Welcome to Mad Libs. This is an irregular debate column where our columnists, contributors, and staff writers (or even you, dear reader) can duke it out over the big ideas we’re discussing in the metaphorical pages of this magazine.
Today’s Mad Libs was inspired by The Argument staff writer Kelsey Piper’s education coverage, which has pointed out, again and again, that American schools are failing kids in ways we already know how to fix. Piper has argued that illiteracy is a policy choice, that Mississippi’s much-ballyhooed reading gains are real, and, most recently, that we have known since the 1960s that direct instruction works.
Freddie deBoer, a committed education pessimist, is a frequent critic of our coverage. He thinks education “miracles” have a long history of turning out to be frauds, that no reform has ever reliably moved kids from one spot in the academic distribution to another, and that American schools were never the world-beaters we imagine they were.
When deBoer first responded to Piper, he wondered whether a magazine premised on the idea that “liberal good vibes can overcome every inconvenient fact and complication” would ever engage with his criticism. So, I invited him to debate Piper directly. You can read their exchange below!
Mad Libs is for paid subscribers, so subscribe to get more content like this in your inbox!
- Jerusalem Demsas, editor-in-chief of The Argument
Freddie deBoer:
The Sold a Story podcast is only the latest vehicle advancing the notion that we could fix our schools if we only choose to.
This contention has been a constant in American educational debates for decades, exemplified by the 2010 documentary Waiting for “Superman” and its triumphalist narrative about charter schools. Unfortunately, this attitude runs into a stark reality: Schools and teachers just don’t have that much impact on relative student outcomes.
I’m afraid that the evidence shows that student academic talent is more or less static. Nothing is more predictive of a student’s future success than their past success. In large majorities, individual students gravitate to a given level of academic performance at an early age and tend to stay there throughout their academic life, as this 2025 meta-analysis demonstrates.
There are many other studies we can pull from. As far back as the 1960s, pioneering large-scale federal studies (like Project Talent, the Coleman report, and Project Follow Through) supported the notion that most students stay in the same performance band over time. Data collected the summer after kindergarten provides useful predictive information about how students will perform in high school; a student’s third-grade reading level is an effective indicator of who will go to college and who won’t. Standardized tests administered at age 12 or 13 tell us not only which students will go to college but which will earn a doctorate, hold a patent, or get tenure. The basic reality is that our rank in the academic hierarchy is largely fixed.
Douglas Detterman has summarized the research record by saying “only about 10% of school achievement can be attributed to schools and teachers.”
Interventions that are often treated as transformative just don’t appear to matter. Promising early results inevitably fade out in educational research, and I use the word “inevitably” carefully.
Phonics-based reading instruction is currently undergoing a period of immense hype, but the advantages students see from phonics at age 5 are dramatically reduced by age 7 and nonexistent by age 11. For years, liberals championed pre-K schooling as a magic bullet, but study after study shows little or no positive effects as children age.
Past hype is no predictor of future performance. Class size reductions were ballyhooed, then over time their effects were revealed to be inconsistent at best. Many high-quality studies have shown that coveted “exam” high schools in large cities make no difference for student outcomes.
The result is a very bare cupboard when it comes to finding educational interventions that “work.” As Rimfeld et al. put it in a 2018 paper, “Individual differences in educational achievement are highly stable across the years of compulsory schooling from primary through secondary school. Children who do well at the beginning of primary school also tend to do well at the end of compulsory education for much the same reasons.”
You can change the school or type of school, and you can change the teacher. But what stays the same is the student. None of this depends on any wild theories or contrarian reading of the available data; perfectly mainstream educational research has provided evidence for these conclusions for decades.
Consider Harvard economist Raj Chetty. He’s been immensely influential among reformers, in large measure because his work gives people the optimistic read on schooling that they’re looking for — a better teacher leads to lifelong higher earnings for students.
I’ve written about the fundamental methodological problems with Chetty’s work in the past. But setting aside those issues, I think it’s essential to understand: Chetty’s work implies that teachers control very little of the variance in student outcomes!
Taking Chetty et al. at face value, a much better teacher (a full standard deviation better) raises a student’s test scores by about one-tenth of a standard deviation, which implies that the difference between teachers explains just 1% to 2% of the variation in that year’s scores.
Chetty’s work is known for its use of future student earnings as a dependent variable. It suggests that, by age 20, a student with a one-standard-deviation-better teacher is less than 1 percentage point more likely to attend college. By age 28, they earn only 1.3% more.
This is… educational optimism?
We could play the same game with CREDO, the influential Stanford University-based research project. CREDO has long been cited as the gold standard when it comes to making the evidence-based case for charter schools; politicians, think tanks, and commentators regularly refer to CREDO’s research to back up the notion that the average charter school is significantly better than the average public school.
Setting aside very real methodological concerns, once we translate from researchers’ absurd, invented “days of learning” metric, we find that charter schools make hardly any difference at all. Again, we don’t have to dispute their numbers; we simply need to express them in the terms that are used in almost all credible education research.
And that’s from a research entity that was hosted by a conservative think tank and funded by pro-charter money! This is a funny basis for arguing the importance of schools and teachers in student outcomes.
This is what frustrates me about the educational debate in our media: There’s no acknowledgment of the profound limitations of what schools can do. We insist on having a conversation that takes as given that schools and teachers have a major impact on student outcomes — an assumption the data just doesn’t support.
Kelsey Piper:
Education indeed doesn’t affect relative student outcomes much. I don’t really care.