Bad Outcomes and Better Systems

Every so often, a new popular explanation for why American K–12 education is not working surfaces. For some years, the problem was standards. Then it was testing. Then it was funding. More recently, the pandemic became the most frequently cited focus of reform conversations. Each story arrives with urgency, dominates the conversation for a time, and gets added to a growing list of very real background grievances. Meanwhile, educational outcomes continue to stagnate or worsen.

The truth is that well before COVID, student achievement in the United States had already flattened out. Between 2011 and 2019, national NAEP scores in 8th‑grade math hovered around 278–281 on a 0–500 scale, with no sustained upward trend, and reading scores fell modestly. After COVID, both scores dropped sharply. By 2022, average NAEP math scores were 5 points lower in 4th grade and 8 points lower in 8th grade than in 2019, the largest single‑cycle declines since NAEP began. The 2024 results show only partial recovery: national math scores remain several points below 2019 levels for both 4th and 8th graders, and reading scores are about 5 points lower than in 2019, continuing a downward trend that began before the pandemic. On the nation’s own assessments, only about a third of students reach the NAEP proficiency benchmark in reading or math, and that share has barely budged since the early 2010s. 

International comparisons make the picture starker. On the 2022 PISA exam, American fifteen‑year‑olds scored 465 in math, statistically similar to but slightly below the OECD average of 472 and far behind top systems like Singapore at 575, with several high performers such as Japan, Korea, and Estonia scoring above 500. In reading, U.S. students scored 504 versus an OECD average of 476, and in science 499 versus 485, but here too the trend over recent cycles is flat rather than rising. Presented alongside the cost, these deficiencies become more troubling. Across primary through upper‑secondary levels, the United States spends roughly 14,600 dollars per student per year, with total education spending from primary through tertiary at about 20,400 dollars per student, compared with an OECD average around 15,000 dollars. Education spending in the U.S. totals about 5.8 percent of GDP, versus an OECD average of 4.7 percent. We are paying more and getting very little, and have been for some time.

When officials point to modest post‑pandemic gains, they are usually comparing current results to the low point of 2022. From that very low floor, national 4th‑grade math scores ticked up by a couple of points between 2022 and 2024, and 8th‑grade math at least stopped falling, but both remain below 2019 and well below earlier highs like the 2013 peak in 8th‑grade math at 281. Reading shows the same pattern: 2024 scores for both 4th and 8th grade are about 5 points below 2019 and 2 points lower than in 2022. Recovery, in this context, means inching back toward a plateau that was already too low. It is evident that there is something fundamentally wrong in how the system is organized. 

And yet, there are schools that have bucked this trend in meaningful and relevant ways. They were able to do this because they made a set of data-driven operational choices which have proved effective.

Consider a high‑poverty elementary school (studied under the pseudonym Cinco Elementary School) in Florida that, for seven consecutive years, earned the state’s top accountability grade. Roughly 64 percent of its students qualified for free or reduced‑price lunch, a demographic profile that usually predicts chronic underperformance; nationally, one study found that only about 1 percent of high‑poverty schools consistently perform in the top third of their state’s performance distribution at all. “Cinco Elementary School” operated under the same basic funding formula as its peers and achieved an A rating year after year, performing in roughly the top 15 percent of all Florida elementary schools and the top 7 percent among high‑poverty schools statewide.​

The school used a shared, standards‑aligned curriculum across grades, with explicit agreement about what proficiency meant in reading and math at each level. Teachers planned backward from those expectations instead of improvising unit by unit. Assessment was constant and consequential. In one case study, baseline reading performance at the school sat around the 12th percentile on a norm‑referenced measure; over three years, reading achievement rose to the 44th percentile, a 32‑percentile‑point climb that moved the school from far below average into the midrange, on the way to its later top‑quartile status. Student work was examined regularly, gaps were identified early, and responses were fast and specific: small‑group reteaching, in‑school intervention blocks, and close progress monitoring rather than waiting for annual test scores to deliver bad news.​

Leadership reinforced this focus. The principal functioned as an instructional leader, constantly in classrooms, giving concrete feedback anchored in student outcomes, and organizing professional development around the school’s chosen curriculum and data. Teachers met weekly in grade‑level teams to align instruction, review assessment results, and plan interventions, turning collaboration from a slogan into a standing practice. Hiring was explicit about fit: the school prioritized teachers who wanted to work with low‑income students, believed all students regardless of background could meet rigorous standards, and were willing to collaborate and be coached. Over time, the staff stabilized around a shared way of working, and this cultural coherence is what made the academic coherence stick.​

The Florida school operated within the state’s existing finance system, with only the modest additional weights that high‑poverty schools receive statewide. What differed was how time, attention, and people were deployed. That is why the model is so compelling: it shows that a traditional public school can beat the poverty odds through disciplined curriculum, instruction, and leadership.​

A different path to improvement appears in the Detroit suburbs. Adlai Stevenson Elementary in Southfield, Michigan, serves a majority low‑income student body, with around two‑thirds of its students considered economically disadvantaged, and just a few years ago it was the lowest‑performing elementary school in its roughly 5,000‑student district. Achievement was weak, chronic absenteeism was high, and enrollment had declined to about 390 students as families with options drifted away. Instead of a state takeover, Stevenson was invited into a small community‑school initiative supported by regional philanthropy and a local civic intermediary. The amounts were modest: roughly $225,000 per school in grant funding, plus about $275,000 pooled regionally to support three peer schools. 

Stevenson committed to a full community‑school design with three big shifts. First, it hired a full‑time community school coordinator whose job was to integrate services and partnerships, not run isolated programs. That coordinator built a team of staff, parents, and community partners who set specific goals around reading, math, and attendance, and then matched resources to those goals. Second, the school made parents central to the turnaround. Staff contacted families, moved events to times they could attend, and invited them into classrooms and planning meetings. Within several years, about 98 percent of parents were attending conferences, an extremely high level of engagement. Third, Stevenson used funding to wrap supportive infrastructure around the academic core: on‑site counseling and health services, transportation assistance, and extended‑day tutoring that moved alongside the daily curriculum.

The results manifested quickly. Between spring 2022 and spring 2023, the percentage of Stevenson’s 3rd–5th graders who met or exceeded Michigan’s proficiency standard in English language arts rose by 14 percentage points, and math proficiency rose by 5 points. In 3rd‑grade reading, 42.3 percent of students were proficient or advanced, slightly above the Michigan state average of 40.9 percent, despite Stevenson serving a more disadvantaged population than the state as a whole. Enrollment climbed from roughly 390 students to nearly 500, and a waitlist formed—including families from outside the school’s immediate zone. A school that had been a last resort became a school of choice.​

Placed side by side, the Florida exemplar and Stevenson illustrate two complementary truths. The Florida school demonstrates that tight instruction, coherent curriculum, and strong leadership can move a high‑poverty school from the bottom of the distribution into the top 10–15 percent of performance, and keep it there. Stevenson shows that when you add integrated supports and parent‑centered design on top of a serious academic core, a failing school can produce double‑digit proficiency gains and dramatically increase engagement in a short period. Both schools are public. Both serve mostly low‑income children. Both operate within the constraints of state funding systems. And both achieved their gains by making operational choices—about curriculum, staffing, schedules, partnerships, and accountability—that most systems talk about but rarely sustain.

This article is not intended to be prescriptive, but to be a demonstration of what is already within reach, and what may rapidly proliferate as school choice policy and funding bring changes in K-12. The Florida school shows that even within ordinary funding formulas, a tightly organized academic model can move a high‑poverty campus into the top tier of performance and keep it there. Stevenson shows that when additional resources do arrive, they are extremely powerful in the hands of schools that organize partnerships, services, and time around family participation as their north star. Different school models are already appearing all over the country. Meanwhile, parents empowered to choose where their child goes to school mobilizes a distributed, highly-informed intelligence network for decision making – adults that know their children better than anyone, and can adjust, aid, and mobilize additional resources right when additional resources are needed.

It is true that both models described in this article have garnered some criticism. High‑performing, high‑poverty schools built around ambitious instruction can be intense places to work and depend heavily on unusually strong principals and willing staff. Community‑school turnarounds can be resource‑intensive and risk mission drift. But focusing only on those tensions misses the larger point, which is that in a national context where aggregate results are flat or worse, international standing in math is sliding, and per‑pupil spending is very high for real outcomes, the most important question is not whether a model is perfect but whether schools are moving the needle for the students they serve. 

In the end, the question is how we maximize learning for as many students as possible. Families, educators, and policymakers all need options that meet their specific needs. School choice, charter schools, and other innovative models shift power closer to families, make performance more transparent, and expand the diversity of schooling options available. The systems that will endure are the ones that adapt dynamically to a changing landscape, improve, drive forwards with facts on their side, and raise our standards.