Teaching Assistants: Choi, Gulati, & Scott on Commercial Boilerplate & Landmines

Stephen Choi (left), Mitu Gulati (below right), and Robert Scott (below left) have collaborated on Commercial Boilerplate: A Review and Research Agenda, which you can find on SSRN.  They aren't kidding about the research agenda thing, because they also have a book in the works about commercial boilerplate.  Mitu shared a draft of the introduction, and so I can offer some surmises in this post on the connections between the research agenda and the book.  The working title is The Contract Production Paradox.

Their scholarship is unique and exciting because, while a lot of us contracts scholars have been concerned with boilerplate contracting, we have focused on asymmetrical contracts in the consumer or employment context in which the dominant party dictates boilerplate terms to the counterparty, who accepts those terms with little or no ability to negotiate.  The Authors focus on commercial boilerplate, and their research turns up all sorts of surprises.

It turns out that boilerplate is ubiquitous in large commercial transactions involving sophisticated parties. Here too, the lawyers do not review the boilerplate, nor do they negotiate over boilerplate terms. Why? Because they are in a hurry. The transactions are complex; the assets being traded may fluctuate in value, and like most of us, they either assume that the boilerplate terms are good enough or the costs of careful negotiation outweigh the litigation risk that perhaps-faulty boilerplate terms might entail. 

Their literature review covers the early discovery in the law and economics literature that even sub-optimal boilerplate terms could be sticky; that is, attorneys continued to use the terms, notwithstanding their faults.  But early scholarship assume that the terms that survived tended to approach optimality.  In complex loan transactions, standard terms meant that one could trade loan instruments quickly without reviewing terms, confident that the effect of the boilerplate provisions was well-understood and that their value had been priced.

A second generation of scholarship discovered that the reality departed from the model.  Sticky terms were used despite their sub-optimality and they were not in fact well-understood and could in fact be challenged by opportunistic litigants.  These provisions came to be known as "black holes," presumably because their meaning was impenetrable and yet they could not be easily removed without causing the surrounding deal structure to collapse.  But it gets worse.  The standard language turns out not to be standard after all, and so one cannot even assume that the standard boilerplate provisions, regardless of their opacity, have some accepted meaning that can be priced.

The Authors then turn their attention to the process whereby the boilerplate is made, and this part of the Review and Research Agenda introduces the main theme of the Authors' forthcoming book on commercial boilerplate.  Inattention to the mode of contract production transforms boilerplate "black holes" into "landmines." Transactional lawyers assume that boilerplate clauses are both fixed and well-understood. They are neither. And as slight changes slip into common boilerplate provisions, opportunistic lawyers can pounce.   

Still, the Authors note that there will always be a trade-off in contract design between the high production costs associated with bespoke contract drafting and the accidental inefficiencies associated with adopting boilerplate provisions, which might not be the right fit for the transaction (see related work on "alien vomit") or might be corrupted in ways that are not easily detectible in the hurly-burly of transactions negotiated under time pressure.  The more common the transaction, the more likely it is to be larded with landmine boilerplate provisions. 

The authors describe the process though which such landmines come into being in a context they have studied carefully, sovereign bond contracts.  They illustrate the effects of such landmines through a discussion of the impact of a misunderstood pari passu clause, a landmine triggered in 2011 in connection with the Argentinian debt crisis.  They have created a typology of landmines: historical holdovers, random errors, subversive accretions, and obsolete provisions.  They conclude with a list of eight emerging areas of research, followed by a dozen pages of references that will no doubt serve as a mandatory reading list for other scholars in this area.

It looks like the related book project will focus on the mechanics of commercial contract production.  The Authors argue that we need to improve our understanding of the tradeoffs between efficiency and tailored contract drafting in order to understand the provenance of boilerplate terms.  They illustrate problems with commercial boilerplate that have arisen in sovereign debt instruments.  They then provide evidence that similar landmines exist in other types of contracts dependent on commercial boilerplate.  The review and research agenda mentions interpretation issues that can arise in connection with boilerplate terms.  Knowing how a landmine got into a contract might indeed be important to a court looking beyond the text to the intentions of the parties. They return to that subject in the proposed book's concluding chapter.

I will note for the record to David Hoffman (right) has also posted about this article on Jotwell, and I wanted to complete my own assessment before reading his.  Having turned my attention to Professor Hoffman's piece after completing my own, I attach the following addendum.  You really should have read Professor Hoffman's take on this first.

In this post, I have refrained from commenting on the possible impact of AI on commercial boilerplate, assuming that one of the authors of Generative Interpretation would take up that challenge. Mere mortal that I am, I can only imagine that AI tools already at hand could now be deployed in a manner consistent with the Authors' work.  Transactional lawyers assume that commercial boilerplate is unchanging and and unchallenged.  It is neither.  Armed with that knowledge, they can use AI tools to efficiently police their boilerplate provisions for variations or even create a genealogy of the provisions and thus perhaps cull the alien vomit.  Following Hoffman and Arbel's work, one could also presumably use generative AI to predict the likely interpretation of boilerplate terms.