Until recently, there was some doubt whether such waivers were enforceable. But last October, the U.S. Senate voted to rescind a new rule promulgated by the Consumer Financial Protection Bureau (CFPB) that would have overridden these now ubiquitous provisions barring consumers from joining any class action. So consumer class actions are pretty much dead in the financial industry, at least for now. But business may regret this apparent victory.

To be sure, the CFPB is itself controversial. It is an odd agency born of the 2010 Dodd–Frank Act and seen by many conservatives as the pet project of Sen. Elizabeth Warren (D–Mass.). It is not overseen by a bipartisan panel of commissioners, but rather by an appointed director. That director was Richard Cordray, who began his tenure via an attempted recess appointment by President Barack Obama and whose statutory successor has since been replaced by President Trump with Mick Mulvaney (who also continues to serve as director of the Office of Management and Budget).

That controversy aside, the Senate vote was no doubt influenced by a Treasury Department report issued the day before the vote, lambasting the CFPB and the rule. The most striking feature of the report is its title, "Limiting Consumer Choice, Expanding Costly Litigation: An Analysis of the CFPB Arbitration Rule."

Limiting consumer choice? Really?

In practice, consumers have zero choice whether to agree to arbitration. Indeed, they seldom know that they have done so until they complain. But in the era of fake news, it seems Treasury is entitled to its own facts.

The Treasury critique may seem persuasive: Only 13% of consumer class actions result in recovery, of which 31% goes to plaintiff attorneys. And just 4% of class members claim their average $32.35 share of successful class recoveries. Treasury says this means that consumers place little value on class actions. But if only 4% bother to seek payment when they are already entitled to it, how many will initiate arbitration?

The likely effect of abrogating the CFPB rule is that financial institutions will escape the need to answer at all for abusive practices. No doubt that was the real goal.

But the CFPB rule was not about compensation or compliance or consumer choice. The idea that consumers will choose a bank or credit card based on whether it comes with arbitration or permits class actions is ludicrous. The rule was about deterring business practices that gouge consumers, so to focus on consumer recovery misses the point. The focus should be on the payout by the wrongdoers and not on who gets the money.

There is no doubt that class actions can be abused by unscrupulous plaintiff attorneys. The mere threat of liability to a class comprising thousands of consumers (or investors) may be enough to induce a defendant to settle even if the claim is frivolous. Indeed, it is quite rational to do so if the case can be settled for less than the cost of seeking dismissal in court. And there are law firms that have made a business of exploiting the math.

On the other hand, there is no doubt that class actions can be beneficial. As the U.S. Supreme Court stated in Califano v. Yamaski (1979), class actions can save "the resources of both the courts and the parties by permitting an issue affecting every [class member] to be litigated in an economical fashion" where "the amounts at stake for individuals may be so small that separate suits would be impracticable"—that is, too costly to try one at a time.

Moreover, an individual plaintiff cannot simply declare an action to be a class action. Under Rule 23 of the Federal Rules of Civil Procedure, the court must approve the lead plaintiff and lawyer (among other things) before the action can be certified to proceed as a class action. If the court does not do so, the case reverts to a simple lawsuit between plaintiff and defendant.

To explain: A class action is an ordinary lawsuit in which the plaintiff asserts claims typical of those of many other potential plaintiffs. So the plaintiff seeks to act as a representative for hundreds or thousands of others with similar claims. One might think of it as a test case in which the claims of absent class members are actually tried and settled. Since the named plaintiff serves as a fiduciary for absent class members, he or she must be found to be worthy: not only to be typical, but also to be an adequate representative. But a lead plaintiff who is subject to an agreement to submit to arbitration—who has no right to be in court in the first place—can hardly be an adequate representative for absent class members (although they may have no right to be in court either). Indeed, defendants routinely oppose class certification on such grounds—and prevail—thus sending plaintiffs back to bilateral trial or arbitration of their individual claims.

Avoiding the Bill of Peace

It is surprising that business has been so hostile to class actions because they provide a way to adjudicate numerous small claims all at once, thus precluding never-ending bother by the Lilliputian swarm. It is telling that when the class action first emerged under English law, it was called a Bill of Peace. One would think that the prospect of dispensing with the claims of thousands of consumers all at once—cutting off the claims of absent class members who may not even know they have a remedy—would be quite attractive.

Alas, abrogation of the CFPB rule returns us to the legal world as it was following the 2011 Supreme Court decision in AT&T Mobility v. Concepcion. There, the plaintiffs had signed up for cell phone service and a free phone. When AT&T sent them a bill for the sales tax on the phone—$30.22—they sued and sought to litigate the matter as a class action. But the contract of sale included an agreement to arbitrate and a class action waiver. The trial court and the Ninth Circuit ruled that the class action waiver was unconscionable and thus unenforceable under a California statute as interpreted by the state's courts because it effectively precluded anyone from suing for such a small sum.

The Supreme Court reversed, holding that California law was preempted by the Federal Arbitration Act (FAA) adopted by Congress in 1925 to assure that contracts to arbitrate would be enforced. In the majority opinion, Justice Antonin Scalia emphasized that an agreement to arbitrate is just that: a bilateral contract in which two parties have agreed to arbitrate any dispute. As a legitimate private contract, it should be enforced as written, particularly in light of the FAA. But Justice Scalia also observed that class arbitration is oxymoronic—that a class action is inherently inconsistent with arbitration (although the Court had previously suggested quite the opposite regarding cases in which the parties agree to class arbitration).

In the end, the argument from contract (and the FAA) proves too much. An agreement to arbitrate is just that: an agreement between two parties to handle a dispute between the two of them. I cannot waive your rights. You cannot waive my rights. So how is it that the rights of similarly situated consumers to join forces under established rules of civil procedure has disappeared as a result of many individual one-on-one contracts that supposedly affect only the rights of the signatories?

To be clear, the CFPB rule would not have affected bilateral arbitration provisions. Rather, it would have prohibited banks and other financial firms from relying on such provisions to avoid litigating a class action in court. But it is not clear that the FAA policy favoring the enforceability of agreements to arbitrate ought to apply to class waivers. A class waiver is not an agreement to arbitrate. It is an agreement to refrain from exercising a right. It is black letter law that a waiver will be strictly construed. As such, a waiver is quite different from a positive agreement to arbitrate a claim. The fact that a class waiver has been attached as a rider to an agreement to arbitrate does not mean the rider falls within the FAA.

Danger or Abuse?

Moreover, the danger of abusive class actions is overblown. Again, the rules require that the court approve both the individual representative plaintiff and class counsel. Thus, the court has the power to remove and replace both plaintiff and lawyer if necessary to protect the interests of the class. In addition, the court must approve the definition of the class, thus determining who is or is not a member and the scope of the action. Perhaps most important, the court must approve the voluntary dismissal or settlement of the action: once filed, a class action may not be dropped because the plaintiff or lawyer has a change of heart (say) as a result of being paid off individually by the defendant.

Thus, the only real danger of abuse comes from failure of the courts to do their job. Indeed, several state courts became havens for class actions, presumably because judges were less than rigorous in enforcing the rules. Thus, Congress passed the Class Action Fairness Act (CAFA) of 2005, which among other things provides for the transfer of a state-court class action to federal court if the amount at stake is $5 million or more.

The bottom line is that a class action does not belong to the individual—or lawyer—who files it. The point is emphatically illustrated by Standard Fire Insurance Co. v. Knowles (2013), wherein the named plaintiff and her lawyer agreed in advance that the class would seek a total award of less than the $5 million limit that would trigger removal of the case under CAFA from an Arkansas state court to federal court. The Supreme Court rejected this tactic because neither the class lawyer nor the class plaintiff can bind absent class members until the court certifies the action as a class action. To repeat: how is it that the rights of similarly situated consumers to join forces under established rules of civil procedure disappeared as a result of many individual one-on-one contracts that supposedly affect only the rights of the signatories?

What about Punitive Damages?

Assuming that we want consumers to be made whole, how should we go about it?

Consider the recent litigation about bank overdraft fees generated by some banks' practice of clearing big checks before small checks, thus resulting in a fee for each overdrafted small check. It is unlikely that the aggregate effect of the practice would even be considered in bilateral arbitration. The bank would likely argue that the dispute is with this customer, who should have read the fine print and been more careful with his money. And even if a few consumers prevail, why would the bank change its ways? In contrast, a class action can address distinct collective interests of consumers.

Proponents of arbitration argue that it is quick and cheap because it can dispense with many of the niceties of litigation in court. But it is also private. There is typically no published opinion. Thus, the public is deprived of information about their rights and remedies. And the law ossifies.

In the absence of a class action, punitive damages or some specified financial penalty may be necessary. If a wrong is difficult to detect or victims may not always sue, the wrongdoer may not be deterred from bad behavior by the prospect of paying up when caught—especially if the wrongdoer gains from the wrong. Thus, the law imposes treble damages for antitrust violations and disgorgement plus treble damages in cases of insider trading.

If there is no such statutory remedy, punitive damages can address the problem. For example, if there is a 20% chance that a consumer will notice an overcharge and a 50% chance that a consumer who notices will sue, it makes sense to award damages equal to ten times the individual harm in order to deter the defendant from overcharging others. Merely to make the consumer whole is no way to deter the seller from continuing the practice.

Surprisingly, the Supreme Court held in Mastrobuono v. Shearson Lehman Hutton (1995) that punitive damages can be awarded in arbitration. Before Mastrobuono, well-settled state law prohibited punitive damages in arbitration. Traditionally, the courts viewed punitive damages as more in the nature of a criminal fine or penalty based on harms suffered by others not present in a bilateral arbitration. But the Supreme Court ruled that this sensible doctrine was trumped by the FAA. Sound familiar? As in Concepcion, the Court ruled that the FAA trumped the common law.

Should we rely on a class action where we calculate the aggregate loss and write everyone a small check? Or should we rely on individuals to sue in isolation and seek punitive damages where the aggregate recovery may be some multiple of the gain to the wrongdoer?

Needless to say, the prospect of a big award of punitive damages will motivate consumers—and their lawyers—to ferret out wrongs. But what is to prevent multiple lawsuits for punitive damages with the possible result that defendant businesses pay out far more than their ill-gotten gains and the losses suffered by consumers—especially when word spreads that the courts are handing out free money? It was just such a feeding frenzy that led the Second Circuit Court of Appeals, in Roginsky v. Richardson-Merrell (1967), to overturn an award of punitive damages in a classic opinion by Judge Henry Friendly (at a time when class actions had scarcely been born). Do we really want a legal system based on the illogic of the lottery? One can well imagine defendant companies crying "Uncle!" and seeking the protection of class action status.

In short, punitive damages are a poor substitute for a class action, and especially so in arbitration. Estimating the chances of detection and recovery is better suited to a court that is attuned to public policy and sees many cases. Moreover, if punitive damages are intended in part to be exemplary—to set a public example—how exactly does that work in arbitration where the reasoning and decision remain private?

To be clear, the rationale for punitive damages disappears in a consumer class action. If all of the victims of a scam are party to the action, there is no need to multiply the damages. Moreover, one would think that business would prefer such a Bill of Peace—settling all claims likely for cents on the dollar—over bilateral arbitration with the possibility of repeated awards of punitive damages. Apparently not. Could it be that business reckons it will escape the need to answer at all for abusive practices because few consumers will bother to file claims?

Purpose of Damage Awards

A lawsuit serves two important functions. One is compensation: to make whole someone who suffers harm at the hands of a wrongdoer. The other is deterrence: to discourage wrongdoing by imposing compensation. Ideally, these two functions complement each other. As Judge Learned Hand explained in the legendary 1947 decision in United States v. Carroll Towing, the law should provide compensation to victims only if the cost to avoid the harm is less than the cost of the harm itself. Ronald Coase made essentially the same point in a classic 1960 Journal of Law & Economics article, "The Problem of Social Cost," for which he got a theorem named after him—not to mention a Nobel Prize.

If compensation is too generous, there will be too many lawsuits and potential defendants will be too cautious. That is essentially what has happened with securities fraud class actions where defendant companies gain little or nothing from the so-called fraud but where the damages that may be claimed by plaintiffs are so generous that almost every case that survives a motion to dismiss quickly settles for the entire amount of any insurance carried by the defendant.

The problem is quite the opposite in most consumer class actions where aggregate compensation is a small fraction of ill-gotten gains. Punitive damages seek to address this imbalance typically by awarding some multiple of compensatory damages to reflect the unlikelihood of detection and legal action. Quite aside from the contradictions inherent in what amounts to a criminal penalty payable to a lucky individual plaintiff rather than the state, the calculus of probabilities is speculative at best. Indeed, the law resorts to broad-brush concepts that seek to capture the intent of defendants (like Initech) who bank on not getting caught or pursued.

It does not help that one often hears the argument from consumer activists that punitive damages are intended to punish wrongdoers—that the idea is to inflict real financial pain because otherwise the defendant will treat the award as nothing more than a cost of doing business. Quite to the contrary, the point of punitive damages should be to assure that businesses as well as individuals bear the full cost of their economic activities. In other words, we want businesses to internalize such costs. Given that many lawyers and judges (including the Supreme Court) seem not to understand how punitive damages should work, it is little wonder that defendants quake at the possibility of an award of punitive damages in a class action, notwithstanding the contradictions inherent in such a result. No doubt such fears were responsible in part for abrogation of the CFPB rule.

The tragedy is that class actions can address the problems inherent in punitive damages. By compensating all individual plaintiffs for their actual losses, there is no need to speculate about how many fail to recognize that they have been harmed or fail to complain. And given that only a small fraction of consumers who understand their rights would seek redress, do we really care that some of the recovery goes to the lawyer who organizes the effort?

Conclusion

The Treasury Department prevailed in the Senate by focusing attention on compensation and poking holes in the idea that consumers are made whole by class actions. But the focus should be on deterrence—the payout—not the recipient.

While it is easy to calculate how much consumers receive, there is no good way to measure the effects of deterrence. How do we quantify the benefit of preventing schemes that do not materialize because a thoughtful businessperson foresees the prospect of a class action?

No empirical study is likely to show the best way to prevent a one-off scam that is unlikely to be repeated. But that is no reason not to get the rules right—or at least to try.