Tuesday, February 7, 2023
Sarah B. Lawsky (Northwestern; Google Scholar) presents Coding The Code: Catala And Computationally Accessible Tax Law, 75 S.M.U. L. Rev. 535 (2022), at Duke today as part of its Tax Policy Seminar hosted by Lawrence Zelenak:
This Article describes a new programming language, Catala, created by a team of computer scientists and lawyers. Catala provides a tractable and functional approach to coding U.S. tax law that offers a more transparent formalization and could potentially hold the government more accountable than the current patchwork of forms, worksheets, and secret programs.
While this Article describes a particular programming language, key characteristics of this particular language could generalize to other programming languages that formalize the law. First, Catala is a domain-specific programming language designed specifically for formalizing tax law. In particular, Catala is structured using default logic, a nonstandard logic that represents the underlying structure of the U.S. tax code more accurately than does standard logic. This structure makes the computer code easier to read, easier to create, and easier to modify when the law changes.
Second, computer code is created in Catala using a well-known approach in the field of computer science (though rarely mentioned in legal literature) called “pair programming,” which, in this implementation, takes advantage of the knowledge of both lawyers and computer coders. Finally, Catala uses literate programming to create computer code that is, among other things, easier to read and that communicates the decisions behind the coding to the user.