Editor’s note: This essay is an entry in Fordham’s 2026 Wonkathon, which asked contributors to answer this question: “What rules should govern the uses of digital technology in schools—and who should set them?” Learn more.
In the not-so-distant past, public school elementary students in your town used shared Chromebooks on a portable cart. Later, each of those students received their own Chromebook in a well-funded push to make educational apps more accessible. Now, parents—like my own neighbors—are signing a petition to put the Chromebooks back onto carts next year.
There was no master plan here. Instead, various layers of decentralized influence on schools were all just reacting to factors in the zeitgeist. We can form different opinions about who should theoretically set education policy, but the truth is, you cannot really top-down implement education policy whole cloth in the United States. A planned garden—especially one with exotic elements—takes a level of care that millions of disparate gardeners are not necessarily willing or able to provide. Instead, the garden ecosystem tends to exist somewhere between default conditions and planned ones.
As Tyack and Cuban explained so well in the now-canonical Tinkering Toward Utopia (1995), changes in American classrooms happen slowly, incrementally, and in cycles—at best. For better or for worse, idealistic (and ahistorical) educational reforms cannot ever truly reinvent the grammar of schooling or force the secular religion of cure-all schooling to deliver on its hefty promises. Teachers, understandably, are reluctant, or unable, to throw away their current methods for the next big thing in education. Instead, they are always adding in, remixing, and making reforms their own. In other words, “tinkering.”
In this climate of teacher hybridization of reforms and limited top-down control, what can deliberate policy actually accomplish in education and ed tech reform? For one thing, this type of system can remove things, or weed them out. It can sometimes implement reforms that leave the classroom core undisturbed. Legibility is key. The change must be easy to monitor, or teachers and their existing practices and priorities will quickly take over.
Rules that stand a chance of sticking include automated, hard-to-fake screen-time audits and software blockers that independent third parties can install (without teacher ability to turn them off). “Independent app audits,” on the other hand, fall between clear, legible reforms and too-squishy-to-matter ones. The system will predictably warp whatever happens in such an audit, for example, through the people selected to conduct it or the scoring mechanism (or even which apps make it into the choice set in the first place).
Stipulations like “teachers should teach wise tech usage” are basically just a distraction—more wish than policy. No one agrees on what using tech wisely means. There’s no way to teach teachers how to teach it clearly at scale, and there’s no way to audit whether students internalize those lessons. It would be better to just leave that kind of task to families.
If the ed tech ecosystem is a garden that we can shift but not design, the first, clearly viable set of interventions (automated screen-time audits, YouTube blocking software) would be like hardy plants that are native to this environment. The medium, iffy set of interventions (app audits) would be like plants that need a little TLC, and then measures like “AI literacy” would be orchids. They require more attention and care than is reasonable to expect in these conditions.
In this inherently unmanageable educational environment, it’s important to keep experimentation and diversity alive and avoid a monoculture. Otherwise, our educational experiments will be contaminated by low-fidelity implementation and lowest-common-denominator practices. Schools that choose different policies may be better able to enact reform within themselves because of existing stakeholder buy-in, transparency where it counts, teachers who opted in, and families who self-selected the model, too.
The overall picture is more like an evolving ecosystem, or a set of factors that sometimes reach temporary equilibria. That doesn’t mean we have no control, though. In the garden, we can control water, light, and soil. For instance, educational technology should be tested under real-world usage conditions, not theoretical ones. We also have some control over what gets reported. Complex systems comply with reporting requirements when they are clear and achievable, perhaps with financial incentives attached (though Goodhart’s law looms too: when a measure becomes a target, it ceases to be a good measure).
Special education exceptions are currently getting the most attention, but other candidates for reasonable exceptions exist too. For instance, one clear candidate for a tech exception should be students who are demonstrably already working above grade level. These students can’t grow because the current age-based grade-level system keeps them root-bound. Since baseline curriculum is already largely a waste of time for them, the downside risk of allowing them to move forward with their strengths on educational apps (hopefully vetted by the same high-quality process that vets general-use apps) is lower. In the past, gifted education sources might have referred to “correspondence courses” as a meaningful outlet for self-motivated gifted students. High-quality apps like Math Academy are simply the correspondence courses of today.
Although many individual teachers certainly have good judgment regarding which tech is high quality and whether it’s being used appropriately overall, we cannot depend on teacher discretion at scale to save the day. As teachers themselves are reporting, systems tend to succumb to entropy, and the lowest common denominator tends to prevail if not directly countered on a continuous basis. Put another way, the weeds come back.
Families shouldn’t write policy, but they provide a meaningful ongoing check on what policy can do and provide a counterweight when it’s overshot the mark. On the one hand, involved parents have a front-row view of whether the apps their children bring home are any good. Families see individual children’s growth in this ecosystem daily. If it happens along clear dimensions, they’ll notice the wilt before any centralized bureaucrat does.
On the other hand, though, we can’t expect parents as a group to be alert to the finer points of app development, like whether the app insists on mastery or incorporates spaced repetition. Parents are already a bit too enthralled by the framing of simple minutes of “screen time,” without sufficient regard for the actual quality of those minutes versus their potential quality. Parents are also sometimes left in the dark about what standardized test scores mean. If they don’t understand or trust standardized tests, they can’t respond appropriately to the tech, or lack of tech, that produced those scores.
If our ed tech ecosystems are truly being smothered by the pollution of education-unrelated computer use or apps that work worse than conventional alternatives, it is reasonable and necessary to demand that we weed them out quickly and clearly. But overall, tending America’s enormous and decentralized education ecosystem ultimately relies on the unruly efforts of millions of educators, administrators, and families who will never unite to implement any specific gardening program from the top down.
