Arbitration Agreements

Brian Wolfman has an interesting post about e-Bay's new arbitration agreement with a class action opt-out.  Curiously, e-Bay's arbitration agreement isn't mandatory, but it is opt-out with a limited opt-out period.  Brian's take is that this opt-out is consumer choice window-dressing:  while there is formally a consumer choice involved, functionally it is meaningless. I agree. 

First, consumers aren't likely to pay attention to the opt-out notice in the first place in this age of information overload.  (That's one reason why I don't like the mandatory annual Gramm-Leach-Bliley Act privacy notice--it contributes to information overload by telling me nothing--basically there are no privacy rights--and lulling me into thinking that all fine-print disclosures by my bank don't matter.)  Second, even if they do pay attention, consumers are unlikely to place much value ex-ante on the right to sue in court or to proceed as part of a class; certainly not enough to bother opting out.  

The problem, it seems to me, with arbitration or class action waiver or forum selection clauses in contracts, even if explicit opt-out provisions are available, is that there's an inherent imbalance between businesses and consumers in the way valuations of the provision are going to work: businesses value arbitration clauses in the aggregate, while consumers value them based on individual transactions. For contract provisions with small value this means that businesses are more likely to value them than consumers, which therefore warps the nature of any sort of bargain. The business is bargaining in aggregate;the consumer is bargaining based on an individal valuation.

e-Bay, as a repeat player, knows that by law of large numbers, if nothing else, that it is going to have a lot of disputes.  While value to e-Bay of arbitration for any single consumer complaint is likely small, in aggregate it may be significant.  

An individual consumer, however, is unlikely to place much value ex-ante on dispute resolution mechanisms.  If the consumer thought a dispute was likely, the consumer would likely avoid e-Bay. A rational consumer will expect a very low likelihood of a dispute, and this is probably correct. Moreover, I'm guessing most consumers don't know whether a judicial vs. arbitral proceeding matters or what a class action waiver means. Put this together and you have a consumer who thinks there is a low likelihood of something that doesn't matter much occuring.  Such a consumer won't take the time to opt-out.  

The problem is perhaps more severe for class action opt-outs. For negative value claims (claims that are worth less than the cost of litigating individually), the right to a class action doesn't gain the consumer much, even if it is important for protecting consumers from abuse. For example, let's say Chase were to hit its cardholders with a $10 annual fee that wasn't part of their cardholder agreement. (Oh wait, Chase did that, before it rolled back the fee after getting sued and regulatory pressure--but there was no class action waiver or arbitration agreement in force there.) Absent the right to proceed as a class, there would be no private litigation over this; no one would sue over $10. (And Chase would be safer and sounder as a result of such an additional fee, so who knows what a prudential regulator would say.)

But how much is the right to sue as a class  over such a fee really worth to an individual consumer? The consumer is unlikely to get $10 back:  there are the attorneys' fees off the top of any recovery, and a settlement is likely to be at a fraction of the fee.  So let's say there's a settlement for 50 cents on the dollar, and the attorneys' get a 20% cut.  The consumer has gotten back $4.  Given the low probability of most businesses doing something so rapacious and the low return from being able to sue as part of a class, is it really worth the time and postage to opt out?  If you value your time at $10/hr and opting out takes 15 minutes, you'd have to think you're getting more than around $3 of value from the opt-out.  If you think the probability of the opt-out mattering at all is 1 in a thousand (.1%), then we're not even close to a rational consumer opting out. 

I'm sure there will be a handful of opt-outs, but those are going to be people with idiosyncratic valuations, or in less technical terms, bones to pick.  

Notice that this has all assumed a totally rational consumer--hyperbolic discounting and other cognitive biases make the problem worse to be sure, but they aren't necessary to recognize that there's an imbalance in valuation of arbitration provisions and other dispute resolution terms between a repeat player (a business) and a one-time player (a consumer).  The business's valuation is an aggregate valuation, while individual consumers' valuations are disaggregated and therefore quite small.  

The aggregate valuation problem strikes me as a fundamental imbalance in contracting that hasn't been thought through very much in contract theory, perhaps because contract theory focuses on the single transaction, rather than understanding how the transaction fits into a larger business model. Put another way, contract theory hasn't figured out what to do with underwriting. There's been lots written about contracts of adhesion, but I think the aggregate valuation problem is separate, if related. Of course, when Justice Scalia write that the contracts of adhesion ship has long sailed--a true, if upsetting line--it makes me wonder if contract law has anything to do with its long-standing theoretical basis of bargained-for exchange or at least if there's a separate law that has emerged for consumers. 

The CFPB has the power to prohibit binding mandatory arbitration in consumer financial contracts, and this is an issue I would like to see the CFPB look into, but I would also note that forum seletion clauses and class action waivers also strike me as potentially falling within the "unfair" prong of the CFPB's UDAAP power when in consumer financial contracts, as forum selection provisions and class action waivers can cause substantial injury to consumers, may not be reasonably avoidable by consumers (query whether the opt-out really changes things), and is not outweighed by countervailing benefits to consumers or to competition.  There's an interesting burden of proof question about finding countervailing benefits, which cannot, I submit, be blithely assumed as the Supreme Court did about forum selection clauses in Carnival Cruise Lines v. Shute. For credit cards, at least, I haven't been able to detect any difference in credit card pricing based on the presence or absence of arbitration provisions in cardholderagreements.  There may be cost-savings via arbitration, but it's not at all clear to me that they are captured by the consumer.