« Some sad news | Main | Supreme Court grants cert in UT Affirmative Action Case »

Monday, February 20, 2012

Stoner Law Reform: Fee-Shifting

This week, I'm going to post some stoner law-reform proposals. Sometimes, you need to remove your own common sense to imagine how the world might be different. And what better way to do that than stoner logic?

First up, consider fee-shifting. Critics complain, and rightly so, that the American rule of each side bearing its own costs is bad for plaintiffs with good claims. They may find it too expensive to vindicate their rights. But while we've picked up a variety of fee-shifting statutes here and there, we've stubbornly resisted the English rule, in which the losing party must pay for the winning party's lawyers. Critics complain, and rightly so, that the English rule encourages overspending and can put unbearable pressure on parties facing a well-financed opponent.

I asked my inner stoner, and he said, "What if the loser pays its own fees to the winner?" In essence, this rule means that the loser ends up paying double its attorneys fees: once to its own lawyers, and once to the winner. The English rule tries to make the plaintiff whole. But that's really hard to get right and it creates weird incentives. A stoner would rather just charge the loser what it paid, call it close enough, and order some pizza.

I think he might have a point. Where the parties are equally matched and spending evenly, loser-pays-double rule is ex post equivalent to the English rule. But ex ante, it dials up the incentive to get the lawsuit done cheaply. Where the parties are mismatched, loser-pays-double looks even better. A pro se party up against a behemoth faces no risk of a crushing fee award. Its wealthy opponent knows that every dollar spent on intimidation only increases the little guy's potential payday. Loser-pays-double also answers the criticism that the English rule can result in wholly disproportionate fee awards: a party's potential fee payout is never more again than it has already spent.

Crazy, or so crazy it might just work?

Posted by James Grimmelmann on February 20, 2012 at 10:37 AM in Civil Procedure | Permalink


William, various United States fee award provisions award fees to attorneys who weren't paid up front. That's how class-action fee awards work in effect. Yes, they come out of the class's "recovery," but that recovery itself came from the defendant.

TJ, I'm not assuming that the parties are generally equally matched. My argument is that where they are, loser-pays-double reduces to the English rule. But where they're mismatched, it tends to create more upside for the David than the Goliath, which matches our intuitions about how the system ought to work. I think you're right that the second-order incentive effects may be a strong argument against loser-pays-double: it incentivizes expenses that force the other side to respond disproportionately. Perhaps one response is that defendants who are confident of their cases wouldn't have a reason to fear this effect, but unfortunately our patent system makes it all but impossible for defendants to feel truly confident in their cases, no matter how strong.

As for your specific example of discovery costs, I have a simple answer (which I considered adding to the Stoner Law Reform series but couldn't turn into a coherent proposal): unilateral disarmament. Yes, there are technical expenses to comply with discovery requests. But a defendant worried about overspending should just ask for a protective order and then not bother with a relevance or privilege review. Shovel everything over to the other side and let them deal with it. My suspicion is that this particular form of document review isn't a terribly good value proposition for clients, and that it's going to be under very strong pressure, especially from learning algorithms. How many years before Fortune 500 companies instruct their law firms to run the half-million emails through a classification algorithm and not to have humans review more than a small sample of the results?

AF, the "whichever is less" variant has been proposed a couple of times in the literature. It creates some surprising and discontinuous incentives -- e.g., once you've outspent your opponent, you might as well keep going. And in any event, I had to make my proposal different so it would be novel. :-)

Posted by: James Grimmelmann | Feb 21, 2012 11:50:00 PM

This is an interesting proposal, but I'm not sure it gets at the heart of the matter. In particular, I'm not sure the American and English rules really have that much impact on litigation spending, once the decision is made to bring suit. There is only so much you can do to control litigation costs, particularly as a defendant; you are obligated to respond to discovery requests, and while you can choose not to bring dispositive motions, that only serves to make the expense of trial inevitable. Plaintiffs, for their part, particularly resource-constrained plaintiffs with relatively small damages claims (ie, those who are theoretically most likely to be affected by fee-shifting), seem to be motivated more by immediate resource constraints than by the prospect of fee awards, even when the applicable law provides for fee-shifting. This is particularly true given that, as others have pointed out, actual awards are typically based on ex-post facto assessments of reasonable fees, rather than on actual expenses.

The more important difference between the two regimes seems to be their effect on the decision whether to litigate in the first place. Relative to the American system, the English system increases the financial risks and rewards of litigating. All else equal -- which of course it is not -- this favors rich parties, who are better able to bear the risk of paying a fee award if they lose, and parties with good claims, who are more likely to have their fees paid than to pay the other side's fees. At the decision stage, however, the cost/benefit analysis is necessary imprecise. The difference between the English rule and the above proposal would seem to fall within the margin of error in most cases.

One thing the proposal does seem to do is make it less risky than it is under the English rule for plaintiffs to bring claims in situations where defending the claims costs more than bringing the claim. The problem, as TJ points out, is that this overcompensates plaintiffs and creates a perverse incentive for plaintiffs to run up the defendants' costs (which is actually easier than running up the plaintiffs' own costs, as it entails less financial risk). Better solutions to the problem of "crushing fee awards" imposed on resource-constrained losers would seem to be one-way fee shifting regimes like we have in some areas in the US, or perhaps a regime where the losing party pays either its own or the other party's expenses -- whichever is less. Or, of course, the American rule.

Posted by: AF | Feb 21, 2012 2:06:36 PM

It is an interesting idea, but I question your assumption that "the parties are equally matched and spending evenly." Take patent litigation, with which I am most familiar. A patent defendant has very high discovery costs because of the enormous volume of documents that must be produced and reviewed for privilege. In contrast, patent plaintiffs generally have low costs because--especially in cases of non-practicing plaintiffs--they have almost no documents to produce, and reviewing the defendant's documents can be done by word searches on computer. At least at first look, therefore, it seems the effect of your proposal in this kind of asymmetric cost structure, which I suspect is more general than just patent law, would be to significantly increase plaintiff leverage in nuisance settlement value.

Posted by: TJ | Feb 21, 2012 3:08:52 AM


The other difference is that under the current English rule, one never has to pay for the other side's contingency fees, because a lawyer only gets a contingency fee if he prevails, but his side only owes attorneys' fees if he loses. (And it's hard to know what the contingency fee of a claim that failed "would have" been if the claim hadn't failed.) But I agree that one of the many other methods could be adapted to capture this.

Posted by: William Baude | Feb 21, 2012 12:59:39 AM

In England they have fee schedules that largely take care of the calculation problem. My understanding is that the lawyers don't get the fees they actually charge but what is determined to be an appropriate fee. This is not dissimilar from how civil rights fees are awarded. Even if a firm does private work, they usually don't get a fee rate that equals what they can get for their time on the market.

I love this idea. It requires some thinking through regarding the asymmetric spending problem. Some things are more expensive for one side than the other (its easier to draft an answer than a complaint, generally favoring defendant; easier to draft document requests than comply with them, generally favoring plaintiff).

Posted by: Aprof | Feb 20, 2012 10:45:00 PM

I know nothing is perfect, but if you're interested in cost-shifting, the English rule seems more appropriate for a dismissal at the MTD stage. I've seen complaints that looked to be written in three or four hours require many (many) times that to decipher and (successfully) oppose. How about you pay your opponent's full costs if you lose on the pleadings (12b6/12c), you pay an amount equal to your costs if you lose at trial, and the judge gets to decide which rule is more equitable at the summary judgment stage.

Posted by: brennan | Feb 20, 2012 9:46:26 PM

This problem isn't unique to loser-pays-double. It also arises whenever a courts is required to decide what portion of a winner's fees are allowable under the standard loser-pays rule. So a court could presumably impute reasonable fees for any work done under a salaried or contingency-fee arrangement.

The difference might be in the reporting incentives. Under standard loser-pays, the party hoping for a fee award will keep careful and detailed logs to substantiate its fee request. Under modified loser-pays, parties will be tempted to do off-the-books work to minimize any award of imputed fees, and that could be harder to police.

Posted by: James Grimmelmann | Feb 20, 2012 3:08:42 PM

This isn't necessarily a refutation, but I wonder how the rule would deal with legal work that isn't done by the hour-- for the defense side, I think government attorneys, in-house counsel, attorneys hired for flat-fee retainers to defend a bunch of cases, etc. For the plaintiff's side, I think of contingency fees. If the plaintiff loses a contingency fee case, his lawyers get zero, so the defense gets zero too.

If, as I suspect, the modified loser-pays rule would result in much less compensation in the case of these kinds of arrangements, presumably it would have the effect of creating an incentive to shift to them. That might not be a bad thing-- lots of people think we should get rid of the billable hour!-- but it strikes me as worth exploring and debating.

Posted by: Will Baude | Feb 20, 2012 1:09:48 PM

The comments to this entry are closed.