Research Overview

Contract theorists are just beginning to systematically unpack the behavioral implications of contract design and implementation. Practitioners regularly struggle with these issues. Take the example of Oeresund A/S -- a joint venture between the Swedish and the Danish state, which was created to construct a "link" between Copenhagen, Denmark and Malmö, Sweden -- and one of its contractors. Given the size of the bridge and the sometimes adverse weather conditions, the bridge required special cables and damper technology to keep vibrations below a given safe threshold that exceeded the current state of art. The company offered its contractor an attractive price with a schedule of deductions to be implemented to the extent that the contractor failed to achieve the stipulated and never-before-achieved threshold. Both the company and the contractor recognized that the likelihood of meeting the threshold was close to zero. Nonetheless, by exerting greater effort, the contractor could reduce the deductions. Contracts, like the one used by Oeresund A/S, where breach is almost a certainty, are sometimes called Cadillac contracts because they call for the highest possible quality level. In hindsight, company officials questioned whether the Cadillac clause undermined the contractor's performance by its very terms, which the contractor reported as unfair (see external page Yale SOM case study).  

Evidence from the field suggests that framing contracts in a manner that makes "losses" more salient than "gains" leads to greater effort in teams. These effects are also present at the individual level, and is driven by two distinct behavioral channels: an endowment effect and a communication of expectation effect (see external page article). There are, however, further findings that the interaction of these two behavioral channels makes loss framing in contracts risky. By stipulating a demanding but realistic “stretching threshold”, a principal can maximize the mean effort of his agents. However, setting a threshold that is either too modest or too extreme might lead to lower effort levels than setting no threshold at all. This has an important implication for contract drafting: Unless the principal has nuanced information about his agents’ abilities and available technology – information which is often private and unobservable – he might not be confident enough about being able to set the right “stretching” threshold. In this case, setting no threshold at all might be the better choice. Moreover, there are instances in which the principal might want to set a modest threshold if he cares about minimizing the variance of agents’ effort levels. These insights not only have applications to contract design, but also to the design of regulatory regimes (prescription drug cost targets for doctors, safety standards), or international agreements like the Paris Agreement on Climate Change (climate change goals) (see external page article).  

In one project we offer an ex-ante view of rescission followed by restitution. It is commonly held that if getting a contractual remedy is costless and fully compensatory, rescission followed by restitution would not exist as a remedy for breach of contract. This claim, we demonstrate, is not correct. Rescission and restitution offer more than remedial convenience. Rational parties, we argue, would often desire a right of rescission followed by restitution even if damages were fully compensatory and costless to enforce. The mere presence of a threat to rescind, even if not carried out, exerts an effect on the behavior of parties. Parties can enlist this effect to increase the value of contracting (for further information click external page here, or external page here).

In a current doctrinal project we revisit the idea of going off contract as a strategy for class actions against internet companies violating their privacy policies. A typical case would involve a privacy policy stating that the company will not sell information to third parties that do not adhere to at least the same level of privacy protection. But then - in violation of this privacy policy - the company sells information to unqualified third parties. The Federal Trade Commission (FTC) is charged with enforcing adherence to these privacy policies (following the safe harbor agreement with the EU and Switzerland). Yet, it seems that the prospect of FTC enforcement insufficiently deters companies. Another route to enforcing privacy policies has been to bring class actions for breach of contract, for common law fraud or because of violations of a statute that gives individuals the right to sue. However, such cases have been unsuccessful for failure to establish harm (see, e.g., the US Supreme Court’s 2016 decision in external page Spokeo, Inc. v. Robins). The direction these attempts at private enforcement have been taking is for plaintiffs’ lawyers to try to convince courts of accepting a broader theory of dignitary harm. So far, this strategy has not been successful. We argue for a hitherto unexplored strategy that sidesteps these difficulties, namely, that plaintiffs should seek remedies off contract. While it may be difficult to prove that a plaintiff suffered harm from a company selling information to third parties in violation of its privacy policy, it might be much easier to prove that a company profited from this violation and convince courts to disgorge the profits derived from this violation to avoid unjust enrichment. Such a strategy would shift the focus, away from proving harm, to proving that companies benefitted from wrongful actions. If successful, this approach would eliminate a significant enforcement deficit in the regulation of one of the most important modern industries.

We argue that proportional liability, a rule that discounts damages with the probability of causation, should be applied to solve difficulties in proving causation in medical malpractice. We show ways of how to correctly calculate damages if applying this rule to medical malpractice (for further information click external page here, or external page here). We also demonstrate in a mathematical model that, under certain conditions, a regime of proportional liability which sets the measure of damages equal to harm discounted by the probability of causation is a very robust rule, that is, it sets the correct incentives in the presence of many imperfections in the legal system.

We study how promisees’ expectations created by previous promises influence the decision of promisors to keep their promises. We present data from a lab experiment that suggests that higher performance expectations lead to higher levels of promise keeping if those expectations were caused by the promise (see external page GEB paper) we demonstrate three related results: First, promisors are more likely to keep a promise, when promisees relied on this promise (and third party observers indicate a higher willingness to punish a promisor who has broken a relied upon promise). Second, promisees anticipate this effect and overinvest in reliance on promises in order to psychologically lock in the promisor. Third, introducing a legal regime reduces overinvestment as parties no longer need to use the extra-legal mechanism of psychological lock-in to create commitment. Reducing overinvestment might therefore be an underappreciated benefit of legal regimes enforcing (click here for the external page JLS paper). We also showed that promisors keep their promises for three reasons: First, they feel that they are duty-bound to keep their promises regardless of whether the promisee expects them to (the "promising per se" effect). Second, they care about not disappointing the promisees' expectations, regardless of whether those expectations were induced by the promise (the "expectations per se" effect). Third, they avoid disappointing promisees' expectations even more when those expectations were induced by a promise (the "interaction" effect). These results are consistent with the centrality of the concept of promising in contract law, and put some prominent philosophical theories of promise keeping to an empirical test (click here for the external page JLE paper). Currently we have a number of research projects how these insight can inform institutional design.

Scholars have also considered costs of these advances, such as the privacy implications of big data methods and ways in which algorithms might have deleterious consequences for equality. However, while there has been much theorizing about the legal implications of algorithms and big data, there has been comparatively little empirical-scientific study of the way in which these methods may affect law. Many of the larger theoretical and policy issues depend on answers to open empirical questions. We explore the experimental science of big data, evidence-based methods, algorithms, and law, particularly in the context of contracts, promises, and agreements.