Catala: A Programming Language for the Law (2103.03198v2)

Abstract: Law at large underpins modern society, codifying and governing many aspects of citizens' daily lives. Oftentimes, law is subject to interpretation, debate and challenges throughout various courts and jurisdictions. But in some other areas, law leaves little room for interpretation, and essentially aims to rigorously describe a computation, a decision procedure or, simply said, an algorithm. Unfortunately, prose remains a woefully inadequate tool for the job. The lack of formalism leaves room for ambiguities; the structure of legal statutes, with many paragraphs and sub-sections spread across multiple pages, makes it hard to compute the intended outcome of the algorithm underlying a given text; and, as with any other piece of poorly-specified critical software, the use of informal language leaves corner cases unaddressed. We introduce Catala, a new programming language that we specifically designed to allow a straightforward and systematic translation of statutory law into an executable implementation. Catala aims to bring together lawyers and programmers through a shared medium, which together they can understand, edit and evolve, bridging a gap that often results in dramatically incorrect implementations of the law. We have implemented a compiler for Catala, and have proven the correctness of its core compilation steps using the F* proof assistant. We evaluate Catala on several legal texts that are algorithms in disguise, notably section 121 of the US federal income tax and the byzantine French family benefits; in doing so, we uncover a bug in the official implementation. We observe as a consequence of the formalization process that using Catala enables rich interactions between lawyers and programmers, leading to a greater understanding of the original legislative intent, while producing a correct-by-construction executable specification reusable by the greater software ecosystem.

• The paper introduces Catala, a domain-specific language using prioritized default logic and literate programming to formalize statutory law into executable code.
• Case studies using Catala, such as formalizing tax and benefits law, successfully uncovered discrepancies and errors in existing legal software implementations.
• Evaluation indicates that Catala can bridge the gap between legal texts and technical implementation, enabling collaboration between legal professionals and technologists.

Overview of "Catala: A Programming Language for the Law"

This paper presents the design, implementation, and potential applications of Catala, a domain-specific programming language created to translate statutory law into executable code. The motivation behind Catala lies in addressing the ambiguities and procedural complexities of legal language, providing a structured format to codify legal algorithms that are embedded in legislative texts. Legal language often outlines computational law—parts of legal statutes that define unambiguous rules akin to algorithms—and the paper proposes Catala as a solution for creating formal, correct-by-construction executable specifications of such legal texts.

Core Design Principles and Language Structure

Catala aims to unify the efforts of programmers and lawyers by using prioritized default logic as its core structure. The authors highlight the unique drafting style of statutory law, which commonly follows a "general case followed by exceptions" format that naturally aligns with default logic. In this logic, preconditions and exceptions can be defined in a structured manner that mirrors legal stipulations.

Catala's syntax and semantics are designed to correspond closely with legal statutes, enabling lawyers to understand and review the translated code effectively. The inclusion of literate programming, where legal prose and its corresponding formal specification are interleaved, enhances accessibility for legal professionals engaged in formalizing law through code.

Compiler Architecture and Development

The Catala language features a multi-stage compilation process, transforming high-level legal syntax into a simple lambda calculus representation. The compiler, implemented in OCaml, translates Catala programs into executable code while preserving the logical structure inherent in the law. The authors emphasize usability and extensibility, allowing for interoperability by generating source code in various languages to accommodate different legal expert systems and legacy environments.

This structured approach aims to create a flexible tool that can integrate seamlessly into existing systems without requiring significant changes to underlying IT infrastructure.

Evaluation and Real-World Application

The paper evaluates Catala through several case studies, including the formalization of complex legal texts such as Section 121 of the U.S. Internal Revenue Code and the French family benefits computation. Through these applications, Catala uncovers discrepancies in existing implementations, such as erroneous interpretations in official family benefits software, demonstrating its capacity to enhance the accuracy of legal computations.

User studies with law students illustrate Catala's potential to bridge the gap between legal texts and their executable specifications: participants were able to understand and verify the formalized code after a brief introduction, suggesting that with minimal training, legal professionals can effectively use Catala.

Implications and Future Work

The authors posit that the failures in implementing legal systems, stemming from inadequate translation of law into computer code, could be mitigated using the structured approach offered by Catala. The paper suggests that future expansion could include more comprehensive evaluations across diverse legal contexts, fostering a greater understanding of statutory law's computational aspects.

The paper does not merely introduce Catala but also triggers a broader dialogue on improving the collaboration between legal experts and technologists. Through this collaborative model, the authors envision creating more robust, transparent, and practically reliable legal software systems.

In a landscape where computational law increasingly intersects with everyday governance, Catala represents a promising step towards optimizing legal system implementations. Its implications span both theoretical and practical domains, paving the way for more rigorous engagements between technology and the law. As AI and legal tech continue to evolve, developments like Catala offer a foundation for future innovation and enhanced trust in automated legal procedures.