Progress has stalled

As Figure 1 from Malkus’ paper shows, chronic absenteeism rates, the percentage of students who miss 10 percent or more of the school year, nearly doubled from 15% in 2019 to 28% in 2022. Those have come down a bit over time, but the rate of progress has slowed. At the national level, chronic absenteeism rates dropped 2.6 points in 2023, 2.2 points in 2024, and 0.9 points in 2025.

The patterns vary across states and schools, but the overall trajectory is not promising and suggests we may soon hit a plateau that’s much higher than we saw pre-pandemic.

It seems to be a systemic problem

Malkus has been following these data for years, but one new wrinkle with this year’s update is that he looked at how chronic absenteeism rates have changed across grade levels. Not every state reports the data in this way, but he found 13 states that did. It allowed him to investigate whether today's attendance problems are merely about the specific cohort of students who lived through school closures, or if the pandemic changed expectations about school attendance for everyone.

The chart below from the AEI report comes from Table 5 Panel B. It shows how chronic absenteeism rates have changed by grade from the 2019 baseline. Looking across the top row, chronic absenteeism among kindergartners in 2025 was 7.2 percentage points higher than it was for kindergartners in 2019. Those children were babies—or not yet born—during the pandemic. Yet their attendance looks the same as it does for older students. Malkus summarizes this by saying, “Chronic absenteeism is elevated in every grade in every post-pandemic year, and there is no clear differential between students who did and did not experience schooling during the pandemic.”

Another way to look at is to follow the stair-step line in the table. For example, first-graders in 2022 had chronic absenteeism rates that were 13.8 points higher than their peers from 2019. As those students moved on to second grades in 2023, their rates improved somewhat and they were now down only 10.5 points. A year later and they were down 7.3 points. In 2025, when these students were in fourth grade, they were 6.4 points behind. These are improvements, but the rate of change has slowed considerably.

In his conclusion, Malkus writes that, “whatever is keeping absenteeism elevated is a property of how American families and schools currently approach school attendance. The pandemic damaged our approach to schooling, not merely a particular generation of students.”

How to shift norms?

There’s a whole lot of blame to go around. Parents, Malkus notes, almost universally acknowledge the importance of going to school on a regular basis. However, parent surveys suggest that they may be wildly out of touch about the depths of the attendance challenge.

Students were more accurate about their own attendance patterns, but about half report missing school for “fun reasons, like going on vacation, skipping school, or traveling for a sport.”

Schools have invested a lot of resources into improving chronic absenteeism rates, and they are understandably leery of being held responsible for something they don’t think is under their control. But a recent news story from Oregon suggests accountability may have a role to play. After Oregon lawmakers eliminated truancy citations statewide in 2021, officials in one small county decided to restore citations as a last resort. That county now has one of the strongest attendance recoveries in the state. District leaders there note that they rarely have to actually issue citations or collect fines. Instead, the threat of legal consequences gives schools leverage to bring families to the table, connect them with services, and develop attendance plans.

The Oregon example is one case study showing how policymakers and school district leaders can prioritize in-person instructional time. Small signals matter here. When adults treat school as something that can be skipped for convenience, students and families notice. Likewise, when school leaders cancel classes, shorten days, or soften the consequences for missing school, they unintentionally reinforce the idea that instructional time isn't that important. But it is. Every missed day is a lost opportunity to learn, and a growing body of research links chronic absenteeism to lower achievement, lower graduation rates, and weaker long-term outcomes.

In other news, I had a fun conversation with about the “decoding threshold” and what it means for students and for policymakers. Check it out —>

The Bell Ringer

New study suggests phonics overdose with little “real” reading

“I love all your interviews. You are always a step ahead.” —paid subscriber Laura…

Read more

18 hours ago · 32 likes · 5 comments · Holly Korbey

Reading List

Oregon Journalism Project: Oregon’s poorest county now posts the state’s best attendance numbers

: The Secret to a Bigger Vocabulary

: How to do a tech detox with your teen

AASA: The Budgetary Impact of Healthcare Costs on School Districts

Kevin Mahnken: Why Have AP Scores Soared?

Collaborative for Student Success: Governors urge American Academy of Pediatrics to give families guidance on early math

Alex Kontorovich: Check out the Rutgers Math Corps

: Are Chromebook carts the answer to the ed tech conundrum?

: Should later school start times be an ed reform priority?

: How states can leverage the federal tax credit program to advance child care and early learning goals

John McWhorter: “What’s racist is pretending the test is the problem.”

I felt this in myself and have been on a personal journey to reclaim my time. I’m sharing it here in case it inspires anyone else…

As a kid, I was an avid reader. My parents left my brother and me home alone a lot, so I spent hours devouring almost every book we owned. I tore through The Hardy Boys, Roald Dahl, and C.S. Lewis, read all the John Grisham novels, even picked up a parenting book out of curiosity, and once attempted the Bible (I didn't get very far). When school held a competition for most books read, I finished second in my class.

But then, I slowly fell out of the habit. In high school and college, I read what I had to for school but not much else, only tucking into a few books when I got really bored in the summers.

In adulthood, with a wife, kids, and a full-time job where I read a lot of research and policy papers, tweets, and news articles, I found myself barely reading whole books anymore. But I began to feel like I was gorging on the reading equivalent of snacks or candy, so in 2018 I set a modest goal of reading at least one book per month. I surpassed that target and kept going. Since the beginning of 2021, I’ve averaged about a book a week.1 (Because I was/am naturally drawn to nonfiction, mostly written by white men, I now have diversity goals as well.) To keep myself honest, I use a simple Google Sheet where I log every book I finish.

I’m sharing this personal anecdote in case there’s anyone else out there who wants to rediscover a lost love of reading. It’s doable! I find that tracking my book reading is another way to manage my media diet, and now I can’t get to sleep at night without reading for at least 10-20 minutes. I prefer to read physical books, and I am a regular user of our local public library, but these are personal preferences and I’d encourage others to find what works for them.

Like any other habit, I find that the more I read, the more I want to read. The country seems to be moving in the opposite direction, but I've learned that habits can change. Mine did. If you've drifted away from books over the years, it might be worth trying to find your way back.

Reading List

Kate Taylor: Algebra-for-all backfires in Cambridge, MA

: Only 140 Massachusetts teachers were deemed to be unsatisfactory last year

: When scrutinizing private school choice spending, ask whether public schools also spend money on that thing

WSJ: Norway’s volunteer soccer coaches “preached a simple philosophy: ‘As many as possible, for as long as possible, and as good as possible.’”

In 2023, the Texas Legislature passed its version of an automatic enrollment policy. Going forward, school districts would have to automatically enroll students who scored in the top 40 percent on the grade 5 state math test into advanced coursework in grade 6, putting them on a pathway to complete Algebra I by 8th grade. The idea was simple: too many capable students were being left out of advanced math pathways not because they couldn’t handle it, but because no one told them they could.

Two years into implementation, the results from the spring 2026 state test (STAAR) are striking. The first cohort of affected students were in 7th grade this year, and the percentage of 7^th graders who took the grade 8 math assessment has doubled since 2023, rising from 16% to 32%. Last year, 127,760 7th graders took an advanced math class, up from 66,463 three years ago.

Source: 2026 STAAR Results presentation from the Texas Education Agency

More notable still: the gains weren’t confined to already-advantaged students. The number of economically disadvantaged 7th graders taking the grade 8 math assessment more than doubled between 2023 and 2026, compared to a 70% increase for their non-low-income peers. In other words, low-income students gained access to the advanced pathway at a much faster rate than everyone else.

The performance data are also holding up. When TEA analyzes math results by students’ enrolled grade level rather than the specific test they took, proficiency rates increased across every middle school grade in 2026.

The full picture on whether accelerated enrollment translates into long-term gains in Algebra won’t be available until this time next year, when the first cohort fully impacted by the law reaches 8^th grade. But the early evidence suggests that Texas may have found a relatively low-cost lever for expanding rigorous coursework. More states should look into creating their own versions!

Reading List

Allister Chang: DC’s high school diploma should mean something

Brookings: New Kentucky research complicates the corequisite course consensus in higher education

If a district has low reading scores and uses instructional practices that are inconsistent with what has been well documented about how children learn to read, the case is clear for switching to an evidence-based science of reading approach.

But Lexington’s young readers are not struggling. They have relatively strong test scores. So, perhaps not surprisingly, the prospect of being forced to change their literacy practices hasn’t sat well with Lexington’s leaders.

What should a state do with respect to districts like this? Lexington is a wealthy district, and it’s reading scores are… fine? On one hand, you could note that only 44% of the district’s low-income 3rd grade students scored proficient in reading last year, and only 25% of its students with disabilities did so. On the other hand, it’s also worth noting that each of these subgroups outperformed their peers statewide by a fair margin.

Could the district be doing even better? Probably, especially if it did a better job of teaching foundational skills in reading.

But should the state micro-manage a district like Lexington? That’s where I’m less convinced. As Rachel Schechter pointed out on LinkedIn, this district happens to have a knowledge-rich social studies curriculum in grades K-5, which may be helping to boost its reading scores. Other readers have noted that it’s quite possible the district’s wealth is contributing to its reading outcomes, not necessarily anything the district is doing instructionally. Regardless, I’d personally rather the state use carrots and financial incentives and tools like a phonics check for 6-year-olds as a way to ensure that all kids were learning how to decode, rather than the state imposing a list of reading programs. To my mind, it’s the student outcomes that matter, not the inputs.

I can see arguments either way here. So go read my full piece, and then weigh in in the comments!

Reading List

: It’s the Instructional Design, Stupid

InsideHigherEd: Brown Professor Suspects Majority of His Class Used AI to Cheat

Bill DeBaun: Class of 2026 Sets FAFSA Completion Record

Samantha Musacchia: The edtech lesson hiding in Toy Story 5

We had been in our old place for a long time. We loved our neighbors, and our mortgage was low. But it happened to be a natural transition year. My youngest is entering middle school, and my oldest is about to start high school.

And as we went looking into our options, we were confronted with a pretty harsh reality. We happened to be zoned for a school that has a “4” rating on GreatSchools. More galling, we lived at the very tip of high school district zone, and there were five other high schools within reasonable driving distance that all had higher ratings. Two are rated as “6s” on GreatSchools’ scale, and three are rated as “9s.”

So we moved. Not far, mind you. But we paid more money to upgrade to a nicer townhouse so we could officially be in one of these other zones.

I feel a bit guilty about all this, but I also recognize my privilege. And by writing this publicly I hope that other people who are making similar decisions can understand what it is that we’re doing and consider how this system affects people with fewer resources.

My friend Brian Kisida wrote about this issue for The 74. He noted that some critics resist choice politically while exercising it for their own kids through private schools or expensive homes in ‘good’ districts. He writes:

For lower-income families, access to better schools through the housing market is often out of reach. Policies that expand school choice — charter schools, vouchers or open enrollment — are among the few mechanisms that allow these parents to exercise the kind of educational agency that affluent families already enjoy. The real debate over school choice is not whether it should exist, but who gets access to it.

Choice critics argue that the focus should instead be on improving traditional public schools. For students assigned to underperforming schools, this means waiting indefinitely for reforms that may never arrive, while viable alternatives are blocked. It also assumes that equalizing school quality is both feasible and sufficient — a concept at odds with decades of uneven reform and a large body of evidence showing that peers and community environments shape long-term outcomes.

Now, I work in education policy. I’ve spent a lot of time looking at GreatSchools’ rating scale, and professionally I’ve worked on accountability and other rating systems. Given the weight of the evidence, I’m quite confident that our kids would have been fine at their old school. Maybe they would have even benefitted from a “big fish in a small pond” effect. But we had the means to buy our way into a nicer school, so we did.

This is how most families exercise school choice here in the USA. School choice isn’t some fringe policy experiment. It’s already the dominant system for families with money. We just don’t usually call it that. If families have the money, they can choose which school district and even which school zone they want to live in. My parents did it for me when I was a kid, and now I’ve done it for my own kids.

Reading List

: Ten Reasons Why Democrats Should Support the Federal Scholarship Tax Credit

: Why New Jersey should consider the federal scholarship program

Greenberg-Ellis and Pimentel: Connecticut Charters Break Through in Historic Legislative Session

NWEA study: Most kindergartners who start school behind never reach proficiency

Reflecting this same spirit, the state legislature in Oklahoma recently dropped chronic absenteeism from its school rating system. See Linda Jacobson’s story from The 74 about how the debate played out. Here’s a key quote:

“I’m not sure that it’s fair to evaluate schools based on something that we cannot control,” said Mike Simpson, superintendent of the Guthrie Public Schools, north of Oklahoma City. Originally in favor of making chronic absenteeism a factor in schools’ A-F grades, he no longer thinks it’s a good way to assess schools.

Can schools affect attendance?

No one disputes the fact that school attendance is highly correlated with student academic performance. It’s obvious that attendance is important, and every day that a child misses school affects their performance. The more days that a student can attend school, the better off they will be.

But the debate breaks down over whether schools can actually do anything about this. Should they just educate the students who show up in their building in the morning, or is there anything they can do to shape those student (and family) behaviors?

I think the answer is yes. While schools certainly cannot control every reason students miss school, they do have some control over student attendance. The most obvious example in front of us is COVID. As expectations around staying home changed, absenteeism surged. Schools have spent the past several years trying to rebuild those norms, with varied success.

Subscribe now

But studies have found that relatively low-cost interventions, such as personalized letters or text messages informing parents about their child's attendance can modestly reduce absenteeism. More intensive approaches, including early warning systems that identify at-risk students, mentoring programs, and efforts to strengthen school climate and student engagement, have also been shown to improve attendance. Sports, tutoring, and other activities can give students a reason to come to school.

In terms of actually implementing these ideas, a recent case study from SchoolStatus showed how they helped 89 districts make substantial improvements in their chronic absenteeism rates.

Importantly, attendance rates also vary across schools serving similar students and neighborhoods, suggesting that school practices and climate influence whether students choose to come to school. While no single intervention can eliminate chronic absenteeism, schools can improve attendance by identifying problems early (see this cool tool from AEI’s Nat Malkus and Sam Hollon), building stronger relationships with students and families, and making school a place where students feel connected and want to be present.

Iowa takes a different approach

Policymakers can affect attendance as well. Rather than throwing up their hands, Iowa has taken a different approach. According to a recent story by Wayne D’Orio, their tough love approach is paying off:

One state that stands out for its progress in beating this trend is Iowa. It passed a wide-ranging law two years ago that requires each school to notify parents if a child’s non-exempt absences hit 10% of a grading period, create a specific attendance plan for that student if missed days reach 15% and those absences are affecting academic progress and push the case to a county attorney for possible misdemeanor charges if a student misses 20% of school days.

In just one year after the law was passed, chronic absenteeism in Iowa fell to just below 16% in 2024-25 from 21.6% in 2023-24. Over the last three years, the state has seen a 9.8 percentage-point decrease in chronic absenteeism — significantly better than the 5.9 percentage point decrease seen nationwide during that time, [AEI’s Nat] Malkus said.

Of course, we can't know how much of Iowa's improvement was caused by the law itself. But the state's experience demonstrates that policymakers are not powerless in the face of a surge in absenteeism. Clear communication around norms, expectations, and consequences can help.

Ultimately, attendance isn't solely the responsibility of a student or the family or the school. Schools cannot eliminate chronic absenteeism, but they can make it better, or they can make it worse.

Accountability systems should recognize that reality. They shouldn't pretend attendance is entirely outside schools' control, nor should they expect schools to solve the problem alone. The question is how much weight attendance deserves alongside everything else schools are trying to accomplish.

Reading List

Aldeman: One Florida District’s Effort to Get 90% of Kids Reading Proficiently

: “Accountability, done well, …is about making sure we do not look away when students are not learning.”

: We Need Accountability, But We Aren’t Going to Re-make the Mistakes of NCLB

: How does Oregon, a state in the top tier of spending, land in the bottom tier of results?

Kelly Sia: There’s more to the i-Ready story

He starts off by noting that:

New data from Stanford’s Educational Opportunity Project confirms what close observers already knew: America’s test scores are slipping. The pandemic worsened the decline, but the slide began years before. In one-third of school districts, students are reading a full grade level lower than they were in 2015.

The part I bolded sounds really bad! Did anything happen in 2015 that could have led to such large achievement declines?

In his next paragraph he tells us that there was, indeed, a big policy change in 2015 that could have been contributing to these declines:

The new data is emboldening calls to restore something like the No Child Left Behind Act, the stringent, test-based accountability policy that defined American education from 2002 to 2015 and imposed penalties on schools whose students did not meet proficiency requirements on state standardized tests.

But weirdly, Wiener is not arguing that we should restore or improve upon those accountability policies. In fact, he suggests the opposite!

Subscribe now

Before going back to Wiener, let me remind you that the No Child Left Behind era coincided with some of the fastest achievement gains we’ve seen for Black, Hispanic, and low-income students, for students with disabilities, and for English Learners. It was by no means the starting point, but it culminated what Education Next authors dubbed “A Half-Century of Student Progress Nationwide.”1

Just to focus on the No Child Left Behind era specifically, take a look at a graph I worked on with Eamonn Fitzmaurice at The 74. I called it “A Tale of Two Eras,” because it compared the achievement gains during the No Child Left Behind era to what came afterwards.

Prior to 2013, achievement scores were rising. Those gains were broadly shared, including for the highest and lowest performing students (in yellow and red, respectively). But now look to the right of the graph. When NCLB was no longer in effect, achievement scores fell, and they fell particularly fast for the lowest performing students (in red).

Now, is this just a coincidence? Maybe? There were other things that were going on at the time that could explain some of these trends. In my piece for The 74 I explored some of those other possible explanations, but I argued that school accountability policy could plausibly explain at least some portion of what we’re seeing. Jim Wyckoff and Nat Malkus have done excellent analyses along similar lines.

But do test scores even matter?

Wiener seems to have made a hard pivot. He used to believe that objective, standardized measures of achievement mattered, and now he doesn’t. He now seems to believe that the act of trying to boost reading and math scores was in fact harmful:

Over time, I became convinced that, with the best of intentions, I and many others in the education reform community had transferred our moral commitment to children over to the standardized tests. We had done this earnestly, not cynically, but we still did damage.

His column barely grapples with why policymakers felt the need to create accountability systems in the first place. Before NCLB, many states simply did not know—or did not publicly acknowledge—how poorly some groups of students were performing. Without objective, comparable data, schools can tell comforting stories indefinitely. This is what we’re seeing now with the rise of grade inflation and high school graduation rates. More students are earning higher grades and graduating from high school, but objective measures suggest they are learning less than previous generations. When they go out in the real world, that lack of knowledge will affect their ability to do their jobs, whatever those might be.

Subscribe now

And this is where Wiener’s essay really breaks down. Per ’s point, Wiener implies there’s a choice between reading and math achievement on one hand, and richer educational experiences with “meaning and purpose” on the other. But there’s no inherent tradeoff. A school can (and should):

• teach kids to read and do math fluently,

• build strong character,

• offer meaningful projects,

• foster civic engagement, and

• create a welcoming environment so kids want to come to school.

These things are not in tension. In fact, solid fundamental skills in literacy and math are often what makes many of the other goals possible. Students who struggle with the basics are less able to participate in project-based learning, civic engagement, career preparation, and almost every other educational aspiration.

The real lesson of NCLB

The lesson of NCLB is not that accountability was a mistake; it’s that accountability systems need to evolve over time. NCLB got important things wrong, from its fixation on proficiency rates to its unrealistic expectation that all students would reach proficiency on the same timetable. It also encouraged schools to focus too heavily on tested subjects at the expense of other important goals.

But we should also be humble about how much progress any policy can deliver. The NCLB era produced some of the fastest achievement gains on record, particularly for historically underserved students, and yet those gains were still not enough to close longstanding gaps.

Reading List

Alex Zimmerman: Lots of buzzwords, little clarity: NY unveils new graduation requirements

Wayne D’Orio: Iowa Declared War on Chronic Absenteeism. Now It’s Gaining Some Ground

Linda Jacobson: This Summer Program Boosts Learning for Tens of Thousands of Kids

Shakeel and Peterson: A Half Century of Student Progress Nationwide

My blood pressure starts to rise just a bit every time I think about Social Security. I’m not particularly close to retirement age, but every year we get a bit closer to disaster.

Did you know that, according to the latest Social Security Trustees Report, the trust fund is expected to be depleted in 2032? That’s six years away. Because of the electoral calendar, this is no longer a distant problem. If the projections hold, the trust fund would run out during the next presidential term. That makes Social Security one of the most important unresolved policy questions facing whoever wins the 2028 election.

What happens then? Well, if nothing changes before 2032, every recipient of Social Security’s old-age benefits would receive an across-the-board cut of 22%. Social Security benefits would not stop entirely, but a 22% cut would make a meaningful difference to millions of recipients out there.

Subscribe now

I’m not sure people are aware of this, or just how soon this could play out. There are a lot of complicated projections around birth rates, life expectancy, and economic growth, and the dollar amounts at stake are enormous. So I turned to ChatGPT to help me simplify it:

Chad: When does the Social Security trust fund run out?

ChatGPT: The retirement and survivors fund (OASI) is currently projected to run out of reserves in late 2032. At that point, ongoing payroll tax revenue would still cover about 78% of scheduled benefits.

Chad: Do we know what month exactly?

ChatGPT: The latest 2026 Social Security Trustees Report is actually fairly specific: The retirement trust fund (OASI) is projected to be depleted in the 4th quarter of 2032.

The report does not pinpoint an exact month. “Fourth quarter of 2032” means sometime between October and December 2032.

One interesting wrinkle: last year’s report had the retirement fund running out in the first quarter of 2033, so the new estimate moved the date about 3–6 months earlier.

For planning purposes, I’d think of this less as a cliff in October/November/December 2032 and more as a political deadline.

Chad: Assuming nothing else changes, what happens in the 4th quarter of 2032?

ChatGPT: If Congress does nothing and the current projections prove accurate, the sequence is surprisingly mundane:

1. The OASI trust fund balance reaches zero sometime in the fourth quarter of 2032.

2. Social Security continues collecting payroll taxes every payday from workers and employers.

3. Because benefits can only be paid from available assets and current revenue, the Social Security Administration would have to reduce payments to match incoming revenue.

4. Retirees, survivors, and dependents receiving OASI benefits would see an across-the-board reduction of roughly 22% (based on current projections). There is no legal authority to borrow indefinitely to cover the gap.

For example:

The exact percentages would vary as revenues and benefit obligations evolve, but that’s the basic idea.

A few additional nuances:

• Benefits would not suddenly stop.

• The reduction would likely affect both current and future beneficiaries unless Congress specified otherwise.

• The cut would probably be implemented administratively by Social Security rather than through individual eligibility reviews.

The political implications are why most observers expect action before then. A 22% benefit cut would affect tens of millions of voters immediately, including many middle-income retirees who depend heavily on Social Security.

The most likely “nothing changes until the deadline” scenario is not that Congress ignores the problem until 2033. Rather, lawmakers wait until the last minute and then pass a package of tax increases, benefit adjustments, or both—much as they did in the lead-up to the 1983 Social Security reforms. The closer the deadline gets, the more likely that becomes.

Chad again: I’m less optimistic than ChatGPT is that a political fix is coming anytime soon. Regardless, now is the time to talk about what we want from Social Security and how to protect and reshape it for the generations to come.

Reading List

Aldeman: When a job list becomes a math policy argument

Dale Chu: Dynamic Assessment or Dynamic Branding?

Lina Eroh and Anu Malipatil: 5 Questions to Help Schools and Districts Make Smarter Ed Tech Decisions

:

In most states, we don’t really have a way to know how enrollment changes across the school year. But because Washington state conducts monthly counts, and uses those counts to drive funding levels, we can get a sense of how this might be playing out in other places.

The chart below shows monthly K-12 enrollment in Washington from 2016 through the current school year. The pattern is remarkably consistent. Enrollment rises from September to October and then gradually declines throughout the rest of the school year. By May, districts are serving about 1% fewer students than they reported in October.

This year, that amounted to roughly 11,000 fewer students.

The COVID-19 pandemic disrupted overall enrollment levels, but it did not fundamentally change the within-year pattern. Before the pandemic and after it, Washington schools typically lost about 1% of their students between fall and spring.

Subscribe now

Most states would never see this pattern. States that rely on a single fall count date may be funding schools based on enrollment levels that are higher than what districts serve for much of the school year. And, not only can Washington see these within-year patterns because they conduct monthly counts, the patterns may be smaller than they are in places with only a single fall count.

That matters for both state and local policymakers. If policymakers are basing funding and staffing models on October enrollment counts, it's a bit like designing a gym around the surge of new members who show up in January rather than the number who are still attending by spring.

The obvious question is what happens to these students. Given the rebound each fall, the graph likely reflects a combination of family mobility, transfers to private schools or homeschooling, and dropout and re-engagement patterns. States that rely on a single fall count may want to understand whether they experience similar within-year declines and how those patterns affect funding and staffing decisions.

Reading List

Jessika Harkay: A Delaware School Once Felt Like a ‘Prison.’ Now It’s a National Model

Anna Egalite: The Quiet Erosion of the Five-Day School Week

Kevin Mahnken: D.C. School Reforms Face Political Test

Jessica Baghian: “Accountability systems aren’t built for statisticians. They’re built for educators.”

The Mind Trust: Indy Summer Learning Labs Kicks Off Its Sixth Summer

: Social Security is not (just) a math or a political problem

: The real problem with “gifted” education

The key trend over the last 10 to 15 years has been a steady decline in student performance across a range of tests, across ages and grade levels, and across a variety of subjects. Moreover, the steepest declines have been among the lowest-performing kids.

But the latest results showed something different. In math, for example, the median score for 9-year-olds rose 3.5 points. For the highest performers (those at the 90^th percentile), scores rose by 0.7 points. In contrast, for students at the 10^th percentile, their math scores rose 7.5 points. Similarly, in reading, the lowest performers gained 9.3 points.

The gains were meaningful, historically large, and widespread across student groups.

You can read my piece for more, including important caveats and comparisons to other data sources.

Reading List

Aldeman: Grade inflation causes students to learn less

Tom Loveless: What Are Reasonable Goals for NAEP Scores?

Hannah Irion-Frake: Why I Stopped Using Running Records and Why You Should Too

: You can’t learn everything through the medium of Taylor Swift

They write that, “inflation outpaced year-over-year wage gains in May for the second month in a row. Inflation-adjusted hourly earnings declined 0.7% on the year in May after dropping 0.3% in April… That means households’ purchasing power isn’t keeping up with increases in the cost of things like rent, groceries and gasoline.”

That’s bad news for lots of workers, but it can be particularly painful for teachers. Unlike many private-sector workers, teachers often work under multi-year contracts that lock in salaries well before anyone knows what inflation will be. That can leave them vulnerable to sudden shocks.

Over the long, long-term, the average teacher salary has mostly kept up with inflation, but no more than that. And teacher salaries tend to lag behind inflation increases. Once inflation comes down, teacher salaries have a chance to catch up.

You can see this in the chart below. Using BLS data that also came out this week, I graphed the change in average teacher salaries versus inflation over the last 10 years. Starting from left to right, you can see that teacher salaries were not growing very fast in 2016 and 2017. But then the “Red for Ed” movement started winning larger salary increases in 2018 and 2019, and those kicked in in 2019 and 2020.

And we all remember what happened next. In the wake of COVID and federal stimulus checks, inflation exploded in 2021-2022, and teacher salaries could not keep up.

As inflation came back down though, teacher salaries were once again rising faster than inflation for the second half of 2023, all of 2024, and most of 2025.

My graph only runs through March of this year, and inflation has ticked up noticeably since then thanks to the war with Iran and the rise in gasoline price.

The good news is that teacher salaries have historically caught up once inflation stabilizes. The bad news is that when inflation rises suddenly, teachers often spend months or even years absorbing the hit before salary schedules have a chance to adjust.

Reading List

Aldeman: High school grade inflation lowers graduation rates, college attainment & future income

Lauren Wagner: How This Oklahoma School Gets Kids Reading at Grade Level

My quick take: Some reasonably good news, especially for 9-year-olds and the lowest-performing students. I’ll have more to say on that soon…

Share

In the meantime, this release updated some important survey data that provides context about how students experience schools. If schools want students to learn more, they're swimming against these powerful currents.

1. Fewer kids are reading for fun

Books continue to lose the battle for attention, for adults and for kids. This week’s data showed that we reached new lows in the percentage of 9-year-olds who read for fun, while the percentage of 13-year-olds reading in their free time held steady.

Why does this matter? Schools can teach reading, but they can't make students choose books in their free time. And there’s research suggesting that books — even children’s books — use more distinct words than the ones that come up in spoken conversations. The more kids read, and the wider variety of books they encounter, the larger their vocabulary is likely to become. And the bigger their vocabulary, the easier it will be for them to access and learn from new material.

2. More kids have no homework

Last month I shared data showing that more kids reported having zero homework. The new results extended that out to 2025, and we saw another big jump. When asked, “How much time did you spend on homework yesterday?” 37% of 9-year-olds and 40% of 13-year-olds said “none.” These figures are at all-time highs since NCES began collecting the data in the 1980s.

Why does homework matter? Because it creates additional opportunities for learning. It’s time for kids to practice the skills they’re learning in school, or to do extra reading they don’t have time to do at school.

Now, the research on homework is mixed, and it’s not always the case that more is better. But there’s some evidence that a small amount of homework that specifically reinforces what kids are learning in school can produce learning gains. It can also help students form productive habits and routines.

3. The rise and fall of 8th grade Algebra

For EduProgress.org, I wrote last year about the rise and fall of 8th grade Algebra. Throughout all of the 1990s and 2000s, the percentage of 8th graders taking Algebra rose steadily.

And then, it peaked and began to fall. Since 2012, the percentage of 8th graders taking Algebra has fallen from 34 to 23%.

Is this good or bad? Well, that depends. My takeaway from the research is that advanced opportunities are very helpful for students who are ready for them. But it can be harmful to accelerate students who are not ready or to just relabel courses without also changing the content students receive.

The trick, then, is to open doors of opportunity for students who are ready to succeed while giving other students more time to get there. That’s why I support automatic enrollment policies.

Whatever the motivation, fewer students today are taking Algebra by 8th grade than they were a decade ago. That’s not a good direction, and, in my view, schools should find ways to identify students who are ready for acceleration opportunities and then give those students a chance to succeed.

None of these trends are determinative on their own. Schools can still help students learn more, read more, and take on challenging coursework. But it's important to recognize the environment educators are operating in. If students are reading less for pleasure, spending less time on homework, and enrolling in fewer advanced classes in middle school, schools will have to work even harder to produce stronger academic outcomes.

Reading List

Aldeman: Oklahoma has a strong charter school sector

New NBER study: The diffusion of the iPhone reduced birth rates

Alex Tabarrok: The Nationalization of American Science

: Tough Times for an Education Budget Hawk

: Obvious, Intuitive, Seductive, and Wrong: Why Education’s “Instructional Illusions” Feel Right

Roland Fryer: “The classic summer job has been disappearing for nearly half a century”

Social Security Trustees: Unless Congress shores up the retirement program, the depletion of reserves would trigger a 22% reduction in benefits in late 2032

Competency-based education is the most promising alternative. Under this model, students only advance when they’ve mastered the material, not when the clock runs out. The logic is intuitive and the motivation is in the right place. But in practice, CBE has largely been confined to the margins, in things like credit recovery or alternative programs, where it often functions as a faster path to a passing grade or a credential rather than a more rigorous one to actual learning. Too often, the practical effect is to accelerate students through material quickly rather than ensure they have learned it deeply.

Moreover, the theory behind competency-based education doesn’t account for how students actually behave. Left to manage their own pace, most students won’t distribute their effort wisely. They’ll defer, delay, and then scramble. In my experience, teenagers are not exactly strong in the planning department, and without external structure, even motivated students will treat a flexible timeline as permission to procrastinate. A system that relies on students to self-regulate their way to mastery will mostly produce last-minute cramming with a competency label stamped on it.

A well-designed competency-based system can get around this issue. It would have students retrieve, apply, and revisit knowledge continuously along the way. By the time students make their way through these types of gateways checks, any endpoint assessment would become a confirmation of learning that’s already happened, not a high-stakes gamble on whether cramming worked.

This structure builds on one of the most replicated findings in learning research. Frequent, low-stakes quizzes and check-ins don’t just measure learning; they help produce it. If those retrieval opportunities are spaced out over time, the benefits will compound into deeper learning. Competency-based models could help students build that type of durable knowledge, but only if they include frequent opportunities to monitor progress along the way.

Reading List

Bruno Manno: AI reinforces the case for strong disciplinary knowledge

: When teachers ditch “productive struggle” for evidence, student performance improves

: Learning Pays More Than Schooling

: Gratitude for (Virginia’s) standardized tests

Pamela Snow: “education has a fraught and not very pretty history when it comes to generating, critiquing, and applying rigorous evidence”

Kerry McDonald: The College Credit Opportunity Few People Know About

Checker Finn is feeling optimistic

3. Redesign schooling so active learning, including project-based, experiential and career-connected learning, is the norm across all grade levels. That means redesigning accountability as well.

4. Ensure students have a solid foundation in literacy, numeracy and civic engagement.

These ideas may sound complementary, but they’re actually in tension with one another. And in the age of AI, schools should prioritize the latter.

For decades, progressive education reformers like Weingarten have argued that schools spend too much time on facts and “rote” memorization and not enough on projects, experiences, and real-world applications. The theory was that, as information became cheap and widely available, students didn’t need to memorize things anymore because they could always look them up.

Artificial intelligence has taken that logic even further. Students can now ask a chatbot to explain a concept, summarize a book, or draft an essay in seconds.

But that’s precisely why foundational knowledge has become more important, not less.

at Chalkbeat sat in on AFT’s AI teacher training, and it suffers from this same disconnect. Here’s the key excerpt from his piece for Chalkbeat:

I sat in on this training, and many teachers I spoke with said they appreciated its message and strategies. But one important idea was largely missing: that critical thinking is directly connected to the content in math, history, and science classes. This is an essential reality often absent from discussions about how schools should respond to the spread of generative AI.

Indeed, the common refrain that teachers should focus on abstract critical thinking skills, disconnected from content, risks de-emphasizing the very thing — fluency with a broad set of facts — that supports critical thinking.

“Domain knowledge is a crucial driver of thinking skill,” wrote University of Virginia cognitive psychologist Daniel Willingham in 2020 for the American Educator, a publication of the American Federation of Teachers. “Critical thinking for open-ended problems is enabled by extensive stores of knowledge.”

Project-based learning can be valuable when students already possess a strong foundation of knowledge and skills. Too often, however, schools treat projects as a substitute for content rather than an opportunity to apply content. The result in some cases is that students spend lots of time doing activities without really learning very much.

The students who will benefit most from AI won’t be the ones who know the least. It will be the ones who know enough to ask good questions, evaluate the answers, and then proceed to build on what they learn.

Reading List

Brian Bannon: America is facing a book-reading crisis

: How schools can “re-center” the book

: Can AI bring objectivity to Project-Based Learning?

:

Early Developments

Early childhood is at an inflection point

Following early childhood policy over the last six years has been a lot like riding roller coaster: The start of the COVID-19 pandemic hit early childhood programs hard, forcing many to close. Even though child care providers reopened much sooner than public schools in many cities and states, they struggled with COVID safety protocols, reduced enrollmen…

Read more

2 months ago · Sara Mead

A whopping 81% of Asian high school graduates immediately enrolled in a four-year college or university. “Immediate” in this case is defined as students who graduated at any point in the first nine months of 2024 and were enrolled in any form of postsecondary institution by October of the same year.

There were also notable gaps by gender. 69 percent of female high school completers immediately enrolled in college compared to 55 percent of males.

Overall, the immediate college-going rate in 2024 was 63 percent, which was down from the 68% in 2014 but not much changed from 2023.

Reading List

Jill Barshay: Inside the latest global research on school cellphone bans

Kunjan Narechania: States are doing both too much and too little on school improvement

The Right to Read Project: “I Was Certified to Teach But Not Prepared To”

University of California math professors: Reinstate the ACT/ SAT requirement for STEM majors

: Ohio 7th graders who have demonstrated readiness deserve advanced math opportunities

A large and growing body of research in cognitive science can give us a clearer picture of what’s happening here and how to learn better. Those findings have real, practical implications for teachers, parents, and coaches.

The Science of Learning report from Deans for Impact unpacks what cognitive science researchers have learned about learning and development. The report was originally written in partnership with cognitive and educational psychologists and is aimed primarily at teacher candidates and early-career educators. Now in its second edition, it was updated this month to reflect the most current research. It also flags common mistakes teachers make when trying to apply the science in practice.

Here are nine of my takeaways from the report.

1. Prior knowledge is the foundation of all new learning

Students don’t absorb new information in a vacuum — they connect it to what they already know. A well-sequenced, knowledge-rich curriculum is essential.

Moreover, teachers shouldn’t assume that just because a student has prior knowledge that they’ll know how to use it. Teachers may need to activate that knowledge through questions, retrieval practice, or analogies.

2. Your brain has limits

The brain can only consciously process so much at one time. When students are overwhelmed with new information, or asked to perform tasks that are cognitively too demanding, learning breaks down. This is why it can be valuable to pre-teach vocabulary words before a lesson and why worked examples are valuable for novices. This “working memory” is a practical constraint teachers must work within every day.

Share

3. Practice is good

Practice is essential for long-term learning, in almost every endeavor we face in life, but the type of practice matters as well. The most effective practice is spaced out over time, requires effortful retrieval from memory, and is woven across different problem types rather than blocked by topic.

Asking students to actively retrieve information rather than simply re-read it is one of the most well-supported study strategies. Crucially, this doesn’t have to feel like a test: think-pair-share, teaching a peer, or fill-in-the-blank exercises all count. As the report puts it, retrieval from memory should be “challenging but doable.”

4. Feedback is also good

Good feedback is one of the highest-leverage tools a teacher has, but not all feedback is good feedback. To be effective, the feedback must be specific and clear, focused on the task rather than the person, and oriented toward improvement rather than just grading in-the-moment performance.

5. Learning is hard

This is an unfortunate one, but if learning feels easy, it probably isn’t sticking. The strategies that produce the strongest long-term retention, such as spaced practice, retrieval, and interleaving, tend to feel more difficult and less satisfying. But they work. Students and teachers alike can mistake how easy something is in the moment for durable learning.

6. Skills may not transfer, at least not automatically

A student can’t just be told to “think critically” and apply that to new domains. In fact, knowledge and skills are deeply tied to specific content areas. A student who learns to identify the main idea in a reading passage won’t automatically apply that habit in a science lab.

Teachers can help this transfer process somewhat by helping students make those connections and recognize when a familiar concept applies in a new setting.

7. Self-regulation is a skill that can be taught

We can learn to actively manage our own learning. The report describes this as a cyclical process of planning, monitoring, and self-reflection.

Teachers can cultivate these skills by asking students to track their own progress, discuss strategies for tackling problems, and reflect on what’s working. When students get stuck, teachers should avoid jumping in with a quick answer. Instead, the report suggests asking students “What else can you try?” or “What resources could you use?” to foster independent thinking.

8. Common myths about learning are still surprisingly widespread

Despite decades of research, several popular ideas about learning turn out to be wrong or misleading. Learning styles — the idea that students are visual, auditory, or kinesthetic learners who need information delivered in their preferred mode — are mostly bogus. The left-brain/right-brain distinction is a myth. The belief that memorization is mere “rote” learning and therefore bad is actively harmful. Instead, building strong knowledge in long-term memory actually frees up the brain’s capacity for deeper thinking.

9. Motivation doesn’t have to precede learning

It might be tempting to assume that students need to be motivated before they can learn. But the relationship runs both ways. When people get better at something, motivation tends to follow.

In other words, teachers can build motivation by breaking complex tasks into manageable steps, sequencing content so early success is likely, and making improvement visible. Teachers don’t always have to manufacture student interest before teaching something.

Those are my takeaways. Read the Deans for Impact report for yourself here.

Reading List

Checker Finn: State leaders should opt in to the 12th grade NAEP tests

SRI Education: Connecting Word Reading to Meaning in Early Literacy

: Why we need both more and less technology

: When “Productive Struggle” Becomes “Destructive Frustration”

Subscribe now

So I asked my friend Michelle Lerner, a communications expert with experience at non-profits and school districts, on how she might draft a more encouraging message.

Here’s what she had to say:

It’s great news that Fairfax County Public Schools implemented automatic enrollment in middle school algebra, especially for students who are economically disadvantaged and students of color, those most frequently left out opportunities like these.

But this smart policy is obscured by less-then-clear communications. It’s a common problem. I write this as a parent of a child in another public school district and an inbox full of district emails I’ve had to read multiple times to understand.

It’s possible for districts to both create and implement great policies and to communicate about them effectively, which families want.

Below is a proposed message that would, for one, make automatic enrollment just that: automatic. The original letter clearly signaled the ways a student could opt out, but didn’t clearly explain the benefits.

In the suggested rewrite below I try to explain the value of the accelerated pathway and some level of personalization. A parent will take a school letter far more seriously if it includes some details about their child and, at a minimum, references their name. The proposed draft also adds excitement about high standards. If schools aren’t geeking about math, who will?

Dear [Student First Last Name’s] Family,

Congratulations! Based on [Student First Name]’s success in Advanced Math, [Student First Name] will be enrolled in Algebra 1 Honors next year. This is a high school-level class that moves at an accelerated pace and will be included on [Student First Name]’s high school transcript and GPA.

Completing algebra in middle school is an important milestone to accessing higher-level math and science in high school, which opens doors for students to find their passion in high school and beyond.

At Fairfax County Public Schools, nearly 85 percent of 8th graders take or already have completed Algebra I, helping to meet our goal that every student has access to high-quality academic programming.

We believe [Student First Name] can be successful in Algebra I Honors so no additional steps are needed to enroll. If you want to learn more about Algebra I Honors, you can find more information online. Please also see the email from Fairfax County Public Schools sent on [DATE].

If you have questions, please contact [Student First Name]’s math teacher at EMAIL ADDRESS, where you can alternatively enroll in Prealgebra Honors.

Thank you for helping us live up to our promise of excellence, equity, and opportunity at Fairfax County Public Schools!

[Sign off by Middle School Principal]

Reading List

Aldeman: North Carolina Has An Implementation Plan to Address Math Gaps

Julie O’Connor: New Jersey schools bought fake ‘Blue Ribbon’ awards

Adam Tyner: Why does it take so long to get Oklahoma school test results?

Will Oremus: Homework Is Starting to Look a Lot Like Candy Crush

: The tech-lash wave has not not peaked yet

: The Global Fertility Crisis Is Worse Than You Think

Jonathan Haidt: Pay Attention

Launched in May 2026, the Opportunity Knocks Scholarship will provide $8,000 annually for 500 students displaced by the Philadelphia school closures, to help families enroll in private schools throughout the city.

This is a clever idea. Janine and Jeff Yass are billionaire philanthropists looking to support education entrepreneurism (mainly in the private sector). By offering these scholarships, they’ll help up to 500 Philadelphia students transition out of one of the 17 public schools currently slated for closure and instead attend a private school they couldn’t otherwise afford. These are potentially vulnerable students. School closures are disruptive and kids can easily get off track socially and academically as they get separated from familiar friends and teachers. It’s a smart move all-around on the Yass’ part.

The Philadelphia public school system is not in a financial position to match their offer. After all, that’s why they’re closing schools!

But you know who could? A new scholarship granting organization could eventually do this for public schools. Thanks to the federal tax credit scholarship program that opens up next year, nonprofits could raise money for “transition scholarships” for students affected by school closures.

Now, this could only happen in Pennsylvania if Governor Josh Shapiro saw the opportunity here and decides to opt his state in to the program. But in states that do opt in, students could use stipends like this to pay for navigators or counselors to help them find new (public) schools, get tutoring, or attend summer or after school programs to make it a smooth transition and meet new people.

I’m probably not being creative enough here, but it could be an interesting use case for some entrepreneurial thinking in the public sector.

Reading List

: The bipartisan tax credit scholarship opt-in chorus grows louder

NBER: Breaking the Early Bell: Lessons from the First Statewide Mandate on School Start Times

: It’s Time for Democrats to Stop Fighting the Schools Their Voters Rely On

A recent mini-study looked at what happened after a subset of schools in Newark, New Jersey adopted Khan Academy as a supplemental math tool. Not only did they see strong usage numbers, but students who used it more frequently saw larger gains, and the gains were particularly large for previously low-performing students.

Across three school years, 13 schools in one part of the district offered Khan Academy as a math supplement to roughly 6,700 students. Of those students, 87% of the students qualified for free- or reduced-price lunch, a proxy measure for income. In the last year of the three-year period, 70% of students met the recommended usage of at least 30 hours minutes per week.

That alone is kind of a big deal. There’s plenty of research suggesting that high-quality apps like Khan Academy are helpful, but only for the kids who actually use them regularly. Laurence Holt dubbed this the “5 percent problem” because so few kids actually engage with the tools enough to see results. So the fact that Newark was able to get 70 percent engagement is quite good.

On average, the Khan Academy supplement helped produce small but noticeably positive gains. The returns were also linear, in the sense that kids who spent more time on the app and completed more lessons learned more. This may sound obvious, but it’s a good reminder that engagement really is the holy grail of learning.

But now look at the right side the chart above. When the authors sorted the effects of Khan Academy implementation across performance levels the largest effects, by far, were for the lowest-performing students (NJSLA level 1).

This is an important finding given that achievement scores across the country have fallen so dramatically for the lowest-performing students. In other words, schools and districts have tools available to address one of their biggest challenges. But students have to actually use those tools to see any benefits.

Reading List

Oreopolous et al: In-school supervised EdTech support produces massive learning gains (in India)

K12 Dive: 12 years later, San Francisco USD to offer Algebra I in 8th grade again

CALDER: The Divergent Paths of High- and Low-Achieving Students in Algebra I After the Pandemic