Wednesday, May 09, 2007

Latest Draft Of ALI Principles Of Aggregate Litigation: Three Steps Forward, Two Steps Back

Regular visitors to our blog may remember that back in February we posted on “31 (or More) Reasons to Watch ALI’s Principles of the Law of Aggregate Litigation.” For reference, here's the link. Basically we didn’t like where this project was headed. We highlighted 31 ways in which the draft principles would have changed current law – all in order to facilitate the increased aggregation of litigation, mostly in the form of more class actions.

Not surprisingly we gave that draft two thumbs down. As defense attorneys, we don’t like aggregated litigation – particularly attempts to try cases on an aggregated basis (there are some efficiencies in aggregated discovery, if done right, that we think are beneficial). For one thing, any time litigation is aggregated, the stakes are raised in direct proportion to the extent of the aggregation. A corporate defendant doesn’t like to bet the company. Claims that on an individualized basis are easily defensible, even so weak as to be ludicrous, become incalculably more dangerous when thousands or millions of them (most of which would never have entered the legal system in the first place) are joined together in one, big mess.

Think we’re exaggerating? Here’s an example, and it’s not even particularly recent.

The Agent Orange litigation is classic cautionary tale of the perils of defending aggregate litigation. There was no way in a million years that the plaintiffs could prove causation in that litigation, not as to product identification (which defendants’ product actually injured them) and not as to their many and varied conditions (exposure to dioxin at the low concentrations involved hasn’t been shown to cause anything). See In re Agent Orange Product Liability Litigation, 611 F. Supp. 1267 (E.D.N.Y. 1985) and In re “Agent Orange” Product Liability Litigation, 597 F. Supp. 740 (E.D.N.Y.1984) (both discussing impossibility of proving causation). Nevertheless, because the trial court insisted on aggregating 600,000 individual “claims,” the defendants settled these meritless cases for over $200 million dollars (back in the mid-1980s when that was real money) because aggregation had so magnified the litigation risk. See In re Agent Orange Product Liability Litigation, 611 F. Supp. 1396 (E.D.N.Y. 1985) (discussing settlement).

The only way to certify a class in Agent Orange was to trash accepted choice-of-law principles by using nonexistent “national consensus” law. Belatedly (after settlement) the Second Circuit thankfully disapproved that ruling. In re Agent Orange Product Liability Litigation, 818 F.2d 179, 182-83 (2d Cir. 1987). But the damage had been done. In fact, if Rule 23 had permitted interlocutory appeals of class certification back then (instead of just extremely limited mandamus review), perhaps the entire Agent Orange litigation debacle could have been nipped in the bud.

As it was, the only way the Agent Orange defendants were willing to settle was to purchase “peace” at the expense of many thousands of unsuspecting veterans who although they hadn’t even been injured yet, conceivably could be in the future (under virtually open-ended causation standards). Unfortunately for the defendants, there’s absolutely no way to give effective notice people who aren’t injured and thus have no reason to pay attention to litigation they don’t think involves them. Whether such “futures” claim even constitute constitutional “cases or controversies” is questionable. Perhaps inevitably, the settlement got overturned more than a decade after the fact as a gross violation of due process. Stephenson v. Dow Chemical Co., 273 F.3d 249 (2d Cir. 2001), aff’d by equally divided court, 539 U.S. 111 (2003). The defendants, of course, never got their money back.

Thus for claims that never had any business being in court for causation reasons, Westlaw shows 107 separate “Agent Orange” entries in the district court (between 1979 and 2005), 22 in the Second Circuit, and a “History” that’s over three Westlaw screens long. This 26-year real-life version of “Bleak House” could not have occurred without improvident (and unprecedented) decisions to aggregate claims. Agent Orange is a poster child for why defendants don’t like aggregation. It turns molehills into mountains.

Enough of this tangent, you say. OK. Back to the ALI’s project. After we panned what was called in ALI-speak “Preliminary Draft No. 4,” the ALI’s reporters (the folks in charge of preparing this sort of thing) came out with a revised version on April 6, 2007 called “Discussion Draft No. 2.” Again, we’d like to post a public link to it, but we can’t because the ALI sells this stuff to support itself – and as loyal ALI members, we want the organization to prosper. So go here and purchase a copy if you’d like.

Our most fundamental disagreement with the current draft remains the same. It’s the Draft’s calculated attempt to increase the amount of aggregated litigation by changing the rules to eliminate almost every existing legal impediment. The Reporters take the position that aggregating litigation is simply a matter of degree, and we disagree. Changing the size of litigation by anywhere from five to ten orders of magnitude creates a change in kind, not just degree.

Aggregating litigation is a lot like astrophysics – if enough stuff piles up in one place, there isn’t just a pile of stuff anymore – it becomes a star. That’s what happens when litigation is aggregated into huge piles. Worthless claims have value. Everything becomes worth fighting about. Details get lost. What should have been resolved in six months ends up taking 26 years.

Thus we really don’t like §§2.03 and 2.04 of the Draft, which give a green light to unlimited piecemeal aggregation in litigation that as a whole is unsuitable for such treatment. It resolves – in favor of a minority position – a dispute over how issue certification works. Even the reporters recognize that currently issue certification under Rule 23(c)(4) operates only in a “more limited” fashion, “within the larger constellation” of the entire matter at suit. §2.04, comment b. The Draft would change that, by adopting a “common issues” test focusing on only whether aggregation will “materially advance the disposition of multiple civil claims.” See Draft §2.03, comment a. That test changes the nature of Rule 23(c)(4) dramatically, by making it potentially applicable to any “common issue” no matter how many other individual issues remain.

The Draft would thus turn the prevailing view of Rule 23(c)(4) on its head. Currently, issue certification is rare – it’s mostly in the nature of a “carve out.” That is, if an action is on the whole suitable for class certification, but has some individual issues, Rule 23(c)(4) permits certification of only the common issues. The Draft would say “don’t look at the entire action at all.” Rather, any “common” issue that’s significant enough to meet the vague “material advancement” test would allow aggregation, even if there are five times as many remaining individualized issues that would make the entire action uncertifiable. Issue certification thus becomes a loophole in the predominance requirement big enough to drive almost any class certification motion through.

Still don’t believe us? Well, it’s time for another example – this time from what we know best, the prescription drug/medical device mass tort. In one of the notable improvements to this Draft the Reporters added §1.02, comment w, which recognizes that:

[T]he class action has fallen into disfavor as a means of resolving mass-tort claims. This development reflects many factors, including concerns about the quality of the representation received by members of settlement classes, difficulties presented by choice-of-law problems, and the need for individual evidence of exposure, injury, and damages.

See also Reporters’ Notes to §1.02, comment ac (noting that in toxic exposure cases, pre-trial consolidation but not “a combined trial” is “desirable). Having so described mass torts in §1.02, however, the Draft does not reiterate the same mass tort point in the substantive sections dealing with settlement, choice of law, or individualized issues.

Well, everyone who litigates mass torts knows that almost inevitably (unless there’s an obvious statute of limitations issue), the complaint will include a claim for breach of warranty. But look at Illustrations 9 and 12 in §2.04 on issue certification. What’s the “issue” that’s suitable for aggregation? None other than whether a product is “merchantable.” The Draft thus invites mandatory aggregation back into product liability mass torts through the back door at the same time it concedes up front in comment w that mandatory aggregation of mass torts is “disfavored.”

The Draft continues to use the “carve at the joint” metaphor from In re Rhone-Poulenc Rorer, Inc., 51 F.3d 1293 (7th Cir. 1995) (e.g., §2.03, comment c & Reporters’ Notes) to justify widespread issue certification, even though Rhone-Poulenc used that phrase not in support of issue certification, but to caution against bifurcating a trial in a way that would have different juries examining the same issue in violation of the Seventh Amendment. 51 F.3d at 1302-03. Ironically, the Draft proposes overruling Rhone-Poulenc on precisely this point. See §2.11, Reporter’s notes to Comment a; §2.04 Reporters Notes to comment b.

Overall, the Draft remains almost as dangerous to defense interests now as it ever was. While the former frontal assault on the concepts of predominance and manageability has been toned down somewhat – given the frosty reception that received at the last ALI general meeting – this is a matter of form rather than substance in the nature of a tactical retreat. Although no longer overtly advocating the overthrow of predominance, the Draft continues to advance issue certification as a means of reducing predominance to the same perfunctory status as commonality and typicality already receive.

With that said, here’s a run through of what the new Discussion Draft 2 has – or more commonly, has not – changed about the 31 instances of pro-aggregation alterations that we discussed in our prior post. We also highlight some other changes to the new draft:

  1. §1.01, Illustrations [new]: – Although, as discussed, the Draft now acknowledges that aggregation of mass torts is “disfavored,” the first three new illustrations of who are “parties” involve a mass tort – asbestos personal injury. If mass torts are a “disfavored” form of aggregation, then it stands to reason that there should not be any more mass tort illustrations than absolutely necessary.
  2. §1.03, Reporters Notes to Comment b: – At first we thought the Reporters had taken out the language about “consciously breaking” with existing terminology (the language of Rule 23, primarily) that used to be in §1.01. No such luck. It’s just been moved. At least the comment no longer makes any specific mention of predominance, but the Draft’s continuing hostility to predominance as “overly formalistic” is still manifest in §1.02, Reporters’ Notes to comment u.
  3. §1.05: – The “objects” of aggregate proceedings are essentially unchanged, and remain stated almost entirely from the plaintiffs’ point of view, with next to no consideration of the perspective of those opposing the aggregation. Do “claimants prefer larger recoveries to smaller ones”? Reporters’ Note to comment a. They sure do – just as defendants prefer the opposite. Are there “massive frauds in the business sector”? Id. Sure are – but there are equally massive frauds in aggregated fen-phen and silica litigation. The problem is that §1.05 consistently ignores the defense position. The description of “objects” in this section remains perhaps the most blatant example of pro-aggregation bias permeating the Draft.
  4. §2.03, Comment a: – As we just finished discussing, the circularly pro-aggregation test for determining when aggregation is appropriate remains. A rather revealing discussion of “marketability” added, stating that, “in situations of low marketability in the absence of aggregation but also low variation between individual claims, procedural law has long regarded aggregation as important, as a practical matter to the vindication of private rights of action.” In plain English, what the Draft is really saying is that that supposedly procedural aggregation operates substantively – as a difference in kind, rather than degree – to create value in what are otherwise worthless (“unmarketable”) claims.
  5. §2.03, Comment b: – We won’t repeat our discussion of issue certification. Suffice it to say that the prior draft’s position is effectively unchanged and would still effectively abolish predominance sub silentio.
  6. §2.03, Comment d [new]: – Actually this isn’t new, we just missed it before. The draft would prevent defendants from avoiding aggregation even by conceding common issues. The Draft gives litigation “efficiency” as its overriding “goal.” E.g., §1.03, comment a. Well, there’s no more efficient way to dispose of an issue than to have it conceded. The Draft’s hostility towards concessions establishes that the goal of encouraging ever more litigation trumps even the goal of efficiency.
  7. §2.03, Comment e: – This comment continues to endorses “creative” procedural arrangements, but at least no longer mentions “trial plans.” Without actual trials, trial plans are just more, often fanciful, pieces of paper. We’re still not happy with the exhortation to “creativity,” though, since creativity in the aggregation context has a history of running roughshod over defendants’ procedural and substantive rights. The history of aggregate litigation demonstrates that procedural “advances” have the same traffic generation effect as advances in highway construction. Build it and they will come – the number of suits filed will expand to fit any more streamlined procedures.
  8. §2.04: – The circular “material advancement” test is still here as well. Although not changing substance, this Draft is a little more balanced in its illustrations, having added three new situations where aggregation is inappropriate.
  9. §2.04, Illustration 4 (formerly 1): – This illustration, which gives a single-point pollution case as an example of appropriate class certification, remains. Actually, it’s even worse than we previously thought, since “the defendant’s negligence” has been added as a “common” issue. Some defendants might well be willing to concede that issue in some cases of this nature in order to avoid aggregation – but as discussed in Point 6, the Draft would prohibit such concessions.
  10. §2.04, Illustrations 9 and 12 (formerly 6 and 9): – These are the merchantability examples discussed above. They are inconsistent with mass tort class actions being “disfavored.”
  11. §2.04, Reporter’s Notes to Comment b: – For once they heard us. The pre-Ortiz/Amchem cases that the prior Draft gave as examples of issue certification have been deleted. Now, if they would only conform to existing law and change the substance of their treatment of issue certification to conform to existing law.
  12. §2.05, and Illustrations 1-5: – They heard us again. Medical monitoring is no longer presented as the paradigm of an “indivisible” remedy warranting no-opt-out, mandatory class certification. Instead there are now five illustrations about medical monitoring, four of them stating that either aggregation, or at least mandatory aggregation, is inappropriate. That’s a good start, but all the new permutations leave us wondering why medical monitoring should be discussed any longer. If it’s an indivisible remedy at all (a proposition we reject) then that’s so only in limited circumstances with lots of exceptions. Unfortunately, medical monitoring is substantively such a hash that trying to generalize about it only serves to confuse matters. This section would be better off if it never mentions medical monitoring.
  13. §2.05, Illustration 4 [new]: – This illustration really addresses whether a medical monitoring claim has substantive merit, since whether monitoring will in fact help “mitigate disease” is an element of the claim in most of those relatively few states that allow medical monitoring as an independent tort. Because it goes to the merits of the claim it’s not really an appropriate illustration here.
  14. §2.05, Reporter’s Notes to Comment b: – We were heard again. The Draft no longer states that an “indivisible” remedy is one where “claimants do not simply receive money that they might use in whatever way they choose.” We don’t even particularly disagree with the Draft’s discomfort with the existing law/equity dichotomy. We just draw the line differently. We’re like Gertrude Stein – a payment is a payment is a payment. As long as the effect of a claim is that “the defendant must fund” something, then it should be treated as a claim for money damages, no matter how many “court-supervised” bells and whistles a plaintiff may seek to add in an attempt to get Rule 23(b)(2) certification. By tying the result to the type of relief that the plaintiff alleges, the Draft leaves the door wide open for manipulative pleadings.
  15. §2.06, Comments a & f and Reporters’ Notes: – They heard us but made this aspect of the Draft worse. We pointed before that all current precedent requires proponents of aggregation to prove their case. In order to cite something in support of the Draft’s proposal to shift the burden of proving choice-of-law problems to aggregation opponents, the Reporters’ Notes now references a footnote in a full faith and credit case that has nothing to do with class certification. We’re used to that kind of thing in opposing briefs, but consider it unworthy of the ALI. This stretch only illustrates how dramatically contrary to existing law this proposal to make defendants affirmatively prove conflicts really is .
  16. §2.06, Comment c: – This comment still gives an unwarranted bow in the direction of the novel argument that a defendant’s principal place of business should control choice of law. In the few months since our last post, we count four more decisions just in the drug/device area that have rejected the principal place of business argument. See Rowe v. Hoffman-La Roche Inc., 917 A.2d 767, 775-76 (N.J. 2007); Kelley v. Eli Lilly & Co., 2007 WL 1238789, at *2-3 (D.D.C. April 27, 2007); Bearden v. Wyeth, ___ F. Supp.2d ___, 2006 WL 4474723, at *2-5 (E.D. Pa. May 5, 2006); Devore v. Pfizer Inc., 2007 N.Y. Misc. Lexis 1247 (N.Y. Sup. Mar. 14, 2007).
  17. §2.06, Reporter’s Notes to Comment b: – The current draft omits the reversed post-Ortiz/Amchem case we mentioned before from its discussion of permissible “patterns in substantive law” approach to choice of law. That’s good, but otherwise the discussion is unchanged. As a result of the deletion, the Draft now relies entirely on doubtful pre-Ortiz/Amchem precedent.
  18. §2.08, Comment j and accompanying Reporter’s Notes: – The Draft’s advocacy of overturning Lexecon Inc. v. Milberg Weiss Bershad Hynes & Lerach, 523 U.S. 26 (1998), is unchanged. We’re agnostic on the point, see here, but it’s another example of the Draft’s continued pro-aggregation posture.
  19. §2.08, and Reporter’s Notes to Comment i: – The draft continues to advocate deciding punitive damages on an aggregate, classwide basis. After Philip Morris USA v. Williams, 127 S. Ct. 1057 (2007), that has to come out, since Williams expressly recognized that defendants are constitutionally entitled to raise all individualized defenses prior to being hit with punitives. We've already discussed that here.
  20. §2.08, Comment k: – This is unchanged. As discussed in our prior post, defendants’ due process rights remain merely an “admonition,” while plaintiffs’ due process rights are enshrined in mandatory, black, letter language. Such positional bias does not become the ALI.
  21. 2.09(A)(3) [new]: – This is something else we hadn’t focused on in our first post. The Draft proposes an interlocutory appeal as of right from “any determination of common issues on the merits.” It’s not so much the proposal we question, as the reason for it. It’s the same “issue certification” problem all over again. To us, the fact that an extraordinary form of appeal is seen as necessary is yet another red flag demonstrating that the entire concept of expanding single issue certification is improvident.
  22. §2.09(b): – This is unchanged. In its black letter, the Draft supports allowing courts to order non-consenting plaintiffs in “other proceedings” to pay “common costs” to lawyers they have not retained. The draft now cites a District Court case allowing the practice. That's not very impressive authority, given that the only appellate authority we know of which actually examined this type of arrangement flatly rejected it as a abuse of power. In re Showa Denko K.K. L-Tryptophan Products Liability Litigation-II, 953 F.2d 162, 166 (4th Cir. 1992).
  23. §2.10, Comment a: – The pro-aggregation opt-in class proposal is unchanged, although the Draft now at least acknowledges that a rules change would be required.
  24. §2.11, Comments a and c: – Contrary to current Seventh Amendment precedent, the Draft continues to advocate allowing “pragmatic” re-examination of jury verdicts in multiple-phased aggregated trials. There is no improvement except the express recognition that a change in constitutional law would be required.
  25. §2.12, and Comment a: – The Draft still would give plaintiffs multiple attempts at certifying identical classes in different jurisdictions, by rejecting the collateral estoppel analysis in In re Bridgestone/Firestone, Inc., Tires Products Liability Litigation, 333 F.3d 763 (7th Cir. 2003). However, in §1.04, Reporters Notes to comment b, the Draft recognizes that the plaintiffs’ “contingent fee lawyer is a real party in interest” in aggregate litigation. That’s enough basis for collateral estoppel for us. The only appropriate exception to Bridgestone/Firestone would be where the adverse certification decision was based upon inadequate representation. If representation in the first certification decision was adequate, then the estoppel is justified.
  26. §2.13: – The draft continues to give inordinate weight to pieces of paper – “trial plans” – in making manageability/feasibility determinations. Whether it’s a complaint (medical monitoring) or a trial plan, we’re against giving significant weight to something that either party can simply gin up for its own purposes. If one wishes to determine whether a “polyfurcated” trial is manageable, the logical thing would be to try a case or two before aggregating anything.
  27. 3.02, Reporters’ Note to Comment c [new]: First, we’re confused. There’s a “Reporters’ Note” to comment c, but no comment c. Second, the Reporters’ Note concerns class-action tolling. We don’t like class-action tolling very much, see here, so we’re very leery of anything the Draft might be planning to say about this substantive area of the law.
  28. §3.04 (formerly §3.05): – This Draft still does away with individual notice to class members as too expensive. We don’t mind cheaper alternative means of personal notice, such as email, but if the entire litigation isn’t even worth the cost of complying with Due Process notice requirements, then suit probably should never have been brought in the first place.
  29. §3.06 (formerly §3.07): – This might be better. While the Draft still would allow settlement classes where individual issues predominate – yet another attack on predominance – it now adds a “cohesion” requirement. If “cohesion” is interpreted in accordance with current Rule 23(b)(2) precedent, this is probably an adequate substitute.
  30. §3.07 (formerly §3.08): – The encouragement of “cy pres” settlements – allowing courts to play Santa Claus with other people’s money – is if anything worse in this Draft. Again we’re against de minimis litigation. If it’s not economically “feasible” even to match class members with their supposed losses, than the matter isn’t one that belongs in court. Private litigation is not a panacea, and some things should be left for responsible law enforcement bodies. If there’s extra money because claimants “chose not to participate,” then that’s the same as if they never brought suit in the first place. Logically, that money properly should revert to the defendant. If returning the money to the defendant defeats the “deterrent” effect of some statute, then the money should go to the governmental body that enacted the statute, not to the court’s or the parties’ favorite relevant charities. Cy pres is bad law and worse policy. The judiciary is not intended to be a eleemosynary institution.
  31. §3.09(c) [new]: We missed this the last time around. If a class settlement is rejected as unfair to the class, the Draft would foist part of the successful objector’s counsel fees on the “defendant.” Our objection is simple. Defendants owe no duties, fiduciary or otherwise, to their opponents in litigation. Thus, there’s no legal basis for assessing fees against the defendant. It did nothing wrong. If a proposed settlement is “unfair” to the class – good for the defendant. A defendant’s object in aggregated litigation (as in all litigation) is to pay as little as possible in return for as much benefit as possible. Only counsel for the class owe duties to the class. If class counsel failed their clients, it’s appropriate to sanction them. It’s not appropriate to sanction the defendant because class counsel made a mistake.
  32. §3.09(c) (formerly §3.09(c)): – This problem is fixed. The new Draft applies the same “frivolous” standard for objectors as for any other litigant. That’s fair.
  33. §3.10, (formerly §3.09) Comment b: – The new Draft quite appropriately deals with the “futures” issue in a separate section. That issue is too important to be relegated to a comment. Substantively, we can’t agree with the Draft’s position because we don’t think pure futures claims (as opposed to medical monitoring) are legally cognizable under most states’ substantive law. E.g., Simmons v. Pacor, Inc., 543 Pa. 664, 674 A.2d 232, 237 (Pa. 1996) (increased risk without present symptoms not cognizable); see Restatement (Third) of Torts: Liability for Physical and Emotional Harm Chapter 2 §4 and accompanying comments and notes (Tentative Draft No.5). Even putting aside “case or controversy,” if “fidelity” to existing law is to be taken seriously, then these claims are simply not recognized by the law. Beyond that, as already discussed, “futures” claims pose insuperable notice problems. We don’t think there’s any fair way to settle (or litigate) claims for injuries that haven’t happened yet.
  34. §3.11: – This Draft goes even further than the last one and would make a second right to opt out almost mandatory in class settlement situations. We have mixed feelings about this, but for present purposes it’s another example of a change in the law to facilitate aggregated litigation.
  35. §3.12: – This, too, may well be a good idea (interlocutory appeals from rejected settlements), but it’s another example of changing the law to facilitate aggregated litigation.
  36. §3.18(f): – This is fixed. They heard us (or heard somebody). The proposal to hold settling defendants liable for the ethical lapses of opposing counsel has been deleted.

Of the 36 items just listed, our problems with three (and most of a fourth) have been fixed. But upon further review, we’ve found some other things that bother us, so we have 32 (maybe 33), rather than 31, points where the Draft would change the law to increase the number of class actions and other forms of aggregated litigation. If that means we’ve moved the goalposts, we’re sorry, but topic is too important to apply any sort of a waiver rule (not that we’d ever shut up, anyway).

We shouldn’t close without also acknowledging what we do like better about this Draft. We’ve already said we like the addition of §1.02, comment w. We’re OK with much of the revision of §2.05 on medical monitoring as well.

We also like the discussion of bellwether trials in §1.02, Reporters’ Notes to comment ak. We agree that bellwethers should always be selected in a way that isn’t biased in favor of either side, and we’re sick and tired of judges letting the plaintiffs simply pick the cases they want (and dismiss cases the defendants want) and calling those “bellwethers.” The Draft rightfully condemns such shenanigans, so in that respect it’s a step in the right direction.

Another change for the better is the Draft’s express recognition that state prohibitions against claim splitting limit the extent to which litigation can be aggregated. §2.03, Reporters’ Notes to comment d.

We also approve of the discussion in §2.05, comment a to the effect that there’s not really any difference between Rule 23(b)(1)(A) and 23(b)(2). We’ve taken the same position ourselves in a prior post here.

We’re also gratified to learn that there are apparently at least some class actions so extreme that even this Draft can’t stomach them. We agree with the various statements in the Draft, such as in the Reporter’s Notes to §2.07, comment a, disapproving Dukes v. Walmart, Inc., 474 F.3d 1214 (9th Cir. 2007).

We also support new comment c to §3.06. There’s law holding that, if defendants concede some elements of class certification for settlement purposes, they’re bound by that concession if the settlement falls through. We’ve had to face this conundrum, and we agree with the Draft that such precedent puts defendants in an unfair position and is a significant deterrent to settlement.

No comments: