(Illustration by Ibrahim Rayintakath)
In their article “A Democratic Vision for Public Schools,” Kent McGuire and Matt Wilka have done a service by calling us to reinvigorate the public purpose of American education. Their diagnosis that three decades of efficiency-first reform narrowed curricula and produced short-term score gains but failed to sustain long-term academic progress is important. We are now seeing a crisis of chronic absenteeism and declining youth mental health that cannot be ignored. Their recognition that we are living between paradigms, with the old framework no longer fitting our current reality and a new one struggling to emerge, is the right place to start.
But the way forward requires more than reinvigorating the public purpose of education. It requires renegotiating the social contract that underlies education—who holds power, who earns trust, who gets to make consequential decisions—and, in parallel, redesigning the system to fairly deliver on that new contract. Fairness here is not a downstream outcome of this renegotiation but a core design principle that must shape how power, resources, and opportunity are distributed from the start. These are distinct but related tasks, and conflating or trying to do only one of them at a time may be one reason reform has stalled. What we cannot afford right now is to continue mistaking schooling for learning and efficiency for fairness.
My organization, LearnerStudio, exists to build a fair, coherent, positive, future-ready learning system, focusing on human flourishing in the age of artificial intelligence. Over the past three years, we have supported emerging tools and models, listened to families and learners, and worked with education innovators building what we believe is the emergent third horizon of the system, designed not for industrial-era efficiency, nor for incremental improvements inside that old box, but to support all learners as individuals, in careers, and for democracy in this modern era. What follows is an attempt to identify the new social contract we need to define and to sketch the system design it requires.
The Old Social Contract Is Broken
The industrial era social contract for education rested on the premise that where you can afford to live drives what you get. “Experts” in the system were trusted to define what children need to learn. The system delivered standardized content and mandated compulsory seat time in schools. Sorting by test scores, zip codes, and credentials was accepted as a reasonable proxy for meritocracy and fairness. Parents of public school students, in this contract, were largely asked to hand their children to the state and wait.
This contract was negotiated implicitly over the late-19th and early-20th centuries, and though its terms weren’t always explicitly acknowledged, they were pretty clear. The state promised free universal schooling, a diploma useful to employers and colleges, and a shot at the middle class for any child who showed up and complied. Families promised deference in return: compulsory attendance, trust in professional expertise, and acceptance of standardization grounded in a system of course grades and multiple-choice tests. The system optimized for efficiency at scale, moving large numbers of children through a standardized process designed for an industrial economy that needed reliable, trainable workers. Over time, additions were layered on—for individuals with disabilities, or small alternative pathways, eventually charter public schools to increase educator and parent agency for a small percentage of families, “college for all” goals—but the core system stayed a standardized and efficiency-based system. Parents were largely peripheral. Individual flourishing, genuine mastery of knowledge or skills, and real equity were never seriously promised. Wealthy families had the agency and power to move their residency or pay for private school as ways to opt out.
The collapse of that old contract is accelerating under the weight of four forces that McGuire and Wilka acknowledge but do not fully reckon with.
The first is the COVID disruption. Pandemic school closures didn't just interrupt learning, they broke trust and laid bare the system’s flaws. Parents watched, often powerlessly, as inflexible and irrational seat-time requirements had children passively receive compulsory content over Zoom. Many were radicalized by the experience. They saw the system's rigidity and flatness up close, and they began demanding something different: more agency, more flexibility, more genuine attention to their child as an individual. That demand has not receded; it has accelerated.
The second is a generational transition and teen crisis of engagement and mental health. Disturbing signals are coming from young people that the system is not working for them: chronic absenteeism reaching 37 percent in some communities, rising rates of depression and anxiety, and deepening social disconnection. When nearly two-fifths of students are regularly absent, the problem is not individual motivation. It is a structural mismatch between a rigid, compulsory institution and what adolescents actually need to thrive.
The third force is artificial intelligence. Artificial intelligence is not merely a new tool for classrooms. It is an “arrival technology,” meaning technology that enters people’s lives before institutions choose or know how to adopt it, forcing systems to react rather than plan. Like smartphones, it is already in learners’ hands, regardless of whether schools are ready. The implications are critically meaningful for policy and design. When a technology arrives before adoption, the design question shifts from “how do we introduce this carefully?” to “how do we build our systems around and reckon with what is already here?” In real time, schools are choosing whether to engage with AI intentionally or to cede that decision entirely.
This shift is reshaping what counts as valuable knowledge, which skills define meaningful work and citizenship, and even what it means to be human. It commoditizes rote content recall at the same moment it demands higher-order capabilities: sustained attention, collaborative problem-solving, ethical reasoning, relational intelligence. The opportunity stakes of AI are as high as the equity stakes. Without intentional design, the families and communities best positioned to navigate new tools will move faster, while learners in under-resourced communities risk absorbing AI’s disruptions without access to its benefits.
It also poses quieter but equally serious threats: By removing relational friction, AI companions condition young people away from the productive struggle of real human relationships and dangerously substitute for the very human connection our society needs to heal growing divides. As one young person put it: "Young people are already developing emotional muscle memory with AI. We vent, ask for advice, process big emotions—and it almost always responds with validation, even when we prompt for nuance or challenge. So what are we practicing, if the only 'person' we talk to never disagrees? That's why it's so important for adults—whether that's parents, educators, or mentors—to create space for safe, constructive disagreement. Because if we're not building that muscle with humans, we risk raising a generation that avoids conflict entirely."
The fourth force is the unbundling of the public education system through a fundamental shift in governance and funding. For decades, a shared federal-state architecture held together a collective commitment to universal public education, with public dollars flowing primarily to public institutions. That architecture is being redefined. The 2015 Every Student Succeeds Act decentralized authority to states, and 2025 marked a turning point as significant amounts of public funding began shifting away from public institutions and toward individuals through Education Savings Accounts (ESAs), voucher programs, and new federal tax credit mechanisms. More than 30 states now fund at least one private school choice program. This shift from a shared public system to a fragmented marketplace of state-by-state choices and privately administered accounts means that, as funding and participation disperse, the institutions intended to build civic identity and social trust are weakening, with no explicit effort to identify or support other ways to deliver on those goals. The Brookings Institution projects traditional public schools could lose another 8.5 million (~20 percent) students by 2050, raising serious doubt about whether the institutions we have historically counted on for delivering on our educational social contract will still exist at sufficient scale.
Together, these four forces don't call for reform of the old educational social contract. They require a new one.
What Families and Learners Are Asking For
Before defining a new social contract for education, we need to listen. Across dozens of conversations with families and learners, the LearnerStudio has heard four broad purposes.
The first is individual thriving, within a community. Parents want their children to reach their full potential; to be safe, healthy, and connected; and to have opportunities to develop both academic knowledge and personal passions. Before COVID, most described this in roughly Maslow-like terms: safety first, then academic quality, then enrichment. The pandemic disrupted that hierarchy. Remote learning exposed the emptiness of seat-time compliance and underscored the fact that belonging, mental health and motivation are not enrichment, they are essential. Families began naming resilience, connection, and purpose as core to their understanding of what it means to be a thriving learner.
The second is career preparation for a disrupted world. Education has always been a pathway to financial sustainability and meaningful work. For decades that meant "college readiness," but cracks in that promise appeared long before the pandemic. Rising tuition and student debt raised questions about return on investment. AI's disruption of traditional career ladders has accelerated the reckoning. What was once an "alternative pathway"—apprenticeships, trades, stackable credentials—is now increasingly understood as essential preparation for a workforce that AI is remaking in real time. At this point, even if we did accomplish what felt like an ambitious goal of “college for all” within the old efficiency framework, college is no longer a reliable proxy for career opportunity, dignity, or adulthood.
The third is values and character development. Families across the ideological spectrum want their children to develop a moral compass and a sense of identity and belonging. For some, this is framed in terms of justice and equity; for others, in terms of liberty and religious formation. But recent Beacon Project research suggests something that transcends this divide: Americans are hungry for what Beacon calls "morally directed agency"—agency for self-determination, yes, but also the felt ability to act together across differences toward a common good. Proficiency in math and ELA is important, but it cannot generate the trust, unity, and sense of shared purpose that society and democracy require.
The fourth is being a contributor to community. Most families, prior to this past year, rarely named "civics" or "democracy" explicitly. But most said something like: "I want my kid to be a productive contributor to society." Increasingly, they are asking for real-world and life skills that include the ability to solve real problems.
These four purposes need not be in tension. They can be a unified vision of what it means to flourish—as an individual, in a career, and as a participant in community and democracy. The old social contract addressed, at best, fragments of this vision. The new one needs to address all of it.
Where the Current Conversation Falls Short
McGuire and Wilka offer important suggestions like field-building, timely research, civics competencies, and local dialogue about shared learning goals. These are solid uses of philanthropic capital. But there are four places where their frame limits the conversation.
The most important is their treatment of educational choice as primarily a market phenomenon and a neoliberal device that fragments the public system. This misses something fundamental. Choice is not just an instrumental mechanism; it is an expression of human agency, which is a basic condition of well-being. As Barry Schwartz argues in The Paradox of Choice, too little agency leads not merely to dissatisfaction but to learned helplessness. Families make educational choices for two reasons: instrumental (to achieve outcomes) and expressive (to express their values or identity). Both matter.
My own family's experience illustrates the difference. We have accessed all four school types—district, charter, alternative, and private—as our daughters' needs changed. For high school, one needed to prioritize belonging and diversity over academic pressure; we enrolled her in a local charter school not because of test score data but because she needed to be with kids who felt like "her people." That was an expressive choice, and it enabled her to thrive in certain ways but not others. Our other daughter wanted rigorous academics and a strong commitment to community service, so she attended the local Catholic high school, though we are not Catholic. She made an expressive tradeoff to achieve an instrumental goal. Having the agency to make those choices increased our sense of responsibility to make them work. We also recognize that our family had access to resources—financial, informational, and relational—that made that navigation possible. Most families do not have that same access. The new social contract needs to enable choice and agency for all families and learners who want it.
When unions responded to charter schools not as a vehicle for expanding parent and learner agency within the public sector but as a competitive threat to be defeated, they may have inadvertently pushed families who wanted more agency toward ESAs, vouchers, and private alternatives. This is not primarily an ideological story. It is a systems story. The growth of vouchers and ESAs signals that families are looking for more responsiveness than the system was offering. A desire for agency doesn't disappear when suppressed; it adapts. When it adapts outside of public structures, it becomes harder to ensure that the common good is still being served. The lesson here is not that choice is dangerous. The lesson is that when systems don’t make room for the human need for agency, that need will eventually find itself being met elsewhere.
The second limitation is framing education as something that happens only in "school." Learning happens everywhere—on YouTube, in community organizations, through apprenticeships, in faith communities. Recognizing, enabling, and crediting learning across these contexts is essential. As Pamela Cantor, Fernande Raine, and Susan Rivers argue in “The Science of Experiential Civics,” real-world civic engagement is far more powerful than classroom instruction about civics. While it’s possible that McGuire and Wilka would agree, their default frame of "school" as the only location of learning is unnecessarily limiting.
Third, their analysis doesn't fully reckon with the power wealthy families have always had to exit the public system, avoiding any obligations intended to serve the public good. Any serious new social contract needs to address this and find ways to ensure that families attending private schools and/or using ESAs, vouchers, and private options are still contributing to, not just consuming, the common good. When public funds flow without civic obligations attached, we risk producing citizens who have never had to navigate difference, collaborate across disagreement, or invest in something beyond their own community, which is precisely what sustaining a pluralistic democracy requires. The consequence will likely be increased isolation and the slow erosion of the civic capacity democracy depends on.
Arthur Schlesinger described the recurring American pattern: periods of intense private focus followed by a turn to a generation that redirects its energy toward collective problem solving. We believe we have an opportunity to get beyond the idea that these pursuits are mutually exclusive. The choice between meeting the individual learner's private good of agentic choices and personalized pathways, and serving the collective civic public good, is a false one. This tension is not inevitable; it is a design failure we now have the capacity to solve. We now have the technological capacity to accomplish both simultaneously. The question is whether we have the creative capacity to design for it, and the will to demand it.
A New Social Contract
The old social contract said more or less: attend an efficiently organized system of schools, accept standardized content and residential assignments, and the common good will follow from common requirements. The new social contract must rest on more humanizing terms, with more mutually reinforcing allocations of trust, power, and control. Below are some possible starting points.
The new contract says:
- Learners and families hold meaningful agency.
- Systems earn trust through responsiveness and results.
- Power is distributed, but bounded by fairness and public purpose.
- Learning builds both individual flourishing and civic thriving.
On trust: Rather than asking families to trust experts in schools to define and deliver what their children need, the new contract asks the system to trust parents and learners with more agency, and to earn trust in return by demonstrating that it can serve each learner's actual needs—in and out of school. Trust flows from authentic relationships, deeper connections, and demonstrated responsiveness, not merely from credentialed authority. One example of this kind of increased trust is a new project by the Georgia Department of Education responding to increased student departures by partnering with the Institute for Self Directed Learning to support school districts in connecting with their communities, understanding parent and learner needs, and building new learner-centered education models to provide both deeper learning for students and to inspire a return to public schools among families who left for other learning environments.
On power: Rather than concentrating decision-making authority in institutions, the new contract distributes more power to learners and families. But distributing power without deliberately designing for equal access to information, resources, and support does not produce fairness; it reproduces the same gaps in a new form. The new contract must be explicit. Expanded agency is only meaningful if it is genuinely within reach for every family, not just those who already know how to find it. Without that commitment, we risk building a more flexible system that still sorts children by circumstance, just in new ways. This requires designing for fairness upfront, ensuring that those historically furthest from opportunity have greater access to resources, support, and high-quality options, not just equal access in theory. This is the distinction between what we might call a "mirage of standardization"—declaring common standards as a tool for achieving equity while failing to achieve them for many learners, producing what could be called "destructive variability" that correlates with income and race—versus achieving genuine "productive variability," where diverse pathways lead to flourishing in many different ways for different learners.
On control: Rather than the system owning the evidence of learning, learners will need to access and leverage their own data so they can chart adaptable paths. Learning and Employment Records (LERs), portable, mastery-based records of what a learner knows and can do, replace rigid institutional transcripts as the currency of the new system. This data sovereignty is not just a technical reform; it is a statement about whose life this is, and it needs to be durable across previous silos of age and schooling in order to create a tech-enabled backbone for lifelong learning. Innovation and multiple pathways are also going to require reframed measurement and accountability as we create new tools and infrastructure to support capturing mastery of a broader set of outcomes we want to achieve. The risk to guard against is allowing this reorientation toward broader outcomes to reduce rigorous, honest measurement of whether mastery is actually being achieved.
A new social contract is not only a set of promises from the system to families. It is also a set of obligations that flow the other way. Families who access public funds carry a responsibility to contribute to the common good, not just consume from it. Institutions must earn trust, and those who have always had the means to opt out must now have reason to invest. A contract that only asks things of those with the least is not a new social contract. It is the old one with new language.
As McGuire and Wilka remind us, education sits within, and is affected by, a larger political ecosystem. As we consider how to craft this new social contract for education in the age of AI, we see two worldviews competing to define the future more broadly. The first, which we at LearnerStudio call techno-nihilism, pretends there is a values-neutral default to technology. There is not. Techno-nihilists center efficiency above all else and assume that anything AI can do faster than humans, it should do. The second, which we call nostalgic humanism, responds to technology's acceleration by calling for a return to a pre-digital world. Both are inadequate. What we need is a new social contract to articulate a middle way, focused on flourishing. This means using technology purposefully, where it genuinely serves human development; protecting and cultivating the distinctly human capacities that AI cannot replace; and centering values like the dignity of human work, human connection, and human agency as non-negotiable core values.
System Design for the New Social Contract
The moment requires not only a new social contract but also a new system architecture. As we have shared previously, a three horizons heuristic helps clarify the design challenge. The first horizon is the old industrial system—still dominant, visibly in decline. The second horizon is reform work that tries to improve the old system but remains constrained by its efficiency rules and infrastructure. The third horizon is the system we need to create, emerging at a small scale in communities across the country, ready to be recognized, resourced, and scaled.
The new system needs to center three pillars: human agency, with young people as active drivers of their learning and humans as drivers of AI, not passive recipients; human connection, where learning is fundamentally relational, with technology enabling and amplifying rather than replacing human bonds; and human sustainability that equips learners with modernized content knowledge and skills, including AI fluency, cross-disciplinary problem-solving, leadership, entrepreneurship, and the morally directed agency they need to thrive and contribute.
These pillars need to connect with four concrete system design commitments.
Flexible and fair funding. We will need weighted student-based funding that accounts for poverty, disability, and other factors affecting learning, and follows the learner rather than the zip code. This untethers opportunity from residential geography. This might also include ways to provide more innovation flexibility like re-invigorating charter public schools by defining new “innovation charter schools” with significantly more freedom from traditional accountability rules, as well as similarly flexible district innovation schools and microschools as ways to strengthen public options that offer agentic learning.
Agency, flexibility, and sovereignty. Learners and families need the power to construct learning pathways that meet their needs—sometimes at the school level, often at the level of individual learning experiences. This means cultivating not only self-directed agency but also “morally directed agency”: the capacity and desire to contribute to something beyond oneself. And it means learners owning their data through LERs—portable evidence of mastery that empowers navigation of both the learning system and the labor market.
A spectrum of civic service. The concept of “civic thriving” advanced by Fernande Raine and colleagues at the History Co:Lab is core infrastructure for a functioning democracy in the age of AI. We believe this needs to be addressed in a developmental arc. It begins with elementary civic knowledge and service-learning, deepens through community-based problem-solving—middle school models like RevX and The Forest School are already demonstrating what this looks like—and we propose it should culminate in universal national service (civilian or military) for every young person between 18 and 24. This "civic utility" would build national cohesion across lines of geography, race, class, and ideology, and also addresses a practical problem. As AI has disrupted the entry-level career ladder, national service is one of the most powerful ways to rebuild it while simultaneously developing the civic skills democracy requires.
Supply-side stewardship. The state's role shifts from managing a standardized system to stewarding a robust, diverse ecosystem of high-quality options: schools, community learning hubs, microschools, lab schools, and stackable learning experiences as varied as the learners they serve, all held to broad standards of quality and access, redefined for flourishing in the age of artificial intelligence.
The Design Challenge
The destination we are designing toward is this: Every young person, regardless of zip code, is inspired and prepared to launch into adulthood with agency, purpose, a pathway to a fulfilling and sustaining career, and the desire and capacity to engage in our diverse democracy, not as a passive subject, but as an active co-creator.
This is a true design challenge, not a policy tweak. It requires what Thomas Kuhn would call a paradigm shift, not refinements to the old framework but the construction of a new one. The old “grammar of schooling,” a learner sitting passively in an age-sorted classroom, in a building called "school," progressing by seat time toward a standardized credential, needs to give way to a new “grammar of learning”: centering human agency, human connection, human sustainability, and providing mastery-based, personalized, relationally rich, credentialed learning in and beyond traditional school walls.
If we succeed, private gain and public good reinforce each other. Young people who experience genuine agency develop the confidence to exercise it in civic life, unleashing innovation that could fuel our economy, freeing learners from a check-the-box, do-as-you're-told mindset to innovate in ways that could create jobs, lift more people out of poverty, and improve our overall quality of life. Citizens who practice collaborative problem-solving build the relational muscle to sustain pluralistic democracy. Individual flourishing and collective thriving are not competing ends, they are the same destination approached from different directions.
If we fail, we will continue the cycle of overcorrection: swinging from standardization in the hopes of equity to privatization in the name of individual freedom, without ever building an integrated vision that can sustain both liberty and justice for all. The age of AI does not have to be an age of human displacement and disconnection. But if we hold to the old social contract, that is the path we are on.
A new social contract is possible. The need is clear. The models, tools, and policy ideas are emerging. What's needed now is the clarity to set a vision, the courage to build and scale, and the investment to ensure it fairly reaches every learner, not just those whose parents have the means to find it on their own.
Read more stories by Kim Smith.