This is revenues, not expenditures. And it doesn’t back out current expenditures versus debt service. Still, it’s a big, symbolic figure.

The Edunomics Lab team has more about how we got here and how the numbers vary across the states.

One more thing I’ll note: Benefit costs, including health care and retirement costs, now add up to $201.7 billion a year. That means benefits are eating up more than 30% of what public schools spend on annual personnel costs. I don’t think this is a great use of resources, and teachers would be much better off if they got more of their compensation in the form of salaries rather than in-kind benefits.

Reading List

Sarah Schwartz: Are high school graduates ready for college math?

Good Reason Houston: More students are meeting college- and career-ready benchmarks through lower-rigor pathways

Jo Napolitano: Maryland District Sheds Remedial High School Math Courses, Sees Students Soar

Erica Meltzer: Curriculum misalignment may cause students to become “intervention lifers”

Sarah Carr: Early intervention services including occupational, physical and speech therapies can improve children’s test scores, even years down the road

Emily Tate Sullivan: Kids in State-Funded Preschools Hit Record High

Kearney and Levine: “The decline in fertility is… broad-based, with strikingly similar trends across women with and without a college degree.”

: If America Is So Rich, Why Are Americans So Sad?

To learn more about these questions, I sat down with Elliot Regenstein, a Partner at Foresight Law and Policy and the author of a new book called Readiness: Preparing State Early Childhood Systems for a Bright Future. I picked his brain about the book, how early childhood fits within the “abundance” agenda, and how states can better serve parents looking for early childhood services for their children.

What follows is a lightly edited transcript of our conversation.

Chad Aldeman: You spend time at the beginning of the book talking about the different goals behind early childhood investments. Can you talk about what those are and how casual readers should think about them?

Elliot Regenstein: Sure, there are a few different rationales for early childhood investments. The book talks about how those can overlap and shape public services in different ways.

For example, one of the major arguments is child development, and there’s a fair amount of evidence that high-quality experiences for young children are impactful in both the short and long term. I often say that there’s pretty strong agreement and understanding that the first five years of child development are super important, and that adults influence that development in meaningful ways. What there’s less consensus about is, what exactly should government do about that?

Another major rationale for early childhood investment is keeping parents in the workforce, particularly when unemployment is very low [and workers are in high demand.] This has been a powerful rationale, because the argument is that if parents don’t have somewhere that they trust their kid will be well cared-for during the day, then one parent, usually the mom, will drop out of the workforce, which will reduce the overall pool of potential employees. And so there have been a few examples where business leaders have taken the lead in pushing for early childhood investment, in part because they’re just trying to get more folks in the workforce.

Different folks within the early childhood field have different views about those arguments, but fundamentally, those are the major reasons, and most early childhood services are designed to meet some combination of those two needs.

Aldeman: One takeaway I had from the book is just how fragmented the early childhood system is, if you want to call it a system. In addition to serving multiple goals, it has different funding sources, different pools of kids who are eligible, etc. We’ll get back to that.

But for now let’s set that aside and say you’re talking to a governor in a random state. If they said they wanted to be known as an “early education governor,” what would you recommend they do?

Regenstein: For a governor to show leadership on early childhood, there are a few things that are really important. One is at the state government level, you’ve got to get your act together to administer the system in a coherent way. If the pieces of the system do not work sufficiently well together, you can’t just add money and expect good results. The fragmentation negatively impacts providers and families, and you have to solve for that in some way. Different states have addressed that differently.

It’s also the case that in early childhood, more investment is necessary. In K-12, every child is required to be enrolled and offered a spot at a public school. Neither of those things are true in early childhood. I’m not arguing for mandatory early childhood. No one has argued for that. But even as a voluntary service, there are a lot of families who would take advantage of it if the opportunities were there, and there currently aren’t options available. Strengthening not only the administration of the system, but also the availability of quality services, is key.

And then a final piece is to emphasize the quality of child and family experience in designing the system. Historically, a lot of the measurement of quality in early childhood has been focused on essentially bureaucratic inputs as opposed to the actual experiences of children. There are a couple of states that have led the way in reorienting that conversation. That quality aspect is an important piece as well.

Aldeman: In your book, you write about South Carolina’s Parent Portal. What is it? How does it work? And what are other good examples of making early childhood supports work well for parents?

Regenstein: South Carolina has been a national leader, but there are a lot of states currently working on the issue of how to make it easier for families to access services.

The South Carolina people would tell you is it’s not strictly a technology problem. Technology can certainly help, but it’s really important to get the human systems in place to support families. Because in addition to the statewide technological infrastructure they’ve built, they also have a network of local-level supports to help families in a human way, to supplement the improved technology.

Historically, and still today in most places, if you were interested in finding child care, you had to do the research on your own. There are some websites that aggregate information, but those are not always comprehensive, and it’s not always set up for parents to contact or apply for the services. With South Carolina’s portal, not only could you identify services in your area, but you could apply for those services. Then, if you fill out one application and start on another, it pre-populates that application with all the information you entered to the first one, so that at the very least, you’re not having to repeat yourself over and over again.

Ultimately, what you’re hoping to achieve is a situation where parents can look and not only see what providers are available, but also see if they have spaces. Could I even enroll my child? Just having a child care center close to you, if it’s over-subscribed with a two-year waiting list, doesn’t really help you.

Aldeman: These types of supports could help in many realms. K-12 education is becoming increasingly unbundled. In early childhood, it’s certainly unbundled, and it can be hard for parents to know where to go. Are there places that have funded some sort of navigational support for families?

Regenstein: In fact, that’s been funded for a long time through a network of what are known as “resource and referral agencies” that are generally funded with child care funds. And there are resource and referral agencies in most states across the country.

Aldeman: Are those working well?

Regenstein: Well, some of them are. Some of them are visible and known to families, and have the sort of comprehensive coverage and ability to support families. Not all of them are, and to some degree that’s dependent on state context and the people involved. The need has been identified, and there are some resources going toward it.

It’s also the case the technology matters to this. Historically, a lot of these things were set up as, say, a phone number you would call, but young parents today are not calling phone numbers. They’re looking on the internet, from their phones. So what you need is more sort of app-based, or text-based, and some places are adapting faster than others. But the general gist is that states know that there’s much more to do here, and they are trying to ramp up those efforts.

Aldeman: Another area I’d like to get your feedback on is the so-called “abundance” agenda. There have been several popular books pushing for abundance in the housing space in particular, but also in energy and healthcare.

But those authors have not touched on K-12 education. They have also not touched on early childhood. But it seems like early childhood is ripe for this conversation as well. Parents say they want more of it, and yet the market is not delivering it in the quantities or at the prices parents can afford.

Meanwhile, there’s also a tension between quantity and quality. How do you see this tension playing out in the early childhood space?

Regenstein: I’ve enjoyed those books. As a brief aside, I really enjoyed Why Nothing Works by Mark Dunkelman.

Aldeman: Yes, me too.

Regenstein: Early childhood is easy to square with an abundance agenda, right? If you want children to have positive early experiences, if you want adults to be engaged in the workforce, then early childhood is really important.

But there is a huge market failure, because early childhood is extremely labor-intensive, particularly for the very youngest children who can’t change their own diaper and couldn’t get themselves out of a burning building. For those kids, it’s really important that the adult-child ratios are kept relatively low. You cannot have too many kids in the room or else it becomes a safety hazard.

But it’s also the case that young families don’t have a lot of spare income. In general, people make more money as they get older, and so if you are just a few years out of high school or college, which most first-time parents are, then you haven’t reached your full earning potential. You haven’t had a chance to save up, and now, all of a sudden, you’re hit with this enormous expense, and it’s very difficult to pay for it.

Historically, the market for early childhood has been dependent on a lot of free labor, so the salaries are also very low, with the exception of pre-K teachers employed by public school districts. The percentage of early childhood personnel who are on some form of public assistance is about half. These workers are in the third percentile of income nationally. Anecdotally, you’ll talk to center directors who lose staff to Walmart and Target or who decide to start driving for Uber full time because they can make more money that way.

So you have a couple of problems politically. One is that it is very hard to get people to pay more on a per-unit basis for a service that they’ve gotten used to pricing at a certain level. For state government leaders to say, “We’re going to invest a lot of money, not just so that more kids can have service, but so that that service is going to be much better and that the staff are going to be better paid” becomes a tough sell politically.

On top of that, a lot of the early childhood workforce isn’t particularly powerful politically. It’s low income, over 90% female, more diverse than the K-12 teacher population, and it doesn’t have unions that are making major political donations and mobilizing people to vote. It’s just not a constituency that policymakers have responded to. There are probably some number of abundance-oriented officials who, if you asked them, would say, “of course, we should be paying these folks better and delivering higher quality and expanding services,” but at the end of the day, they’re not going to get punished by voters for failing to do that.

Aldeman: My mind goes to advocates who push for bachelor’s degrees for all early childhood workers, in the hopes of professionalizing the workforce and boosting quality, but which costs more. Or, you note in the book that the Obama Administration passed Head Start regulations that will require them to raise employee wages to match that of local school districts by the year 2031. That’s going to cut the number of seats available at Head Start programs, no? They’ll have to scale back to serve fewer kids. How should we think about this?

Regenstein: There’s actually a point about Head Start that I want to come back to, but this has been a very lively debate within the field.

There’s a group that feels that, for early childhood professionals to be treated on equal footing with K-12 teachers, that’s going to require having the same kinds of credentials that K-12 teachers do. Degrees aren’t the only thing that define a profession, but they’ve historically been an important element. And if we’re going to deliver a quality service and take the field seriously, it needs to have degree requirements.

There’s a competing school of thought that the that degrees are not proven to improve competency. And, particularly in a diverse workforce, when many of the current workers don’t have bachelor degrees, putting in a degree requirement will make it almost impossible to hire, particularly for positions that are so low paid. Moreover, a lot of the skilled people already in the workforce will be pushed out, and there will be no one there to replace them.

There are also concerns about what happens when states or now the federal government start evaluating bachelor’s degree programs by the income of their graduates. At the very moment we’re talking about the need to expand early childhood services, higher ed would have a strong incentive to get out of the business of training people for it. For early childhood, that’s dangerous territory.

I’m not fully resolved on how to navigate these questions, but if I were a leader of that optimal state you described earlier, I would get stakeholders together and have a coherent strategy for what we want this to look like in 5-10 or 20 years. What pieces are we putting in place to make that happen? Because if the state’s not dealing with these questions, no one will.

Aldeman: Last year, the federal government created a tax credit scholarship program for K-12 education. But why not expand that to cover all kids from ages 0 to 18? That could potentially open up funding for scholarships at early childhood programs. What are your thoughts on that way to get money into the system?

Regenstein: Well, I don’t know if I agree with her broader philosophical point, but when Iggy Azalea said, “Never turn down nothing,” I think there was something to that.

As an early childhood advocate, when there are moving trains, there is a value to getting on them. That includes education savings accounts or tax credits from the state or federal levels. These may be imperfect vehicles for the kind of thoughtful system expansion that I advocate for, but it’s also the case that I’d rather have an imperfect world with more money than a perfect one with less.

If there is the opportunity to bring new resources into the system to serve families, I wouldn’t want to reflexively say, “No.”

And, if a state has a coherent thesis and strong leadership, they’re going to be able to figure out how to work an ESA or a tax credit into the existing fabric in a way that is additive. If a state doesn’t have that strong leadership, it may be more of a struggle, but one way or another, if there are opportunities to get more money, I think we should at least be open to that and be part of those conversations.

Aldeman: You wanted to go back to Head Start?

Regenstein: Yeah, and it’s about a tension within the early childhood field, which is, what age range should we focus on?

When Pre-K Now did its big campaign roughly 20 years ago, the focus was mainly on 4-year-olds. But there are a lot of people in the field whose emphasis is on birth through age three, because developmentally, that’s the most critical time period. But for the most part, when states have invested, they invest more in those later years than the earlier years.

One of the things that’s happened in Head Start is that in some of the states with very large state-funded pre-k programs for 4-year-olds, Head Start has adapted and started serving 3-year-olds, or even converted to Early Head Start to serve kids 0-3. Even though converting from a Head Start slot to an Early Head Start slot is a process that’s not easy or automatic, in most cases, the federal money is more flexible than the state money when it comes to serving infants and toddlers. So where states invest in 4-year-olds, Head Start often adjusts to serve other children.

If you look at the data on Georgia and Oklahoma, for example, those Head Start programs have de-emphasized 4-year-olds to focus on younger children. That is incrementally more expensive on a per-child basis, but it’s positive from a system perspective. The idea that these programs should be complementary rather than in conflict is a really powerful notion, and it would be great to see a future iteration of Head Start that accounts for changes at the state level.

Aldeman: For my last question, let’s go back to my hypothetical. If you could build a Frankenstein model of elements from multiple states, what core components would you like to see?

Regenstein: Well, I’m not a deep expert on every state, so my answer is going to potentially omit some good work happening in some places…

But I’ve actually been using Virginia as one of my primary examples. First of all, it emphasizes just how bipartisan this work is. There are a lot of blue states that have done great work in this area. There are a lot of red states that have done great work too. But in Virginia it’s been both. A Democrat brought in Jenna Conway and her team, Republican Governor Youngkin promoted them, and then {newly elected} Democratic Governor Spanberger promoted them all again.

There are a few things that are really core to Virginia’s success. One is they have unified state leadership. They have the core programs under the same roof. And in Virginia’s case, it’s not a standalone agency, it’s the Department of Education. They made the choice to bring the programs together so that they can be run as a holistic unit, not as multiple different things administered by different mid-level managers.

They also have a relentless focus on data and quality. They are one of the states that has a definition of quality focused on child experience, and more importantly, when it comes to both quality and access, they are scarfing up all the data they can and analyzing it to inform their decisions in a way that a number of states are working toward, but that not everybody is there yet. Notably, Virginia has also increased its investment in the last few years.

It’s also important to give credit to the states like Oklahoma and Georgia and West Virginia, which have had a longstanding commitment to access and quality. Alabama, Rhode Island, and Michigan are also known for their focus on quality. These states insist on ensuring quality in their pre-k programs. So again, there are a lot of different states that are doing some aspect well.

We’ve clearly made enormous progress in the last 10-20 years, in terms of access and quality and system design. And my hope is that when we look back in another 20 years, we’ll see that we’ve continued to make substantial progress, because there is still a long way to go.

Aldeman: That’s a great point to end on, Elliot, and hopefully your book will help lead us in that direction.

Reading List

Chalkbeat: Indiana restarts child care vouchers with one-time $200 million boost

Urban Institute: DC’s Early Childhood Educator Compensation Program Helps Stabilize Child Care Businesses

That framing is misleading, or at least incomplete.

The program is not a direct spending initiative that pulls money out of public school budgets. It operates through the tax code, offering credits to donors who contribute to scholarship-granting organizations. Left on its own, the program may well tilt toward private school tuition. But policy design—not just intent—will determine whether that’s the only outcome. That distinction matters, and its effects will ultimately depend heavily on how the program is structured, regulated, and used.

And that’s where the repeal effort falls short. By treating the program as nothing more than a traditional voucher, critics risk overlooking how it could be deployed to support a broad array of educational services, including tutoring, afterschool programs, and summer learning opportunities that benefit students in public or private schools. With some slight entrepreneurial thinking on the part of Democrats, this program could direct substantial new resources into those areas.

The question, then, is not whether the program should exist at all, but whether leaders of all political stripes will be willing to step up and use it in ways that align with their priorities. Right now, too many Democrats are focused on repeal. That could backfire and leave a lot of money and opportunities on the table. A more productive approach would be to utilize the policy and improve it so it better targets need, protects students, and expands access to high-quality learning opportunities.

Subscribe now

As currently designed, the program falls short on several of those fronts. But those shortcomings point to opportunities for reform, not reasons to abandon the effort altogether. Here are four areas where policymakers could focus their attention:

Prevent discrimination by scholarship-granting organizations

Right now, the law is silent on which types of programs are eligible for funding. But we’re talking about a federal tax credit here that could cost the Treasury billions of dollars a year in foregone revenue—the least Congress can do is require that scholarship granting organizations (SGOs) not discriminate based on a student’s race, religion, sex, national origin, and disability status. You want a tax credit? You can’t discriminate.

Target scholarships to families with financial needs

When drafting the original legislation, the Republican Congress did set an income limit, but it’s laughably high. Recipients must be able to document that they earn within “300 percent of the area median gross income.” In Northern Virginia, where I live, the maximum cutoff is someone earning $491,700 per year. That is far too high. People earning that much money shouldn’t qualify for a federally subsidized education “scholarship.” That’s not a wise use of tax dollars, and Democrats should work to lower the income cap to focus on truly needy families.

Expand eligibility to cover early childhood services

The current program focuses on K-12 educational expenses. It’s a broad list of allowable expenditures including tutoring and other supplemental services, but it is limited to K-12. But why stop there? Many of the largest gaps in access occur earlier in a child’s life. Extending eligibility to cover early childhood services—from birth through age five—would open up a new stream of funding for childcare providers and early learning programs. That would not only support children’s development but also help more families afford care and participate in the workforce.

Give states the flexibility to adjust the program as they see fit

The law was written in such a way that the Treasury Department could have allowed states flexibility about the types of SGOs they wanted to let operate in their state. The Trump Administration interpreted it more rigidly and seem to be taking an “all-or-nothing” approach. We’re still waiting on the final rules, but Democrats could demand that states get a say. Federal rules could allow states greater flexibility to define eligible providers, set accountability standards, and direct funds toward priority areas. Some states may choose to emphasize private school options, while others might prioritize tutoring, summer learning, or evidence-based interventions within public school choice systems. Allowing that variation would enable states to align the program with their own policy goals while maintaining appropriate guardrails.

Lawmakers and advocates can continue to debate the program in familiar, polarized terms, or they can engage with the details and shape it into something that better serves students. Focusing exclusively on repeal risks ceding that ground.

A more constructive path would be to treat the program as a starting point—one that can be refined to better target need, expand opportunity, and ensure that public resources are used responsibly. Democrats who ignore those possibilities are doing a disservice to the children in their states.

Reading List

Jeb Bush: State leaders need to eat more broccoli

Rachel Canter: States can’t pursue Mississippi-lite versions and expect the same results

Before I show you the graph, take a moment to envision where you think new teachers come from. Are they 22-year-olds coming directly from a teacher preparation program at a college or university?

And take a moment to think about the number of newly licensed teachers versus the number of new teacher hires. Which one would you expect to be higher?

You might assume that these lines would run pretty much in tandem. Or maybe you might think that the state’s teacher preparation programs are preparing a few more teachers each year than eventually get hired. That is, not every teacher preparation candidate can end up with a teaching job in-state.

Now look at the chart below, from the Texas Education Agency. It compares the number of newly certified teachers in Texas (in blue) versus the number of new teachers hired by public schools (in gray).

Not only are there more new teacher hires, the numbers aren’t particularly close! In fact, the gap has been widening more over time.

Who are these people? Is this just unique to Texas? Yes and no.

The next chart comes from the same state dashboard. As you can see, the largest source of new hires are actually re-enterers. These are people who taught for a while, took some time off, and then came back into the profession. Having a large number of re-entrants is not unique to Texas and is often a large source of teachers for schools all across the country.

But the second biggest source of new hires was unique to Texas. These are people with “No Texas Certification/Permit.” This source of 12,000 teachers is unique to Texas, which, for a few years, allowed schools to hire teachers without any prior certification. The state has recently pulled this back due to research suggesting these teachers weren’t as effective in the classroom and tended to have higher turnover rates. Still, it will be interesting to watch how Texas schools fill that gap in the future.

Before I leave off, take a moment to look at all the other categories here. More teachers enter Texas schools with an alternative “intern” certificate than a “standard” teacher preparation program. The remainder of the list includes things like teachers who were working part-time, who came from out of state, or who earned a certification in the past but hadn’t found a teaching job yet (“lagged” entrants).

Again, Texas is unique in some ways, but the next time you think about the teacher preparation pipeline in your state, I hope you’ll remember that it often looks different than the standard story of college—>teacher certification—>teaching job.

My quote of the week award goes to Andy Rotherham, near the end of this passage on AI and education:

So for schools, whether you like AI or think it’s good for education matters less than how you respond to it. In the same way that whatever you think about cold matters a lot less than dressing appropriately when it’s going to be freezing out.

The second is carpentry. I’m a mediocre carpenter. I have friends who do it professionally or as serious hobbyists and they’re not mediocre. They take shortcuts, move quickly, and can skip steps. I can’t do those things. I take my time, I follow the sequence. The same is true of cooking: amateurs follow exact recipes, pros can improvise. To some extent this is also true in music, which is why pros can sit in and play together in ways amateurs cannot.

You can see where I’m going. The same logic applies to learning in school. If you already know how to write, you can sometimes skip steps. If you’re a skilled reader, the same is true. And in areas where you’ve accumulated a lot of knowledge, you develop a kind of “horse sense” that translates to efficiency and allows you to use some tools differently. But none of that is helpful while you’re still learning. You need to follow the steps — and the shortcutting we now call “cognitive offloading” doesn’t help you learn. It makes you worse off.

…

The idea that we live in such revolutionary times that you don’t need to know anything — you can just look it up — is a seductive, stubborn, and bad, idea that lives zombie-like across the sector.

Subscribe now

Reading List

Aldeman: When Students Earn a High School Diploma But Still Can’t Do Math

Jessika Harkay: High-Need Connecticut School District Doing ‘Things People Don’t Believe Are Possible’

Upshot: What New Federal Loan Caps Will Mean for Graduate Students

: Will tutoring save us?

: Education research is weak and sloppy.

Jed Wallace: Three Policy Changes to Accelerate the Marginal Revolution

: “you have to actually build up a teen’s Real Life Substitutes for screentime”

I have this (perhaps irrational) hope that as student enrollment continues to fall (which it almost certainly will, given the recent significant drops in births), state and local budgets will be able to free up resources to invest in early childhood.

Is it crazy? So far, district leaders have mostly tried to hide from enrollment losses. They have been reluctant to close under-enrolled schools and have used financial windfalls to boost staffing numbers. But can they be more proactive and actually divert some of that money to invest in younger kids? What would that take?

Theresa and I sat down recently to talk about her idea. What follows is a lightly edited version of our conversation.

Chad Aldeman: We’re talking today about the problem, or opportunity, of communities with declining enrollment thanks to declining birth rates. If they’re serving fewer kids, can funds be put to better uses? You came to me with this idea. Can you explain what you’re thinking?

Theresa Hawley: As I look at it, we’ve seen a dramatic decline in births in this country, and that’s going to translate over the next couple of years into dramatic declines in the number of school-age children, especially in the early grades. Depending on the state, we’re talking about declines of 15% over the last 10 to 15 years of the number of kids being born. That’s the equivalent of more than a grade’s worth of children.

We also have a financing system for K 12 education based on local property taxes and state investments. That’s likely not going to go down in terms of the revenue stream that’s available to the system, but the number of kids that need to be served is going to go down a lot.

So I look at this as an opportunity for early care and education to say, “How can some of those resources be put towards serving more children at younger ages? How can we use some of those resources to meet the age-old problem in early childhood, which is that we need to start educating our children younger?” We absolutely need to invest more in early care and education. My whole career I’ve been told, “Unfortunately, we just don’t have the resources.”

Aldeman: As additional context, could you also talk about your team’s comparison of spending in K 12 education versus early childhood?

Hawley: Unfortunately, no matter how much we believe that it’s important to invest in the earliest years, our country doesn’t. We spend 21 cents on the dollar for preschoolers and only 11 cents on the dollar for infants and toddlers compared to what we spend on kids once they enter the door at kindergarten.

That ranges quite a lot across different states, but even the very best states are spending a fraction on kids at age four that they’re spending on them when they turn age five and enter kindergarten.

This is an opportunity for us to widen our frame a little bit and see the education system starting a bit younger, and getting resources into the system to invest earlier. We know that would make a positive impact on kids’ long-term development.

Aldeman: What you’re asking is for governors and mayors to look out a little bit further in the future than they’re normally accustomed to. Rather than just looking at two- or three-year enrollment projections, they should be looking five, 10, or even 15 years out. My fear is, if they don’t do that, they’ll continue to be reactive. They’ll just keep doing short-term fixes. Does that sound right to you, and how do we change that mentality?

Hawley: This is absolutely my fear. These [enrollment shifts] are going to be slow and gradual, 1% a year, 2% a year. In that scenario, the resources that are in the K-12 system will just continue to be absorbed by K-12 schools, and there won’t be any room for investments in early childhood.

So I think it’s important for us to zoom out and to make the case. If we’re serious about the need for our country to invest in kids at younger ages, we have to look at the revenue systems we already have and opportunities that might be coming in them to be able to make new investments.

Aldeman: Talk to me about some of the benefits of doing this. One might be just that more parents would have a place for their child to go. Potentially, you could also have better alignment from early childhood into K-12. And three, we could see better services for kids. Is that right, or what else would you add to the list?

Hawley: I think there are two big questions here. One is the question of the money, and the other is the question of which entity should provide the services to children and families.

Let’s start with the money. Just because I’m saying that the resources may be there thanks to a shrinking need in K-12, that doesn’t necessarily mean that the public schools must be the ones who have to serve even younger children.

We’ve seen a lot of really good models out there. Many schools do a great job of providing pre-K education to 4-year-olds, sometimes to 3-year-olds. But public schools could also partner with community-based organizations and pass funding through to those organizations so that they can serve kids with full-day, year-round services, in settings that are a little bit smaller in scale and more intimate and more conducive to the needs of young kids. Those kids are more likely to have to take a nap in the middle of the day than your average elementary students would be.

There are benefits to seeing early care and K-12 services as part of a larger education system. That includes aligning the curriculum that’s used in a 3-year-old and 4-year-old program with what that child is going to encounter when they enter kindergarten and first grade. Ideally that becomes a seamless continuum, so that kids are building up language skills that move straight into reading readiness, they are building up quantitative skills that lead right into math readiness.

There’s tons of research that shows that the partnership between public schools and community-based settings where younger kids are being served makes a huge difference in how effective those investments are in shaping the long-term trajectory of child development.

Aldeman: Are there any models here that you’re thinking of? What places have done this well already?

Hawley: Yes, you can look at Washington, D.C. as a city that led early on in terms of investing in 3- and 4-year-old education. They’ve been taking that seriously for more than a decade now. Kids receive services not just in schools but also in community-based settings, and in charter schools. Regardless of the different kind of settings that kids might be in, they have universal access to those services.

New York City is has made tremendous strides as well. Again, it uses a mixed delivery system across multiple settings, but the city and the school system is investing in those younger grades.

Aldeman: I appreciate the point about being neutral to who is actually providing the services.

I also want to throw in a new study out of the Urban Institute that looked at universal pre-k in D.C. They found that it actually boosted enrollment for the K-12 schools as well, because the families were more committed to the system. So there’s another upside for the systems that are part of it.

Hawley: For sure, and actually I think that’s why a lot of cities do invest in 3- and 4-year-olds.

It’s part of why cities have been in the vanguard of this, because there’s an understanding that if they can get kids into the school system and keep families in the city when their kids are three and four, they may be able to keep them when their kids are five, six, seven, and eight. Or, if a younger sibling is three or four and receiving services, that may keep those kids and families in city schools. You see that in Washington, D.C., for sure, and I think you’ll see it in New York over time as well.

Aldeman: Are there any downsides here? Who would oppose this?

Hawley: I think exactly what you first said. Someone needs to actually see the opportunity for what is as a long-term play, because this transition is going to be pretty gradual as the younger grades start losing children. The temptation is going to be to keep the funding for all of the K-12 parts of the system exactly the same, regardless of the smaller number of kids that are being served.

There is also a fear on the part of some folks in early childhood that if the public schools become too involved in the funding of early care and education, then that will crowd out the providers that are serving families now. In particular, I don’t think many people believe that the schools are going to be a great place for infants and toddlers. The funding system right now is set up in a way that in order to serve infants and toddlers, you have to serve preschoolers too, because you can only make money on the preschoolers, and that helps subsidize the infants and toddlers.

If we’re going to take this approach and capitalize on the opportunity, we’re going to need to do it in a way that comes up with strategies for funding the entire birth-to-five system. The investments that schools may be able to make, especially in 3- and 4-year-olds, could be an important part, but it can’t be the whole thing.

Reading List

Melissa Kearney: More on that misleading NYTimes piece about fertility delays

Nat Malkus and Sam Hollon: Participation in structured, voluntary extracurricular activities improves attendance

Sarah Mervosh: “Mississippi is one of the best places in the country for a poor child to get an education.”

Vlad Kogan: How the Columbus, OH school district went broke

Andrew Goldman: Is AI improving learning—or just performance?

Tom Toch: How Democrats Can Win on Education Again

Parents: 58 activities for kids who say “I’m bored”

Check out the table below, courtesy of the Penn-Wharton Budget Model. They tallied up federal outlays and then did their best to allocate that spending based on the age of the recipient. All told, the federal government spent $449 billion on children and young adults under the age of 26. The bulk of that money went toward health care and income security (including child nutrition subsidies and childcare assistance). Given the size of this population, the total spending works out to about $4,300 per person under the age of 26.

But now look at retirees. They’re getting Medicare and Social Security benefits, plus retirement savings subsidies, Medicaid long-term care, and housing assistance. Those add up to $2.7 trillion in annual expenditures, or about $43,700 per person over the age of 65.

In per-person terms, Washington spends about ten times as much on retirees as it does on children and young adults.

In fact, direct spending on retirees now outpaces the spending that can’t be directly connected to someone’s age (what the chart calls “all ages residual”). This latter category includes things like interest on the national debt and the military, as well as transportation, justice, and community investments.

The Wharton authors conclude that, “An age-based view makes clear that older adults receive the largest assignable share of federal outlays… and that spending on younger people is concentrated in a narrower set of means-tested and education-related programs.”

Reading List

Kevin Huffman: “The difference between tutoring that works and tutoring that doesn’t is often whether students actually receive enough of it.”

Fordham: Industry-recognized credentials produce positive earnings gains, but mainly for males who pursue construction, manufacturing, or transportation

Blueprint Labs at MIT finds that, “test score value-added is a better predictor of longer-term school effectiveness than qualitative information reported by either students or experts.”

Opportunity Culture schools are producing exceptionally strong student growth rates

Recent stories about grade inflation and college students struggling with basic math hint that these gaps exist. But when I started digging into the numbers, I was startled by just how large they were.

Consider just a few examples:

• In Florida, the state reports a 90% graduation rate while just 44% of students scored at level 3 or above on the state’s end-of-course exams in Algebra and Geometry. The state warns that students performing at this level “may need additional support for the next grade/course.”

• In Connecticut, 89% of high school students graduate within four years. Yet, just 31% of the state’s 11^th graders met college-ready benchmarks on the SAT math portion. Rhode Island’s gap is even wider: 84 versus 23%.

• In Washington, D.C., 86% of students graduate within four years, but only 15% of students met or exceeded the state’s expectations on their Algebra I, Algebra II, or Geometry exams.

• Tennessee schools can boast of a 92% graduation rate while just 29% of high school students met the state’s expectations in Algebra and Geometry.

These “graduation gaps” are larger in math than they are in reading. They’re also larger in states using some externally validated benchmark exam like the SAT or ACT or the Smarter Balanced assessment. The gaps also tend to be even larger for low-income students, English Learners, and students with disabilities.

The solution is not to “fix the glitch” by simply lowering graduation rates to match proficiency. Nor is it to lower the bar for math proficiency, as a couple states have done (I’m looking at you, Texas and Virginia).

As I write for The Collaborative, “it’s a good thing for more kids to graduate from high school. But it’s disingenuous for states to continue to grant high school diplomas and send students off into the world as if they’re fully prepared for college or a career.”

These gaps should force policymakers to confront an uncomfortable question: what does a high school diploma actually certify?

If states continue awarding diplomas while large shares of students remain far below grade level in math, the credential itself loses meaning. That demands a more honest conversation about readiness, timely intervention, and what students need before they walk across the stage.

In the meantime, you can find your state’s graduation gap data here.

Reading List

Jim Cowen: A Growing Gap Between The Reality And Fiction Of High School Graduation Rates

Minnesota graduation rates at record high, but are students ready?

: Against the Smartphone Theory of Everything

Matt Barnum: Access to special education services improved students’ academic trajectories

Reading Reimagined: A gap in advanced decoding skills underpins the literacy crisis for older readers

Michigan basketball coach Dusty May picked up lessons from educators like and

This is the key education policy trend over the last 10 years, and I’ve been writing about in piece after piece after piece.

Look at the graph below from my piece at The 74 on the changing distribution of 12th grade NAEP math scores. The red line represents the distribution as of 2005, and the green line shows scores as of 2024. The main story is that the entire curve shifted to the left, and a lot more kids are earning lower scores.

I haven’t spent as much time on it, but recently picked up on the fact that high-achieving students have actually been doing ok. In the version of the graph below, I’ve added an arrow pointing to the increase in the number of very high-achievers.

As another example from the same piece, consider the 12th grade reading scores. Again, the entire distribution shifted to the left, in pretty horrifying ways, but there was a small increase of very high-achievers at the top end.

This trend shows up in a number of places and in a number of ways, but it’s important to hold both of these stories in mind. The bottom is falling, and the top is either staying the same or even rising a bit. This matters in terms of how policymakers should think about improvements. It’s important to define the problems correctly.

Subscribe now

So who are the students on either side of these charts? What does their daily life look like?

Which kids are flourishing?

I’m a big fan of ’s work at the Center for Teen Flourishing. In recent case studies, he showed that one way to keep teens off their phones is to give them something productive to do with their time, whether that’s track and field or a job.

This makes sense to me. When I want to eat healthier, I’m not great at restricting myself away from snacks or sweets, but adding fruits and vegetables to my plate seems to work better. Addition is easier than subtraction.

But for some reason this logic breaks down when it comes to teenagers. Can we really ask kids to add more? Aren’t kids overscheduled enough already?

The answers to these questions depend on which kids we’re talking about. It’s true that some kids are busy. They’re taking AP classes… and volunteering… and doing a sport or theater or other afterschool club. If you’re reading my wonky Substack, these could very well be the types of kids you’re most familiar with. My own kids are too young to be in high school, but they’re on this path.

But these are the kids on the righthand side of the graphs above. They’re doing just fine, flourishing even.

But the kids on the lefthand side of the achievement graphs above are not flourishing, academically or otherwise. They aren’t overscheduled. They don’t have as much homework as their peers did in the past. They don’t have jobs. They’re not volunteering. The students on the left side of the achievement distribution aren’t busy in any traditional sense—they’re more likely to be disengaged.

I’m putting this crudely and overgeneralizing here, but there’s a real disconnect in how teenagers are spending their time.

To put some data behind this, consider the nationally representative survey released by 50CAN last month. They found that higher-performing students spend their time at home differently. Kids who earn A’s and B’s are doing more homework, spending more time with friends, and doing more free reading. They’re doing it all.

In contrast, students earning D’s and F’s and more likely to play video games and spend time scrolling on their phones and accessing social media.

The 2024 version of the 50CAN survey also looked at other out of school activities. The participation trends were similar. Kids who were getting mostly A’s and B’s at school were far more likely to be in organized sports, arts and music, summer programs, religious programs, community service, and afterschool programs.

The story here isn’t just falling scores. It’s divergence. Some students are engaged in school and out of it. They’re busy. They’re studying, practicing, competing, and participating. They’re doing more, not less.

But the students at the bottom look very different. They have less structure, fewer commitments, and fewer demands on their time.

That’s the gap that matters.

And it points to a simple but uncomfortable truth. The students who are struggling the most don’t need us to take things off their metaphorical plates. They need parents and policymakers to ask more of them.

Reading List

Todd Collins: More on 8th grade reading results

Willingham and Hirsch: Knowledge as the Key to Reading

Adam Ezring: Governors are not using their platforms to elevate math as a priority

Charlie Barone: “Democrats must come to an understanding with teachers unions and white progressives that unconditional opposition to innovation, accountability, and public school choice is a political and moral dead end.”

: “The same people who say schools are broke, families need help, and programs are getting cut are working overtime to kill or stigmatize a donor-funded tax-credit mechanism because they are terrified someone in their coalition will call it a voucher.”

Thanks to a combination of benefit cuts, plus a surge of new contributions, it has dramatically improved the health of its teacher pension plan.

For example, the system’s unfunded liability, essentially the difference between how much it had promised and how much it had saved toward those promises, shrank from $10.4 billion in 2010 down to $6.1 billion last year. Its funded ratio — a comparison between its assets and its liabilities — has improved from 47% in 2010 all the way up to 80% as of last June.

A big part of that progress has come from a huge surplus of contributions, thanks to special levies on the state’s sales taxes, cigarette taxes, corporate income taxes, individual income taxes and lottery proceeds. This extra state contribution came out to $456 million last year, and now state legislators are wondering if they can cut back just a bit.

Read my full piece for more on how they would use that money.

Subscribe now

But are all those pension costs good for workers?

Here I want to talk about one other aspect, which is what all this money is paying for. Namely, Oklahoma’s teacher pension plan is quite costly, but it also kind of sucks? In fact, Oklahoma offers far worse retirement plans to K-12 teachers than it offers to other types of public servants.

This might surprise some people. But I created a simple table below to help compare the retirement benefits that Oklahoma offers its K-12 teachers versus the retirement plan it offers University of Oklahoma employees and other state workers.

First is the matter of choice. Do employees get a choice over what type of retirement plan they get? In K-12, the answer is typically no. In higher ed, the answer is often yes. And at the University of Oklahoma, new employees get a choice between the same defined benefit retirement plan that K-12 teachers have, or they can choose a defined contribution plan (a 401k).

How long do workers have to stay before qualifying for at least some share of their employer’s contributions? This is called the “vesting” period, and it’s set at seven years for K-12 teachers. That is, a teacher who leaves after five years gets no employer-provided retirement benefits. Both defined contribution plans are shorter, and a new bill would drop it entirely for state employees.

But what about costs and benefits? These are a bit tricker to compare across plan types, so I’ve broken out the costs of the defined benefit plans into two different components. The first is called the plan’s “normal cost” of benefits, or the plan’s own estimates about how much it needs to contribute today in order to pay the promised benefits in the future. Essentially, this is the plan’s estimate for how much its benefits are worth, on average across all of its members.

This is where it becomes obvious that K-12 teachers are getting a raw deal. On average across all K-12 teachers, the value of their employer-provided benefits comes out to just 3.8% of salary—roughly what a basic private-sector 401(k) might offer.

In contrast, the public-sector defined contribution plans are much more generous. State employees get a 6% match, while University of Oklahoma employees get 9%.

But defined benefit pension plans also carry a second cost when they’re not fully funded. These unfunded liabilities are essentially a form of debt, and they reflect the plan’s estimate for how much it should be contributing today in order to reach full funding in the future.

Note that defined contribution plans do not have this type of debt. They make no promises about the future, so the full employer contribution goes toward employee benefits.

Across all of these categories, Oklahoma’s higher education employees have the best retirement plan options. Not only can they pick a plan that’s right for them, they at least have access to a low-cost, high-quality defined contribution plan. If I were choosing, I’d want that one.

Reading List

Lauren Wagner: This Texas Elementary Is Achieving High Reading Scores a Million Words at a Time

Ashley Zanchelli: When did school book fairs stop selling books?

: We remember stories and are likely to think more deeply about information when we encounter it in a story.

: The broken feedback loop keeping parents, students, and colleges in the dark

: “the highest returns [to education] come from bringing new learners into the system, including low-income students, rural populations, and first-generation learners.”

: “For the first time in two decades, the reported average Massachusetts teacher salary has dropped below per capita income in Massachusetts.”

Tim Harford: Many people are sending texts that should have been emails

That’s true…ish.

But what these pieces leave out is that no state is really covering themselves in glory in 8th grade reading scores. Nationally, we’re down 10 points since 2013. We even declined two points from 2022 to 2024, when math scores rebounded somewhat.

Not only that, but the declines have been worse for the lowest-performing students. From 2013 to 2024, the scores for the top ten percent of students fell 3 points. They fell 8 points for the median student. But they fell a whopping 19 points (!) for the bottom ten percent.

Share

In other words: 8th grade reading scores are a national problem, not a Southern one.

Is anyone bucking the national trends? Kind of? The Department of Defense Education Agency (DoDEA) again stands out here, with their scores rising by 5 points. The District of Columbia is also up 4 points—they’re a quiet success story that deserves more attention. And then there comes Mississippi, eking out a gain of 0.4 points over this time period.

These are the only places with gains in 8th grade reading over the last decade. For what it’s worth, Louisiana comes next, with a decline of 0.9 points. Compared to the national decline of nearly 10 points, these “Southern Surge” states are among the best performers in the country—even in 8th grade reading.

What about our leading states as of 2013? Can we just copy what they’re doing? No, please don’t do that. Massachusetts, for example, tied for the lead in 2013 with the DoDEA schools. But while DoDEA rose, Massachusetts scores fell 9 points. (Yes, DoDEA schools are now leading the pack by a wide margin.) As you can see in the graph below, Massachusetts’ entire score distribution shifted in the wrong direction. Whatever they did in the past is now history.

Among other leading states as of 2013, New Jersey scores fell 10 points by 2024, Connecticut was down 11, Vermont 17, and New Hampshire 10.

Which states brought down the national average the most? The biggest declines came from Vermont, Delaware, Maryland, Alaska, Maine, Oregon, Nebraska, Oklahoma, Pennsylvania, and Florida.

What about states with the biggest gap growth? That is, which states had their highest-performing student pull further away from their lowest-performers? That dishonor goes to Maryland, which saw its 90th percentile pull nearly 28 points further ahead of its lowest-performing 10th percentile. Other notable gap growth states include New Jersey (26 points), Texas (24 points), New York (23 points), Delaware (22), Virginia (22), Wyoming (21), California (21), Rhode Island (20), and Indiana (20). These states all declined on average, but they let their lowest-performing students fall even more.

Should we care more about 4th or 8th grade reading scores?

Both 4th and 8th grade results matter, but they tell us different things. On one hand, the 8th grade scores reflect more about a student’s total body of knowledge up to that point. They are also more indicative of what a student might be ready to do for their college and careers.

But if you care about whether school systems can systematically improve outcomes—especially for struggling readers—the 4th grade results may be the more actionable signal. That’s because early advantages (and disadvantages) can compound over time. This is especially true in reading, where scores are so heavily tied to things like background knowledge and language skills.

For schools in particular, they have the most control over whether students are mastering basic skills. This is why it’s so much easier for interventions to move the needle in math versus reading and earlier versus later ones.

TLDR: Don’t dismiss the Southern Surge states just because their 8th grade reading scores haven’t moved as much. There’s still a lot to learn from how they’ve raised the floor on reading performance.

Reading List

Linda Jacobson: Should states receiving federal literacy grants have to follow the science of reading?

Nora Gordon: Snow days, moral hazards, and Montgomery County’s request for an exception

: Worry less, do more

Bill Gurley: How to Succeed and Thrive in a Career You Love

As I write in my review, the book is NOT about phonics:

When people hear the “science of reading,” they might (mistakenly) equate it with phonics, but the authors spend little time on those core foundational skills. In fact, they take systematic phonics instruction in grades K-3 as the assumed starting point for literacy instruction and note that their book is about “the science of reading beyond phonics (emphasis theirs).”

Instead, it’s a book intended to help teachers implement all the other important—but often neglected—aspects of the science of reading. That starts with fluency but also includes vocabulary, content knowledge, and writing.

Most importantly, the authors want kids to be reading and engaging with high-quality texts, as early as and as often as possible. How to do that? Here’s more from my review:

It’s not as simple as putting good books in front of students, because if [students] can’t read the words on the page quickly and easily, they will struggle to comprehend and make meaning out of the text. The authors cite an Italian study that found, “reading fluency predicted all school marks in all literacy-based subjects, with reading rapidity being the most important predictor.”

In response, the authors suggest that, “The best way by far to improve fluency is to provide students opportunities to hear, read and reread text aloud.”

The book is full of practical tips and embedded QR codes that take readers to videos of teachers putting those strategies into practice.

Read my full review here, or buy the book here.

Reading List

Aldeman: When high school grads can’t do math

: What If Our Approach to Intervention is Just Making Things Worse?

and Ben Austin: America faces an eduction depression. Why are Democrats silent?

Elizabeth Statmore: San Francisco’s Algebra plan is the opposite of equity

Todd Collins: “So which states are the leaders in fourth grade NAEP math improvement for economically disadvantaged students? Here’s a surprise: It’s also Mississippi and Louisiana.”

InsideHigherEd: “the net price of four-year tuition continues to drop or remain steady for all but the highest-income students”

: Fiscal pressures are heating up in some of America’s largest cities

Within public education specifically, job openings were down 24% last year compared to 2022. Hires were down 18%.

2022 was also the peak year for employee separations in public education. They’ve fallen each year since then, and by 2025 separation rates in public K–12 and higher education were 11 percent lower than their 2022 peak.

That cooling shows up in the latest employment numbers as well. The newest data, released last week, include final totals for the 2025 calendar year. They’re… not great. Here are the year-over-year changes within education:

• Public K-12: Up 0.7% (+58,000 jobs)

• Public higher ed: Down 1.6% (-42,000 jobs)

• Private childcare services: Up 0.4% (+4,000 jobs)

Other state and local government employers also didn’t add many new jobs. State governments added just 3,000 jobs outside of education (a gain of 0.1%). Local governments beyond K-12 looked the strongest, adding 88,000 jobs (a gain of 1.3%).

Subscribe now

To put the education numbers in historical perspective, the job growth in public K-12 looked about like it did in the late-2010s. Positive but modest. For higher ed it was one of the worst years, outside of COVID, since 1977. The tepid growth in childcare employment looked similar to the years in and around the Great Recession period of 2008-2013.

These are not flattering comparisons. I’ll keep watching the numbers closely, especially as we head into education’s typical summer hiring rollercoaster.

Reading List

Jessica Baghian: What “good” looks like in practice

WSJ: AI seems to be, “increasing the speed, density and complexity of work rather than reducing it”

: Ten things in education that look right but aren’t

Josh McGee and Gema Zamarro: Raising base salaries increases teacher retention rates in Arkansas

: Cool charts on earnings distributions by educational attainment

The requirements for homeschooling in Virginia are minimal. Parents must either be licensed teachers, hold at least a high school diploma, or provide “evidence that the parent can provide an adequate education.” We chose the last option because it was the easiest to demonstrate. After looking for samples online, we discovered we only needed to provide a declaration of what subjects we planned to cover. We submitted a simple list saying our kids would receive grade-level instruction in:

- Math

- Reading

- Social Studies

- Science

- Physical Education

- Foreign Language

- Art and Music

That’s it.

No one asked which curriculum we were planning to follow or press us on the specifics of how we were going to do this. At the end of the year, we had to prove that our kids scored at or above the “fourth stanine” on a nationally normed standardized test. That essentially meant that our kids had to score at or above the 24th percentile nationally.

Of course, no one gave us any money to do this. The state had invested in public schools. By opting out, we were choosing to forgo that investment.

That arrangement made sense. We were on our own, and the state wasn’t funding our decision. But the policy question looks different when the state is providing public funding.

Subscribe now

Where should states set the bar?

Most states have minimal standards for public schools. Thanks in part to federal requirements, every state administers reading and math tests to all students in public schools grades 3-8. Publishing those results provides at least a minimal amount of transparency. Anyone can go to state report cards or GreatSchools to see how any public schools is faring.

And, while it varies by state, there is usually some sort of accountability for persistent poor performance. If students in a particular school or district do particularly poorly, year after year, the state will step in and provide extra scrutiny or potentially even take over district operations. State takeovers happen in both red and blue states.

But, as I wrote earlier this year, many states have forgotten about these minimal safeguards when it comes to private school choice programs. Most states with vouchers or ESA programs don’t require students to take any tests, let alone the same ones required by public school students. In other words, they have at least some minimal standards for public schools but nothing for private schools that take public money.

Iowa is an exception to that rule. It requires students to take the same state tests in exchange for $7,988 in an education savings account. In a piece last month for Informed Choice, compiled the results. In one clear chart, he showed that Iowa’s ESA students were outperforming other public school students.

Now, higher achievement levels do not mean that the ESA program caused any performance differences. Given that parents had to apply to the ESA program, it’s likely that its beneficiaries were different in some ways than non-applicants.

But Iowa goes one step further. It also calculates Student Growth Percentiles for ESA program recipients. As Iowa explains it, “a student growth percentile (SGP) describes a student's growth compared to other students with similar prior test scores (their academic peers).” Across all students, the median SGP is 50.

The graph below shows the SGP math scores for Iowa’s ESA participants. Across the entire group, the median for ESA program participants came in at 49. According to this metric, Iowa’s ESA program participants overall are learning just a little bit less than other similar students across Iowa.

But particular groups of students made slightly more progress with the ESA. English Learners, students with disabilities, and low-income students who received an ESA all made slightly more progress in math than their peer groups across the state.

ESA participants showed similar growth rates in reading.

Iowa’s approach isn’t perfect. It still doesn’t provide school-level results. If I were a parent in Iowa, I would want to know which schools and programs were getting the strongest results, but the state doesn’t provide that level of transparency.

However, Iowa taxpayers can at least be confident that their ESA program participants are learning something and making roughly comparable progress in reading and math. Most states with private school choice programs can’t make that claim, but Iowa can.

Reading List

Linda Jacobson: National, State Data Point to Slow Pace of Pandemic Recovery

NWEA: Math and reading scores lag even among 1st and 2nd graders

Patrick Graff: The positive effect of competition on public schools is quite large

Per my piece last week, EdWeek finds that Arkansas teachers have the highest morale

Zach Groshell: The only question that matters is whether students learn

: The inside story of fixing the FAFSA

…while New York pays 70 percent higher teacher salaries than Louisiana on average, its entry-level salaries are only 7 percent higher. That doesn’t come close to compensating for the higher cost of living. If you ask where “teacher” counts as a decent-paying job for someone just starting out, Mississippi and Louisiana look good and New York looks terrible, despite there being much higher average salaries in New York.

That observation raises a lot of interesting questions. But today I want to explore what happens when a state or district dramatically increases starting teacher salaries.

Subscribe now

Let’s start with the state-level data. Consider the graph below. It shows the average teacher starting salary by state (left to right) versus the top salary (up and down). As the graph shows, states with higher starting salaries tend to also offer their teachers higher maximum salaries. Washington, California, and the District of Columbia pay the highest salaries regardless of a teacher’s experience level.

None of these are adjusted for cost of living. But now I want you to focus just on the starting salary aspect. I’ve added a red line to show something interesting. As of last year’s data, Arkansas schools offered nearly identical starting salaries ($50,031) as New York ($50,077). That’s a striking comparison given the huge differences in cost of living between the two states.

In fact, Arkansas’ starting teacher salary is now higher than it is in states like Illinois, Rhode Island, Delaware, Virginia, and Connecticut. All of these states have higher costs of living than Arkansas does.

Arkansas is an extreme case. Among all the states with data, Arkansas ranks 14th in terms of starting salaries but last in terms of maximum salary.

How did Arkansas get to this point? And why?

The short answer is a piece of legislation passed in 2023 called the Arkansas LEARNS Act. Among other things, it lifted the floor of teacher salaries in the state up from $36,000 to $50,000 and guaranteed all teachers a raise of at least $2,000 if they were above the minimum. That was a big change, and the state paid for it all, ultimately costing $183 million. Because many teachers in poorer, rural districts had been closer to the old minimum, much of the money flowed into those areas.

So what happened? A team of researchers at the University of Arkansas led by Gema Zamarro has been tracking the results. According to their most recent analysis, presented at the CALDER conference last week, they found that, “the reform largely eliminated the previously negative and significant association between starting teacher salaries and district poverty and rurality.” Moreover, the salary increases boosted teacher retention and reduced retirements and transfers to other districts. The gains were particularly large for teachers who earned salary increases of $6,000 or more (aka, those in poor, rural areas). The Arkansas researchers are still exploring how the LEARNS Act affected teacher recruitment, but the results so far are positive.

The Arkansas results are consistent with a broader emerging literature on starting salaries. In a recent working paper, Prasiddha Shakya looked at what happened after a union-led effort pushed school districts to raise base salaries in New Jersey. The campaign led to higher starting salaries, as intended. But, without any new money attached, the districts had to find a way to pay for them. And they did so by increasing class sizes slightly and generally reallocating money away from other staff.

Shakya found that, despite these trade-offs, boosting early-career salaries led to some achievement gains, particularly in math, and boosted graduation rates. The districts did not appear to be poaching workers from each other. Instead, the gains seemed to have come from bringing in new talent or increasing effort in some way.

Put all this together and a clear pattern emerges. Raising starting salaries appears to change the teacher labor market in meaningful ways. Places that concentrate more resources on early-career teachers may see real benefits for recruitment, retention, and ultimately student outcomes.

Reading List

: Cut kids’ screen time by getting them in sports or finding them a job

: Massachusetts education is #1… For some kids.

Yours truly: Illinois’ “Comprehensive Numeracy Plan” Has Very Little Math In It

Then, they were asked a series of questions about the passage. One of those questions asked them about the use of the word “conform” in the context of the passage.

The correct answer was B, “he does not act and think like other mice.” Just over half of the test-takers got it correct (54%), and the rest (46%) got it wrong. The wrong answers were fairly evenly distributed across A, C, and D.

Subscribe now

But why? Was it because they didn’t understand the question, didn’t know the meaning of the word “conform,” or got misled in some other way? The truth is we don’t know.

When students miss a comprehension question, the problem often isn’t about higher-order thinking skills. It’s that they can’t read the words fluently and accurately enough to understand the passage in the first place.

Why do kids struggle with reading comprehension?

The question about “why” kids couldn’t get a question correct isn’t just an academic exercise. It should animate what schools do to improve their students’ ability to read and understand text.

One strategy might be to say, ok, students are bad at reading comprehension, so we should just drill them with lots of similar exercises. Under this way of thinking, students would read a lot of short excerpts and practice answering questions about them.

Schools today are doing a LOT of this. My kid came home last week and said his homework was to read two passages. One was about the Berlin Wall and the other about Earth Day. These are both fine topics, but they were not integrated into history or science lessons; they were part of an English Language Arts curriculum. But why these two? What did they have in common? Nothing. Nothing at all.

To be clear, there is some research that instruction on comprehension strategies can help students on standardized tests, but the effects don’t translate into gains from more and more practice. Once kids have been reminded to “find the main idea,” they don’t need to practice it all year long.

So what does matter? A reader recently pointed me to the 2018 NAEP Oral Reading Fluency Study. It explores these questions about why students are struggling, and it turns out that bad readers are bad at several important skills. It notes that:

When there is weakness in oral reading fluency (evidenced by errors in reading, hesitation, or inappropriate pauses that signal the student is challenged by the text), it is often explainable in terms of the difficulties a child is experiencing in word reading, phonological decoding, vocabulary, or grammatical structures of the English language.

Let me repeat that first part back to you: Bad readers struggle because they literally cannot read the words.

Many kids simply don’t know how to decode letters on a page into English words. This is a big part of my work at ReadNotGuess, and I’m optimistic that recent policy changes around curriculum, teacher preparation, and early screening will help boost students’ early decoding skills.

In the NAEP study, they tested this by looking at whether students could decode what are called “pseudowords” or “nonsense” words like “bep” or “glork.” These aren’t actual words but they should be decodable anyway if kids know how the English language works. As you can see in the graph below, the kids who score the lowest on NAEP reading comprehension questions are the worst at reading these nonsense words. Better readers are better decoders.

This is one reason I’m such a big fan of England’s phonics check. They ask all 6-year-olds to read 40 words out loud to their teachers. Half of the words are real words and the other half are nonsense words. Students earn a passing score if they get at least 80% correct, and the percentage of England’s 6-year-olds who can pass the phonics check has risen from 58 to 80% in recent years. It’s a simple, effective check on a core reading skill.

Letters build into words, and words build into sentences. And bad readers happen to be slow readers. The lowest-scoring students read about half as slow as the highest-performing ones.

But it’s not just about speed. Speed without accuracy is worthless. Bad readers also struggle to read with expression. They don’t know to pause when they come to a comma or a period, and they struggle with appropriate intonation, rhythm, and emphasis.

Bad readers also tend to guess, skip words, or make other mistakes. The next graph shows the words correct per minute (WCPM) by performance level. Again, bad readers tend to be slow and to make a lot more mistakes.

The lowest-performing students weren’t just slightly slower—they read roughly half as fast and with far more errors. The gap between the very lowest readers and the next tier up was larger than the gap between Proficient and Advanced students. The biggest opportunity for growth isn’t at the top—it’s among students who still struggle to decode and read accurately.

So why do kids struggle with reading comprehension?

Once you start to decompose reading comprehension into its component parts, the picture becomes much clearer. Students struggle not because comprehension is some mysterious higher-order skill, but because they lack the foundational abilities that make comprehension possible in the first place.

That starts with phonics and decoding. The point of phonics isn’t phonics for its own sake; it’s to allow students to reliably translate print into spoken language. When students cannot decode words automatically and accurately, their cognitive energy is consumed just getting through the text. Research has also shown that weak decoding skills can also limit middle and high school students’ access to more complex academic material.

Subscribe now

Decoding alone, however, is not enough. Students also need the vocabulary knowledge to attach meaning to words. A sentence can be perfectly decodable and still incomprehensible if too many of the words are unfamiliar. Vocabulary, like decoding, accumulates gradually through systematic instruction and sustained exposure to coherent, content-rich text.

Fluency sits at the intersection of these skills. In the NAEP oral reading fluency study, all of these measures were captured by listening to students read aloud. Students who read slowly, inaccurately, or without appropriate phrasing were far less likely to answer comprehension question correctly. Yet most comprehension assessments—and much of the practice students receive—are entirely silent, masking the underlying source of difficulty.

This helps explain why drilling students on comprehension strategies or isolated passages produces limited returns. Strategy practice cannot compensate for weak decoding, limited vocabulary, or lack of fluency.

When a fourth grader misses a question about the word conform, the problem may not be that the student failed to think deeply about the passage. It may be that the student never fully understood what was being asked of them. If schools want more students to answer those types of questions correctly, the solution begins earlier and runs deeper: ensuring that all children can read words accurately, fluently, and with understanding.

Reading List

: Student effort is way down. It’s not their fault - it’s ours.

Philadelphia: Student attendance craters on half days

More research on the importance of high school quality

More research on the racial effects of teacher licensure tests in Texas

Dan himself has been doing a series on New England states, or I worked with The 74 to show how the 12th grade reading and math scores have evolved over time. The result: Some cool charts showing ugly results.

Subscribe now

So today I’m going to return to everyone’s favorite: Mississippi’s 4th grade reading scores. As I noted when the scores first came out, Mississippi is the only state in the country that saw scores rise for its lowest performing students. In fact, Mississippi was able to increase scores at the 10th, 50th, and 90th percentiles.

But those are just words, Chad, show me a picture!

Ok, here’s what it looks like visually. The blue curve represents Mississippi’s 4th grade reading scores in 2013, and the red curve represents those scores as of 2024. As you can see, the average rose by 10 points, which is great, but the entire curve shifted to the right. There were fewer kids who were very low scoring and a lot more kids on the higher end. It was a success all around.

Yay Mississippi!

So how did they do it? For more on the “how” question, read my original article in The 74, or see more from Karen Vaites, Kelsey Piper, or Rachel Canter.

Reading List

: “Fragile systems rely on great teachers. Durable systems assume ordinary teachers and build routines strong enough to support them.”

ERS: How Alliance College-Ready Public Schools reimagined teacher planning time and dramatically boosted student outcomes

New research: NAEP math scores are linked to income, teen motherhood, incarceration, and arrest rates

Yours truly: The Maryland School District ‘Doing the Improbable’ in Teaching Kids to Read

Yours truly: 5 Things the Government Can Do to Help Make Reading Cool Again

: “When the federal government offers states money to help working families, ideological objections are illegitimate. Take the money and help your constituents.”

Brandon Brown: Indianapolis is putting charters and traditional public schools on the same footing

Sarah Mervosh: The new Texas private school choice program has received more than 16,000 applications for preschool

Annie E Casey Foundation: “There is every rea­son to think that when nation­al data for 2025 becomes avail­able, it will show that youth crime con­tin­ued to decline.”

Despite a poverty rate that falls somewhere in the middle of the pack, Worcester County leads the state in 3rd grade reading proficiency. It also leads the state in reading scores for low-income students and for students with disabilities, and for non-low-income students and students without disabilities.

In other words, it’s doing a great job serving all of its students.

Subscribe now

So what is Worcester County doing differently? Their secret sauce seems to be a combination of determined leadership, a focus on high-quality core instruction for each and every kid, and strong partnerships with parents.

As I write in the piece:

…they are religious about giving kids lots of time to practice.…During the school day, kids are typically reading physical books that help them build phonics skills or engage in sustained independent reading. Students are expected to complete two 15-minute blocks of reading at school — and then read for 30 additional minutes per day at home. This regimen may vary based on the child’s age and skill level, but kids have to log what they read and then have their teacher or parent sign off.

Families, in fact, are the third key component of Worcester’s reading plan. At the beginning of the school year, they’re asked to sign a “home coach contract” saying that they will check and monitor their child’s reading. Throughout the year, kids are expected to read for half an hour at home five days a week. Over the course of a 180-day school year, that could add up to 900 extra minutes of practice.

Check out the full piece for more on what we can learn from Worcester County.

Reading List

Carrie Hahnel and Alex Spurrier: An Abundance Agenda for K-12 Education

XQ report: 10 policy actions states can take to bolster high school innovation

Nicholas Bagley and Robert Gordon: Blue-state governors and mayors need to get serious about delivering excellent public services

Stanley Litow: P-TECH gives low-income students a path to a diploma, an associate degree and good-paying jobs.

Kate Rix: High-impact tutoring is producing large reading gains in San Francisco

Why? Because there’s a lot of money potentially at play, and there are very few parameters on what might qualify.

Subscribe now

Let me back up. Last year’s federal tax bill created a new tax credit scholarship program for this current tax year, and it’s EXTREMELY generous. If you need a primer on the program, see this piece from —>

Center for Educational Progress

The $1,700 Question

Read the full analysis below, or start with the key takeaways…

Read more

2 months ago · 8 likes · Joshua Dwyer

Anyone who donates to an SGO will receive a dollar-for-dollar credit up to $1,700.1 That is, if you donate $1,700 you will reduce your taxes by $1,700. Even if you don’t itemize your taxes, you can still reduce your tax bill dollar-for-dollar.

That’s a big deal, and that’s far more generous than the way the tax code rewards other types of charitable giving.

Ok, but isn’t this just for private schools? No, no it is not. The money can be used for any “qualified elementary or secondary education expense.” What is a “a qualified elementary or secondary education expense?” Here’s the relevant code language:

The term “qualified elementary and secondary education expenses” means—

(i) expenses for tuition, fees, academic tutoring, special needs services in the case of a special needs beneficiary, books, supplies, and other equipment which are incurred in connection with the enrollment or attendance of the designated beneficiary of the trust as an elementary or secondary school student at a public, private, or religious school,

(ii) expenses for room and board, uniforms, transportation, and supplementary items and services (including extended day programs) which are required or provided by a public, private, or religious school in connection with such enrollment or attendance, and

(iii) expenses for the purchase of any computer technology or equipment or Internet access and related services, if such technology, equipment, or services are to be used by the beneficiary and the beneficiary’s family during any of the years the beneficiary is in school.

Clause (iii) shall not include expenses for computer software designed for sports, games, or hobbies unless the software is predominantly educational in nature.

So far most of the public activity has been on the school voucher / private school choice side, but read through this list again. It specifically includes public schools. It specifically calls out tutoring. The phrase “supplementary items and services (including extended day programs)” opens the door to a wide variety of expenditures. It’s a very broad list.

So, we have a very generous incentive on the donation side, plus a very flexible rule on the spending side. That’s a recipe for a bonanza.

Who can benefit? And who can run an SGO? The Treasury Department is still drafting regulations to implement the new law, but the law itself sets few parameters on how an SGO must operate:

- It must be a nonprofit;

- It must allocate at least 90% of its revenues toward scholarships;2

- It must verify the income of the scholarship recipients;3

- It cannot allow donors to steer funds directly to particular students; and

- It must award scholarships to more than 10 students who do not all attend the same school.

This list is not very onerous. If you run a school system, it’s possible your community foundation could operate an SGO and give out afterschool or summer program scholarships. If you already run a summer or tutoring program, your organization could probably benefit. Any community group that provides before- or after-school programming, such as a church, a YMCA, etc. could potentially benefit. And any charitable organization that already raises money for these types of activities (think American Way) could launch its own SGO and offer its donors a more tax-efficient way to give.

I’m probably not being creative enough here.

Now, an SGO can only award scholarships in states that officially opt into the program. There are a lot of political questions there. Some Democrats (rightly) fear that this program will supercharge private giving. As currently constructed, it has very few accountability safeguards and may or may not let states set their own oversight rules. We’ll have to wait and see exactly where the Treasury regulations land.

But if you’re a state policymaker, you should be prepared for a LOT of people who could potentially benefit. Any accountant worth their salt will be recommending their clients take advantage of this credit. I suspect TurboTax and H&R Block will be looking into this as well.

And now that this program has been created, anyone who’s even peripherally involved in serving school-age kids should be looking into how they can make it work for them.

Gerrymandering Lobster Claws

I’m not a fan of gerrymandering. A few years ago, I supported a constitutional amendment in my home state of Virginia to move to a nonpartisan districting process.

If you’ve been following the national political news, you may have seen that Democrats in the state legislature are trying to create a new congressional map. In a state that is currently 6-5 in favor of Democrats, the new map could potentially lead to a 10-1 map in favor of the Dems.

They get there by squeezing every Democratic vote in northern Virginia and creating some pretty crazy-looking maps. Here’s what it would look like for my district. You can see my current district, shaded in purple at the far right. The new map would carve out an entirely different footprint (in black).

That picture doesn’t do the full map justice. Consider the proposed “lobster” district. Again, the purple is the current district and the black is the proposed new version:

Needless to say, if the Democrats’ redistricting plan makes it past the legal challenges and goes to a public vote later this spring, I will be voting no.

Reading List

Jude Schawlbach: Open enrollment is increasingly popular

Jessica Baghian: “Every state is a local control state. And still, every state has levers they can pull to improve outcomes for kids.”

Helen Baxendale: What American Education Reformers Can Learn from England

: “[Higher education] debt cannot simultaneously be a life-ruining burden and an essential tool that students must have unlimited access to.”

And yet, kids are doing less reading today than they did in the past. Long-term NAEP data show a steady decline in how often kids read for fun. Since the mid-1980s, we’ve been asking kids “How often do you read for fun on your own time?” For 13-year-olds, these numbers fell steadily throughout most of the last few decades, but they took a noticeable dive from 2012 to 2020 and fell again in 2023.

Just 14% of 13-year-olds now say they read “almost every day.”

Subscribe now

The numbers for 9-year-olds have always been better, and they remained fairly consistent for a long stretch of time. In 2012, for example, the exact same percentage said they read almost every day (53%) as their peers did in 1984. But those numbers have fallen sharply since then and were down to just 39% in 2022.

What about assigned reading? Those numbers are down too, although not quite as noticeably. Teachers are assigning fewer pages of reading at school, and they’re assigning a lot less homework as well.

A lot of this has to do with broader societal trends. Some, though not all of it, can be blamed on screens and technology.

Schools and parents shouldn’t accept these trends as inevitable. I interviewed a school district recently that expected all of its elementary school students to read for 30 minutes a day at home. Kids had to keep a daily reading log and parents were engaged as partners and asked to sign off on their child’s work. That kind of sustained daily practice is almost certainly part of why they perform so well. I’ll have more to say about that outlier district soon…

In the meantime, if reading volume is falling, we shouldn’t be surprised when achievement follows. The simplest lever we have is also the most obvious: get kids reading more.

Reading List

New research: Virginia had particularly large COVID math declines

LAUSD Board to Vote on 3,200 Layoff Notices Amid Budget Crisis

Bronx charter school set to open 12 hours a day, 7 days a week

Biting, kicking, wandering: Teachers see rise in misbehavior even among the littlest kids

San Francisco teachers can’t afford rent, but at least they get free healthcare

: Public education is a goal, not a fixed set of institutions

Sunday, January 25: Snowstorm drops 5-8 inches of snow, followed by another 1-2 inches of freezing rain.

Monday, January 26: Schools are closed.

Tuesday, January 27: Schools are closed.

Wednesday, January 28: Schools are closed.

Thursday, January 29: Schools are closed for pre-scheduled teacher workday.

Friday, January 30: Schools are closed for pre-scheduled teacher workday.

Monday, February 2: Schools are closed.

Tuesday: February 3: Two-hour delay.

Wednesday, February 4: Two-hour delay.

Thursday, February 5: Two-hour delay.

Friday, February 6: Two-hour delay.

…

Friday, February 13: All parents at my kid’s school get an email reminding parents of the importance of consistent attendance. Here’s the message:

We ask for your continued support in reinforcing the importance of daily attendance and arriving to school and each class on time. In middle school, every class period matters. Missing instruction or arriving late can make it difficult for students to keep up with assignments, projects, and assessments.

Encouraging consistent routines at home such as getting adequate rest, organizing materials, and being prepared for the school day, can make a significant difference in your child’s success.

Look, I totally understand what the school is trying to do here. Chronic absenteeism rates soared during the pandemic and have remained stubbornly high. And the school leaders had no say over the snow day decisions.

But geez, the words from the school and the actions from the district are telling two very different stories. Is school essential and important? Well, that depends on the day and who’s asking!

Reading List

Kara Arundel: DC’s citywide effort to boost math achievement

New research: “increased funding for counselors led to large increases in high school graduation and measures of school climate”

More new research: “[Communities in Schools] boosts test scores for struggling students, and that these improvements persist, ultimately increasing high-school completion and adult earnings”

Ten Miles Square

The Stunning Collapse of KIPP DC

Before the pandemic, KIPP DC was a middle-of-the-pack charter network. Its 11 campuses tested about 2,900 students, making it by far the largest charter operator in the District. Its proficiency rates — 37 percent in ELA, 40 percent in math — were roughly in line with the charter sector average and not far from DCPS. Some individual campuses, like Promi…

Read more

2 months ago · 13 likes · Matthew Yglesias