will be opting out of the class's efficiencies. This is a classic collective action problem; I don't see much of a commons or anticommons angle to it.
On the other hand, you're saying that plaintiffs have an incentive to go it alone, undermining the class's benefits. This story is plausible when some class members can free ride on others' investments (so it's the dual of (a) above). In this case, the collective-action problem is exacerbated; you could also frame it as a problem of spillovers and excludability from the nonrival benefits of litigation by the class. I'm still not sure this is a commons or anticommons problem; it's a straightforward problem of underinvestment in a nonexcludable good. I don't see (b) above leading to any commons/anticommons problems either (correct me if I'm wrong). The third place this angle could come from would be if class members can affirmatively blow things up for each other by going it alone. They could deplete the defendant's assets before the rest of the class can get at them (a commons problem, and one very familiar in toxic tort litigation); they could also potentially destroy a settlement at all if the defendant thinks it won't really get global piece (which is more of an anticommons problem of the holdout variety).
Is this a fair summary of the moving pieces?
Posted by: James Grimmelmann | Nov 15, 2010 10:01:06 PM
James and Ohwilleke, thanks for your comments. Let me try to address them briefly. As to the "reverse auction," the concern is that the defendant can effectively shop for class actions to settle their global liability. But a mandatory class action does not perpetuate this problem, and, with other design features, would solve it. Under my proposal, the winner of the race will get to certify a class action that binds all members. Moreover, following Tobias Wolff's suggestion in his Penn Law review piece on anti-suit injunctions, such a class attorney would be allowed to enjoin any other competing class actions that seek to undercut it. Thus, if the class is certified relatively quickly, then class counsel would be incentivized to seek the total expected recovery for the class, since it can threaten a class trial otherwise.
As to the scope of my argument, it applies equally to small-value and large-value litigation. In fact, the need for a mandatory class action in the large-value context is greater, since class members with high value claims will want to defect and pursue claims alone, destroying the economies of scale provided by the class action. As to scholarship on "informal aggregation," I am aware of this literature, and would admit that claim inventories and coordination among plaintiffs' firms have improved plaintiffs' prospects considerably. However, mandatory class actions are better, since it avoids the high transaction costs and strategic behavior endemic to informal aggregation. Think of it this way - informal aggregation is just a manifestation of Coasean bargaining, in which the plaintiffs aggregate their rights to sue to match the scale efficiencies of the defendants. But if that's the case, then it is clear that the assumptions needed to achieve "Coasean irrelevance" as to individual entitlement allocations (no transaction costs, no strategic behavior) are simply not present in the mass tort setting. Judicial intervention in the form of a mandatory class action is needed.
As to ohwilleke's comment, I really like it! I am a big fan of auctioning off rights to bring private suits. It would make the enforcement function of the litigation clearer, and suggests ways in which the "license," to use your term, could be used to coordinate private and public enforcers. Thanks for your thoughts!
Posted by: Sergio J. Campos | Nov 15, 2010 9:26:53 PM
Class actions generally flow from situations where a large institution makes a mistake that is minor in any individual case that by virtue of its scale impact a lot of people. They are a matter of bureaucratic quality control.
Ideally, we would like the discovery of an instance of lapsed bureaucratic quality control to (1) very rapidly stop the mistake from happening in the future, (2) cause the institution to acknowledge the mistake and pro-activity try to remedy it with existing customers or known past customers, (3) provide some opportunity for remedy for unknown past customers, and (4) deprive the institution of a benefit for having erred.
The NTSB and regulatory system organized recall processes for automobiles and food products do a quite good job in their narrow domains at addressing (1) and (2) and (3) and with fines have the potential do deal with (4) as well.
The question posed by private class action litigation is (1) why are regulatory agencies in so many industries failing where the NTSB, autombile recall system and food product recall system largely succeed?, and (2) when would it be and not be more fruitful to try to improve the regulatory system as opposed to relying on private class action litigation.
The story that justifies private class action litigation answers that regulatory failure is a result of regulatory capture, which private class action litigators are less prone to experience, and that a regulatory approach is largely doomed in the long run.
But, there are real costs to a private class action litigation system. Most events that can conceivably give rise to a class action give rise to myriad suits and a lot of what makes that litigation expensive, slow and complex is the process of consolidating those different suits related to the same set of events. This also prompts a race to the court house that rewards the firms that take action as soon as a half-baked case is ready. The mere potential for a large number of parties that can bring suit also prompts firms that could be on the receiving end of them to avoid admissions of fault at all costs, because no one has the authority to settle a suit completely at any point until the class action is settled (and not even then if there are many opt outs).
Perhaps a middle ground would be to auction a limited number of rights to bring class action suits in any given jurisdiction something like seats on the stock exchange or notaries in civil law countries, with bidder qualifications that would require bidders not to have any conflicting financial interests and secured revenues solely from their class action litigation activities (so that the rights to be a private attorney general couldn't be bought by defendants to shut down the entire system), and to provide a swift and routine arbitration system amongst private attorneys-general who have won one of these licenses to quickly establish a single legitimate representative of injured persons in a given incident giving rise to potential class actions.
Businesses would get someone with real authority to deal with them so that they could get the issue behind them promptly, courts and injured parties would get confidence in the competence and good faith of the class representative, and class action suits would get the benefit of a process that could focus on whether there really was collective harm and the merits of the collective issues, rather than having to spend a lot of time consolidating multi-district litigation and second guessing the lawyer for the class.
The fees paid for the privilege of being private attorneys-general could be used to pay for the justice system administrative costs of maintaining the system.
Posted by: ohwilleke | Nov 15, 2010 2:34:22 PM
What about the reverse auction? How do you keep easy mandatory classes from exacerbating the race among plaintiffs' attorneys to file and settle? Relatedly, how much is your reasoning confined to low-dollar-value claims? I thought that one of the profound insights of the late-1990s scholarship on mass torts was that when the individual verdicts were large enough, plaintiffs' law firms could build extensive portfolios of cases and then crank them through in a strategic, self-financing way that recoups investments by increasing the value of the firm's other cases. The class action then enters that picture out of the defendant's desire for global peace and the plaintiffs' attorneys desire to force their rivals from the stage, rather than because plaintiffs' attorneys are under-investing in the suits and need to join forces.
Posted by: James Grimmelmann | Nov 14, 2010 11:35:33 PM
Brad, again thank you for your comments. I'll respond really briefly. First, although net plaintiff recovery is important, I prefer deterrence, which I will explain in my next posts. Second, the point of mentioning the "common fund" doctrine is to show that the class action context is different from the real estate context. As to your substantive point, you are arguing that depending on a class attorney's portfolio of cases (or possible cases), he or she may choose to settle a case cheaply in order to free up time to work on another case. I am skeptical of this argument, but I will say that in the context of negative expected value suits, there is at least one proposal to require filed cases to proceed to summary judgment prior to settlement to avoid strategic use of litigation costs to extract settlements for meritless claims. See http://www.virginialawreview.org/content/pdfs/90/1849.pdf Perhaps this "design" proposal can address your concerns, since it would require a class attorney to litigate a suit all the way to summary judgment, and not just collect cheap settlements.
Posted by: Sergio J. Campos | Nov 12, 2010 11:04:31 PM
I aware that federal law allows for cy pres and injunctive relief to be included in the calculation of legal fees, and take note that ALI which, it must be said is made up of attorneys, has endorsed the concept. But if the concern of your project is the maximization of the total value to the plaintiffs than surely the proper measure of value is from the point of view of the plaintiffs. It is an empirical question whether the courts are correctly calculating the value to the plaintiffs of these non-monetary rewards. Are you aware of any studies which have attempted to answer this empirical question?
As a frequent class member (judging by the number of settlement notices I receive), most recently as a user of GMail I find no satisfaction in cy pres payments to an organization I've never heard of or a consent decree that codifies steps that were already taken when the agreement was signed.
As for the Real Estate example, I'm not sure I understand how the common fund doctrine plays in. Although expenses in a class action settlement are sometimes paid "off the top" attorney hours never are. So each marginal hour an attorney spends increasing the total recovery costs him an hour of his time but only nets him a third of the increase in total recovery. Which leads to the last point. Whether a class attorney is willing to work to squeeze the last bit of recovery from the defendant depends on his alternatives. If his case acquisition costs (sales and marketing) are low than the it may be more profitable to work on a series of cases with 'sweetheart' settlements, because the marginal hour spent on a sweetheart deal is more profitable than the marginal hour spent on a hard fought deal. Perhaps a hypothetical will make this clear:
A firm files a class action against company Y and spends 500 attorney hours getting to the first offer which is $10,000,000. It estimates that with an additional 1500 attorney hours it can get that offer bumped to $20,000,000. So the rate of earnings for the first 500 hours is $6,666 per hour, whereas the rate of earning for the next 1500 hours is $2,222. If the firm has another similar case in the queue, or is confident it can find one, it would be entirely rational to take the first offer and move on.
Posted by: brad | Nov 12, 2010 10:00:56 PM
Many thanks for your comments Brad. As to your first point, CAFA limits attorneys' fees to the value of the coupons actually redeemed, and otherwise subjects coupon settlements to increased scrutiny, so as to avoid the problem you describe. See generally 28 U.S.C. § 1712. As to cy pres payments, they are typically included in the total recovery to calculate attorneys' fees, and the ALI has blessed this procedure. See ALI's Principles § 3.13 cmt. a. As to injunctive relief, although injunctions are difficult to monetize, the value of such relief can be determined for purposes of fees. See id. (noting that percentage fees should be based on "actual value" of "equitable relief."). Thus, my argument would still apply. Moreover, actions involving injunctive relief can be certified as mandatory class actions anyway. See Fed. R. Civ. P. 23(b)(2).
As to your second point, the Real Estate Agents discussion in Freakonomics was meant to highlight the problem of asymmetric information between an agent and a home seller, and, if I remember correctly, the problem of quick sales arose from the fact that the agent bore all of the costs of marketing but only got a small portion of the sale. Keep in mind that under the "common fund" doctrine, the costs of litigation are apportioned to the class as a whole, so the concern with the suboptimal outcomes that may arise where the class attorney bears all costs but only gets a portion of the recovery is not present. In any event, the point of using a percentage fee is to incentivize the attorney to maximize total recovery. It is unclear to me why a class attorney would work less even though he or she could make more money by making further investments.
Posted by: Sergio J. Campos | Nov 12, 2010 9:12:08 PM
In a mandatory class action, the class attorney is partially incentivized to maximize total compensation on which his fee is based. But this total compensation can be comprised of: injunctive relief, cy pres payments to third parties, coupons all of which are often of little to no value to the actual class (where value is measured using the traditional economic metric of willingness to pay) even after aggregation.
Furthermore, just like the Real Estate Agents discussed in Freakonomics, the class attorneys have an incentive to settle quickly and move on since they get paid a fixed percentage rather than on an hourly basis. Half the fee for a tenth or a hundredth of the work is a pretty good deal!
Posted by: brad | Nov 12, 2010 7:08:40 PM
The comments to this entry are closed.