Monday, February 22, 2010
Allocating power between courts and arbitrators - and why scholars of federal courts should care
The Federal Arbitration Act makes arbitration agreements as enforceable as all other contracts. In April, the Supreme Court will hear argument in Rent-A-Center v. Jackson, which concerns the question of who - court or arbitrator - decides a claim that an arbitration agreement is unconscionable and thus unenforceable. In this case, the arbitration agreement itself assigns (or at least purports to assign) that power to the arbitrator. The Ninth Circuit, however, held that unconscionability was an issue for the court. This case holds obvious interest for those who study ADR, consumer law (most consumer contracts have arbitration clauses, whether or not you know it), and employment law (this case is an employment discrimination suit). What I hope to show you is that it is just as interesting for those who study federal courts and judicial politics. Beneath the surface, the case is, in a sense, more Bush v. Gore than Williams v. Walker-Thomas Furniture.
To see why the case is so intriguing, one has to appreciate what one might call its strategic context. The Supreme Court is strongly pro-arbitration. Some state and federal courts are not quite so enthusiastic, at least when it comes to consumer and employment contexts with their largely adhesionary contracts. (Please note that I'm not discussing whether the Court's decisions in this area, and its broader pro-arbitration stance, reflect sound interpretations of the relevant statute, good policy, etc.) Over the course of the last couple of decades the Supreme Court has shut off most avenues for challenging arbitration agreements at the wholesale level - state law cannot declare particular fields like consumer transactions off limits from arbitration, courts cannot deem arbitration per se violative of public policy, etc. All such arguments are preempted by the Federal Arbitration Act. What remains, though, is the possibility for retail-level challenges to particular arbitration clauses under section 2 of the Act, which allows ordinary contract defenses that would invalidate any contract. So arbitration itself cannot be questioned, but a particular arbitration clause might be invalidated as the product of duress, fraud, etc.
In the last few years, as other routes for challenging arbitration have been closed off, unconscionability has become a surprisingly common and surprisingly effective way of attacking arbitration agreements. The challenges do not attack arbitration per se - federal law favors arbitration - but instead target various aspects of a particular arbitration process: a given clause might forbid class arbitrations, bar punitive damages or otherwise restrict remedies, sharply curtail discovery, require a consumer to pay hefty arbitrator's fees, etc. There have been many cases on these topics in recent years, and a good number of them sustain the challenge to the arbitration clause.
These cases have generated lots of petitions for certiorari in which businesses, assisted by amici like the Chamber of Commerce, charge that some state courts and lower federal courts are using unconscionability to discriminate against arbitration in violation of federal law. That is, while ordinary contract defenses can be used to invalidate an arbitration clause, those doctrines are not supposed to be used differently in the arbitration context in order to disadvantage arbitration. Perhaps surprisingly, given its pro-arbitration stance, the Supreme Court has for years let these petitions go by. My suspicion is that the Court has avoided these cases because it feels ill-equipped to resolve whether a lower court is discriminating against arbitration. First, unconscionability analysis often requires a fact-intensive inquiry. Second, and more important, determining whether a lower court is using unconscionability differently when it comes to arbitration requires an engagement with the details of state law and a comparison of lots of prior unconscionability cases. Third, and maybe most important of all, a holding that the lower court is applying unconscionability unfairly, especially when the lower court says it is applying the same analysis it applies elsewhere, carries with it some serious expressive baggage. Essentially, it requires the Supreme Court to say that the lower court is being dishonest. That happens, but when it does so, it is a big deal (think cases like Bush v. Gore or the cases from the 50s/60s rejecting supposed procedural defaults in the state courts).
The nice thing about a case like Rent-A-Center v. Jackson, at least from the point of view of a pro-arbitration court, is that it concerns not the merits of an unconscionability challenge but instead the allocation question - a question of who decides. That doesn't require diving into the weeds of state law and the record. If the Court assigns the issue to the arbitrator, that will be a very easy rule to monitor for compliance (unlike deciding whether the lower court applied unconscionability correctly). All of those unconscionability cases out there will instantly become not wrong but irrelevant - because courts won't be deciding the issue anymore. And it won't matter whether some lower courts can be trusted to apply unconscionability correctly, because they will be cut out of the picture.
All of the analysis above suggests reasons why the Supreme Court would find it useful to assign unconscionability to the arbitrator. On the other side, there are some reasons for it not to do that. Notably, it might be that judicial review for unconscionability operates as a sort of safety valve that makes arbitration of consumer and employment disputes palatable.
To this point I haven't said anything about the more narrowly doctrinal aspects of the case. I think that, as with many cases that reach the Supreme Court, the conventionally authoritative legal materials leave enough room here that acceptable legal reasoning could get you to either result. In brief (and this compressed summary probably won't make sense to non-specialists): On Jackson's side, one might say, as the Ninth Circuit did, that the Prima Paint separability doctrine does not apply here because this is a challenge specifically to an agreement to arbitrate. (Indeed, this was a stand-alone arbitration agreement, not a larger contract that included an arbitration clause.) On the other side, I can imagine Rent-A-Center arguing that there is a separable agreement to arbitrate here. After all, the agreement to arbitrate is not itself unconscionable - it can't be. Rather, the problem is the limitations on remedies and so forth. So one could imagine that there is an underlying theoretical agreement to arbitrate that remains uninfected, and then Jackson is supposed to go to arbitration to try to get the arbitrator to invalidate the offensive restrictions. That would be slicing separability pretty thinly, of course.
TrackBack URL for this entry:
Listed below are links to weblogs that reference Allocating power between courts and arbitrators - and why scholars of federal courts should care:
Excellent post. Do read Aaron's longer NYU LR article too.
Posted by: Don Philbin | Feb 22, 2010 4:16:42 PM
I enjoyed reading your very thoughtful and insightful post on this controversial case. I will have to make time to read your NYU Law Review article, as Don suggests!
What is particularly interesting is how both sides will probably rely on the severability doctrine in support of their cases. But the point you make about why the Court has been so reluctant to take on a unconscionability case that concerns the merits of a challenge -- not just its arbitrability -- is an excellent one.
I'll go out on a limb and predict that a majority of the Court will find in favor of Rent-a-Center. I see the force of your argument about the need for judicial review of unconscionability decisions. Here, however, the employee clearly and unmistakably agreed to arbitrate arbitrability, and the unconscionability challenge is not directed at that agreement to arbitrate arbitrability, but to the stand-alone arbitration agreement as a whole. Not separately enforcing the agreement to arbitrate arbitrability would effectively limit the ability of arbitrators to decide arbitrability to cases where there was a stand-alone pre- or postdispute agreement to arbitrate arbitrability. I do not believe the "clear and unmistakable exception" set forth in First Options and AT&T was intended to be so narrow, and I suspect the Court will not construe it so narrowly.
Posted by: Philip J. Loree Jr. | Feb 23, 2010 9:13:47 AM
I guess I'm having a tough time seeing how an arbitrator could ever be trusted to determine whether the agreement to arbitrate is unconscionable. The unconscionability of the agreement is correlated with the likelihood that an arbitrator will find the agreement to be proper. Let's take the Fourth Circuit's 1999 Hooters case as an example. If Hooters selects the arbitrators, are they likely to find the agreement unconscionable? I suppose there could be post-arbitration review by the courts, but then the employee/consumer has to go through and pay for the unfair arbitration and then hope that the radically strict review standard for arbitration decisions is somehow met. Sure, I bet the Ninth Circuit would review those cases. But they'd be distorting the arbitration review standard, and they shouldn't have to.
If the FAA and the Supreme Court are going to make arbitration a matter of state contract law and only state contract law, then the Court needs to stay out of it when the lower courts are applying state contract law. This may not be what the Court would prefer as a policy matter, but too bad. If five justices decide that lower courts cannot apply state law in the same way that courts apply this law to all other contracts -- simply based on additional boilerplate that "clearly and unmistakably" assigns the question to an arbitrator -- then this decision, in my view, would be almost as intellectually shoddy as Bush v. Gore. But I'm open to counterarguments!
Posted by: Matt Bodie | Feb 23, 2010 11:42:36 AM
To address the point in Matt’s first paragraph -
I think I agree that there are certain types of issues that the arbitrator shouldn’t/couldn’t be asked to decide. As you say (citing the Hooters case), if the claim is that the clause is unconscionable because the arbitrator is biased, it would be odd to require that claim to be presented to the arbitrator. Another example might be a clause that says that the arbitration will take place in Reykjavik. If the claim is that it is unconscionable to hold the arbitration there, it would be strange to require someone to go to Reykjavik to argue that it is unconscionable to be required to go to Reykjavik. There might be clever ways to deal with these cases after the fact, but, as you suggest, it probably makes more sense just to say you can go to court in the first instance. So I think we are on the same page so far.
I believe, though, that the more common situations - limitations on remedies or arbitral procedures - are different. There it would not be wholly illogical to say that there exists an underlying, separable agreement to present such matters to the arbitrator.
Note that I said it wouldn’t be illogical. It might still be bad policy and so forth. Maybe arbitrators can’t be trusted to rule on such matters fairly (though my limited understanding of the evidence shows that they are perhaps more willing to strike down remedial/procedural restrictions than one might guess). As I hope I am making clear, I’m just trying to show that plausible readings of the relevant texts and precedents would allow the Court to reach either result in this case.
Posted by: Aaron Bruhl | Feb 23, 2010 3:01:47 PM
By the way, Lawrence Cunningham at Concurring Opinions also did a post on this topic yesterday.
Posted by: Aaron Bruhl | Feb 23, 2010 3:53:38 PM
Thanks for the thoughtful response. I guess what I'm looking for is the legal doctrine through which the Supreme Court could declare it a matter of federal statutory law as to whether the arbtration was "Reykjavik-level" unconscionable or merely "borderline" unconscionable. Where is that in the FAA? Unconscionability is a matter of state common law, and its application should be left to courts applying state common law, no? I can't see any way of separating out the unconscionability of various arbitration provisions from the unconscionability of the agreement to arbitrate. It's not a claim that falls within the contractual relationship, as in Prima Paint -- it's a claim that the agreement to arbitrate itself is unconscionable. Those seem totally different to me.
Posted by: Matt Bodie | Feb 23, 2010 5:51:59 PM
Maybe the best answer to your very good questions is a) who knows? and b) none of this is in the FAA!
More seriously, I find the interaction between federal law and state law in this context really interesting and complicated. I don't claim to have it all sorted out. It seems to me that state law contract defenses and other state contract doctrines operate, but only within the bounds set by federal law (e.g., even-handedness, separability, various allocation rules). And while one could try to extract the federal law doctrines from the FAA's text, it is probably more forthright to say that they are creatures of federal common law (it lives!)
Posted by: Aaron | Feb 24, 2010 9:04:13 AM
I agree that if a lower court looks like it's using state law disingenuously in order to defeat federal law, then the Court can step in. As you point out, though, "[e]ssentially, it requires the Supreme Court to say that the lower court is being dishonest." The Supreme Court has done this to the NLRB on a few occasions -- it should do it here if it thinks it's an issue. But frankly, I don't think the Court could make that claim believably in this case, as unconscionability doctrine has always given common law courts a great deal of discretion. I don't see the Ninth Circuit's unconscionability decisions as disingenuous. If the Court decides that ("clear and unmistakable" boilerplate) + (less severe unconscionability claim) = arbitral issue, I just don't see how even-handedness, separability, or various allocation rules could justify such an approach.
Posted by: Matt Bodie | Feb 24, 2010 10:46:38 AM
I appreciate your comments, Matt. We might not be able to come to any agreement on this, but perhaps we can clarify what exactly the disagreement is.
Suppose a consumer’s agreement with his or her cell phone company says something like the following:
“Both sides agree to arbitrate all disputes concerning the telephone service. The arbitrator lacks the power to award punitive damages. No discovery of any kind will be permitted. The proceedings in arbitration may not proceed on a class basis. Both sides also agree to arbitrate any disputes regarding the enforceability of this agreement to arbitrate.”
The consumer says the company ripped him or her off, and the consumer wants to sue in court, arguing that the arbitration clause is invalid/unconscionable, on account of the restrictions on damages, discovery, and class proceedings. Maybe he or she is right. Who decides?
I think we agree that the unconscionability analysis is supplied by state law (though, as I think we also agree, within limits provided by federal law). I believe there is nothing illogical in saying, as a matter of federal common law, that the arguments for invalidity (which themselves are state law arguments) have to be presented to the arbitrator for decision, because here the allegedly unconscionable problems with the arbitration (no punitive damages allowed, etc.) are separable from a basic agreement to arbitrate all disputes, including over enforceability. Put differently, it would be a plausible interpretation of Prima Paint to so hold. To be sure, it would be slicing things pretty thin and making a very subtle distinction. Certainly the separation is harder to see than in the usual Prima Paint situation, where the defect that supposedly invalidates the whole contract has no particular relationship with the agreement to arbitrate.
Things are different in the Hooters or Reykjavik situation, not because those are more unconscionable in degree, but because there we don’t have an arbitral process that can be separated from the allegedly unconscionable problem with the process.
If you want to know what I really think we should do about these kinds of situations, in terms of policy, I’d say that Congress should enact some version of the current proposals for substantially restricting the use of binding pre-dispute arbitration clauses in a broad class of consumer or employment disputes. Can we agree on that?
Posted by: Aaron | Feb 24, 2010 12:24:06 PM
Yes! I totally agree that there needs to be some sort of uniformity on what constitutes fair process, and state common law is a particularly poor way of doing that. Congress should either ban arbitration, limit it, or put process protections in place that assure more uniformity.
In the meantime, I do not buy the distinction between "the allegedly unconscionable problems with the arbitration" and "a basic agreement to arbitrate all disputes." I read Prima Paint to say that if you are alleging the nonarbitral aspects of the contract to be unconscionable, then that has to be decided by an arbitrator. So, for example, if it's a consumer contract resembling the Walker-Thomas Furniture case, then the arbitrator would make the call as to whether the substantive terms of the contract are unconscionable. But I don't agree that you can separate out the decision to arbitrate from the aspects of the arbitration itself. They go hand in hand.
More importantly, perhaps, I don't see how you can distinguish the Reykjavik/Hooters examples by saying that "there we don’t have an arbitral process that can be separated from the allegedly unconscionable problem with the process." Why not? Why couldn't an arbitrator get briefs on whether a Reykjavik arbitration would be unconscionable or not? The location is just as "separable" from the arbitration decision as the question of class actions or punitive damages. So if the Supreme Court says that some terms are "separable" and others are not, it will essentially be deciding which state law unconscionability claims can be made and which cannot. Which is essentially making state law. Are we going to overturn Erie?
Anyway, thanks so much for engaging in this sustained debate. I hope you come back to report on the oral argument and the ultimate decision.
Posted by: Matt Bodie | Feb 24, 2010 1:01:34 PM
"[I]t might be that judicial review for unconscionability operates as a sort of safety valve that makes arbitration of consumer and employment disputes palatable."
There is a lot of merit to this observation. Colorado used to have a progressive judge (he died 1910) called Robert Steele (after whom an elementary school and a street in Denver are named), whose obituary noted that he was the most powerful judge on the Colorado Supreme Court despite the fact that he routinely wrote dissents as a minority of one. This is because his dissents were routinely adopted by the legislature in the form of statutory changes to the law.
A dissent in Rent-A-Center v. Jackson (which is a domestic consumer case involving an industry with a reputation for exploiting consumers and catering primarily to vulnerable or less able clientele), from a finding that unconscionability may be found by an arbitrator, rather than a court, is just the sort of case that could produce a legislative response.
Lots of otherwise normal and legal contracts have arbitration clauses. In some industries, this is the norm. They are routine signed or accepted without signature, without being read, without being part of the bargain-in-fact.
The simple reality that the vast majority of transactions do not lead to litigation makes this likely to remain true indefinitely. If there are problems only 2% of the time, it takes a very dire problem in the dispute resolution process of which consumers are aware for the dispute resolution process to have much of an economic impact, and even then, the effect on the transactions is limited only to those with knowledge of the problem which may not happen until a friend is impacted or there is a high profile case (in the functional equivalent of a "first bite" rule). If the arbitration clause contains a gag order and the arbitrator holds that it is enforceable and valid, this feedback loop in the economy is restricted even further.
The recent collapse of one of the major consumer credit card dispute arbitration firms in the face of widespread evidence of improper conduct (and the discontinuation of credit card arbitration by another major arbitration firm not implicated) also gives weight to the argument that arbitrators cannot universally and in all cases be trusted to adjudicate disputes regarding the validity of the source of their own authority.
A finding that the arbitrator makes the decision comes too close to the kind of situations presented in the Lemony Snicket books, "A Series Of Unfortunate Events" in which a room full of people sees the contract entered into improperly given effect and they are forced to accept an interpretation of the agreement that permits obviously improper, unfair and overreaching conduct towards a party to it, since it is authorized by the agreement. In this distopian world with a formalistic view of the law, a signed instrument can make improper conduct legal, even in situations where common law contract invalidity doctrines apply. As long as one has a crooked judge to determine the issue of unconsionability (and those are the cases where the arbitrator is most likely to be crooked), the federal courts have to aid and abet this theft by arbitration clause.
From an incentive perspective, a bad actor has too much to gain and too little to lose from drawing up an unconscionable arbitration provision (or contract) which predictably leads to the appointment of a bad arbitrator who has no personal liability for a biased decision and a financial incentive to keep making decisions that will get the arbitrator employment.
If the arbitrator is free to uphold and administer an arbitration contract that a court would find to be unconscionable, the temptation for bad actors becomes too great, since there is no meaningful review of the arbitrator's authority before the case is heard, or the correctness of the decision, after, the arbitrator hears the case.
Essentially, this would remove the last safety valve from an arbitration system where review even for clear mistakes of fact, mistakes of law, statistically or structurally demonstrable bias, appeal within an aribtration system, and more are not available.
When FAA arbitration's intense insulation from the court system allows aburd results to flow simply to a court interpretation of the FAA, a mere statute, legislative action is likely. Legislative action is particularly likely at a time when Democrats hold majorities in the House and Senate, and a Democrat is the President. The legislative action is likely to eliminate the only meaningful court safety valve from the system is almost certain to lead to either the prohibition of arbitration entirely in a large class of cases, or much heavier court supervision of arbitration.
Posted by: ohwilleke | Feb 25, 2010 2:40:11 PM
This is a great thread. Thanks to Aaron for starting it and thanks to Matt Bodie for thinking of and asking most of the big questions I have. I'm not a contracts guy, my experience with this issue is almost entirely limited to employment cases, and I apologize if the following questions are addressed in the full article. With those caveats. . . .
First, my instincts are similar to Matt's: a decent number of cases in the employment realm (including, but not limited to litigation involving Hooters) have arbitration agreements unconscionable precisely because the arbitrator would likely be biased. Further, even when arbitrators are not obviously biased, they arguably have a financial incentive to favor, at least at the margins, arbitrability -- after all, the more types of agreements they find valid, the more work for them generally.
Second, where is the role of precedent here? Courts, presumably, can work out rules about what sorts of agreements are unconscionable in a given state. Arbitrators not so much. Is making all those state law cases irrelevant really a plus?
Third, Aaron, when you look at arbitration cases generally, do you get the sense that state courts are using unconscionability doctrine more aggressively in arbitration cases than in other contract cases? Again, I'm not a contracts guy, but the cases I know that strike down arbitration agreements at least usually do so for what seem to me to be pretty good reasons.
Finally, is it hard to answer that question because it's hard to compare arbitration agrements to other types of contracts? Unconscionability analysis in a normal pay-for-goods-or-services contract, or even in a pay-to-perform-work contract would seem to me to necessarily have to be a least somewhat different than unconscionability analysis regarding a contract to substitute a different process for enforcing rights a federal statute already gives you.
Anyway, I look forward to reading the article.
Posted by: Joseph Slater | Feb 25, 2010 6:10:40 PM
I'm glad this post has generated so many thoughtful and sophisticated responses. Just a few thoughts ...
Regarding the intuition, voiced by multiple commenters here, that the arbitrator cannot be trusted to decide issues concerning arbitrability -
In my view, if the claim is that the arbitrator is biased, and that's why the agreement is unconscionable, then it seems like you can't (or at least shouldn't bother to) ask the arbitrator to rule on that. My sense of things is that probably most - and certainly very many - claims of unconscionability are not of that sort. (Admittedly, what we can readily perceive and tabulate is just the tip of a mostly hidden iceberg.) When we are dealing with things like limitations on remedies that supposedly make the arbitration clause unconscionable, I have argued that it would be a legally plausible (not legally required)interpretation/application of the separability principle to say that there can be a valid agreement to have an arbitrator decide whether those restrictions are enforceable. Suppose I'm right about that. Nonetheless, the commenters' basic intuition that we just shouldn't have the arbitrator ruling on that sort of thing still resonates with me. (Maybe it will resonate with a majority of the Court, who knows.) There is an air of self-dealing and being judge in one's own case.
Yet, over time, I've come to think that we can't assume that the arbitrator will take the most pro-arbitration position and refuse to strike down various limitations on remedies or procedure. As a case in point, some companies have been burned by arbitrators' willingness to let cases proceed as arbitration class actions. So now the companies are writing their clauses to get back into court class actions!
So should we allocate these rulings to arbitrators, all things considered? Bottom line is I don't know. I've just been arguing that the Court could so rule while producing a legally respectable and logically coherent opinion that represents one faithful reading of the FAA, Prima Paint, First Options, etc.
Regarding whether courts are using unconscionability differently against arbitration clauses (and how we would know if they were) -
Joseph is absolutely right that it is hard to know. Surely it doesn't suffice to point out that courts in a certain state had 20 reported decisions last year invalidating arbitration clauses as unconscionable and 2 invalidating anything else as unconscionable. Maybe the arbitration clauses were more unconscionable! You want to try to get apples-to-apples comparisons. For instance, let's find cases that are identical except one arises in an arbitration context and another arises in an ordinary litigation context. A forum selection clause specifying litigation out of state vs. one specifying arbitration out of state. A clause banning class actions vs. one banning class arbitration. That would look promising, but you aren't going to have very many cases that are the same except for the variable of interest, and there isn't a huge N here. Worse, you still might not have a true apples-to-apples comparison. For example, whether banning class proceedings is unconscionable depends on how easily somebody could proceed with their claim on an individual basis; but that depends on the cost of the individual proceeding, which will differ between arbitration and litigation.
Now, if I'm a strongly pro-arbitration Court and I suspect that certain other courts aren't applying unconscionability correctly, and yet it is nearly impossible to prove it (and it would be expressively difficult to accuse them of it), whatever could I do? My analysis suggests the Court might impose an allocation rule and cut those "unreliable" courts out of the picture.
Posted by: Aaron | Feb 25, 2010 9:57:44 PM
"Yet, over time, I've come to think that we can't assume that the arbitrator will take the most pro-arbitration position and refuse to strike down various limitations on remedies or procedure."
Arbitrators get paid for arbitrating cases. As a result, arbitrators have a strong economic interest in holding arbitration clauses to be valid -- whether or not the challenge to the clause is based on arbitrator bias. Arbitrators therefore are not neutral decisionmakers with respect to arbitrability. This seems rather elementary to me. We wouldn't countenance judges deciding cases in which one outcome would result in their being paid large sums of money and the other would result in their losing business.
If the dispute is with regard to the validity of procedural provisions short of the arbitrability of the dispute as a whole -- available remedies, etc. -- then this problem may be diminished. In the individual case, for example, an arbitrator may benefit financially from permitting a class arbitration to go forward, because class arbitrations are more elaborate and therefore more lucrative proceedings. On the other hand, arbitrators who gain a reputation for favoring expansive remedies such as class arbitration and punitive damages may lose business in the long run.
Either way, the economic interests of arbitrators should never be overlooked, and when they clearly point in one direction with respect to an issue, the arbitration is not a neutral decisionmaker on that issue.
Posted by: AF | Feb 28, 2010 1:39:13 PM
Not being a lawyer may be helpful or it may be a handicap in my reading of this case.
Hving said that, I would bet that the Court will rule that since these arbitration forms were willingly signed the people who signed them are out of luck. The Liberals will disent but the Court is so pro- business that they will find a way to uphold the arbitration argument. I see Scalia, Thomas and Roberts and Alito as some of the worst "Judges" the Scotus has ever had. One might read
"Garner v. Tennesse" to see what I mean.
Alito did not vote on that case as he was not on the Supreme Court yet, but it is easy to find his memo about the case an see where he stood. For anyone to argue that it is O.K. to shoot and kill a fleeing 15 year old, unarmed youth for a purse snatching boggles the mind. Lenny Bruce said it best."In the halls of justice the only justice is found in the halls"
Posted by: Pete Peterson | Apr 29, 2010 6:07:17 PM
For those who have read this far, note that there is additional, post-argument commentary on this blog here:
Posted by: Aaron | May 3, 2010 9:22:13 AM