image description

The Class Action Fairness Act: A Wolf in Wolves' Clothing

by Jay L. Himes
Class Action Litigation Report |
Federal court jurisdictional statutes can perhaps be likened to high-tech computer software. But what about CAFA procedure? That's not rocket science-more like plumbing.

The Class Action Fairness Act was designed to remedy "a perceived increase in plaintiffs' class actions filed in state courts," writes attorney Jay L. Himes in this Analysis & Perspective.Himes examines interlocutory appeals of federal court remand orders, and "the intersection between CAFA-covered class actions and cases that State Attorneys General bring, typically, on behalf of victimized consumers." The author then "step[s] back and comment[s] on CAFA's broader implications."

The Mischief to Be Remedied by CAFA

Congress enacted the Class Action Fairness Act (CAFA)(1) in 2005 in response to a perceived increase in plaintiffs' class actions filed in state courts, based on state law.(2) CAFA proponents argued that state courts were more favorably inclined to grant class certification than were their federalcounterparts-even willing to certify classes of victims nationwide. These state courts, CAFA proponents asserted, were inappropriate forums to handle cases of "national importance."(3) Indeed, CAFA proponents identified various individual state courts, which were sometimes referred to, in technical jargon, as "judicial hellholes."(4) To fix this, Congress greatly expanded the opportunity for federal court jurisdiction over class actions that are based solely on state law claims. I will, first, summarize what Congress did, and how the statute operates to produce federal jurisdiction. Then, I will address the matter of appealing federal court remand orders, and the intersection between CAFA-covered class actions and cases that State Attorneys General bring, typically, on behalf of victimized consumers. Finally, I will step back and comment on CAFA's broader implications.

The Statute Summarized

Here, in a nutshell, is how the CAFA jurisdictional provisions work:

  • CAFA requires only minimal diversity of citizenship.(5) If any class member and any defendant are citizens of different States, that's minimal diversity, and that's enough to trigger federal jurisdiction.(6)

  • CAFA also permits aggregating individual class member damages for jurisdictional purposes.(7) So long as the total for all class members exceeds $5 million, that suffices for jurisdiction.(8)

  • Removal procedure was also made easier by eliminating the requirement that all defendants consent to removal,(9) and by eliminating the one-year removal deadline that otherwise exists.(10)

The federal jurisdiction thus created is available:

► to plaintiffs, who may file their state law class action in federal court, or

► to defendants, who may remove if the case is begun in state court.

CAFA also has various exclusions-actions against state officials, for example, or actions covered under the federal securities laws, which have their own rules.(11) It further has what are sometimes called the "home state" and "local controversy" exceptions, which permit or require the court to remand a removed action back to state court under various conditions.(12)

In addition, CAFA defined something called a "mass action," over which the district courts also may assert jurisdiction. A mass action exists where "monetary relief claims of 100 or more persons are proposed to be tried jointly" by reason of common questions of law or fact.(13) However, the "mass action" jurisdiction reaches "only . . . those plaintiffs whose claims . . . satisfy the [diversity] jurisdiction amount requirement,"(14) currently $75,000.

Congress thought this sort of litigation, which "involve[d] a lot of people who want[ed] their claims adjudicated together," often produced "the same abuses as class actions."(15) Thus, according to Congress, "mass actions" were "simply class action in disguise."(16) In our environment of heightened security, we can all appreciate the risks of permitting class actions in disguise to roam the halls of state courthouses throughout the country. Therefore, Congress brought mass actions within the law's reach as well.

Let us, then, take a look at what our elected officials in Washington, D.C. did to help fix the mischief that they believed state court class actions had wrought.

The Mischief Created by the Law

A. Cases of "National Importance," an Underpinning for CAFA

First, I can accept that some class actions do present matters of national importance, and that they are, arguably, suitable for federal court litigation. Recognizing that, assume with me the following facts:

(1) Scam Auto, a new and used car dealer, is a Connecticut corporation, with its principal place of business in Connecticut.

(2) Scam Auto's largest retail location is in Providence, Rhode Island, however.

(3) Scam Auto runs a sales program, and-true to its name-it places false and misleading ads in various Providence area newspapers.

(4) Lots and lots of Rhode Island residents, much more than the 100 necessary for CAFA, are duped into buying cars at Scam Auto's Providence store, and are injured by the company's fraud.

(5) The injury suffered by all these Rhode Island residents exceeds $5 million-CAFA's threshhold.

(6) Several of these victimized Rhode Islanders bring a case in Rhode Island state court on behalf of a class consisting only of Rhode Islanders who bought at Scam Auto's Providence location.

If Scam, Inc. wants to litigate this case to the federal district court for the District of Rhode Island, it can.Scam Auto can remove. And nothing in CAFA authorizes-much less compels-the Court to remand.That is because the potentially relevant removal exceptions require, for remand, that a defendant be a citizen of the State in which the action is brought. Here, the corporate defendant, Scam, Inc., is not a citizen of Rhode Island, where the state class action was brought.

Pre-CAFA, this is, of course, a Rhode Island state court case-pure and simple. Although there is diversity of citizenship, no class member sustains injury that reaches the necessary $75,000 jurisdictional amount,(17) and the individual injuries cannot be aggregated in these circumstances. CAFA, however, treats this as if it were a case of "national importance" that simply must be litigated in the national courts if that is the corporate defendant's wish.(18)

This is not a law professor's exaggerated hypothetical.The Sixth Circuit, in a recent 2-1 ruling, upheld CAFA jurisdiction where a class of landowners living in a single county in Tennessee sued a Delaware corporation that operated in North Carolina, across the border from Tennessee.(19)

But granted, these jurisdictional scenarios can get intricate.Federal court jurisdictional statutes can perhaps be likened to high-tech computer software. Nonobvious "bugs" in operation are, therefore, inevitable.But what about CAFA procedure? That's not rocket science-more like plumbing.

Consider remand motions, made after a defendant sued in state court removes to federal court.

B. The Remand Mess

It was important to Congress that CAFA assure the opportunity to litigate cases of national importance in the national courts. Therefore, unlike a run-of-the-mill case, where a district court order on a remand motion is not appealable, an aggrieved CAFA litigant can apply for Court of Appeals review.(20) And to emphasize the need to get jurisdiction decided promptly, Congress, by law, directed the appellate court to rule within 60 days.(21)

Under CAFA, this appeal motion may be made "not less than 7 days after entry" of the remand order.(22) That sounds fine if you say it fast. But not if you read the words that Congress enacted-not "less" than 7 days. This CAFA language calls for a fixed waiting period, whereas Congress plainly intended to impose an appeal deadline.(23)

Take, as an example, the aggrieved defendant, who is being sent back to state court. That defendant is required to wait 6 days, and may, seemingly at any time after that, file its petition to appeal. Put another way, a defendant ordered back to a state court hellhole can wait until hell freezes over to challenge the order.

What's a "textualist" jurist to do with this one? Judge Easterbrook and his fellow panel members in the Seventh Circuit enforced the statute as written.(24) Five other appellate courts have ruled that the CAFA statute has a typo. "Less" means "more." The defendant is supposed to begin the appeal proceedings within 7 days-and not wait 6 days before doing so.(25)

Not only is CAFA riddled with bugs, its plumbing leaks as well. Unless Congress fixes this, the issue would seem to require Supreme Court resolution. However, think, for a moment, of the attention that this clumsiness has captured-the judges, the lawyers, the academicians and law students, and even the casual bloggers, all laboring over this most careless of mistakes.

C. CAFA and State Attorney General Litigation

Next, I wish to look at the interface of CAFA and cases brought by State Attorneys General. Here, too, we find that things get messy.

Nothing in CAFA creates any special jurisdictional rules for cases where a State Attorney General files suit.But during CAFA's consideration in the Senate, there was concern that the law might preclude a State Attorney General from suing in state court. An amendment was introduced to exclude State Attorney General cases from the definition of the term "class action."(26) Although the Senate rejected the change, during the debates several Senators stated that the amendment was not needed because CAFA was directed to "class actions," and unless an Attorney General invoked class action provisions, the law did not apply.(27) For example, Sen. Hatch stated explicitly that CAFA "applies only to class actions, and not parens patriae actions" by State Attorneys General.(28)

And, so, when the first case came along where a defendant tried to remove an Attorney General parens case in New Jersey, the district court held there was no CAFA jurisdiction.(29) That is not the end of the story, however.

After hurricanes Katrina and Rita, the Louisiana Attorney General filed a parens patriae lawsuit in Louisiana state court, alleging that a group of insurance companies violated state antitrust law by colluding to underpay policyholder claims. The insurers removed under CAFA, and remand litigation ensued.

On the remand motion, the insurers argued what I call the "walks like a duck" theory of CAFA jurisdiction.The Attorney General sought to represent the interests of lots of Louisiana policyholders, and to secure a money recovery for their benefit. Never mind that the Attorney General did not plead class action allegations - but invoked recognized parens patriae authority. This thing walked, quacked and looked like a "class action," and that's what it therefore must be for CAFA purposes.

To be fair, however, the insurers' argument was aided by the fact that the State Attorney General engaged private class action attorneys to appear with him on the complaint. And, those attorneys had already filed class litigation raising essentially the same antitrust allegations in state court, which the insurers likewise had removed under CAFA.

As something of a throw-away point on the remand motion, the insurers also asserted that, if you don't like our "walks like a duck"argument, at the very least the Attorney General was suing for more than 100 persons.So, the insurers argued, maybe this state case was a mass action in any event.

The district court, in a bench ruling, pierced the "parens patriae veil," and held the case to be a class action for CAFA purposes. The Louisiana Attorney General appealed. The Fifth Circuit, in a 2-1 ruling, affirmed.(30) The panel majority decided this was not a "class action," but rather a CAFA "mass action," and hence properly subject to federal jurisdiction. In reaching that result, the majority:

  • Ignored the fact that there was only one case filed, not 100+ identifiable individual claims, as the mass action provisions require.

  • Disregarded the absence of any allegation that any policyholder had sustained damages exceeding the $75,000 required for mass action jurisdiction.

  • Said not a word about the express mass action exclusion for any case where "all of the claims in the action are asserted on behalf of the general public (and not on behalf of individual claimants or members of a purported class) pursuant to a state statute specifically authorizing such action ... ."(31)

Compare this majority appellate ruling to one in the district court in Washington, D.C. There, the court held that mass action exclusion for complaints asserted on behalf of the public meant that CAFA did not apply to an action brought under the District's deceptive practices statute by an individual as a "private attorney general."(32)

Where Are We?

This much is clear: although the state of the economy is producing lay-offs left and right, CAFA is a "stimulus package" for lawyers. We have an explosion of litigation on matters of procedure, directed to such issues as:

► Whether the law applies.

► Who has the burden of proof on establishing federal jurisdiction.

► What standard applies to determining whether that burden of proof has been met.

► Is there an exception that permits remand?

► Or, suppose the same plaintiffs file several state court class actions, each arising from the same course of conduct, but covering different time periods or different defendants. Individually, none of the cases meets CAFA's $5 million threshold. May the Court treat these separate class actions as if they were one case, aggregate the damages alleged, and accept jurisdiction?

Moreover, not just the district courts, but also the courts of appeal get to play regularly in CAFA's procedural sandbox.

Beyond litigation itself, of course, there are opportunities for lawyers and non-lawyers alike to compile data on the operation of the statute, and to offer scholarship dissecting the law, its underpinnings, and its impact.(33) The statute is breeding and nurturing a sub-culture of CAFA junkies.

Stripping Away the Trappings

All kidding aside, however, CAFA's effects go beyond relatively harmless bickering and writing about procedural matters. Plainly, CAFA has driven class actions to federal court. Federal Judicial Center studies show class action diversity filings-comprised of both original proceedings and removed actions-to have at least doubled nationwide after CAFA.(34) The increase nationally masks even larger increases in various Circuits.(35) The increase in original filings is both greater, and more consistent, when broken out on a district-bydistrict basis, than that of removed cases.(36) That suggests that plaintiffs' class action attorneys have figured out that they can either file in federal court to begin with-or will find themselves removed there by defense attorneys.

By shifting jurisdiction to the federal courts, CAFA transfers responsibilities for construing state statutes, private contracts, and even common law doctrines to the national courts - while simultaneously denying the state courts opportunities to develop their own state law. And CAFA accomplishes this in major litigations, particularly ones involving relatively small amounts of damages sustained by consumers.

A good example is the on-going price-fixing antitrust case involving "DRAM," a computer memory component, pending in the Northern District of California. Before CAFA was enacted, indirect purchasers filed pricefixing class actions against DRAM cartel members in various state courts. Post-CAFA, plaintiffs counsel agreed to stay the state court cases, and filed a similar case in the Northern District. In this umbrella federal case, they alleged antitrust, unfair competition, and consumer protection claims arising under the laws of nearly two dozen States.

Congress having assigned the federal courts the task of dealing with these many state law claims, the federal district court is doing so.(37) And there is before the Ninth Circuit an interlocutory appeal from two dismissal decisions, which will review these many rulings on various state's laws-with the prospect of Supreme Court review following after that, at least in theory. This scenario is not unique to antitrust; CAFA extends to diversity cases arising under state consumer protection, tort, environmental, and employment relations laws, to name just a few.(38)

To be sure, there are potential safety values available.For example, a federal court may abstain when presented with "difficult questions of state law bearing on policy problems of substantial public import whose importance transcends the result in the case then at bar."(39) Virtually all states also have a procedure by which a federal appellate court may "certify" to the state supreme court for resolution of unsettled state law questions. The state supreme court must also accept the certification.(40) However, I am skeptical that these procedures can be invoked with sufficient frequency, or with sufficient alacrity, to respond adequately to the transfer of decision making authority that CAFA accomplishes.(41)

In sum, as the DRAM experience illustrates, there is here a serious message going forward. What role will state courts-including state supreme courts-have to weigh in on the meaning of their own state laws? As the "laboratories of democracy,"(42) the states need to have their own judicial branches available to evaluate the experiments.Under CAFA, however, the states look more like kids fiddling with chemistry sets in the basement until their parents find out what they're doing, and then stop it.(43)

Indeed, there is open debate right now over whether the federal courts should construe CAFA to over-rule Erie Railroad Co. v. Tompkins,(44) which directs federal courts to apply state substantive law in diversity cases.(45) Authored by Justice Brandeis, Erie empowered progressive era legislation, which sought to assert state regulatory authority over industry, often in the face of hostile treatment by federal courts crafting federal common law under the 1842 precedent of Swift v. Tyson.(46) Despite clear legislative history disavowing any such intention,(47) these commentators argue that CAFA's overarching purpose is to protect national markets from state interference. To achieve that objective, some maintain, the federal courts need to free themselves of Erie's shackles, and set out to develop national law in the class actions governed by CAFA.(48)

On the other side are those who contend that CAFA is unconstitutional, or that Erie, properly applied, requires federal courts to follow state procedural practices where failure to do so would impair state-created substantive rights, and where no uniquely federal interest justifies the impairment.(49)

Thus, in the current short term, the CAFA bickering is largely about matters of procedure. Long term, however, we can expect litigation over the substantive law applicable to multi-state class actions based on diversity, and over how federal judges go about recognizing and developing that law where there is no pre-ordained state rule of decision.(50) There is tension between the Erie doctrine and CAFA. And whatever CAFA's proponents may have said for the legislative record, it is hard to believe they did not know it.


To conclude, is CAFA a good candidate, in this new administration, for "Change that We Can Believe In"? Emphatically so. But am I optimistic this will happen? I am not for several reasons.

First, the radical transformation of the federal-state role that CAFA threatens is not only hard to measure, but probably more important, currently masked by the law's "procedural" nature. "Procedure" is a good thing for courts, rather than legislatures, to tinker with. Second, CAFA has only been around four years. Whatever strange results it may produce, nothing catastrophic has happened yet to command legislative attention. Finally, there is a sense in which CAFA itself is old wine in new wine bottles. The law reflects the ebb and flow over the role of the federal courts in refereeing the commercial marketplaces of the nation-a struggle that began with the adoption of the diversity jurisdiction as part of the Constitution itself. And one that is not readily susceptible of resolution.(51)


1 Pub. L. No. 109-2, 119 Stat 4 (2005).

2 See generally Edward A. Purcell, Jr., The Class Action Fairness Act in Perspective: The Old and the New in Federal Jurisdictional Reform, 156 U. Pa. L. Rev. 1823 (2008).

3 CAFA, § 2(a)(4)(A), 119 Stat. at 5.

4 We are indebted, for the term itself, to the American Tort Reform Association, which in 2004 released a paper entitled "Judicial Hellholes 2004," available at reports /hellholes/2004/hellholes2004.pdf.

5 CAFA thus departs from the "complete diversity" requirement, established as a matter of statutory interpretation by Chief Justice Marshall's opinion in Strawbridge v. Curtiss, 7 U.S. (3 Cranch) 267 (1806). The Supreme Court held that complete diversity is not constitutionally required in S tate Farm Fire & Cas. Co. v. Tashire, 386 U.S. 523 (1967) (upholding the federal interpleader statute, which requires only minimal diversity).

6 By contrast, Supreme Tribe of Ben-Hur v. Cauble, 255 U.S. 356 (1921), held that to determine complete diversity in a class action, the citizenship of only the named representative plaintiffs had to be compared to that of the named defendants.By ignoring the citizenship of class members, Ben-Hur increased the circumstances in which complete diversity existed, thereby expanding the opportunity for federal jurisdiction.

7 Permitting aggregation departs from Zahn v. International Paper Co., 414 U.S. 291 (1973), which required each individual claim to exceed the amount-in-controversy requirement in order to satisfy diversity. Zahn meant that most state law-based class actions had to be litigated in state court. That state of affairs prevailed until Congress enacted the supplemental jurisdictional statute, 28 U.S.C. § 1367, in 1990. In Exxon-Mobil v. Allapattah Services, Inc., 545 U.S. 546 (2005), the Supreme Court held that Section 1367 effectively overruled Zahn. Thus, so long as one named plaintiff satisfied the amount in controversy, the district court could exercise supplemental jurisdiction over the other class members' claims, even though they were below the jurisdictional amount.

8 28 U.S.C. § 1332(d)(2).

9 28 U.S.C. § 1453(b).

10 28 U.S.C. § 1453(b).

11 28 U.S.C. § § 1332(d)(5)(A) & (9)(A).

12 28 U.S.C. § § 1332(d)(3) & (4).

13 28 U.S.C. § § 1332(d)(11)(B)(i).

14 Id. See 28 U.S.C. § 1332(a).

15 S. Rep. No. 109-14, Sen. Comm. on the Judiciary, 109th Cong., 1st Sess., at 47 (Feb. 28, 2005) ("Senate Report"). More specifically, "[o]ne impetus for this fiction was that Mississippi did not have class actions, but large numbers of cases could be joined and treated in many ways as class actions." Class Action Fairness Act and the Federalization of Class Actions, 238 F.R.D. 504, 518 (2007) (footnotes omitted).

16. Senate Report, supra n. 16, at 47.

17 28 U.S.C. § 1332(a), currently $75,000.

18 See generally Stephen B. Burbank, T he Class Action Fairness Act of 2005 in Historical Context: A Preliminary View, 156 U. Pa. L. Rev. 1439, 1527 (2008); Justin D. Forlenza, Note, CAFA and Erie; Unconstitutional Consequences, 75 Ford. L. Rev. 1066, 1103-04 (2006).

19 Freeman v. Blue Ridge Paper Products, Inc., 551 F.3d 405 (6th Cir. 2008).

20 28 U.S.C. § 1453©(1).

21 28 U.S.C. § 1453©(2). See also Senate Report, supra n. 16, at 49 (the appellate review provisions are designed "to develop a body of appellate law interpreting the legislation without unduly delaying the litigation of class actions"). The CAFA provision also permits the appellate court to grant itself one 10-day extension "for good cause shown and in the interests of justice," or "for any period of time" if all parties agree to the extension. 28 U.S.C. § 1453© (3).

22 28 U.S.C. § 1453©(1).

23 Senate Report, supra n. 16, at 49 ("subsection 1453© provides discretionary appellate review . . . but also imposes time limits. Specifically, parties must file a notice of appeal within seven days after entry of a remand order").

24 Spivey v. Vertrue, 528 F.3d 982 (7th Cir. 2008). To cap the maximum period for moving at 30 days, Judge Easterbrook read the CAFA provision with Rule 5(a) of the Federal Rules of Appellate Procedure. Rule 5(a) directs that a petition for permission to appeal "must be filed within the time specified by the statute or rule authorizing the appeal or, if no such time is specified, within the time provided for by Rule 4(a) for filing a notice of appeal." The difficulty here is that a "time" is "specified" in the CAFA provision. It just never stops running. Judge Easterbrook finessed away the difficulty by construing the "time specified" language to include the circumstance "when there is no limit." Id. at 985. See generally Adam N. Steinman, "Less" Is "More"? Textualism, Intentionalism, and A Better Solution to the Class Action Fairness Act's Appellate Deadline Riddle, 92 Iowa L. Rev. 1183 (2007) (arguing for this approach). Professor Steinman contends that, although the CAFA provision specifies a time "after" which a motion to appeal may be made, it does not specify a time "within" which to move. Therefore, Rule 5(a)'s default provision applies. Id. at 1232, 1234-35. In other words, he construes the word "within" to require a beginning and an end period.

25 See Pritchett v. Office Depot, Inc., 420 F.3d 1090, 1093 n.2 (10th Cir. 2005) (concluding that the statute contains a typographical error); Amalgamated Transit Union Local 1309, AFL-CIO v. Laidlaw Transit Servs., Inc., 435 F.3d 1140, 1146 (9th Cir.) (same), rehearing en banc denied, 448 F.3d 1092 (9th Cir. 2006) (six-judge dissent, characterizing the decision to disregard CAFA's words as "an abuse of our judicial power," id. at 1095) See also Morgan v. Gay, 466 F.3d 276, 278 (3rd Cir. 2006) (statute as written contains a typographical error); Miedema v. Maytag Corp., 450 F.3d 1322, 1326 (11th Cir. 2006) (application of plain language would lead to absurd results); Estate of Pew v. Cardarelli, 527 F.3d 25, 27-28 (2nd Cir. 2008).

26 151 Cong. Rec. S1157-58 (Feb. 9, 2005). The amendment was directed to those States where the attorney general lacked parens patriae authority, and thus sometimes sued as a class representative on behalf of state residents. See id. at S1159.

27 See 151 Cong. Rec. S1161-62 (remarks of Sen. Cornyn: "no power of the State attorney general is impeded by virtue of S.5, or will be once it is signed into law") and S1162; id. at S1163 (remarks of Sen. Grassley: S.5 "will not affect" those "very unique attorney general lawsuits authorized under State constitutions or under statutes"; "they do not fall within this definition" of "class action"), and S1163 ( parens patriae actions "are not class actions"). See also id. at S1157-59 (remarks of Sen. Pryor and letter of 46 State attorneys general); S1160 (remarks of Sen. Specter), S1161 (remarks of Sen. Carper), S1164 (remarks of Sen. Pryor).

28 151 Cong. Rec. S1164 (remarks of Sen. Hatch: "[t]his statutory definition . . . applies only to class actions, and not parens patriae actions").

29 Harvey v. Blockbuster, Inc., 384 F. Supp. 2d 749, 753-54 (D.N.J. 2005). See also Antitrust Modernization Commission, Report and Recommendations 272 (Apr. 2, 2007).

30 Louisiana v. AllState Insurance Co., 536 F.3d 418 (5th Cir. 2008). See also Georgene Vairo, CAFA and AG Suits, Nat.LJ., Nov. 24, 2008, at 25, col. 1 (criticizing the ruling sustaining jurisdiction); Louisiana v. AAA Insurance Co., 524 F.3d 700 (5th Cir. 2008) (CAFA reaches an Attorney General case that alleges both class action and non-class action claims).

31 28 U.S.C. § 1332(d)(11)(B)(ii)(III).

32 Breakman v. AOL LLC, 545 F. Supp. 2d 96 (D.D.C. 2008).

33 See, e.g., Adam N. Steinman, "Less" Is "More"? Textualism, Intentionalism, and A Better Solution to the Class Action Fairness Act's Appellate Deadline Riddle, 92 Iowa L. Rev. 1183 (2007) (50-page article, with 313 footnotes, written on the "remand time" issue, outlined above).

34 See generally Emery G. Lee III & Thomas E. Willging, The Impact of the Class Action Fairness Act of 2005 on the Federal Courts - Fourth Interim Report to the Judicial Conference Advisory Committee on Civil Rules 6 - 9 (Fed. Jud. Center Apr. 2008) ("Fourth Interim Report").

35 See Terry Carter, A Step Up In Class, ABA Journal-Law News Now (summarizing FJC data) (May 2008), available at print/.

36 Fourth Interim Report, supra n. 15, at 10-11.

37 See In re Dynamic Random Access Memory (DRAM) Antitrust Litigation , 516 F. Supp. 2d 1072 (N.D. Cal. 2007), and 536 F. Supp. 2d 1129 (N.D. Cal. 2008). Besides these private class actions, dozens of states have also sued in federal court. State of California v. Infineon Technologies AG, No. C 06-4333 PJH (N.D. Cal.) (multi-state complaint); New York v. Micron, Inc., No. C 06-6436 PJH (N.D. Cal.) (action by New York, transferred from the S.D.N.Y. under 28 U.S.C. § 1407).

38 As a example of a non-antitrust case presenting similar multi-state issues, see Ronat v. Martha Stewart Living Omnimedia, Inc., 2008 WL 4963214 (S.D. Ill. Nov. 12, 2008) (declining to certify classes in an action alleging sale of a defective product sold nationwide, and thereupon dismissing the case as lacking subject matter jurisdiction under CAFA).

39 Colorado River Water Conservation Dist. v. United States, 424 U.S. 800, 814 (1976) (explaining Louisiana Power & Light Co. v. Thibodaux, 360 U.S. 25 (1959)).

40 See Eric Eisenberg, Note, A Divine Comity: Certification (At Last) in North Carolina, 58 Duke L. J. 69, 71 & n. 13 (2008) (noting that only North Carolina and Missouri,whose certification procedure has been held unconstitutional, lack this mechanism).

41 See id. at 79 & n.70 (summarizing data showing that state supreme courts answered, on average, 6.6 certified questions during the years 1990-94); William G. Bassler & Michael Potenza, Certification Granted: The Practical and Jurisprudential Reasons Why New Jersey Should Adopt a Certification Procedure, 29 Seton Hall L. Rev. 491, 513-14 & n. 117 (2000) (summarizing data on certification to the Georgia and Florida Supreme Courts, and in the Ninth Circuit) & n. 119 (reporting a poll of federal judges finding that certification was used "sparingly").

42 New State Ice Co. v. Liebmann, 285 U.S. 262, 311 (1932) ("It is one of the happy incidents of the federal system that a single courageous state may, if its citizens choose, serve as a laboratory; and try novel social and economic experiments without risk to the rest of the country") (Brandeis, J., dissenting).

43 See Thorogood v. Sears Roebuck & Co., 547 F.3d 742, 745 (7th Cir. 2008) (noting "the tendency, when the claims in a federal class action are based on state law, to undermine federalism."; citing authorities) (Posner, J.).

44 304 U.S. 64 (1938).

45 See also Rules of Decision Act, 28 U.S.C. § 1652, which provides in pertinent part, that in diversity actions, "[t]he laws of the several states ... shall be regarded as rules of decision in civil actions in the courts of the United States."

46 41 U.S. (16 Pet.) 1 (1842). Until overruled by Erie, Swift authorized the federal diversity court to apply the substantive law thought to be appropriate, without regard for state law, thereby permitting federal common law to develop. The result was to permit dual systems of law to develop whose application turned on citizenship. See, e.g., Black & White Taxicab Transfer Co. v. Brown & Yellow Taxicab Transfer Co., 276 U.S. 518 (1928) (upholding corporate reincorporation to secure the benefit of favorable federal law).

47 See, e.g., Senate Report, supra n. 16, at 49 ( "the Act does not change the application of the Erie Doctrine"); see id. at 61 (CAFA "is . . . a procedural provision only. As such, class action decisions rendered in federal court should be the same as if they were decided in state court-under the Erie doctrine, federal courts must apply state substantive law in diversity cases"), 66.

48 See, e.g., Suzanne Sherry, Overruling Erie: Nationwide Class Actions and National Common Law, 156 U. Pa. L. Rev. 2135, 2142 (2008): What CAFA has done is redefine the relevant community. When it comes to national-market activity, the relevant community is now the nation rather than the individual states. Accordingly, the substantive principles of justice ought to be federal as well. Geoffrey C. Hazard, Jr., Has the Erie Doctrine Been Repealed by Congress, 156 U. Pa. L. Rev. 1629, 1648 (2008) (CAFA's "enactment might invite reconsideration of various applications of the Erie doctrine in light of congressional policy").

49 See Adam N. Steinman, What Is the Erie Doctrine?(And What Does It Mean for the Contemporary Politics of Judicial Federalism?), 84 Notre Dame L. Rev. 245, 249-53, 299-306, 327-30 (2008); see id. at 330 (suggesting that CAFA's expansion of federal judicial power may be justified by "the uniquely federal need to ensure uniform standards for nationally marketed goods or to avoid state interference with national markets"; footnote omitted). Cf. Brown v. Western Railway of Alabama, 338 U.S. 294, 297 (1949) (rejecting a strict state law pleading requirement in a state court case asserting a federal claim : "Strict local rules of pleading cannot be used to impose unnecessary burdens upon rights of recovery authorized by federal laws").

50 See, e.g., Edward A. Purcell, Jr., The Class Action Fairness Act in Perspective: The Old and the New in Federal Jurisdictional Reform, 156 U. Pa. L. Rev. 1823, 1925-26 n. 423 (2008) (noting that, "[i]n the longer term, CAFA may make Erie an inviting target for narrowing or even overruling by the Court-or for preemption by Congress-if pressures for uniform federal law for interstate class actions, or for interstate commercial law generally, become sufficiently great"); David Marcus, Erie, The Class Action Fairness Act, and Some Federalism Implications of Diversity Jurisdiction, 48 Wm. & Mary L. Rev. 1247 (2007) (arguing that CAFA betrays Erie's notion of proper governance because it empowers federal judicial preferences that interfere with the federalism balance in diversity cases); Samuel Issacharoff & Chatherine M. Sharkey, Backdoor Federalization, 53 UCLA L. Rev. 1353, 1418-20 (2006) (arguing that, while CAFA's short-term battles line will be over class certification, if national class actions are to be recognized at all, long-term CAFA will result in the federal courts creating and developing national law); Justin D. Forlenza, Note, CAFA and Erie; Unconstitutional Consequences, 75 Ford. L. Rev. 1066, 1068 (2006) (arguing that because CAFA "forces federal courts to create state substantive common law," the statute "conflicts with the core constitutional holding of Erie").

51 See generally David Marcus, Erie, The Class Action Fairness Act, and Some Federalism Implications of Diversity Jurisdiction, 48 Wm. & Mary L. Rev. 1247 (2007) (surveying the role of the federal courts in diversity cases, with emphasis on the regime that emerged under Swift and on the circumstances leading to CAFA's enactment).