2023 Theses Doctoral

Predicting damage awards: Toward economic efficiency and judicial consistency

Giaoui, Frank S.

From the perspective of both plaintiffs and defendants, the measurement of damages quantum is obviously of the utmost importance. Therefore, it is surprising to see this process left entirely to the court’s discretion — especially since the quantum is traditionally considered a question of facts. The result is that each litigation becomes a unique case calling for a sui generis outcome.This approach leads to a structural uncertainty that is detrimental to the legitimate expectations of both parties. In practice, it deeply corrupts the fundamental principle of full recovery. In order to address this issue, I generate descriptive damages models and argue for a model in which the valuation of damages will be a question of both facts and law that follows rules and methods whose application will be reviewable. I begin to explore specifically damages for breach of contract when they are difficult to quantify, using two simultaneous methodologies that would later be supplemented by several extensions.

The first methodology is a comparison between French civil law, American common law and international commercial law (Chapter 4), and the second is an empirical study involving both qualitative interviews with practitioners and the quantitative analysis of a proprietary sample of cases in which damages are difficult to measure (Chapter 5). Then I select two clusters of cases to implement these novel methods for Breach to Agreements to Agree/Negotiate (Chapter 6) and Non-Pecuniary Harm to Reputation (Chapter 7). Building upon these works, I implement four methodological extensions (Chapter 9) that use technological advances to address issues of sourcing and coding legal data by identifying the key variables of each legal case and automating mass extraction of data with an emphasis on scalability and transferability to other bodies of law and comparative jurisdictions.

I conclude with recommendations for judicial practices and a discussion of the possibility of predictive justice through shared compensatory damages schedules and artificial intelligence models.