"An Unholy Synthesis of Caplan and Fryer"

Conversation on student learning with a reader

A reader recently sent me an email with the subject line “An unholy synthesis of Caplan and Fryer.” Here’s the whole exchange, with the reader’s name redacted. I start with his email, then alternate between my brief replies and his responses. The last email is the best, so read to the end!

P.S. For a little more background, see my Iron Laws of Pedagogy.

---

Bryan,

I recently caught Roland Fryer on EconTalk (Oct 2022 episode) and was impressed by the idea that pedagogical ROI for the marginal dollar might be higher for paying kids to learn material than for subsidizing further supply of education via teachers’ salaries, better textbooks / supplies, nicer buildings, smartboards in the classroom, etc.

I thought I’d get your reaction to the narrow thought as it was reading The Case Against Education that made me receptive to the idea that online education had effectively solved the supply side of education for those who want it — as high quality instruction on just about any subject is available at the price of an internet connection and/or a very modest fee relative to tuition cost (adding LLM-enabled feedback and explanation for the student who’s stuck, and it might seem that the only reasons we still have teachers are (a) status quo bias, (b) employment/political-economy considerations and (c) to fulfill the babysitting functions of K-12 education).

I’ll side-step briefly the fact that the education itself may not be relevant to any end job performed in the economy, will not be remembered by the student, and is largely only remunerated for its signaling value. If we simply assert that it is desirable to have students master a subject, then it is at least valuable to know (if that is indeed what we know) that paying students directly to master material is much more effective than paying other people to offer free education to students who are completely unpaid in the near term for being compelled to encounter the material.

One could imagine an extreme synthesis of Caplan and Fryer that says the state is primarily interested only in teaching those skills you’ve called truly general purpose —literacy and numeracy — and to achieve student learning in these fields we have devised a system of payments to students that are contingent on reaching micro-milestones (e.g., what one might reasonably learn and demonstrate mastery of after spending 60 minutes on Khan Academy) in progress towards mastery of arithmetic, basic algebra, phonics, and reading comprehension. If students find it most cost-effective to earn those payments by subcontracting to tutors and educational coaches who help them reach these milestones (or even on-demand traditional in-class lectures if preferred), then we will primarily see the growth in supply of pedagogical methods which are most capital efficient relative to a desired learning outcome.

If one thinks there are sufficient positive externalities from additional educational gains that enhance one’s labor market viability (vocational training, STEM training, etc.), then one can imagine a world where there are similar payments available for those who achieve milestones on their way to demonstrated mastery in a field like plumbing, carpentry, mechanical engineering, accountancy, etc.

All of which is to ask the simple question of if you could vote yes/no on the proposition that if we simply cannot quell the irresistible urge to waste lots of money paying for education, would you prefer that we instead opted to pay students directly for demonstrated mastery of content as opposed to paying for more supply of education when supply is not the bottleneck?

Hope you’re well — still hoping to have an excuse to pass through Fairfax and would love to catch lunch at that time if the offer still stands.

Best, [redacted]

---

Paying for results >> paying for hours.

Though it would be better to pay periodically for continuing good scores to avoid mere cramming!

— Bryan

---

A great addition! 

I have wondered why schools don’t do this regardless of payment setup — my best guess is just logistical difficulty: how to administer periodic tests of retained knowledge to students who each have a different background set of coursework they should have retained, and what do you do if someone fails? Send them back to the piece of the course they forgot? (Difficult in a sequentially taught curriculum).

Exciting to think of what system could be built, discouraging to consider the feasibility/political constraints. Regardless, appreciate the exchange.

[redacted]

---

The logistics are hard, but even if logistics were maximally easy, they don’t want clear-cut evidence that students forget almost everything they learn.

— Bryan

---

Probably the bigger point.

Illustratively, my brother teaches at a private school, and once (after reading The Case Against Education) ruffled some feathers at a teaching enrichment session when he administered a test covering basic material from different high school subjects (factor a polynomial, label all subordinate clauses, demonstrate basic history fluency) and showed that teachers knew very little about the subjects they didn't directly teach. 

The teachers didn't seem all that bothered by the conclusion that neither were the skills they learned in high school necessary for later professional success nor did they remember them and offered the somewhat lame defense, "our job is to expose students to different subjects so that they know what they want to study later." To your point, teachers already know retention is bad and have made peace with it in a way that taxpayers may not.

Thanks again, [redacted]