Paul L. Caron
Dean





Friday, July 9, 2021

Weekly SSRN Tax Article Review And Roundup: Speck Reviews Lawsky's Teaching Algorithms And Algorithms For Teaching

This week, Sloan Speck (Colorado; Google Scholar) reviews a new work by Sarah Lawsky (Northwestern; Google Scholar), Teaching Algorithms and Algorithms for Teaching, 24 Fla. Tax Rev. ___ (2021).

Speck (2017)

In Teaching Algorithms and Algorithms for Teaching, Sarah Lawsky identifies and elaborates what she denotes as the “algorithm method” for teaching tax. A corollary or companion to the problem method, the algorithm method unpacks complex statutory language by “ask[ing] students to work through unambiguous problems that have right and wrong answers.” Although Lawsky’s terminology is novel and useful, she describes the algorithm method as fundamentally “unremarkable, uncontroversial, and common” in tax instruction. Her article carefully connects the algorithm method to in- and out-of-class learning in the context of “flipped” classrooms and her outstanding exercise-generating website, Lawsky’s Practice Problems. This context illustrates the importance of delineating and deploying the algorithm method in legal pedagogy.

Lawsky cites three potential benefits to the algorithm method: it helps students to connect statutory text to real-world action, it deepens students’ knowledge of the underlying law and policy, and it reveals aspects of the law not otherwise apparent from a more abstract approach. She also warns of “essentially training the student to be a very slow, buggy version of Microsoft Excel” (not her words). One could imagine that typical student outcomes range along this continuum, and Lawsky gives guidance and best practices that might push students away from the slow and buggy pole. At the very least, the algorithm method yields departure points that facilitate more sophisticated engagement, while also giving some purchase to struggling learners.

The algorithm method, of course, shows up in tax and transactional practice. Indeed, algorithms and algorithmic thinking are deployed often enough that one might view their explication in the classroom as a life skill. These instantiations range from financial models to § 199A optimizers to calculators that prepare partnership returns with targeted allocations. Then may not even be automated: I can recall pages of copious calculations on legal pads when stress-testing layer-cake allocations in partnership agreements. Sadly, “slow” and “buggy” might describe a significant proportion of these practice tools (and possibly all things created in Excel). Sometimes ye olde § 382 spreadsheet is misprogrammed or corrupted or poorly tailored, and things just keep falling apart until an enterprising lawyer digs into the thing’s guts to figure out what’s wrong. I suspect that students steeped in the algorithm method show a greater willingness than the average lawyer to get their feet wet or their hands dirty, and that’s a good thing.

In Lawsky’s article, an extended example involving § 453(e) motivates the algorithmic method’s revelatory function, in which a computational approach to novel numeric facts leads to a more nuanced understanding of the law’s purpose and operation. In short, § 453(e) addresses tax gaming in which related parties engage in an installment sale, then quickly sell the subject property (with full basis) to a third party for cash. The provision accelerates gain to the extent that the related parties end up with more cash than they started with. Lawsky outlines how a sale outside the relationship for a loss sheds light on this antiabuse rule’s metes and bounds: she describes the provision as seeking to “prevent a certain type of tax planning but not to punish those who engage in that tax planning.” Some taxpayers might disagree about the “punish” point, especially if there’s bad blood between members of a family or partnership relationships that result in unexpected reattribution. Tax tends to treat happy and unhappy families alike, and contractual protections may prove to be cold comfort. But this dynamic in § 453(e)—fair in quantitative consequences, perhaps overbroad in scope or mechanical application—may emerge more easily with the algorithmic method to set up, reveal, and socialize the provision’s structure.

A final question: how many problems are too many problems? Lawsky notes that access to a practically unlimited number of procedurally generated problems may cause students to “become unmoored from the statute itself” (and, if done over a long weekend or reading period, probably a bit unmoored from reality more generally). She also gives thoughtful and practical advice on how to mitigate this risk. But some will choose curated practice over unlimited, and traditional classrooms over flipped. Implicit in Lawsky’s article is a degree of pluralism and experimentation in tax teaching: many classrooms mix pedagogical styles in creative ways that enhance learning outcomes, and it seems unlikely that any particular methodology dominates for all teachers and all students. Lawsky’s article advances this conversation in a truly meaningful way.

More than a piece on legal pedagogy, Lawsky’s article connects to her ongoing work on formalization, computing, and probability in tax law. And, more than an engaging extension of an established scholarly agenda, Lawsky’s article reads as a paean to classroom teaching. Her passions and commitments to effective teaching are evident, and, for me, also restorative. More generally, Lawsky’s article is important and valuable reading for tax teachers in and outside of law, as well as law teachers who deal with statutes and regulations of any stripe.

Here’s the rest of this week’s SSRN Tax Roundup:

https://taxprof.typepad.com/taxprof_blog/2021/07/weekly-ssrn-tax-article-review-and-roundup-speck-reviews-lawsky-teaching-algorithms-and-algorithms-for-teaching.html

Scholarship, Sloan Speck, Tax, Tax Scholarship, Weekly SSRN Roundup, Weekly Tax Roundup | Permalink