Roland Fryer on Educational Reform
What this covers
Roland Fryer and Russ Roberts discuss what Fryer has learned from experimental work in Houston's failing schools about narrowing the racial achievement gap. The conversation centers on a core claim: the knowledge needed to close the gap already exists, documented through randomized controlled trials and implemented in practice, but scaling these reforms faces obstacles that are political and organizational rather than intellectual. Fryer walks through his evidence for five distinct, replicable practices—more instructional time, human capital strategies around teacher selection and development, data-driven instruction, high-dosage tutoring, and a culture of high expectations—and describes how their combination in Houston's worst twenty schools closed the math achievement gap within three years while placing every high-school student in a college.
The conversation spans multiple dimensions of why these reforms work and why they haven't spread. Fryer details experiments showing that incentivizing student inputs like homework produces large gains, while paying for outputs like test scores does not—a distinction rooted in the substitution effect rather than damage to intrinsic motivation. He argues that high-dosage tutoring, delivered at scale, is non-negotiable, and that what separates effective schools is their willingness to implement short-cycle assessments with real instructional response rather than testing once yearly. A revenue-neutral version of these reforms—giving principals training and a handbook without restructuring personnel or budget—yielded roughly half the effect, pointing to managerial skill as a lever. Roberts and Fryer explore why implementation stalled despite evidence, settling on a diagnosis: the reforms require sustained, unpopular effort including teacher turnover and extended school days—work that generates political resistance. The discussion also touches on parent choice, the design of incentives facing politicians and principals, and whether closing achievement gaps can coexist with popularity.
Fryer argues that we already know how to close the racial achievement gap in failing schools—through five identifiable, replicable practices and well-designed incentives—and that the real obstacle is not knowledge but the courage and hard work required to implement them despite political unpopularity.
- Randomized incentive experiments show paying students for inputs (reading, homework) works while paying for outputs does not, and effects largely persist after payments stop
- Five best practices implemented experimentally in Houston's worst 20 schools closed the math achievement gap and got every high-school pupil admitted to college
- The reforms failed to spread not because they don't work but because they require intense effort and unpopular human-capital changes
Implementation requires sustained monitoring with incentives, not just policy announcement or good intentions.
- A key lesson of economics and management is that passing legislation or announcing a policy does not produce the intended impact; monitoring with incentives is often necessary, as illustrated by teachers quietly ignoring a new curriculum after a workshop because 'the new one doesn't work.'
“passing legislation does not always have the impact that you think it does, and that monitoring it is often necessary with incentives.”
- Citing Milton Friedman, better outcomes come less from finding innately better people than from designing systems where even less-than-ideal actors face incentives to behave well; the incentives politicians (or principals) face are the real challenge, not their innate goodness.
“what you want is a system where bad people have the incentive to be good. The incentives that politicians face are the challenge, not necessarily their innate goodness or badness.”
- The recording took place on September 14, 2022, with guest Roland Fryer of Harvard University.
“Today is September 14th, 2022, and my guest is Roland Fryer of Harvard University.”
Teacher effectiveness is learnable through structured practice and feedback, but incentives are needed to sustain improvement at scale.
- While there is an innate component to being a great teacher, most of it can be learned—but doing so requires structured practice and feedback (without which teachers won't improve any more than a novice handed a golf club) plus sustained effort that incentives, not just love of the craft, are needed to elicit at scale.
“there is an innate part of being a great teacher, but most of it is learned... being a great teacher can be taught, is my view.”
Decades of failed inner-city schools justify experimenting with diverse alternatives like charters and vouchers rather than defending status quo.
- For three generations America failed inner-city kids with horrible schools, and almost no one defends the status quo, which is why Russ supports trying a wide range of alternatives—charter schools, vouchers, and creative experiments—rather than clinging to existing approaches.
“For three generations, we failed inner-city kids with horrible schools... The status quo is awful. Almost no one defends the status quo.”