Legal scholars often are criticized for devoting hundreds of pages to theorizing about mountains that turn out to be molehills when examined in the light of data drawn from practice. Scholarship at the intersection of contracts and civil procedure, to which we have contributed, has been accused of falling into that trap.
Six years ago, we published a law review article on a phenomenon called “contracting for procedure,” defined as the practice of using contracts drafted prior to the onset of any dispute to set out procedural rules for disputes adjudicated in the public courts. We argued that it is important to consider how this practice affects public interests as well as the narrower private interests of the contracting parties. This means taking into account, for instance, how the workloads of taxpayer-funded courts are influenced by contractual provisions such as forum-selection clauses, and how public perceptions of the perceived accuracy and fairness of judicial decision-making are influenced by clauses that alter rules on pleading, discovery, jury trial rights, or class actions.
Our article appeared among a flurry of other articles on the topic, some of which covered contracting over procedures for arbitration as well as litigation. Many (but not all) of these articles were, like ours, typical of traditional legal scholarship: long on discussion of the relative importance of values like party autonomy, efficiency, and public interests, but short on facts.
Fortunately, the predominantly normative and theoretical interventions provoked empirical responses. A handful of enterprising law professors actually took the time to investigate how much contracting for procedure in fact goes on, and to the extent it does, what form it takes. So far, their findings suggest contracting for procedure is not very common at all. This in turn suggests that concerns about contracting for procedure are overblown and legal scholars should find more pressing topics to debate.
We are not convinced. We do not necessarily disagree with the authors’ analyses of their data, but we are not convinced that they have looked at the right data. We are blogging about this topic to draw attention to a frequently overlooked feature of our original article: a call for courts to collect data on the extent to which parties before them have engaged in contracting for procedure.
We are particularly intrigued, and impressed, by the empirical studies conducted by David Hoffman and Mark Weidemaier. Both these studies rely on datasets consisting of executed contracts. Both Hoffman and Weidemaier examine contracts filed with the U.S. Securities and Exchange Commission pursuant to issuers’ obligations to file material contracts. Everyone acknowledges that material contracts are not representative of the entire population of contracts subject to resolution in United States courts. At the same time, a dataset of material contracts is not a bad place to look for evidence of contracting for procedure; if there is any context in which it is likely to be worth going to the effort of specifying procedural rules prior to a dispute it is in the context of a high-stakes agreement. Hoffman also analyzes a dataset of consumer credit card agreements.
The two studies have similar findings, but for the sake of convenience we will focus here on Weidemaier’s article. He finds that “almost 90% of contracts modify the background rules of litigation in some way.” However, the vast majority of these modifications involve arbitration and forum-selection clauses. Waivers of the right to jury trials and loser-pay attorney’s fees are also common, but few contracts in his dataset address matters such as discovery, pleading, or evidence. In other words, it was rare for contracts in his dataset to modify rules of judicial procedure “in the classic sense,” meaning rules concerning pre-trial and trial practice. We do not define contracting for procedure so narrowly, since we believe that clauses such as forum-selection clauses and jury trial waivers touch upon important public interests. For present purposes, however, we will work with the narrower definition.
In light of their findings, both Weidemaier and Hoffman suggest that scholars ought to devote more effort to understanding why contracting for procedure, in the narrow sense, is so rare. This essentially means analyzing why contracting parties may or may not find it in their interests to contract for procedure. Along these lines, Hoffman and Weidemaier speculate that contracting for procedure is rare because it is generally costly for contracting parties to specify procedural rules in advance, as opposed to leaving those rules to be specified by the adjudicator, especially when it is unclear whether provisions will be enforced.
Studies based on databases of executed contracts have an inherent limitation—we do not know what proportion of those contracts end up in court. This is not a problem for researchers interested in the private benefits of contracting for procedure. If the research question is, “How do private parties balance the costs of drafting procedural provisions against the benefits of contracting for procedure?” then a dataset of executed contracts is the right place to look. A database of executed contracts is less appropriate, however, for assessing how contracting for procedure affects public interests — the focus of our original article. (It also does not illuminate how a related practice is affecting other areas of law, the topic of a more recent article that one of us has written together with Marcel Kahan on the emergent role of forum-terms in corporate charters and bylaws.) To the extent we are interested in the impact of contracting for procedure on courts, then we need to focus on the subset of contracts that actually affect courts. Sometimes, those will be the contracts that you least suspect. For instance, in a recent study of federal court dockets, we, together with Nathan Yaffe, discovered a number of cases in which parties who had agreed to resolve disputes through arbitration before foreign tribunals turned to United States courts for assistance with discovery.
When we wrote our original article, we knew too well how little we knew about the prevalence of contracting for procedure—the practice is private, the relevant documents generally are not available for public analysis, and when judges write opinions they do not necessarily specify whether they have been asked to apply privately supplied procedural rules. That is why just about the only firm conclusion we drew was that the matter deserved further study. We also made a concrete proposal about how to conduct that study: require courts to collect data. Our proposal is administratively simple. We recommended that the civil cover sheet, a form filed in connection with the commencement of every lawsuit in federal court, be amended to require the plaintiff to disclose whether the parties have opted out of any public rule of civil procedure.
If implemented, this proposal would produce a dataset more suitable for studying the impact of contracting for procedure on courts than any dataset based only on executed contracts, for it would be limited to contracts that generate disputes filed in court — the ones that affect most directly the public’s interest in courts. We hope that policymakers take notice.