Every few years, education seems to discover something new that will finally fix schools—a new framework, a new approach, a new way of thinking about teaching and learning. It arrives with urgency and conviction, spreads quickly, reshapes professional development and classroom practice, and then fades away, either replaced by the shiny new thing or layered on top of it. Twenty-first century skills, trauma-informed pedagogy, flipped classrooms, 1:1 devices—all promised to succeed where the last one fell short.
Ask a veteran teacher to list the major instructional initiatives they’ve been trained on over the past decade and you’re likely to get a weary laugh before you get an answer. Discipline systems cycle from zero tolerance to restorative practices; “data-driven instruction” yields to “personalized learning,” which is now being rebranded yet again in the age of AI. Each shift arrives with urgency and moral clarity. Each requires retraining, new materials, and a reorientation of practice. Spend enough time in schools and the pattern becomes impossible to ignore. Which raises an uncomfortable question:
Why is education so damn fad-prone?
The easy answer is also the most insulting—that educators are uniquely susceptible to trends, quick to abandon what works, and too eager to embrace whatever comes next. But that answer is wrong. Classroom teachers are typically the least enthusiastic participants in these cycles, having learned through experience how quickly today’s “transformational” idea becomes tomorrow’s abandoned initiative.
Education isn’t fad-driven because the people in it lack judgment. It’s fad-driven because the system they work in makes churn not just common, but rational. Four structural forces, in particular, push schools toward constant reinvention:
1) Weak feedback loops. In most sectors, failure reveals itself quickly. Customers leave, revenue falls, and performance problems become unmistakable. In education, by contrast, the signal is slow and noisy. Instructional changes may take years to show results. Student cohorts turn over annually. Outcomes depend on factors well beyond the classroom, such as attendance, family circumstances, peer effects, and more. Even when results improve or decline, attribution is murky. There are too many moving parts to say with reasonable certainty that any single input was determinative. Under these conditions, it’s impossible to offer decisive proof that a given approach is or isn’t working. This makes education unusually vulnerable not just to bad ideas, but to the premature abandonment of good ones.
2) Leadership legitimacy requires visible change. School systems churn through leaders with striking regularity. Superintendents, principals, and state chiefs shouldn’t buy green bananas. The onus is on each new leader to demonstrate that he or she is, in fact, leading. Many will come to the unfortunate conclusion that leadership is signaled not through stewardship, but through action: launching initiatives, unveiling strategic plans, introducing new frameworks, reorganizing priorities. Anyone who has spent time around school systems has seen the pattern: A new leader arrives, announces a bold vision, rebrands existing efforts, and introduces a new set of priorities. Three years later, often before results are fully visible, that leader departs (school superintendents tend to last a single contract cycle) and the wheel turns another revolution. A leader who says, “We’re going to keep doing what we’re doing, but do it better,” risks appearing passive or directionless, even if the system is performing well. It is a structural expectation. In education, visible change is how leadership signals its worth.
3) Low barriers to new ideas. In fields like medicine or engineering, new approaches must pass through layers of validation before they reach widespread adoption. Education has far fewer guardrails. A new framework can be published, marketed, and adopted by districts in rapid succession. Professional development cycles spread new ideas quickly, consultants package innovations into turnkey programs, and procurement systems often treat instructional approaches as interchangeable. The result is a highly permeable system—one in which ideas can enter and scale rapidly, often long before their effectiveness is firmly established.
4) Moral urgency. Education is not a typical service sector. It concerns children’s lives and futures, and that reality creates a constant sense of moral urgency. If a proposal claims it might help struggling students succeed—especially disadvantaged students—the pressure to act is immense. Waiting for perfect evidence can feel ethically unacceptable; trying something new, even with incomplete proof, feels compassionate. This isn’t foolish or irresponsible. It’s what happens when moral responsibility collides with uncertainty. But it creates a powerful bias toward action—and, by extension, toward constant change.
Taken together, these forces produce a system in which reform cycles are not accidental but predictable. Slow, ambiguous evidence makes it hard to know what is working. Leadership incentives reward visible change. New ideas face little resistance to adoption. Moral urgency pushes systems to act rather than wait. Under those conditions, stability is not the natural condition. Change is. Indeed, stability can easily be mistaken—and often is—for complacency or indifference.
The problem is not that education experiments with new ideas. Some experimentation is necessary and healthy. The problem is that the system struggles to sustain success once it finds it. Schools that improve often do so through unglamorous means: adopting a coherent curriculum, building teacher expertise, reinforcing consistent instructional routines, and maintaining focus over time. None of this is flashy. None of it lends itself to prizes or glowing media profiles. And all of it is fragile.
Recent reforms in Houston’s schools offer a useful contrast. Rather than chasing novelty, the district has focused on something far less glamorous: tightly specified lessons, routine checks for understanding, and instructional systems designed for the teachers schools actually have—not the ones reformers wish they had. Whatever one thinks of the model (and its critics are many and voluble) its premise runs against the grain of the broader system: Improvement comes not from constant reinvention, but from disciplined execution.
A change in leadership, a shift in priorities, or the arrival of a “promising” new approach can disrupt years of accumulated progress. Elsewhere, I’ve argued that the real miracle in education is not that some schools succeed. It’s that any manage to keep succeeding. The four factors enumerated above help explain why.
If we want to break the cycle, the solution is not to scold educators for chasing new ideas. It is to realign incentives so that stability and execution are valued as forms of leadership. That means treating implementation fidelity as an achievement, not an afterthought, and creating political and institutional cover for leaders who choose continuity over novelty. It means building systems that measure and reward long-term improvement, not short-term activity, and elevating professional norms that prize mastery over constant reinvention.
In short, we need to make competence visible. Because until we do, the system will continue to reward the appearance of change over the reality of improvement. So yes, education is fad-prone. Just not for the reasons we usually assume. We don’t chase reform because we forget what works. We chase reform because the system makes standing still look irresponsible—even when standing still is exactly what success requires.
This is an excellent post. I'd like to comment on two of your points.
First, your heading "Low barriers to new ideas" understates the problem. You are of course correct that innovations, including very large-scale ones, are often implemented with no real evidence evaluating either their efficacy or their side-effects. But worse, the educational establishment as a whole often even resists evaluating these innovations after they are implemented. Many of the key people want to show they are on board with the latest trend and don't want to be the skeptic demanding evidence. There are always a few exceptions. For example, Ed Reidy, who was Deputy Commissioner in KY during the huge KIRIS program, said publicly that the department owed it to kids to have people evaluate their program, and he allowed independent researchers to conduct evaluations, even evaluations that could produce results threatening to the program as designed. That is not the norm, however. IMHO, this is more than a recipe for failure. It's unethical. Major innovations should not be imposed on kids without evidence.
Second, in your conclusion, you wrote "In short, we need to make competence visible." I could not agree more. The difficult question is HOW. This is a particularly thorny problem in education, and I can't address it adequately in a short post, but I'll make a few points. As everyone knows, the system reflects a distrust of professional judgment, in part for very good reasons. In most districts, most teacher evaluations are pro forma and let poor teaching continue. (An anecdote: in the middle school my kids attended, generally considered one of the best in the state, one English teacher was so incompetent that her corrections actually introduced grammatical errors into kids' texts. She remained, year after year.) As some economists have noted, this is a common problem in the public sector because manager don't have money on the table; keeping a marginal or incompetent teacher often imposes less of a cost than getting rid of them. However, the system's primary responses for decades has been a counterproductive overreliance on objective measures. This is one of the core rationales for test-based accountability, which is premised on the idea that you can improve things by looking only at outputs, not at practice. It works for sulfur dioxide emissions from powerplants, but not for teaching. Yet for years, the people in charge insisted, in the face of persuasive evidence that this isn't true, that increased scores can be trusted without bothering to look to see how scores are raised. The partial retreat from that policy, which was largely but not entirely a failure, was to expand the system to incorporate other objective measures. They can be informative, just as test scores can be, but they are insufficient. For example, there are a number of studies showing that on some of the leading objective measures of classroom practice, US math classes score better than Japanese math classes--despite the fact that Japanese kids are on average far better in math that US kids.
This is a very difficult problem, and there is no silver bullet. IMHO, any good accountability system has to include observations of practice, but how to do that is not entirely clear. There is a good argument for including standardized measures as part of that, but that isn't a sufficient response. Complementing this with potentially more subjective evaluation is tough. For example, the Dutch system relies in part on evaluations by trained inspectors, but some Dutch scholars have found evidence that teachers have found ways to inflate inspection ratings, just as American teachers have found ways to inflate test scores. Some readers here may know that there has been a great deal of controversy in recent years about the practices of Ofsted, the British inspection system. The Singapore system relies on evaluations by both in-school personnel and Ministry personnel, but it isn't clear how well it could be adapted to fit the US context, which has far more fragmented authority.
I think that exploring potentially more effective (and presumably more complex) ways of holding educators accountable should be a priority for education research. However, getting that to happen will require two things beyond the interest of researchers. First, there would have to be funding for this work, which would be a tough row to hoe even before the carnage imposed by the Trump administration. Second, and far more vexing, this sort of work would require the agreement and cooperation of SEAs and LEAs. In my experience, this wouldn't be easy to obtain, in part for reasons in your post.
From my perspective, Robert is one of the few genuine Renaissance thinkers in education. I have long respected his ability to cut through fashionable rhetoric and name the system’s deeper issues, ranging from the role of family structure to the future of public education to curriculum and instruction. He introduced me to Substack by publishing my first article on his platform. I initially planned to write a direct response to his latest essay, but the issues he raised warranted a closer look at the implications of what he had written, so I published my response as a two-part Substack essay: “Stop Admiring the Problem: Escaping Education’s Groundhog Day” https://substack.com/@douglascarnine/p-193524718 and “Part 2: Start Building the Profession”. https://douglascarnine.substack.com/p/part-2-start-building-the-profession?r=20zkdr&utm_campaign=post&utm_medium=web&showWelcomeOnShare=true