American education is experiencing an historic shift. After more than a century of consolidation and standardization, a wave of pluralism is sweeping the country. Charter schools, private-school vouchers, homeschooling, microschools, and education savings accounts (ESAs) have all gained traction, most recently with a new federal tax credit to fund K–12 scholarship programs. These changes have been years in the making, but the pandemic accelerated them, creating a sudden opening for institutional innovation. For the first time in memory, a decentralized and diversified education system seems possible.

The opportunity before us is not just for variety. It is also for dynamism—the kind of productive, experimental diversity that improves organizational performance over time (schools might measurably improve their ability to educate students) and enables societies to adapt to social and economic change, such as shifts in technology, labor market demands, and demographic trends.

Yet despite these promising signs, decentralization could stall without intentional and ongoing support, and this nascent moment of educational pluralism could revert to uniformity and stagnation.

Why dynamism matters

America’s education system has not consistently delivered on either of its foundational goals: cultivating citizenship or promoting academic excellence. Academic excellence matters for all of us. Cognitive capacity, as captured by achievement tests, has had a pronounced impact on the wealth of nations. Yet, according to the 2024 National Assessment of Educational Progress (NAEP), only 39 percent of fourth graders and 28 percent of eighth graders performed at or above the “proficient” level in mathematics. U.S. eighth graders’ most recent performance on international math exams is their worst since testing began in 1995, ranking the U.S. nineteenth out of forty-five countries—on par with Romania, Cyprus, the United Arab Emirates, France, and Azerbaijan. Lower math performance in the U.S. has historically been accompanied by significant social costs, including lower incomes, higher rates of teen pregnancy, and higher rates of incarceration.

These outcomes are not too surprising, considering America’s increasingly centralized and politicized school bureaucracy. Children have differing needs. The evidence is clear that boys, for example, are underserved by traditional school models, and that low-income kids have fallen far behind since 2012. Giving families access to diverse educational settings can help address these challenges in ways that centralized mandates cannot. Moreover, a single, centralized delivery model is a recipe for interest-group capture and organizational inefficiency. Indeed, as centralization has increased, it has been accompanied by a rise in political conflict, bureaucratization, and stagnation.

Recent battles over curriculum and culture, from California’s ethnic studies curriculum to Texas’s K–5 reading curriculum, further illustrate the dangers of one-size-fits-all schooling. Such curricula meet the needs of some parents while raising significant concerns for others. And the political battles over curriculum can hurt everyone. Research by our Ohio State colleague Vladimir Kogan demonstrates the harm caused by these culture wars. In a working paper and forthcoming book, Kogan shows that ideological conflict among adults leads to lower academic performance among students. Such conflicts distract from learning, impose burdens on teachers, and undermine community trust in schools.

Allowing greater educational pluralism could reduce political tension and improve system-wide outcomes. The Cato Institute’s Neal McCluskey has long argued that school choice is not only a tool for academic improvement but also a pressure-release valve for culture war conflict. When families are empowered to choose the best-fitting program for their children, the stakes of any one curricular decision are lowered. Schools can then serve their communities more directly, without becoming battlegrounds for ideological control. Importantly, as scholarship by Patrick Wolf and his colleagues has shown, such pluralism can also strengthen democratic values.

Pluralism also equips the education system to adapt to a rapidly changing world. As David Deming’s research shows, the labor market increasingly rewards “soft skills” such as leadership abilities and decision-making acumen, alongside or even instead of the traditional “cognitive skills” captured by test scores. Deming’s findings suggest that policymakers cannot predict with reasonable certainty which educational pathways will be most valuable in ten or twenty years. The best hedge against this uncertainty is a decentralized system capable of evolving in many directions simultaneously.

Schools offering a classical curriculum, for example, could help students respond to societal changes by imparting problem-solving and communication skills. And a renewed focus on civic virtue and historical understanding might help adapt civic life to technological disruption. But some students will be better served by learning to work with the tools of that very disruption, especially in STEM fields. Educational pluralism can yield dynamism on both dimensions and avoids the trap of ideological monoculture.

A rare policy opening

The present moment is especially promising for educational dynamism. After decades of increasing centralization, the institutional landscape is shifting. The expansion of charter schools in the 1990s and 2000s showed that alternative governance structures were politically and practically viable. Passage of the Every Student Succeeds Act (ESSA) in 2015 marked a retreat from strict federal oversight and gave states more discretion over accountability.

These reforms were significant, but they were slow-moving. The pace of change accelerated rapidly during, and in the immediate aftermath of, the pandemic. Families encountered remote learning and discovered both the limitations and possibilities of alternative education. They disenrolled from traditional public schools in large numbers. Many turned to private schools and homeschooling. Others organized microschools. Legislatures in red states responded by dramatically expanding school choice programs. Eighteen states have enacted universal choice programs since 2021. All students in those states are now eligible for private school scholarships or ESAs, and Congress has just marked Independence Day by enacting its own tax-credit scholarship program.

We have entered a new era in which pluralism is not theoretical but operationalized. More families are opting out of the traditional district model. Public charter enrollment continues to grow, while homeschooling numbers remain elevated well above pre-pandemic levels. Microschools and hybrid programs are expanding, often led by entrepreneurial educators or parent cooperatives.

A rare technological opening

Concurrent with changes in policy, this is a unique moment for technologically driven transformation in education. AI provides opportunities for hyper-customization of curriculum, either to allow for greater parental involvement or to better align offerings with the needs, abilities, and interests of particular students.

With AI presenting opportunities for streamlining and improving lesson planning, instructing, correcting, and grading, the conventional classroom model appears to be on the verge of disruption. The opportunities for customization to improve student learning are obvious. AI tools can provide students with highly customized instruction that dynamically advances as they gain mastery of new concepts. It can provide just-in-time scaffolding, support, information, and resources to students at the moment a challenge or knowledge gap is encountered. Students with the motivation to advance in their studies could accelerate past other students, without being bound to the pace of a conventional classroom. As Frederick Hess has noted, “it doesn’t take a lot of imagination to see how (in theory) 24/7, reliable, personalized AI-powered tutoring could have a profound impact.”

The advent of earlier technologies similarly led to significant promises of educational disruption, usually without much discernable impact. There are signs that the promise of large language models (LLMs) is fundamentally different. Credible studies have begun to document potential improvements in K–12 lesson planning and student learning when teachers make good use of LLMs, for example. And students’ extensive use of LLMs in completing assignments is already prompting changes in pedagogical approaches. There appears to be some consensus around the urgent need to teach students how to harness this technology toward productive ends, while retaining the capacity to think deeply and communicate effectively—and fending off cheating.

This moment also promises to allow for a different kind of technological deployment precisely because of the decentralization of schooling and the break-up of the public-school monopoly. As Larry Cuban argued, organizational environments—the incentives that schools and teachers face—may be the biggest roadblock to realizing technology’s potential in schooling. Recent changes to those incentives could unleash an explosion of educational experimentation. The political forces that induce risk aversion, rule following, and organizational stagnation and conformity appear to be giving way. This moment’s fusion of technological and organizational disruption opens the possibility for AI to feed educational dynamism.

Why dynamism still isn’t guaranteed

The current momentum toward educational diversification is encouraging, but diversity in how education is delivered—through parents, privately-run schools, or government-run schools—does not automatically generate diversity of educational approaches. There are tremendous forces pushing all educators toward conformity. This is the problem of isomorphism: even when structural reforms take place, underlying norms and incentives eventually reassert themselves and pull schools back toward a central model.

One reason is the continued presence of political incentives to standardize and micromanage. Even when funds are portable, regulations often follow them. Political interests, operating at all levels of government, have strong incentives to impose increasingly burdensome regulations to ensure schools pursue their preferred societal outcomes. To be sure, some government oversight is necessary to ensure that parents are choosing among quality school options and public funds are not wasted or used to enrich scammers. But, over time, political incentives to micromanage may render even private schools into heavily regulated public schools. To combat this, policymakers should create a clear bifurcation in regulatory regimes between default public schools—which should be held to a high and consistent standard of delivering core competencies employing proven instructional methods, and would have guaranteed, ongoing, funding—and private schools that are encouraged to experiment bravely and, when their experiments don’t work, to fail quickly.

There are also broader and more subtle social forces at work. To avoid controversy, charters, private schools, and homeschoolers may become exceedingly cautious in their experimentation and feel pressure to adopt conventional curricula. Without a deliberate effort to normalize differentiation and experimentation, the system’s default settings may continue to suppress substantial dynamism. If we want the civic and economic benefits that true educational pluralism can generate, we must convince one another to allow for meaningful difference.

Seizing the opportunity

There are considerable forces working against the educational pluralism that could bring so much benefit to U.S. economic and civic life. If we want decentralization to result in the type of pluralism that can yield educational dynamism, we must find a way to do the following.

First, preserve institutional autonomy. Private schools (religious and secular) and homeschool cooperatives must be allowed to operate according to their distinctive missions and student populations, and they should have considerable freedom to experiment and fail. States should resist the impulse to standardize these institutions through excessive regulation. They should continue to enforce minimum standards of operation that guard against fraud and protect children’s health and safety, and they should require that students acquire basic skills and some foundational knowledge (e.g., with respect to civics, literacy, and numeracy) if services are paid for with public funds. But, beyond that, state governments can let schools and vendors set the learning outcomes they wish to pursue and then collect and publish report cards that evaluate entities with respect to those goals.

Second, build a culture that promotes pluralism. Preserving institutional autonomy requires changing the way we think about public education in the U.S. Ashley Rogers Berner has noted that fellow democracies, such as the Netherlands, Australia, and Sweden have successfully supported diverse types of schools (religious, pedagogical, secular) while still holding them all to common academic standards and assessments. Critics argue that the picture in Europe is not so rosy and that allowing individuals to segregate according to their identities could be damaging—particularly for students at society’s margins. Activating such fears may undermine calls for pluralism. But catch-all public schools will remain an option for students who, due to disability, religious objection, or other belief or identity cannot be effectively served by locally available private options, and eligibility for programs can be conditioned upon the broadest, legally defensible requirements to accept all qualified students. The post-pandemic environment has presented the opportunity to make Americans comfortable with new schooling structures and diverse educational options.

Third, provide parents with information and assistance. Policymakers tend to focus on potential market failures—the ways in which lightly regulated markets may yield schools that systematically fail to meet parental and societal goals. That is a legitimate concern. We are perhaps more concerned with the pronounced government failures in school provision (notably, the stifling of schools’ institutional autonomy and diversity), but we are also realistic about parents’ lack of time and knowledge to make informed educational choices. Families need trustworthy metrics that are comparable across schools. Beyond test scores and graduation rates, states should collect and disseminate information on other dimensions of human flourishing (perhaps captured via parental satisfaction surveys) and longer-term workforce and college outcomes. Critically, parents need support—so that comparing schools on relevant dimensions and enrolling one’s child is quick and painless. Combining transparent data tools, simple apps, and human navigators—in addition to the basic operational and academic standards we reference above—can go a long way toward addressing the market failures to which critics of school choice frequently point.

The current moment brims with promise. Decades of centralization have given way to new models and new energy. But if this promise fails to yield true dynamism, American education may remain mired in mediocrity, albeit with more flavors.

Dynamism is not just about innovation. It is about resilience. A pluralistic education system allows a nation to learn from itself, to respond to new challenges, and to honor the diverse convictions of its citizens. The tools are now within reach. The question is whether we have the courage to use them.