Overview of the Economics of Appeals

• The Gap in Literature: Professor Hylton explains that while the law and economics literature has long studied the incentive to litigate at the trial court level, there was no basic, foundational model explaining the private incentives to appeal a judgment.
• The Model: The researchers developed a skeletal framework to model why plaintiffs and defendants choose to appeal. The model posits the trial court judgment as an "option" or safety net that disappears once an appellate court renders a decision.
• Incentive Conditions: The decision to appeal is based on rational cost-benefit analysis. For a plaintiff, the incentive to appeal depends on the probability of success on appeal multiplied by the expected award, minus the cost of the appeal, compared to the judgment already obtained at trial.
• Settlement: The model also accounts for settlement, noting that parties can choose to settle at any stage—before or after reaching the appellate court—to avoid the risks or costs of an appeal.

Empirical and Behavioral Considerations

• Constraining the Trial Court: Professor Hylton discusses whether appellate courts effectively serve as a "leash" on trial courts. If parties are sufficiently pessimistic about their chances on appeal, they will not appeal, even if the trial court decision was objectively wrong, meaning the appellate court fails to act as a constraint in those instances.
• Rationality vs. Risk Preference: Addressing the common critique that litigation can be "irrational," Professor Hylton suggests that many cases labeled as irrational may instead stem from different risk preferences (e.g., risk-prone or risk-averse behavior) or different valuation systems, which remain consistent and rational within the actor's own framework.
• Future Research: Professor Hylton views his paper as a launching point. Future research could expand the model to include information asymmetry, risk aversion, agency costs (e.g., lawyers pushing for appeals against the client's best interest), or multi-stage litigation models. He and his co-author are currently working on an empirical application of this model.

Law and Economics in Academia

• Field Status: While law and economics remains a strong research area with a powerful set of tools, Professor Hylton acknowledges that it faces a backlash at some elite law schools, partly due to the rise of movements like "Law and Political Economy".
• Antitrust Perspective: Regarding antitrust regulation, Professor Hylton notes that scholars in the field hold diverse viewpoints. However, he emphasizes that the strength of the field is a shared analytical framework and common understanding of economic terms, which is often missing in other areas of legal scholarship, such as constitutional law.

https://www.bu.edu/law/profile/keith-n-hylton/

Elliott Ash is an economist at ETH Zurich, adjunct professor at NYU Law, visiting professor at Columbia Law, and scientific lead at the Swiss AI Institute. He holds a PhD in economics and a JD from Columbia, a BA from UT Austin, and an LLM in international criminal law from the University of Amsterdam. His work sits at the intersection of economics, law, and machine learning, with a focus on judicial behavior and AI in the courts.

In this episode:

•     Inside the Manne Program (c. 1976–1998): how training federal judges in economics shifted their reasoning and language in judicial opinions.

•     Discussion of changes in the field of law and economics since its early beginnings as a field of study.

•     Early insight into a nationwide AI rollout to nearly half of Pakistan’s judges: faster opinions, better writing, smaller backlogs — and the risks of bias, hallucination, and mismatched values.

•     trace.law: using AI to verify filings rather than generate them, and why almost every case contains at least a minor error.

•     Why “the AI is not the judge” — keeping human judgment, agency, and accountability at the center.

Featured AI Legal Tool:

Elliott built a court-filing verifier that scans filings and opinions for verifiable errors - from typos and mismatched section numbers to swapped party names and, most seriously, citations to the wrong statute, a case that doesn’t say what’s claimed, or a case that doesn’t exist at all. With hallucinated-citation sanctions hitting even large firms, the pitch is simple: verifying with AI is a different, safer use than generating with it. The tool is available to lawyers at trace.law, with more detail in his Law360 article.

Links & resources

•     Trace Law — AI court-filing & citation verifier: trace.law

•     Article: Ideas Have Consequences: The Impact of Law and Economics on American Justice — Elliott Ash, Daniel Chen & Suresh Naidu

•     Elliott Ash’s Law360 article on AI-assisted filing verification

Enjoyed this episode?

Subscribe to the Law and Economics Podcast so you never miss an episode, and leave a rating and review — it genuinely helps others find the show. Share this episode with a colleague who thinks about law, economics, or AI, and connect with Elliott Ash to follow his research.