Tuesday, June 02, 2009
Discovery, burdens, risks, and Iqbal
The defense-attorney-bloggers at Drug and Device Law have a detailed post in favor of the Court's new approach to pleading in Twombly and Iqbal (H/T: Civil Procedure Prof Blog) (see my discussion of Iqbal here, here, and here) (see other criticisms here and here).
The point of departure between D&DL and critics of the new pleading regime (including me) ultimately comes down to placement of litigation burdens and risks.
Looser pleading recognizes that plaintiffs often lack critical information at the outset of litigation and need discovery to learn key information that shows the merit (or lack thereof) of their claim. But it burdens the defendant with costly and wide-ranging discovery. D&DL argue that Conley ceased to be workable with the broad changes to the discovery rules that began in 1966 and 1970 (the procedural flipside to my argument about the expansion of substantive federal law post-1938). Looser pleading produces so-called Type I errors--morecases get past pleading into discovery, at cost to the defendant, only to be found to lack merit.
Stricter pleading relieves defendants of the burden and expense of discovery in cases in which there likely is no there there, as indicated by the complaint. But this burdens (unfairly, in my view) the plaintiff with 1) the often-impossible task of framing a detailed complaint when she only can get the necessary evidence through discovery and 2) the risk of dismissal when she is unable to provide the necessary detail in the pleading (this is what was going on in Twombly and, potentially, Iqbal). Stricter pleading produces so-called Type II errors--potentially meritorious cases are deterred or dismissed with the plaintiff never having had a chance to fully explore the evidence in support of the claim.
So the point of departure is at the policy level. Do we prefer Type II errors imposing burdens on plaintiffs or Type I errors imposing burdens on defendants? And the answer will not be value-neutral. But the presence of a non-neutral policy question highlights two final points.
The first goes to a different aspect of the evolution of law, and thus procedure, since 1938. Seventy years ago, parties were largely interchangeable. A person or business entity was as likely to be a defendant as a plaintiff. So big business saw less of a need to push a defense-favorable view of the procedural rules, because a business might find itself as a plaintiff enjoying the benefits of notice pleading. There is far less interchangeability today--corporations and government are almost always defendants (and repeat defendants at that) who know they will almost exclusively enjoy benefits from a defense-favorable pleading regime.
The second goes to a procedural criticism of Iqbal and Twombly. To the extent the choice between the old Conley system and the new Iqbal/Twombly system is a policy determination, it should not come about through a decision in a case, but should be established through the notice-and-comment, increasingly democratized and political, Rules Enabling Act process.
TrackBack URL for this entry:
Listed below are links to weblogs that reference Discovery, burdens, risks, and Iqbal:
D&DL's critique, as I read it, lends weight to greater control of, or even rule changes to, discovery --- not a change in pleading.
Their main point, which as a former defense attorney I saw as well, is that discovery is expensive and is often used as a strategic weapon. Let us assume that is true sometimes.
Changing pleading standards seems an odd prescription for this ailment. Especially when so many of our statutory causes of action require proof of defendant's intent regarding issues that are not obviously inferred from surrounding facts (compare negligence in a car accident with racial animus in a reduction in work force). There is simply no way a plaintiff can ever have solid facts regarding a defendant's racial animus, say, without discovery.
If discovery is the problem, fix discovery. For example, if a Title VII case appears not well founded, a judge might be well served to limit discovery in the early stages or adopt a practice of hearing multiple Rule 56(c) motions. That is, give plaintiffs in such cases a quick (read cheap for defense) shot at producing some evidence of defendant's state of mind. Disallow any other discovery until at least some evidence on that score comes in. This would further the policy of on-the-merits review that permeates the FRCP.
Of course, this expands the role of the managerial judge. But we are already well down that path.
In the end, I agree with Howard. If Iqbal is the best approach because discovery is too expensive for institutional defendants, then that evidence should be presented. It was not in Iqbal. The best forum for that type of discussion is the quasi-legislative forum of the rules committee.
This is a complex area requiring the legislative forum. For example, one study of nearly 1000 civil cases involving some discovery expenses found discovery responsible for, on average, half of the total cost of litigation. See Thomas E. Willging et al., Fed. Judicial Ctr., Discovery and Disclosure Practice, Problems, and Proposals for Change 15 (1997). (I am not sure whether that is high or not, given that cases only go to trial 1% of the time.) In any event, this 50% tab is not consistent across types of cases. For example, complex patent cases often run discovery costs of 13% of the value at issue (a better measure of cost I think than raw litigation expenses). While follow up suits in mass torts cost very little in terms of discovery, because information from the lead case is reused. Simple tort cases are cheap on discovery. And on and on and on.
All this is to say, a simple "discovery is too expense" response may be true --- but it is not universally true, nor do I believe it has been proven to be true most of the time. Now it might be too expense in certain practices, say patent law, but we should prove that. All these questions cry out for a solid base of statistically data that I am unqualified to provide. All the more reason for this to be done at the committee.
Great post, Howard.
Posted by: Lumen Mulligan | Jun 2, 2009 9:09:06 AM
"[C]orporations and government are almost always defendants (and repeat defendants at that) who know they will almost exclusively enjoy benefits from a defense-favorable pleading regime."
Literally speaking this is clearly untrue. Corporations and governments are the predominant plaintiffs as measured by cases filed. But, the cases typically brought by corporations, which also make up a large share of cases typically brought by government, are collection/eviction/foreclosure cases which require little discovery and very frequently produce default judgments or very modest scale litigation, mostly due to how substantive contract law determines what is relevant and allocates burden of proof between plaintiffs and defendants.
Moreover, many civil cases brought by governments which are not of this nature (civil forfeiture, civil tax and securities enforcement cases, termination of parental rights, etc.) are brought by quasi-prosecutorial bodies that substantial pre-case commencement powers to compel adverse parties or third parties to provide them with information.
There are also large classes of litigation in terms of volume of cases filed (e.g., domestic relations) where corporations and governments are almost never parties of any kind.
Corporations and governments are predominantly defendants in many types of tort and employment cases (those typically brought by what are colloquially called trial lawyers, despite the fact that government lawyers and criminal defense lawyers and collections lawyers actually conduct the lion's share of trials these days). But, it is worth recalling what one really means when one says that a rule is plaintiff or defendant friendly
Posted by: ohwilleke | Jun 5, 2009 2:17:58 AM