According to prevailing conceptions, the primary role of contract law is to give effect to the parties’ will (the so-called will theory of contract), thereby enhancing overall human welfare (the standard law and economics perspective). Thus, the law may legitimately intervene in the content of contracts, or otherwise try to influence the contracting process and its outcomes, only if there is some flaw in the contracting process (such as duress) or if there is a market failure (such as a monopoly or an acute information problem). In recent years, the notion of market failure has been extended to encompass behavioral market failures as well—that is, deviations from the assumption that people are invariably rational maximizers of their own utility. However, it is still commonly believed that the need for compulsory (mandatory rules) or choice-preserving interventions (nudges) is limited to transactions with relatively weak and unsophisticated parties—such as consumers, employees, and tenants. When it comes to commercial transactions in competitive markets, the very fact that a contract does or does not contain a given term is perceived as a proof that that term (or its absence) is optimal—that is, maximizes the joint surplus of the parties. Otherwise, why would sophisticated parties include (or fail to include) that term in the contract? In fact, some scholars have grounded an entire theory of contract law on such strong belief in the rationality and competence of commercial contracting parties (Schwartz & Scott).
Until recently, people’s beliefs about these issues were primarily based on anecdotal evidence, personal experience, and ideological inclinations. With the advancement of empirical legal research, observational and experimental studies offer more reliable and systematic evidence about such matters (which does not mean, of course, that ideology ceases to play a role, as people often do not allow themselves to be confused by the facts). In his intriguing new research, Julian Nyarko uses cutting-edge methods of machine learning to study the inclusion or non-inclusion of choice-of-forum clauses in hundreds of thousands of contracts contained in a dataset of commercial agreements reported to the U.S. Securities and Exchange Commission (SEC). He then examines what factors might explain the inclusion or non-inclusion of such clauses in any agreement.
The study found that, while 75% of the contracts in the dataset include a choice-of-law provision, only 44%—less than half—include a choice-of-forum clause. (P. 32.) The latter figure is surprising, given the potentially large practical importance of choice-of-forum clauses (Pp. 10–19) and the prevalent opinion among practitioners that failing to include such a clause verges on malpractice. (P. 63.) Surprisingly, firms do not show consistency in this regard. While very few firms consistently include (or fail to include) such clauses in all (or any) of their agreements, most firms display considerable variance in this respect: the consistency measure—ranging from 0 (choice-of-forum not included in any of the firm’s contracts) to 1 (choice-of-forum included in all of the firm’s contracts)—is normally distributed around 0.5. (Pp. 33–34.) Similar inconsistency is revealed when breaking down the data by industry. (P. 35.) Unsurprisingly, choice-of-forum clauses are more prevalent in some types of agreements than in others (e.g., 61% in joint-venture agreements, compared with only 47% in transportation agreements). (P. 36.) Such clauses are also more prevalent in contracts drafted by leading law firms. (Pp. 38–40.)
If neither the consistent preferences of firms nor the types of industry or agreement explain the decision as to whether or not to include a choice-of-forum clause in an agreement, are there other observable variables that can explain it? It turns out that whether or not a choice-of-forum clause is included in the contract is largely determined by the lawyers that draft the contract. Moreover, in this regard lawyers do not appear to prioritize their own interests over those of their clients. Rather, there is strong evidence to suggest that “whether or not the final contract includes a choice-of-forum provision is determined almost exclusively by the template that a law firm uses.” (P. 7.)
Admittedly, given the observational (rather than experimental) nature of the study, there was an outside chance that “reverse causality” might be at play—namely, that firms hire lawyers who use the most beneficial template for each of their deals. However, the study largely rules out this theoretical possibility by examining the impact of two “external shocks.” One is the fact that some law firms collapsed during the period of observation, thus forcing firms and clients to hire new external counsel. (Pp. 44–46.) The other is significant changes in the legal default rules concerning the courts’ jurisdiction, which appear to have had no noticeable impact on the inclusion or non-inclusion of choice-of-forum clauses. (Pp. 50–58.)
The author readily (and commendably) concedes the limitations of the study. (Pp. 58–63.) For one thing, the study pertains to a particular type of term, which usually does not attract the attention of the negotiating parties, or even their lawyers. More studies are necessary to examine the generalizability of the findings with regard to both primary and secondary contract clauses. For another thing, the study only analyzes observable variables—primarily those that can be extracted from the dataset—and one cannot rule out the possibility that the inclusion or non-inclusion of choice-of-forum clauses in contracts is determined or mediated by other, unobservable factors.
Notwithstanding these limitations, the findings are intriguing, and their potential policy implications important. First, the findings highlight the role of lawyers in commercial transactions—thus questioning the common treatment of each contracting party as a unitary entity. The findings also cast doubt on the attempt to derive normative conclusions from observations of prevailing provisions in commercial contracts (see, e.g., Benoliel).
Furthermore, the strong evidence suggesting that whether a choice-of-forum clause is included in a contract is determined by the language of the boilerplate used to prepare the first draft (rather than by any rational deliberation), lends support to the observation that cognitive heuristics and biases—in this case, the status quo and omission biases that result in a default effect (Zamir & Teichman, Pp. 48–50)—are not limited to laypersons making mundane decisions (see Id., Pp. 114–17). They affect even top-tier professionals who charge large sums of money to handle transactions worth millions of dollars. This observation must be taken into account even if one is solely interested in maximizing overall social utility, to the exclusion of any other normative concern. Among other things, it bears upon the design of legal default rules, and on the appropriate division of labor between the contracting parties and the law (see also Zamir; Ayres).
Finally, Nyarko (P. 65) aptly offers a more general lesson for legal theory in contract law and beyond: rather than trying to theorize away gaps between expectations and reality, we should try to understand these gaps.