Journal of Empirical Legal Studies最新文献

This paper studies whether plaintiffs' lawyers matter in securities class actions. We use inverse propensity score weighting (IPW) to compare the results in cases led by top-tier firms against those brought by lower-tier firms. This technique addresses case selection effects by using all of the cases led by a top-tier firm and then weighting the cases led by lower-tier firms based on how similar these cases are to the cases led by top-tier firms. We do find that top-tier lawyers obtain better outcomes for shareholders in a subset of securities class actions, specifically the cases against the larger (although not the very largest) companies. Outside of these cases, we find that most of the difference in the results obtained by top- and lower-tier firms disappears when we balance observable characteristics using the IPW technique. Although the top-tier firms do not get better results in most cases, they do invest more hours and money into their cases.

Entitlement assignment is unimportant if transaction cost is sufficiently low, as post-litigation bargaining can redress allocative inefficiency, or so goes the Coase theorem. Ward Farnsworth, based on interviews with lawyers, argues that animosity created during litigation, a key mechanism to (re)allocate entitlement, will hinder the conclusion of any deal following litigation. Using a laboratory experiment, we test whether animosity generated before negotiations reduce the rate at which deals are successfully concluded and find evidence for a lower deal rate under one of the treatment conditions (the raw difference being three percentage points). The small practical effect may be attributed to rationality carrying the day and/or the limited degree of animosity we can generated in the lab with human subjects. The Coase theorem holds, while Farsworth's observation should not be ignored.

How does the authority of case law evolve over time? On the Dworkinian legal formalist view, cases increase in authority as they become more embedded in the “chain” of legal precedent, but on the Holmesian legal realist view, each case's authority is proportional to its ability to predict future legal outcomes. In this article, I show how modeling the citation network of U.S. Supreme Court case law not as a chain novel (à la Dworkin) but instead as a Markov chain (à la Holmes, or so I argue) unlocks an intuitive measure of case authority—called HolmesRank—that outperforms the existing approach in a variety of validation tasks. I then demonstrate how the authority scores produced using this Markov machinery empower the analysis of two important normative questions: (1) the ideological basis of lasting precedential authority and (2) the causal effect of the Supreme Court's citation choices on lower court compliance.

Rapidly improving artificial intelligence (AI) technologies have created opportunities for human–machine cooperation in legal practice. We provide evidence from an experiment with law students (N = 206) on the causal impact of machine assistance on the efficiency of legal task completion in a private law setting with natural language inputs and multidimensional AI outputs. We tested two forms of machine assistance: AI-generated summaries of legal complaints and AI-generated text highlighting within those complaints. AI-generated highlighting reduced task completion time by 30% without any reduction in measured quality indicators compared to no AI assistance. AI-generated summaries produced no change in performance metrics. AI summaries and AI highlighting together improved efficiency but not as much as AI highlighting alone. Our results show that AI support can dramatically increase the efficiency of legal task completion, but finding the optimal form of AI assistance is a fine-tuning exercise. Currently, AI-generated highlighting is not readily available from state-of-the-art, consumer-facing large language models, but our work suggests that this capability should be prioritized in the development of legal AI products.

Trademark law protects marks to enable firms to signal their products' qualities to consumers. To qualify for protection, a mark must be able to identify and distinguish goods. US courts typically locate a mark on a “spectrum of distinctiveness”—known as the Abercrombie spectrum—that categorizes marks as fanciful, arbitrary, or suggestive, and thus as “inherently distinctive,” or as descriptive or generic, and thus as not inherently distinctive. This article explores whether locating trademarks on the Abercrombie spectrum can be automated using current natural-language processing techniques. Using about 1.5 million US trademark registrations between 2012 and 2019 as well as 2.2 million related USPTO office actions, the article presents a machine-learning model that learns semantic features of trademark applications and predicts whether a mark is inherently distinctive. Our model can predict trademark actions with 86% accuracy overall, and it can identify subsets of trademark applications where it is highly certain in its predictions of distinctiveness. Using an eXplainable AI (XAI) algorithm, we further analyze which features in trademark applications drive our model's predictions. We then explore the practical and normative implications of our approach. On a practical level, we outline a decision-support system that could, as a “robot trademark clerk,” assist trademark experts in their determination of a trademark's distinctiveness. Such a system could also help trademark experts understand which features of a trademark application contribute the most toward a trademark's distinctiveness. On a theoretical level, we discuss the normative limits of the Abercrombie spectrum and propose to move beyond Abercrombie for trademarks whose distinctiveness is uncertain. We discuss how machine-learning projects in the law not only inform us about the aspects of the legal system that may be automated in the future, but also force us to tackle normative tradeoffs that may be invisible otherwise.

The unprecedented occupational risks posed by the COVID-19 pandemic prompted employers to boost wages and federal authorities to propose hazard pay policies. This article estimates a market-based compensating differential for workers facing elevated risks through contact with the public using CPS employment data for 2019–2020 and occupational characteristic data from the US Department of Labor's Occupational Information Network. The estimated premium for exposure was roughly $820 overall and $1000 for essential workers. These premiums fall short of those proposed—but not enacted—by the federal government and are more commensurate with estimates of the value of a statistical life than were the federal proposals.

Using network analysis, we study referrals among plaintiff-side lawyers handling medical malpractice cases in Indiana. The referral network is stratified, with a few highly connected “hub” firms. Firm connectivity follows a power law distribution—suggesting that new entrants tend to associate with already well-connected firms, rather than starting a new network. Regression analysis shows that, for a given firm, connectivity (i.e., node degree) in the referral network and being loyal to a smaller number of firms both lead to better outcomes in non-referred cases. The referral network also became more concentrated over time. The stratification of the market for plaintiff-side representation is reinforced through these processes.

The plain language movement waged a silent revolution in the last generation, passing nearly 800 laws nationwide with little public debate. The movement asserted that it could scientifically show that there is a widespread readability crisis in legal documents, particularly contracts, that are unreadable to most adults. This article presents the largest empirical analysis of these claims to date, utilizing a dataset of 2 million contracts spanning multiple decades and industries and applying machine learning techniques. The study challenges fundamental tenets of the plain language movement. Contrary to prevailing beliefs, consumer agreements have median reading scores almost indistinguishable from those of daily news articles. A critical evaluation further exposes that readability tools endorsed by the movement are shoddy and manipulable and can produce grade-level differences of up to 4.6 years for identical texts. Moreover, the movement's core belief that Americans cannot read past the level of an eighth grader is exposed as an unsubstantiated myth. These findings fundamentally challenge the premises and effectiveness of one of the central consumer protection policies. These results call for a radical rethinking of legal access strategies, suggesting a shift from superficial readability metrics to addressing substantive issues in market dynamics and focusing on truly vulnerable populations. More broadly, this case study serves as a cautionary tale about the propagation of myths in legal scholarship and the potential for well-intentioned reform movements to divert attention and resources from more effective interventions.

The web of over 3000 Bilateral Investment Treaties (“BITs”) is the primary body of international law regulating cross-border investments. Research suggests that these treaties may have had a limited impact on promoting new investments, but that they still may have helped to improve countries' political relationships. In this paper, we document that this pattern was reversed for one of the most prolific signers of BITs: China. Using a stacked-event research design, we find that Chinese BITs are associated with an increase in Bilateral Foreign Direct Investment Flows but a divergence in voting patterns at the United Nations. We then explore two explanations for why the Chinese BIT program led to increased investment while also producing foreign policy divergence: that the domestic political costs of economic engagement with China push countries away, and that there are offsetting international pressures that have stronger pulls than China's efforts. We find no support for the domestic political costs explanation, but we do find evidence that the countries that received increased aid from the United States after signing a Chinese BIT had greater foreign policy divergence with China.