STEM education has made great strides across America since the turn of the century, permeating the public lexicon, driving school innovation and boosting interest in STEM professions. Most young learners, though, are still missing out. One of the more nettlesome grand challenges facing STEM education is expanding the opportunity for tried-and-true innovations of the STEM brand. Understanding the nature of those innovations, the inertia to their scaling and remedies readily deployable might just tip the scale toward STEM for all.
What’s Worth Scaling in STEM Education
By now, our grand experiments with STEM policies and programming have yielded ample evidence for scaling. Whether classroom curricula, competitions, community collaborations, teacher training, STEM academies, STEM stackable credentials, STEM awards or other vetted proofs of concept, they all factor into fueling America’s future workforce innovators.
Hundreds of effective learning and teaching strategies funded by the National Science Foundation, the U.S. Department of Education and scores of other public and private R&D investors (e.g., STEMworks) now fill the quivers of educators on the hunt for scalable bullseyes. From AI Professional Development Week to Zearn Math, Traffic Safety Drones to Statistics and Probability in Sports, and CompuGirls to Science Buddies and Teachers in Industry, the shelves are stocked with transformational approaches to STEM education. While these models and many more have scaled modestly, it’s still a small proportion of youth and schools that partake. Despite tantalizing gains among those who enjoy quality STEM programs, we’re not yet seeing systemic transformation accompanied by a talent tsunami for high demand career fields. STEM is not yet at that scale.
Inertia to Scaling STEM
Good ideas are scaled reflexively in the private sector. Vibe coding, introduced by Andrej Karpathy at OpenAI in 2025, is now ubiquitous across the IT sector in just over a year. A slower burn but more entrenched scaling is Kaizen, the continuous improvement strategy Toyota adopted half a century ago now transcending business globally. Then there’s the debatable good idea of dynamic pricing American Airlines honed in the 1980s and now universally applied by hotels, restaurants, rideshares, entertainment, retail, theme parks and more as surge pricing. The point is an effective innovation in commerce goes viral.
Why not in education?
Because Newton’s 1st Law—objects resist changes to their current state of motion—is strong in education. Inertia. Scaling a new idea in education is frustratingly incremental and uneven. It takes an external force to overcome inertia, said Newton. In education such force may come as funding (e.g., the National Defense Education Act of 1958, sigh), or a new law (e.g., the Literacy-Based Promotion Act of 2013 credited for the Mississippi Miracle) or widespread public demand (e.g., AI in education). The business world, on the other hand, has an overwhelming force: market-driven competition.
In their 2024 report on scaling K-12 STEM innovation, the National Academies hit the mark on how we got here: Our decentralized schooling system favors local autonomy. The next-big-thing in STEM education has a mountainous 13,000 school district climb to scale across America, but with no external force in sight (i.e., no “significant investment” as the Academies report calls for), scalers of innovation must go internal. What that looks like are educators, administrators and school boards overcoming inertia by finding what works, then mobilizing the resources to bring it—the materials, training, assessments, partners and, most importantly, buy-in on the part of teachers and communities—to scale. Scale-facilitating roles on the part of government, researchers and developers and constituents are critical. This magic does happen occasionally. Getting to 13,000 will require more than tweaks to the formula.
Overcoming Scale Inertia
The prevailing model for scaling is that an education researcher identifies an intervention or innovation that improves learning, then publishes the result in journals and presents the findings at conferences. The study might get replicated or modified for local nuance or outright adoption by practitioners. A commercial vendor may package the innovation for market. Outward-looking school administrators or go-getter teachers spot it and write a small grant to try it out. From there, innovation bakes in locally or burns out against local constraints like personnel turnover or shifts in district focus.
Every stage of that model is encumbered by inertia. Each point of resistance, of which there are five, can be overcome.
First and foremost, being a go-getter teacher or administrator can be a lonely existence. Education systems that build a culture of innovation do so by expecting a go-getter mentality of everybody. It’s a key professional development topic, best driven by an in-house working group of go-getters who sustain through turnover. An innovation expectation is prominent in performance review criteria of the best schools and is, in fact, a policy directive in Texas, Tennessee, New York and other states.
Second, those go-getters need better databases from which to harvest innovations. At the federal level, imagine a readily searchable and transparent (right now we don’t get to see how grant outcomes rate) STEM.gov akin to AI.gov. The NSF’s STEM award search is uncharacteristically clunky at serving up winning K-12 innovations. And the What Works Clearinghouse at the U.S. Department of Education can consume many precious hours to arrive at a practical idea (e.g., the search “AI and STEM” will leave a go-getter going elsewhere for ideas). What works well is the STEMworks platform, for example.
Third, what education researchers choose to study may have little relevance to on-the-ground practitioners. And if it does, the message can often get lost in academic-speak narratives. This is a call for college of education deans and department heads to guide the R&D faculty under their watch toward relevant work. And the same for funding organizations, grant agencies and award programs: Incentivize applicability and impact.
Fourth, in addition to or even in place of academic conference presentations and niche journal publications, aspiring scalers of research and development may do well to engage the public sphere. Blogs, podcasts, op-eds, social media and popular press reach exponentially more audiences than traditional dissemination and in understandable terms.
Fifth, and finally, some of the most successful scalers in K-12 STEM sidestepped the slogging research/evaluate/market path to thousands of classrooms by pitching to gatekeepers. Well-known models for computer science, high school engineering and career exploration have scaled in many U.S. states by impressing governors and state legislators. While this mode can trade fairness for efficiency and rigor for rapidity, advocates justify the risk for the reward of innovation over education inertia.
Come Scale Away
The United States is poised for the opportunity to provide all learners with a quality STEM learning experience. Readily deployable models dot the nation. Five modest changes to components of the educational scale system will overcome inertia, no new dollars necessary. Come scale away.