Math matters.

I’ve written several columns about K-12 schools’ failure to effectively teach foundational reading skills. But a recent report about the decline of first-year university student math skills at one of California’s most selective universities was even more shocking. Its second sentence tells the sad tale.

“Between 2020 and 2025, the number of students whose math skills fall below high school level has increased nearly thirtyfold; moreover 70% of those students fall below middle school levels, reaching roughly one in twelve members of the entering cohort.”

The report rightly drew widespread attention from several news outlets across the ideological spectrum. That’s because the modern world runs on math. Every major public policy debate is rooted in or better understood with math. It’s the language of logic and problem-solving. And, no, you can’t rely on AI for help.

I recently tasked a leading AI app to do a tedious calculation. Reviewing its answer, I saw it was obviously wrong and noted the error to the chatbot. It apologized then made a different error the next time. It wasn’t an important question, and I noticed the mistakes. What happens when false answers to important questions go unnoticed?

I noticed them not because I’m a math wizard — because I’m not. I noticed them because, as I wrote in my last column, I learned math when we still taught it effectively and I can quickly see obvious errors even before I’m certain of the right answer.

If incoming college freshmen in California lack even middle school math skills, imagine the deficits across the rest of their age group around the country.

We don’t have to imagine. I reached out to Ohio State and according to university spokesperson Chris Booker, for the past five years, roughly 10-12 percent of first year students admitted to the Columbus campus require remedial math instruction.

I respect that my alma mater and Ohio’s flagship public university is committed to providing opportunities for students who might be otherwise qualified for college-level studies, but the reality is that many of those students really aren’t up to it and fail to graduate. That doesn’t serve either the degree-less students or the public. Students who lack foundational knowledge and skills in one subject area are often deficient in others. And remedial education simply doesn’t work at the scale necessary.

Learning loss during Covid, grade inflation, declining standards, and objections to standardized testing are all part of the problem but they stem from much more fundamental cultural and pedagogical shifts.

Until the last few decades, most teaching — especially in K-6 — was in the form of “explicit” instruction. The teacher, imbued with superior knowledge, explanatory skills, and clear authority supported by parents and administrators, conveyed the requisite knowledge to students. Some teachers derisively called it “drill and kill” — as in killing the passion for learning.

But whatever you called it, the practice worked, at least for most kids.

Regardless, many teachers and students hated it. So schools began to adopt different methods of “constructivist” or “exploratory” learning — using structured play and exercises to allow students, especially younger kids, to develop conceptual understanding.

Though different than the way I was taught, some of these methods apparently have  good theoretical and evidentiary support, as I learned in a conversation with Dr. Terri Bucci, associate professor of math in OSU’s College of Education. As a life-long learner, I’m glad to have access to the resources at Ohio State to acquire further insights, and Dr. Bucci was generous with her time.

Theory is useful but evidence in actual schoolchildren is what matters. It remains to be seen if the longer-term outcomes show the newer methods will work any better for learning math than they did for reading. Ohio State got the latter spectacularly wrong, continuing to support a widely discredited reading methodology that’s now prohibited under Ohio law.

For now, the evidence in math is overwhelming that whatever we’ve been doing for at least the past two decades isn’t working.

Given their life-long impact on kids, new and existing teaching methods and curricula should be subject to the types of randomized trials that did so much to improve outcomes in medical care. Few have. And those trials should occur before any new approaches are implemented at scale.

When I’ve suggested an education version of evidence-based medicine, I’m usually met with same excuses resistant physicians used: Teachers (doctors) are professionals and know what’s best for their patients (students). Children (patients) aren’t machines and the same methods don’t work for everyone. And, it’s unethical to withhold potentially beneficial interventions from patients (children) as part of a randomized study.

All are true — to a point. But the fact is that some things actually do work better than others for the overwhelming majority of both students and patients, regardless of the practitioner’s opinion and properly controlled studies can illuminate those facts.

As for ethics, we’re already withholding effective interventions from children — while subjecting them to ineffective ones. We’re just doing it with guesswork, rather than evidence.

Some things — like multiplication tables — simply must be memorized for instant recall (what cognitive scientists call automaticity) or every subsequent math skill becomes far more difficult to learn. Ohio standards already call for that achievement by the end of 3rd grade but ask a random handful of middle and high school students to rattle off a few of the 7’s or 12’s and don’t be surprised how few can do so.

In many cases, parents or school administrators push back against a child’s deserved bad grades or their need to repeat a year. That’s often because of misplaced concerns for the child’s self-esteem, or, in later years, the grades necessary to get into college. Unearned achievement doesn’t build self-esteem, it guarantees future disappointment

These problems must be solved at the source before we fail another generation of kids.

First, we should recommit ourselves in principle and in law that the purpose of K-6 education is to instill foundational knowledge of reading, writing, math, science, history, geography, civics, music, and the arts. That’s it. That’s the job. The only social justice initiatives that belong in public schools are those that assure every student actually graduates with that foundation.

The nation’s largest teacher’s union obviously disagrees.

The earliest years should, of course, be filled with fun. And we have a public duty to help children with special needs. There’s also plenty of nuance and things that are nice to have here and there. But overall state support must be explicitly directed toward assuring every student at least starts with a firm foundation on which to build further.

This isn’t a rap on teachers. Far from it. During my eight years of school board service, I developed a great appreciation for the professionalism and genuine care for kids that the overwhelming majority of our teachers and administrators — most of whom started as classroom teachers — bring to school every day. I have little doubt the same is true in most districts.

Instead, this is very much a problem for legislative leadership. Ohio has a long history of local control of public schools, which is appropriate given that most school funding comes from local property taxes. But every Ohio taxpayer contributes to the more than $8 billion in state funding for K-12 schools.

The next problem is curricula. Our lawmakers have made significant strides, first by requiring school districts to use curricula aligned with The Science of Reading. Then last month, the Ohio Senate passed SB19, intended to drive similar reforms for math. While both of these measures are positive steps, they fall short in providing solid guidance to help districts make better curricular choices.

Conversely, another piece of legislation, SB144, goes in the opposite direction by trying to solve a teacher shortage problem by relaxing standards. Under that bill, districts can hire any grade-level certified teacher for any subject, regardless of that teacher’s subject matter expertise. It’s exactly the reverse of what should happen.

The options in the curriculum buffet line are many but few have solid empirical evidence of efficacy. Ohio also chose to largely outsource curriculum review to EdReports, an independent non-profit education resource organization. While that might have seemed like a good idea, a blistering report about the deficiencies of EdReports reviews by education writer Karen Vaites calls that choice into serious question.

In an Atlantic article and her book, Natalie Wexler does an excellent job of explaining the need to pair explicit instruction with rich content. That’s where curriculum meets pedagogy and where teachers and schools need the most help — and direction.

Just as a return to the direct instruction of phonics is essential for reading, the same is true for teaching foundational concepts in math.

Ohio should instead conduct a systematic, evidence-based curriculum review of its own for both reading and math, building on the work of other states, and substantially narrow the list of approved choices.

The legislature should also further tighten the exceptions to the standards of achievement already established for reading and math. Retention — holding students back from 4th grade shouldn’t be optional. The evidence from the “Southern surge” of student achievement in Mississippi shows that retention is a powerful tool. It serves students directly by assuring they’re ready to move forward. Just as important, it focuses all stakeholders — students, parents, teachers, and administrators — on helping students actually being ready to move forward instead of making excuses why they aren’t.

Professor Bucci reminded me about how few people would willingly admit to being bad readers, but many are perfectly comfortable proclaiming their math misses. It’s almost worn as a badge of honor. Yet, innumeracy can be as harmful to future success as illiteracy and the language of math is no more difficult than learning to read English.

It’s clear; if we fail at building their foundations, we fail our children before they even begin. Those failures are more evident with every new report. It is truly a national crisis that demands resolution. We must all be resolute in demanding it.