Grimmelmann Reviews Lawsky's Coding the Code: Catala And Computationally Accessible Tax Law

James Grimmelmann (Cornell; Google Scholar), When Law Is Code (JOTWELL) (reviewing Sarah B. Lawsky (Northwestern; Google Scholar), Coding the Code: Catala and Computationally Accessible Tax Law, 75 S.M.U. L. Rev. 535 (2022)):

Sarah B. Lawsky’s Coding the Code: Catala and Computationally Accessible Tax Law offers an exceptionally thoughtful perspective on the automation of legal rules. It provides not just a nuanced analysis of the consequences of translating legal doctrines into computer programs (something many other scholars have done), but also a tutorial in how to do so effectively, with fidelity to the internal structure of law and humility about what computers do and don’t do well.

Coding the Code builds on Lawsky’s previous work on formal logic and its advantages for statutory interpretation. (Formal logic, sometimes called “symbolic” or “mathematical” logic, involves the precise and rigorous analysis of symbolic expressions representing arguments, such as “p & ¬q” to mean “p is true and q is not true”.) In her 2017 A Logic for Statutes, [21 Fla. Tax Rev. 60 (2017) (reviewed by Alice Abreu (Temple) & Richard Greenstein (Temple) here),] she observed that many statutory provisions have a characteristic structure: rules subject to exceptions. A typical rule says that WHEN certain conditions are satisfied, THEN certain consequences follow, UNLESS one of several exceptions applies. Exceptions have exceptions of their own: interest payments are deductible, unless they are personal, unless they are mortgage payments.

Lawsky’s great insight about law and logic is that this characteristic structure of nested exceptions is most naturally modeled using a branch of formal logic called “default logic.” 

Default logic, unlike standard “monotonic logic,” allows for tentative conclusions. On the basis of what I know now, this is a nondeductible personal interest payment, but let me investigate further, and oh, I see that this is qualified residence interest, so I am withdrawing my tentative conclusion and replacing it with another tentative conclusion that the payment is deductible. And so on, until there are no more clauses of the statute to check, no more exceptions to explore, and the most recent tentative conclusion becomes a definitive one. It is a process of successive refinement, converging on certainty. Monotonic logic, by sharp contrast, requires ruling out all possibilities before drawing a conclusion, which remains valid for all time once drawn. ...

It is unusual for an established law professor to go back to school for a PhD. In philosophy. With a dissertation on formal logic. But Coding the Code, published five years after Lawsky submitted her (highly technical) thesis, shows the great value for legal scholars of the approach she developed in her PhD. It refines her distinctive approach to statutory analysis—which mixes careful legal reading with technical tools from formal logic and computer science—in a way that has great potential to help other lawyers and legal scholars be more precise about what tax laws say. All they need to do is talk to computer scientists, and Lawsky provides a roadmap for how. There is no epistemic trespassing in Sarah Lawsky’s work. Everywhere she goes, she is a welcomed guest.

Editor's Note:  If you would like to receive a daily email with links to tax posts on TaxProf Blog, email me here.