The Stanford Researcher Whose Work on Bottom-Up Education Innovation Finally Has the Evidence to Back Its Biggest Claims

Almost every American school has a teacher who has figured something out somewhere in the building. Perhaps it’s the way she arranges math ideas to help the students who are having trouble. Perhaps it’s the way he leads the conversation that makes the quiet kids talk. These things are effective. The principal who stops by for observations is probably aware of it, as are the students and the teacher. Then, remarkably consistently, it does nothing at all.

For years, Rebecca Wolfe has been observing this. She is a researcher who has worked at Stanford, Harvard, and organizations like Jobs for the Future and KnowledgeWorks. She has dedicated a significant amount of time to researching why innovative ideas in education die in the same hallways where they originated. The Hoover Institution published her new white paper in February 2026, which is the most organized version of an argument she has been developing for some time. It comes at a time when the evidence is finally too strong to ignore.

Key Information: Rebecca E. Wolfe & The Framework

The paper, “Can’t Get There from Here,” is released as the 2025 National Assessment of Educational Progress confirms what many in the field of education have long suspected: student outcomes have not improved despite decades of top-down reform. They’ve made things worse in a few instances. That is what the numbers indicate; it is not a fringe interpretation. However, the majority of the time, the policy response stays largely unchanged: more regulations, more requirements for compliance, and more centrally planned solutions that are handed down to teachers who were not involved in their creation and are not allowed to modify them.

According to Wolfe’s framework, this is structural rather than coincidental. She claims that the system is designed with compliance rather than improvement in mind, and that decision has repercussions. Teachers who experiment with new methods frequently do so without institutional support, time, or networks. Even when the status quo is clearly failing, the culture in many schools has become so risk-averse that it feels actually safer to stick with it than try new things. Even after it is clear that a policy is ineffective, it remains in place for years, consuming the time and effort of those who should be doing something better. It is a cycle that is nearly impossible to break from the top and difficult to break from the outside.

AVID, or Advancement Via Individual Determination, is the case study at the heart of the paper. It began in the early 1980s with a single teacher in California managing a program for thirty students. It is currently active in 47 states. Neither a foundation-funded rollout nor a federal mandate led to that growth. It occurred as a result of the model’s design, which allowed educators to modify it to fit their own situations while maintaining its fundamental ideas. Wolfe refers to this implementation as “tight but loose”—a description that may seem straightforward, but in reality, it is the opposite of how most education reform operates. The majority of reform is tight. Early on, the loose part, the teacher’s judgment, the local adaptation, and the contextual flexibility are squeezed out.

Reading the paper and the related research gives the impression that the field is reaching a certain kind of reckoning. Researchers at Stanford studying artificial intelligence in the classroom are discovering that tools that boost student performance while in use frequently result in gains that disappear as soon as the technology is taken away. This raises the unsettling question of whether the tool was truly teaching anything or merely doing the thinking for the students. Many impressive-sounding claims that, upon closer examination, are supported by a limited number of rigorous studies are emerging from the larger AI conversation in education. This is consistent with Wolfe’s claim that the discourse surrounding education reform has for too long confused motion with progress and activity with evidence.

Her suggestions are direct. Adaptive teaching strategies should be rewarded by accountability structures rather than strict compliance measures. Instead of being consulted at the conclusion of a process that was already created without them, educators must be actively involved in the creation of policies. Additionally, knowledge-sharing systems must be developed—a true infrastructure for documenting and disseminating what educators truly learn in the classroom, as opposed to leaving schools to perpetually relearn the same lessons from the same mistakes.

It’s difficult not to feel a little frustrated for the educators who have spent decades conducting covert experiments in their classrooms while the policy debate takes place far above their heads. Wolfe’s framework identifies the problem but does not address it. “If we don’t change the barriers to an education-innovation ecosystem in a significant way,” she says, “we will find ourselves ten years from now making the same complaints about the same problems.” That’s not pessimism. It’s merely observation. It remains to be seen if those who create accountability systems are open to hearing it.