MEMORANDUM OPINION AND ORDER
DAVID H. COAR, District Judge.The lead plaintiffs in this multidistrict litigation filed suit in the wake of the CPSC recall of Aqua Dots, a popular children’s toy tainted with the so-called date rape drug GHB. Now that eleven individual cases have been transferred and consolidated for pretrial purposes in this court, see 28 U.S.C. § 1407, the lead plaintiffs propose to represent fourteen Rule 23(b)(3) classes seeking damages and injunctive relief against the manufacturer, distributor, and three major retailers of Aqua Dots. For the reasons set forth below, their motion for class certification is DENIED in its entirety.
BACKGROUND
The Aqua Dots craft kit is a popular toy that allows children to create multidimensional designs by assembling small, brightly colored beads that fuse together when sprayed with water. Various Aqua Dots products are marketed nationwide for children ages four and over at prices ranging from $17~$30. Aqua Dots are manufactured by defendant Moose Enterprises and distributed in North America by defendant Spin Master. Defendants Wal-Mart, Toys “R” Us, and Target are among the two-hundred plus retailers in the U.S. that sell Aqua Dots products.
In early November 2007, the CPSC received two reports of children falling comatose after swallowing large quantities of Aqua Dots. The evident cause of their comas was the adhesive coating on the beads: it was manufactured from 1,4 butanediol, an industrial solvent that the human body converts into the illegal drug gamma-hydroxy butyrate (GHB) upon ingestion, whereas it was supposed to have been manufactured from 1,5 pentanediol, a nontoxic compound found in glue.1 Spin Master, in conjunction with the CPSC, accordingly announced a voluntary recall on November 7, 2007 of approximately 4.2 million Aqua Dots products man*380ufactured for sale in the United States. Of those, 1,335,151 had already been sold to consumers. The CPSA recall notice instructed consumers to take all Aqua Dots products away from children immediately and to contact Spin Master to return them for free replacement beads or for a comparably priced replacement toy. The recall and the safety threat posed by Aqua Dots were widely publicized; many major national news programs, television morning shows, and newspapers such as the New York Times covered the story. Happily for the lead plaintiffs, they have no occasion to allege that their children ingested GHB, or were otherwise injured, by using their Aqua Dots toys.
As word got out, roughly 600,000 consumers returned their Aqua Dots, usually for a refund. Wal-Mart has refunded 154,372 Aqua Dots products; Target, 129,012; Toys “R” Us, 105,376; and other nondefendant retailers, 125,043. Spin Master has reimbursed the retailers for them outlays, and although it has provided only 66 refunds directly to consumers, there is no evidence that Spin Master has ever denied a cash refund to a consumer who asked for one. As the recall notice indicated it would, Spin Master has provided free replacement beads — rebranded under the name Pixos, after a roughly six-month delay — to 77,793 consumers, and it has exchanged 2,918 Aqua Dots products for replacement toys of comparable value. In all, refunds have been provided for 513,869 Aqua Dots products and remain available to this day: any consumer who returns an Aqua Dots toy to one of the defendant retailers, or who contacts Spin Master and provides adequate proof of purchase, can still obtain a cash refund for any Aqua Dots product subject to the recall. Not one lead plaintiff, however, has ever tried to return an Aqua Dots toy to Spin Master or to the store from which it was purchased.
Instead, the lead plaintiffs seek certification of fourteen proposed Rule 23(b)(3) classes, each consisting of consumers “who purchased or received one or more Aqua Dots toys, that were manufactured and/or distributed by Defendants subject to the recall announced by the CPSC on November 7, 2007.”2 The proposed classes fall into three general groups:
(1) A nationwide class alleging violations of the reporting requirements under the Consumer Product Safety Act;
(2) Nine single-state classes of consumers in Florida, Oklahoma, Pennsylvania, Texas, Illinois, New York, Kentucky, New Jersey, and California, asserting state-law claims under express and implied warranty theories, unjust enrichment, and state consumer protection statutes;3
(3) Four multistate unjust enrichment subclasses, each purportedly composed of states with materially identical requirements, as follows (emphasis given to transferor states):
Subclass # 1: Colorado, Connecticut, District of Columbia, Hawaii, Indiana, Iowa, Michigan, Mississippi, Nebraska, New Hampshire, Vermont, and West Virginia;
Subclass # 2: Alabama, Alaska, Georgia, Idaho, Illinois, Kansas, Kentucky, Maine, Nevada, New Mexico, Oregon, Rhode Island, South Carolina, South Da*381kota, Tennessee, Virginia, Wisconsin, and Wyoming;
Subclass # 3: Maryland, Missouri, New Jersey, Ohio, Utah, and Washington; and
Subclass # p. Arizona, Arkansas, Delaware, Louisiana, Massachusetts, Minnesota, Montana, North Dakota, and Texas.4
The lead plaintiffs do not seek certification of a personal injury class. They do, however, request punitive and/or treble damages, as well as declaratory and/or injunctive relief. These requests appear to be made on behalf of all classes.
CLASS CERTIFICATION STANDARD
The proposed class representatives bear the burden of establishing that all requirements of Rule 23(a) and one of the requirements of Rule 23(b) are met. Oshana v. Coca-Cola Co., 472 F.3d 506, 513 (7th Cir.2006); see Fed.R.Civ.P. 23(a)-(b). Rule 23(a) permits certification of a class only if (1) the class is so numerous that joinder of all members is impracticable; (2) there are questions of law or fact common to the class; (3) the representative parties’ claims or defenses are typical of those of the class; and (4) the representative parties will fairly and adequately protect the class members’ interests. Id. 23(a). Where, as here, a plaintiff seeks to certify a class under Rule 23(b)(3), common questions of law or fact must predominate over any questions affecting only individual members, and the class-action mechanism must be “superior to other available methods for fairly and efficiently adjudicating the controversy.” Id. 23(b)(3).
The court may certify a class only if, after a “rigorous analysis,” it is clear that the proposed class representative has satisfied all applicable requirements of Rule 23. Gen. Tel. Co. of the Southwest v. Falcon, 457 U.S. 147, 161, 102 S.Ct. 2364, 72 L.Ed.2d 740 (1982). In undertaking this analysis, the court “should make whatever factual and legal inquiries are necessary” rather than accept the allegations of the complaint and the assurances of counsel at face value. Szabo v. Bridgeport Machs., Inc., 249 F.3d 672, 675-76 (7th Cir.2001). Lastly, district courts have broad discretion in deciding motions for class certification. Reiter v. Sonotone Corp., 442 U.S. 330, 345, 99 S.Ct. 2326, 60 L.Ed.2d 931 (1979); Payton v. County of Carroll, 473 F.3d 845, 847 (7th Cir.2007).
ANALYSIS
Superiority
The court begins its analysis with the superiority requirement because it finds this issue dispositive for all of the proposed classes. A threshold legal question is whether a defendant-administered refund program may be found superior to a class action within the meaning of Rule 23(b)(3), which permits the court to certify a class only if it finds, among other things, that “a class action is superior to other available methods for fairly and efficiently adjudicating the controversy.” The Seventh Circuit has not yet answered this question, and other authorities are divided between what the court will call, for ease of reference, the “textual approach” and the “policy approach.” This court joins the majority of district courts and prominent commentators in adopting the latter approach.
The textual approach was articulated in Turner v. Murphy Oil USA, Inc., 234 F.R.D. 597 (E.D.La.2006) and Amalgamated Workers Union of Virgin Islands v. Hess Oil Virgin Islands, 478 F.2d 540 (3d Cir.1973). In Turner, the plaintiffs sought damages arising from an oil spill in New Orleans in the days following Katrina. 234 F.R.D. at 601. At the time of the litigation, the defendant oil company had already implemented a private settlement program, the details of which are left hazy at best, probably because the court found them irrelevant: “the analy*382sis is whether the class action format is superior to other methods of adjudication, not whether a class action is superior to an out-of-court, private settlement program.” Id. at 610; see also Amalgamated Workers Union, 478 F.2d at 543 (reaching same conclusion about administrative remedy). As the Turner court noticed, it makes little sense to describe an out-of-court remedy as an “adjudication” of a claim. See Black’s Law Dictionary (‘adjudication’ means “[t]he formal giving or pronouncing of a judgment or decree in a court proceeding; also the judgment or decision given”). And that is perhaps why the advisory committee focused on the comparison of a class action to individual suits, not to alternative nonjudicial remedies. See Amalgamated Workers Union, 478 F.2d at 543 (discussing advisory committee notes). Since nonjudicial remedies cannot be deemed superior to a class action on the textual approach, it should be obvious that this is the analysis favored by the lead plaintiffs.
In contrast, the policy approach emphasizes that the animating purpose of the superiority requirement is to ensure that the court’s resources are put to efficient use; thus, “[t]he court need not confine itself to other available ‘judicial’ methods of handling the controversy in deciding the superiority of the class action,” since a nonjudicial alternative may well obviate the need for court involvement in the first place. 7AA Wright, Miller, and Kane § 1779 (criticizing Amalgamated Workers Union for relying on “an overly restrictive reading of Rule 23(b)(3)”). So when a defendant is already offering an effective remedy for putative class members through out-of-court channels, a class action threatens to consume substantial judicial resources to no good end.
That was the court’s conclusion in the seminal case of Berley v. Dreyfus & Co., 43 F.R.D. 397 (S.D.N.Y.1967). In Berley, the SEC suspended trading in an unregistered security that the defendant had induced investors to purchase through misrepresentations. Id. When the defendant offered to refund the purchase price of the security, most of the investors accepted the refund without further ado. Id. at 399. A few, however, opted for a class action, but the court declined to certify their proposed class, since class certification in these circumstances “would needlessly replace a simple, amicable settlement procedure with complicated, protracted litigation.” Id. The court acknowledged and rejected the textual approach in favor of a broad view of the policy underlying Rule 23(b)(3): “Although Dreyfus & Co.’s offer to refund the purchase price to its customers is not quite ‘another method for the fair and efficient adjudication of the controversy,’ we think that subparagraph (b)(3) read as a whole reflects a broad policy of economy in the use of society’s difference-settling machinery.” Id. at 398. Since refunds were readily available, a class action would do little more than “create[ ] lawsuits where none previously existed.” Id. And that is very nearly the reverse of Rule 23’s intended effect.
More recently, the plaintiffs in Chin v. Chrysler Corp., 182 F.R.D. 448 (D.N.J.1998), sought compensatory damages and restitution because Chrysler had installed defective anti-lock brakes in their vehicles. Id. at 451. Chrysler implemented a voluntary recall, subject to NHTSA oversight, and it reimbursed prior and current owners of vehicles with defective ABS systems for necessary repairs. Id. at 454. In finding that that a class action was not superior to the reimbursement program, the court observed that “it is unclear if there is any useful remedy this Court could fashion;” nor could the court see “what expenses could be the subject of a restitution order, if Chrysler has already reimbursed Plaintiffs.” Id. at 463.
Recent MDL courts have agreed with Berley and Chin. In In re Phenylpropanolamine (PPA) Prods. Liab. Litig., 214 F.R.D. 614 (W.D.Wash.2003), the court held that an effective recall and refund program precluded a finding of superiority under 23(b)(3). Id. at 623. Pursuant to an FDA health advisory and request for a recall, the defendant drug manufacturers recalled over-the-counter pharmaceutical products containing PPA and instituted refund programs for purchasers of the recalled products. Id. at 614. Despite these programs, the lead plaintiffs sought judicially orchestrated refunds for putative *383class members under theories of unjust enrichment and breach of implied warranty. Id. at 615. (As in the present case, the plaintiffs did not seek certification of a personal injury class.) The court declined to certify the proposed classes because “[i]t makes little sense to certify a class where a class mechanism is unnecessary to afford the class members redress.” Id. at 622 (citing Chin, 182 F.R.D. at 462-65; Berley, 43 F.R.D. at 398-99). As in Berley, the court was concerned that “certification of the proposed class would merely serve to create lawsuits where none previously existed.” PPA 214 F.R.D. at 622.
Lastly, although the ConAgra court rested its holding on the plaintiffs’ inability to show predominance, it provided a thorough analysis of superiority in case an interlocutory appeal was taken. In re ConAgra Peanut Butter Prods. Liab. Litig., 251 F.R.D. 689, 699-701 (N.D.Ga.2008). ConAgra involved the widely publicized recall of the defendant’s peanut butter following an FDA warning linking it to several cases of salmonella poisoning. Id. at 691. Although postrecall testing showed that the vast majority of the recalled peanut butter was uncontaminated, ConAgra immediately offered refunds to purchasers of the recalled peanut butter. Id. In all, ConAgra refunded nearly $3,000,000 directly to consumers, representing the retail price of almost 1,000,000 jars of peanut butter; and it reimbursed retailers more than $30,000,000 for inventory subject to the recall. Id. Still, the lead plaintiffs filed suit, seeking to recover economic losses because the class members’ peanut butter “was rendered unusable and valueless.” Id. at 691-92. As in PPA the court declined to certify an unjust enrichment class after concluding that it was not necessary to afford the class members redress; indeed, they could easily get their money back from ConAgra. Id. at 700. Thus, “the important policy concern that individuals were not or will not be able to recover because of small individual recoveries [was] not heavily implicated.” Id. Moreover, disgorgement of the defendant’s profits would leave each class member with less than they could get by simply asking ConAgra for a refund — even before factoring in class counsel’s inevitably sizeable fees. Id. The refund program could therefore provide a better remedy to putative class members, in much less time, and without the expenditure of substantial judicial resources for no obvious benefit. It was, in a word, superior to a class action.
Where available refunds afford class members a comparable or even better remedy than they could hope to achieve in court, a class action would merely divert a substantial percentage of the refunds’ aggregate value to the class lawyers. For this reason, among others, rational class members would not choose to litigate a multiyear class action just to procure refunds that are readily available here and now. Not so for the class lawyers; here, as elsewhere, there is “a much greater conflict of interest between the members of the class and the class lawyers than there is between an individual client and his lawyer. The class members are interested in relief for the class but the lawyers are interested in their fees....” Thorogood v. Sears, Roebuck & Co., 547 F.3d 742, 744 (7th Cir.2008). This conflict underscores the virtue of the policy approach to superiority, which allows the court the ensure that a putative class action is grounded in the realistic prospect of a remedy that class members could not otherwise obtain. See ConAgra, 251 F.R.D. at 700; PPA, 214 F.R.D. at 622; Chin, 182 F.R.D. at 463; Berley, 43 F.R.D. at 398-99. In making this determination, courts are not merely enforcing sound judicial economy; they are also protecting the interests of class members. See Thorogood, 547 F.3d at 745 (“the judge who presides over the class action ... is charged with responsibility for preventing the class lawyers from selling out the class”). In contrast, the textual approach permits (or even requires) the court to certify class actions that, at best, offer no advantage for the class members, and at worst, benefit class counsel at their expense. The pressure to settle quickly once classes are certified, arising from class counsel’s and class-action defendants’ shared incentive to “trade small damages for high attorney’s fees,” see id. at 744-45, only exacerbates these familiar conflicts. Therefore, to safeguard the interests of class members as well as to minimize needless litigation, this court *384now holds that it may assess the superiority of a class action relative to the out-of-court remedies already available to the class.
Applying this analysis, the court must first inquire into the bona fides of any purported out-of-court remedy, as an illusory promise of private redress obviously will not defeat a request for class certification. Here, the defendants’ standing offers to refund any tainted Aqua Dots toy do in fact offer redress to any class member who demands it. It is telling, in this context, that not one lead plaintiff has ever asked any defendant for a refund; notably, counsel told at least one of the lead plaintiffs, shortly after the recall was announced, not to return his son’s Aqua Dots to Toys “R” Us or Spin Master for a refund. So it is unbelievable that the lead plaintiffs did not know, or could not have known, about the refund offers. Indeed, at least 513,869 consumers managed to procure refunds — all without the assistance of counsel or the court. See PPA, 214 F.R.D. at 622 (about 47,000 refunds was a “significant number” and evinced sufficient publicity to accept refund offer as superior to class action). To be sure, the lead plaintiffs repeatedly decry the defendants’ refund offers as “hit-and-miss, secret, or half-secret efforts” that are only availing to the “lucky and resourceful,” but these bald allegations have no discernable basis. Hundreds of thousands of refund requests have been honored, and there is no evidence that any legitimate requests have ever been refused. Thus, the fact that the rest of the consumers who purchased tainted Aqua Dots have not sought refunds, even though it is widely known that refunds are available, “may well demonstrate that certification of the proposed class[es] would merely serve to create lawsuits where none previously existed.” Id. (citation omitted).
Next, the lead plaintiffs contend that the refund offers are inadequate because some consumers apparently received a few dollars less than what they paid for their Aqua Dots. More specifically, some consumers were refunded the current store price of Aqua Dots at the time they sought their refunds rather than the price they had previously paid, even though the store prices had fallen in the meantime. A class action, however, is not likely to produce vastly different results. Since the purchase price of Aqua Dots- — • which establishes the measure of damages for the class — varied from one store to the next and from one moment to the next, an “estimate of the average damages [each class member] had sustained would be a sensible and legally permissible way to proceed.” Thorogood, 547 F.3d at 748; see also Carnegie v. Household Int’l, Inc., 376 F.3d 656, 661 (7th Cir.2004) (“Rule 23 allows district courts to devise imaginative solutions to problems created by the presence in a class action litigation of individual damages issues.”). Thus, the lead plaintiffs are mistaken when they implicitly suggest that individualized damages inquiries will be necessary to ensure that each class member receives a penny-for-penny refund. Rather, each class member would likely receive a uniform estimate of damages, and there is no reason to think that this estimate would diverge substantially from the average value of the refunds already dispensed. Once again, it is difficult to see what improvement a class action would offer.
The lead plaintiffs mount several more arguments of this order of magnitude, none of which succeed. The refund offers provide the remedy they purport to provide, and nothing that the class could hope to receive in court under the guise of compensatory damages or restitution would be superior to that remedy — or even comparable, since counsel would, at all events, have to be paid out of the damages award. But as the lead plaintiffs point out, they also seek injunctive and/or declaratory relief, as well as punitive and/or treble damages, on behalf of the class, and these remedies are obviously not available through any out-of-court refund program.
A request for injunctive or declaratory relief does not automatically render a class action superior to an out-of-court refund program. If it did, a plaintiff could easily fashion a request for an injunction that would accomplish nothing more than the out-of-court remedy already in place, thereby turning the superiority requirement into a trivial pleading hurdle. Thus, neither the *385Chin nor the PPA court thought that the plaintiffs’ requests for injunctive relief supported a finding of superiority. See Chin, 182 F.R.D. at 451 (request for injunction ordering defendant to offer rescission of sales contract or to recall all allegedly defective products for retrofitting); PPA, 214 F.R.D. at 615 (request for injunction in the form of notice to consumers still in possession of the recalled products). Likewise, given the efficacy of the recall and refund programs the defendants have already implemented here, it is hard to see what meaningful relief would be achieved by, e.g., a restitution order or an injunction ordering the defendants to replace any extant recall program with a new one in which consumers could “easily participate.” In fact, it is hard to see how any of the injunctions or declarations listed in the complaint’s prayer for relief request meaningful, realistic relief for the class. It follows that these requests do not provide any basis for a finding of superiority. Nor is the theoretical availability of punitive damages — which at best offers a remote possibility of a marginally greater recovery per class member — enough to support a finding of superiority when a putative class action has nothing else to recommend it.
At bottom, this is a suit to recover the purchase price of tainted Aqua Dots. Since the defendants will provide a refund — without needless judicial intervention, lawyer’s fees, or delay — to any purchaser who asks for one, there is no realistic sense in which putative class members would be better off coming to court. From their perspective, a class action is not the superior alternative. The court therefore declines to certify any of the proposed classes.
Unjust Enrichment Subclasses5
As the court explained in some detail at oral argument, the unjust enrichment subclasses are fraught with procedural and choice-of-law problems that further preclude certification. To begin with, the lead plaintiffs have failed to put forth an internally consistent and procedurally coherent proposal. The problems in this regard are too numerous to rehash in full here; an example will illustrate the general problem.
The lead plaintiffs suggest, appropriately enough, that a transferor court in each subclass would handle the entire subclass on remand, given the purported uniformity of unjust enrichment law within each subclass. But when it comes to Kentucky’s placement within the subclasses, for instance, the lead plaintiffs are in a muddle. In their original motion, they put Kentucky in subclass # 1. When the defendants pointed out that Kentucky’s unjust enrichment law was materially at odds with the law of other states in the subclass, the lead plaintiffs conceded in their reply brief that Kentucky belongs in subclass #2. Yet they continue to proffer Plaintiff Walker from Kentucky as the representative for subclass # 1. So one problem is the lead plaintiffs’ inability to provide consistent representations as to what their subclass groupings actually are. But that’s not all. After the lead plaintiffs’ concession that Kentucky belongs in subclass #2, subclass # 1 no longer includes a transferor court or forum state, whereas subclass # 2 now includes two (Kentucky and Illinois). Lacking a forum state, there is no forum law to apply to subclass # 1, and problems such as this would wreak havoc with post-MDL proceedings in the remand courts-or for that matter, with this court’s own task at summary judgment. Therefore, the lead plaintiffs have not shown that certification of the unjust enrichment classes will be manageable. See Fed. R.Civ.P. 23(b)(3)(D) (court must consider “the likely difficulties in managing a class action”); Eisen v. Carlisle & Jacquelin, 417 U.S. 156, 164, 94 S.Ct. 2140, 40 L.Ed.2d 732 *386(1974) (manageability requirement “encompasses the whole range of practical problems that may render the class action format inappropriate for a particular suit.”).
Furthermore, the unjust enrichment subclasses pose insurmountable choice-of-law problems. As other members of this court have pointed out, the law of unjust enrichment varies too much from state to state to be amenable to national or even to multistate class treatment. Muehlbauer v. General Motors Corp., 2009 WL 874511, at *5, No. 05-2676, 2009 U.S. Dist. LEXIS 26971, at *14 (N.D.Ill. March 31, 2009) (“the cases explain that multi-state class actions for unjust enrichment are inappropriate because the individual states’ laws regarding unjust enrichment are too nuanced to lend themselves to class treatment”); see also Vulcan Golf LLC v. Google, Inc., 254 F.R.D. 521, 533 (N.D.Ill.2008) (collecting cases rejecting multistate unjust enrichment classes); In re Sears, Roebuck & Co., Nos. 05-4742 & 05-2623, 2006 WL 3754823, at *1, 2006 U.S. Dist. LEXIS 92169, at *1 n. 3 (N.D.Ill.Dec. 18, 2006) (“It is clear just from our review of Illinois law that unjust enrichment is a tricky type of claim that can have varying interpretations even by courts within the same state, let alone amongst the fifty states.”). A small sample of the myriad difficulties with the subclasses will once again illustrate the general problem.
Subclass # 1 is allegedly composed of states that do not require either (1) wrongful conduct by the defendant as an element of the claim, or (2) that the plaintiff lacks an adequate remedy at law. Hawaii, however, requires a plaintiff to lack an adequate remedy at law. Porter v. Hu, 116 Hawai'i 42, 169 P.3d 994, 1007 (App.2007). And so does Colorado. Mahoney Mktg. Corp. v. Sentry Builders of Colo., Inc., 697 P.2d 1139, 1140 (Colo.App.2004); Harris Group, Inc. v. Robinson, 209 P.3d 1188, 1205 (Colo.Ct.App.2009).
The lead plaintiffs maintain that the states in subclass # 2 require the defendant to have knowledge of the benefit received; other than that, the requirements in subclass # 2 are said to be identical to those in subclass # 1. Thus, the plaintiff need not lack an adequate remedy at law in these states, according to the lead plaintiffs’ analysis. Now, Illinois law is to govern subclass # 2, and there is considerable confusion about the law of unjust enrichment, including the no-adequate-remedy-at-law requirement, in Illinois. See, e.g., In re Sears, Roebuck & Co., 2006 WL 3754823, at *1 n. 3, 2006 U.S. Dist. LEXIS 92169, at *1 n. 3. While many intermediate courts in Illinois routinely say that, “because it is an equitable remedy, unjust enrichment is only available when there is no adequate remedy at law,” Guinn v. Hoskins Chevrolet, 361 Ill.App.3d 575, 296 Ill.Dec. 930, 836 N.E.2d 681, 704 (2005) (citation omitted), the Illinois Supreme Court has said that restitution, i.e., money damages under an unjust enrichment theory, is a legal remedy. HPI Health Care Servs., Inc. v. Mt. Vernon Hosp., Inc., 131 Ill.2d 145, 137 Ill.Dec. 19, 545 N.E.2d 672, 678-79 (1989) (“The doctrine of unjust enrichment underlies a number of legal and equitable actions and remedies, including the equitable remedy of constructive trust and legal actions of assumpsit and restitution or quasi-contract.”); see also Partipilo v. Hallman, 156 Ill.App.3d 806, 109 Ill.Dec. 387, 510 N.E.2d 8, 11 (1987) (“equitable defense based upon an adequate remedy at law is unavailable” where plaintiff seeks “money damages” on “unjust enrichment” theory). The correct Illinois rule appears to be a narrower one: unjust enrichment is unavailable where a specific contract governs the relationship of the parties. See, e.g., Shaw v. Hyatt Int’l Corp., 461 F.3d 899, 902 (7th Cir.2006) (citing Guinn v. Hoskins Chevrolet, 361 Ill.App.3d 575, 296 Ill.Dec. 930, 836 N.E.2d 681, 704 (2005)). Here, the lead plaintiffs allege express warranties, and it is well settled that “express warranty claims ... sound in contract.” Cipollone v. Liggett Group, 505 U.S. 504, 525 n. 23, 112 S.Ct. 2608, 120 L.Ed.2d 407 (1992). In contrast, the Supreme Judicial Court of Maine has held that a plaintiff may not pursue an unjust enrichment claim unless he lacks an adequate remedy at law full-stop; thus, not only an adequate remedy in contract but also an adequate remedy in tort will bar an unjust enrichment claim. Wahlcometroflex, Inc. v. Baldwin, 991 A.2d 44, 49-50 (Me.2010). At a minimum, then, Maine and Illinois differ on a *387potentially outcome-determinative requirement.
Subclass # 3 is said to comprise states that require (1) that the plaintiff lacks an adequate remedy at law and (2) that the defendant has knowledge of the benefit received. This says nothing about the presence of any wrongful-conduet-by-defendant requirement, which predictably varies between states in the class. In Ohio, for instance, “[t]he benefit conferred by the plaintiff must be in response to a fraud, misrepresentation, or bad faith on behalf of the defendant.” McCamon-Hunt Ins. Agency, Inc. v. Med. Mut. Of Ohio, No. 07 MA 94, 2008 WL 4444631, at *4, 2008 Ohio App. LEXIS 4338, at * 12 (Ohio Ct.App. Sept. 26, 2008) (citing Nat’l City Bank v. Fleming, 2 Ohio App.3d 50, 440 N.E.2d 590, 599 (1981)). But as the lead plaintiffs point out, Maryland does not require wrongful conduct by the defendant. Hill v. Cross Country Settlements, LLC, 402 Md. 281, 936 A.2d 343, 352 (2007) (“Although the basis for recovery in unjust enrichment is not based on fault, misconduct or fault of one of the parties is sometimes considered in determining whether to permit recovery.”) (emphasis added).
The lead plaintiffs describe the requirements for subclass #4 as “wrongful conduct by defendant, and/or knowledge of the benefit, and/or plaintiff must lack an adequate remedy at law.” As purportedly uniform classifications go, this one is deficient on its face, and the lead plaintiffs themselves make clear that subclass # 4 is what it appears to be, namely, an amalgam of four sub-sub-classes with the following alternative sets of requirements: “(1) wrongful conduct by defendant (but intent not required); (2) ■wrongful conduct by defendant AND plaintiff lacks adequate legal remedy; (3) wrongful conduct by defendant AND knowledge of the benefit; (4) wrongful conduct by defendant AND knowledge of the benefit AND plaintiff lacks adequate legal remedy.” On the lead plaintiffs’ proposal, Texas law is to govern, and Texas law inevitably conflicts with the law of other states in the subclass. For instance, Texas requires a plaintiff to lack an adequate remedy at law, whereas Arkansas does not. Compare Tex. Carpenters Health Benefit Fund, IBEW-NECA v. Philip Morris, Inc., 21 F.Supp.2d 664, 678 (E.D.Tex.1998) with Thompson v. Bayer Corp., No. 07-00017 JMM, 2009 WL 362982, at *4, 2009 U.S. Dist. LEXIS 15190, at *16 (E.DArk.2009). Likewise, Texas requires wrongful conduct by the defendant, whereas Arkansas does not. Compare Burlington N. R.R. Co. v. Southwestern Elec. Power Co., 925 S.W.2d 92, 97 (Tex.App.1996) (“Unjust enrichment is typically found under circumstances in which one person has obtained a benefit from another by fraud, duress, or the taking of an undue advantage.”) with Thompson, 2009 WL 362982, at *5, 2009 U.S. Dist. LEXIS 15190, at *10. As it turns out, then, Arkansas does not fit any of the alternative sets of criteria for membership in subclass # 4.
The lead plaintiffs request that the court simply amend the subclasses if it determines that some particular state belongs in a different subclass. This the court declines to do. It is the lead plaintiffs’ burden to demonstrate that class certification is proper, Oshana, 472 F.3d at 513, and “[n]o class action is proper unless all litigants are governed by the same legal rules.” In re Bridgestone/Firestone Tires Prods. Liab. Litig., 288 F.3d 1012, 1015 (7th Cir.2002). That is emphatically not the case here, and any attempt by the court to tinker with the composition of the subclasses to satisfy this requirement would be futile anyhow. The unjust enrichment subclasses therefore cannot be certified.
CONCLUSION
For the foregoing reasons, the lead plaintiffs’ motion for class certification is DENIED.
. It is now known that JSSY, a Chinese factory to which Moose outsourced production of Aqua Dots, produced the tainted beads — allegedly by substituting 1,4 butanediol for 1,5 pentanediol without Moose’s knowledge. JSSY, in any event, is not a party to this motion, so the details of these allegations are immaterial for present purposes.
. In response to the court’s observation that the initial class definitions included consumers who had already returned their Aqua Dots for a refund and the court’s corresponding request for narrowed class definitions, the lead plaintiffs proposed revised class definitions that excluded these consumers. While numerous problems with the breadth of the proposed classes remain, the court rests its holding on other grounds. Thus, the differences between the two proposed class definitions are immaterial for present purposes.
. Florida (S.D. Fla., No. 08-2368, Plaintiff Bertanowski), Texas (N.D. Tex., No. 08-2371, Plaintiff Botsch) Illinois (N.D. Ill., No. 07-6387, Plaintiff Williams), Kentucky (W.D. Ky., No. 08-2760, Plaintiff Walker), New Jersey (D.N.J., No. 09-2153, Plaintiff Fierro), and California (C.D. Cal., No. 08-2368, Plaintiff Cosgrove; No. 08-2429, Plaintiff Soderstedt) are the forum states of the transferor courts in which the member cases originated. Oklahoma is the home state of Plaintiff Williams, who filed her complaint in this court; Pennsylvania is the home state of Plaintiff Soderstedt, and New York is the home state of Plaintiff Cosgrove, both of whom filed in California.
. This subclass grouping is from the lead plaintiffs' supplemental brief. The original subclasses were different: Kentucky was in # 1 instead of # 2; Illinois was in # 3 instead of # 2; New Jersey was not in any subclass; Florida was in # 2 but now is not in any subclass. In any event, as of now, for subclass # 1, Plaintiffs propose Walker from Kentucky as the class representative, even though Kentucky is now in subclass # 2; for # 2 they propose Ford from Illinois; for # 3, Pierro from New Jersey; for # 4, Botsch from Texas.
. In the court’s view, the following problems provide a sufficient alternative ground for denying certification of the unjust enrichment subclasses. There are further reasons to deny certification of the single-state classes, too, but in light of the court's analysis of superiority, and the fact that many of these problems were not fully explored at oral argument, the court does not take them up here. The problems with the unjust enrichment subclasses, however, deserve further comment for two reasons. First, they provide the most vivid illustration of the generally unwieldy and slipshod state of the lead plaintiffs’ proposed class action. Second, the case law is replete with recent efforts by class-action attorneys to certify multistate unjust enrichment subclasses. The recurring problems with this strategy therefore warrant analysis.