But new studies bolster the evidence that done well, virtual models can be just as effective at moving students forward as face-to-face instruction.

In Massachusetts, first graders who spent 15 minutes a day online with a tutor from stayed on track a year later without additional tutoring, according to exclusively with 蜜桃影视. Students gained, on average, at least five additional months of learning over their expected growth. 

Another virtual program, , produced positive results for the lowest-performing students in the Kansas City, Missouri, schools. Students who received one-on-one tutoring from certified teachers made greater progress than those who didn鈥檛 receive the extra help, .听

鈥淰irtual models are getting stronger,鈥� said Amanda Neitzel, an associate professor at Johns Hopkins University and author of the Ignite Reading study. 鈥淚f you go back just a few years, we had no examples of evidence-proven models and now we are getting them.鈥�

In addition to following Ignite Reading for two years, she recently published a study showing that elementary school students in Texas and Louisiana who received virtual tutoring from , outperformed their peers and gained nearly three additional months of learning.

Results like those have broadened the conversation about how to bring students who are missing critical reading skills up to speed. 

鈥淭utoring can work in many ways and in different settings,鈥� Kevin Huffman, CEO of Accelerate, said earlier this month at the nonprofit鈥檚 annual conference. 

When the organization began funding tutoring research four years ago, there were doubts, he said, about whether virtual programs could compete with in-person models. There鈥檚 more confidence in online versions now, but as with tutoring in general, progress depends on whether providers feature the components of a high-dosage program 鈥� meaning they were offered for roughly 90 minutes a week, during the school day with a trained tutor. Ensuring kids get all the tutoring hours a program is designed to deliver is also key.

鈥淲e obsess over student attendance,鈥� said Jessica Reid Sliwerski, Ignite Reading’s founder. Now in 24 states, the program focuses on building phonics skills and reading fluency.

In the Johns Hopkins Ignite Reading study, which focused on 13 Massachusetts school districts, 85% of students who mastered foundational reading skills 鈥渄uring the crucial first grade window鈥� were still keeping up at the end of second grade, Neitzel wrote. But if students didn鈥檛 meet expectations on time, they couldn鈥檛 catch up. Some were just too far behind.

鈥淢any kids start our program still not knowing basic kindergarten skills, like letter names and sounds,鈥� Sliwerski said. That means tutors have two years of content to get through.

To Sliwerski, the findings demonstrate that third grade, when many states decide whether students are strong enough readers to advance, is too late to intervene. If kids struggle to decode unfamiliar words, they won鈥檛 be able to comprehend more complex reading assignments. 

鈥淲e are so caught up in 鈥榬eading by grade three鈥� that we aren鈥檛 honoring that kids are actually supposed to have fully cracked the code and be able to fluently read grade-level text at the end of first grade,鈥� she said. 鈥淲e act like kids have all the time in the world, when they don鈥檛.鈥� 

The 5,700-student Chelsea Public Schools was among the Massachusetts districts using Ignite Reading as part of a project funded by One8, a nonprofit that helped schools get high-dosage tutoring off the ground. The state the program.  

At first, 鈥渙ur teachers were a little skeptical,鈥� said Superintendent Almi Abeyta, a former kindergarten and first grade teacher. 鈥淭hey were like, 鈥榃e just got off of remote learning. Why are we going to put kids on a computer again?鈥� 鈥� 

Then they saw the data. Students made similar gains on DIBELS, a widely used early literacy assessment, whether they were Black, Hispanic, English learners or had a disability, the study found.

鈥楢 great opportunity鈥�

Results like those are why the Fallbrook Union Elementary School District, near San Diego, California, is now spreading the program to all of its elementary schools as part of its First Grade Promise initiative. 

In a pilot, Fallbrook STEM Academy, which serves a high-poverty population, enrolled 20 second graders in the program. Many of the students speak Spanish at home, didn鈥檛 attend preschool and lack access to books, flash cards and other early reading materials, said Principal Ana Arias. She called each parent to ask that they get their children to school a little early so they could meet with a tutor.

鈥淚 phrased it as an opportunity 鈥� a great opportunity 鈥� but I needed their commitment,鈥� Arias said. 鈥淲e have so many kids in the classroom and there’s so much need. It’s very rare to have a teacher meet one-on-one with a student every single day.鈥� 

At the beginning of this school year, the 20 students were reading at a kindergarten level. By November, 19 had advanced to a first grade level, and she鈥檚 hoping they鈥檒l be on par with their peers by the end of the school year. 

鈥楾ranscend time zones鈥� 

The latest findings build on those that Harvard University and City University of New York researchers published last year. Whether tutoring is remote or in-person, , matters less than whether the tutor is well qualified and students attend sessions regularly.

Virtual models even have some advantages over in-person programs, experts say. Schools have to pay an in-person tutor whether or not the student is present. But virtual programs 鈥渢ranscend time zones,鈥� Sliwerski said, and can redeploy a tutor to meet with another student.  

If the tutor is absent, 鈥渨e have a substitute ready to go,鈥� she said. 鈥淭he technology underpinning the program ensures the child receives the exact lesson they were supposed to get.鈥�

In Kansas City, consistency was key to the strong results. Students in first through fourth grade across 14 schools met with their tutors for 30-minute sessions at least three times a week for 20 weeks during the 2024-25 school year. The more sessions completed, the stronger the growth. Some students gained more than two months of additional learning and were less likely to be placed in special education. 

On average, the students who participated in the Hoot program and those in the comparison group began the school year two grade levels behind. While many are still struggling readers, their progress was significant, said Carly Robinson, a senior researcher at Stanford University and a co-author of the study.

鈥淭his wasn’t a boutique pilot,鈥� she said. 鈥淚t’s tutoring operating inside a district system that is messy, and it still proved to be effective.鈥�

The district had to contend with technical glitches and unexpected snow days that forced students to miss some sessions.

Not all virtual programs have been able to overcome disruptions. 

In a large suburban district in Texas, some students meeting with virtual tutors during the 2021-22 school year did worse in reading than their peers who didn鈥檛 receive the intervention. Scheduling conflicts, like school assemblies, and tutor turnover, contributed to the disappointing results.

鈥楢 higher bar鈥�

Those challenges grow even more complex in the middle grades with electives and block schedules where students don鈥檛 have the same classes every day. But Rahul Kalita, co-founder of Tutored by Teachers, said maintaining relationships between tutors and students is essential. 

He hopes to contribute to the research base on virtual tutoring by participating in a randomized controlled study, funded by Accelerate and focused on math in two large Indianapolis middle schools. 

鈥淚t felt like the right opportunity to test our model under a higher bar of rigor,鈥� he said.

On top of virtual programs refining their practices, districts, he said, 鈥渉ave also become more sophisticated buyers of tutoring.鈥� Multiple districts across the country pay providers higher rates if students make measurable progress or pass state tests. 

In addition, there鈥檚 growing agreement that literacy tutoring, whether virtual or not, is more effective if it’s part of a strong early reading program that includes a curriculum based on the science of reading and screening students for dyslexia or other learning difficulties. 

鈥淵ou can鈥檛 throw tutoring at the problem,鈥� Sliwerski said at the Accelerate conference. 鈥淚t has to be part of a very intentional system.鈥�

Over $120 billion in federal were expiring, leaving school leaders and tutoring providers uncertain whether programs would continue. The incoming administration was focused on slashing Department of Education spending, not issuing new grants. 

鈥淲e didn’t know if this administration would put anything into education,鈥� said Loeb, a Stanford University professor who . 鈥淲e were worried that all of the experimentation that had been going on and that access to tutoring would drop precipitously.鈥� 

That didn鈥檛 happen.

When researchers, district leaders and tutoring providers convened earlier this month in Washington, it was clear that worries over tutoring being nothing more than a pandemic fad had turned to optimism. A growing number of states expect districts to integrate tutoring into the school day and have committed funding and staff to make it happen. Several require tutoring for students scoring below grade level and are vetting providers so districts don鈥檛 have to. And in a recent round of literacy , totaling $256 million, federal education officials signaled that access to tutoring should be a fixture in the nation鈥檚 schools. 

鈥淗igh-dosage tutoring has evolved from a concept into a proven, evidence-based strategy and then into a reality for thousands of students in thousands of schools,鈥� Kirsten Baesler, assistant secretary for elementary and secondary education, told attendees at the annual Accelerate conference. 鈥淚t is a foundational strategy for improving student outcomes.鈥� 

Even before the new federal grants were announced, a by Loeb鈥檚 team showed that nearly half of all states either offer tutoring grants or use their school finance formula to help districts pay for programs. 

Arkansas, which she described as 鈥渟trategic and ambitious,鈥� is one example. Its 2023 LEARNS Act created two tutoring programs.

One provides grants to . To measure the return on investment, the state鈥檚 now flags whether a student receives tutoring during the school day.

鈥淚f policymakers want results, they have to invest in the structures to get those results,鈥� Amy Counts, director of curriculum projects at the state education agency, said during one of the Accelerate sessions. 

Another Arkansas initiative up to $1,500 to spend on tutoring if their children don鈥檛 meet reading standards. Initially, teachers weren鈥檛 fond of the idea that families received the extra money instead of schools.

鈥淭hey didn鈥檛 push the program because of that. But we said, 鈥業f you help parents use that program, that benefits you,鈥� 鈥� Counts said. The other challenge, she said. was that some parents of struggling readers wouldn鈥檛 spend the money 鈥渂ecause they’ve never had to engage in securing services for their child.鈥�

To Accelerate President Nakia Towns, the federal grants represent an important shift. 

鈥淟ook Mama, we made it,鈥� she told attendees.

Accelerate received one of those 24 grants to work with the Oklahoma State Department of Education. They鈥檒l test how factors like group size, the frequency of sessions and whether tutoring is delivered virtually or in person affects results. 

Seven of the awards went directly to state education agencies that are working to scale up tutoring programs, especially in rural areas. Loeb鈥檚 team, for example, will evaluate Arkansas鈥� efforts to , a virtual program. The study will also compare results when tutors are college students versus trained educators.

鈥業mportant step forward鈥�

Accelerate has launched some of that state-level activity through its , and this legislative session, CEO Kevin Huffman is tracking 12 tutoring-related bills in eight states. They include:

• A that would require high-impact tutoring for students with a reading or math 鈥渄eficiency.鈥�

• A to establish a competitive grant program for tutoring.

• An that would require for students scoring at the lowest levels in math and reading. The Senate passed the bill, but it鈥檚 still pending before a House education committee.

• A that would expand an existing program for elementary students through eighth grade.

鈥淎ll of this feels like an important step forward,鈥� Huffman told 蜜桃影视. At the conference, he said 鈥渕omentum is different鈥� because states aren鈥檛 supporting tutoring just because they have one-time federal dollars to spend.

One policy expert recently questioned whether tutoring has produced 鈥渢oo little bang for too much buck.鈥� In , Michael Petrilli, president of the conservative Thomas B. Fordham Institute, said he hasn鈥檛 been able to 鈥渕uster much enthusiasm鈥� for tutoring and suggested that it has been an insufficient way to address the 鈥渄isastrous aftermath of COVID-era school closures.鈥�

Loeb agreed that while pandemic relief funds allowed states and districts to test different models, those early examples didn鈥檛 always produce gains. Some states and districts moved too fast, and implementation challenges, like infrequent sessions and high turnover of tutors, hindered students鈥� progress. Research shows that a mismatch between the material tutors cover and the curriculum in students鈥� regular classes can also contribute to poor results.

鈥淪ome of it worked, and some of it didn’t,鈥� Loeb said. 

But during this month鈥檚 event, Antoinette Mitchell, state superintendent for the District of Columbia, said investments in tutoring have paid off. Her office, which oversees both the District of Columbia Public Schools and charters, contracts with CitySchools Collaborative, a nonprofit, to manage tutoring logistics. It handles scheduling and finds space for tutoring sessions so principals don鈥檛 have to. 

, more than 42% of DCPS students scored at the highest levels in reading, exceeding pre-pandemic results. In math, the percentage of students meeting expectations grew by over 4 percentage points, the largest jump since 2015. With one of the federal grants, CitySchools Collaborative will expand its work into Maryland and Virginia. 

More recent research findings, about the importance of offering tutoring and , have also allowed districts to learn from past mistakes. 

鈥淵ou can actually do this at a decent scale,鈥� Loeb said, 鈥渁nd give students this personalized attention.鈥�

That鈥檚 when a large suburban school district in Texas, flush with COVID relief funds, signed a contract with a virtual tutoring provider to deliver extra help to students in 28 schools who had fallen below grade level. Research showed that could produce significant gains for students and was far more effective than on-demand models.

But the district鈥檚 program , according to a recent study from Stanford University鈥檚 National Student Support Accelerator, which focuses on studying and expanding effective tutoring. Students even lost ground in reading and would have been better off with 鈥渂usiness-as-usual鈥� support, like small group instruction or using a computer program for extra practice. 

Experts view the findings as a cautionary tale of how tutoring can go wrong.

The district had to wait on background checks for tutors, many students were still chronically absent and the tutoring sessions often conflicted with other lessons or special events. As a result, students didn鈥檛 receive the 30 hours or more required under a mandating tutoring for those who failed the annual state test. Instead of five days a week as planned, 81% of the students attended tutoring three or fewer days, and most students worked with a different tutor every time they attended a session.

The findings reinforce the importance of protecting the time students are supposed to receive tutoring, said Elizabeth Huffaker, an assistant professor of education at the University of Florida and the lead author of the study.

High-dosage models 鈥� featuring individualized sessions held at least three times a week with the same, well-trained tutor 鈥� can still 鈥渄rive really significant learning gains,鈥� she said, 鈥渂ut in the field, things are always a little bit more complicated.鈥�

For parents, the Stanford study can help explain why children might not make gains, even when their district offers extra help, said Maribel Gardea, executive director of MindShiftED, a nonprofit advocacy group and network of about 5,000 parents in the San Antonio area. Despite the billions states received in relief funds, many students still haven鈥檛 reached pre-pandemic levels of performance.

鈥淲e knew that high-dosage tutoring was one of those things that was proven,鈥� Gardea said.  鈥淭here was research, but we never saw those results.鈥�

She urges districts to include parent groups like hers in planning tutoring and choosing providers. But she added that too many parents are unaware their children are behind, much less equipped to judge whether a program is set up for success. 

鈥淭he trust has been lost for such a long time,鈥� she said. 鈥淧arents just send their kids to school and they hope for the best.鈥� 

鈥業t鈥檚 logistics鈥�

The results add to a growing body of research at a time when tutoring has shifted from being viewed as an emergency stopgap to an ongoing teaching strategy, according to released last week from Whiteboard Advisors, a consulting organization. 

The authors鈥� interviews with state and local education leaders, researchers and tutoring providers showed that while many schools lean toward in-person tutors, 鈥渆ffective virtual models persist鈥� in many districts. Going forward, they expect more schools to use tutoring as a pipeline for recruiting and training new teachers.

Districts have learned a lot about tutoring since that first, full year back after school closures, one in which districts saw staff shortages, record levels of absenteeism and disruptive behavior from students. have passed legislation to support tutoring or provide at least some short-term funding to keep programs running now that federal relief funds have expired. Some districts, including , are designing contracts that reward tutoring providers with more money when students pass tests or make other significant gains.

Recent shows an increase since December 2022 in the share of schools offering high-dosage tutoring, from 37% to 42% 鈥� especially in the South. But the results of the study show that just giving tutoring a high-dosage label doesn鈥檛 mean students will receive the help they need.

鈥淚t鈥檚 logistics,鈥� said T. Nakia Towns, chief operating officer at Accelerate, which funds research on tutoring and other recovery efforts. 鈥淵ou have to have the scheduling. You have to have the identification of the students.鈥�

High mobility, absenteeism

To encourage the tutoring provider and the Texas district to participate in the study, the researchers didn鈥檛 identify them. But an official with the district, who spoke on background, told 蜜桃影视 that one reason tutoring didn鈥檛 start until the middle of the school year was because leaders waited for winter test data to ensure they were selecting students who needed the most help.

The state required tutors to pass federal background checks, a process that added delays, and it took time to find bilingual tutors and those with special education experience. Students who were furthest behind academically 鈥渨ere also the same students who had high mobility or high absentee rates,鈥� the official said. 

School assemblies interfered with the tutoring schedule, and some principals, the official said, were less supportive of virtual tutoring in general. Now, he said, the district offers in-person afterschool tutoring as one option, but also builds intervention time into the school day for all students.

Tutoring during school hours increases the chances that students will actually get the service, but the model creates some challenges, Huffaker said. Tutoring is now 鈥渃ompeting with other instructional practices during the school day.鈥� 

That includes lessons that teachers are presenting to the whole class and don鈥檛 want students to miss, the district official added.

Recent findings from another tutoring study, the , provides further proof that the more tutoring students receive, the greater their gains. But the 鈥渂ad news,鈥� according to the researchers, from the University of Chicago and MDRC, was that students often didn鈥檛 receive as much tutoring as originally planned.

鈥淐onversations with the operators suggest schools felt they simply had too many competing demands on limited instructional time,鈥� the authors wrote.

Another takeaway from the Stanford study is the 鈥渃ritical role鈥� of relationships between tutors and students, said Rahul Kalita, co-founder of Tutored by Teachers, a virtual provider with a network of over 6,800 certified teachers. In the , one of its largest clients, students are approaching pre-pandemic levels in reading, and nearly 70% of third graders passed a reading test this year required for promotion to fourth grade. 

Without 鈥渃onsistent, human-to-human connection,鈥� Kalita said, results will be similar to on-demand 鈥渆dtech tools鈥� that researchers have found to be ineffective.

鈥楽tart with the curriculum鈥�

Not only did Texas students not receive enough tutoring, the research team found a weak relationship between their sessions and the material they needed to know for tests. Tutors covered about a third of the math standards and only about half that in reading. 

But this is an area where some tutoring companies have shown improvement, said Towns, with Accelerate. More successful providers, she said, 鈥渞eally start with the curriculum,鈥� and hire experts with 鈥渄eep knowledge around literacy or math.鈥� 

now show that remote tutoring can be just as effective as in-person programs. That鈥檚 why she encouraged districts not to give up on virtual models.

鈥淐oming out of the pandemic,鈥� she said, 鈥渆verybody was just like, 鈥楲et’s try anything. Anything is better than nothing,鈥� and in fact that’s not true.鈥� 

For a year and a half in 2023 and 2024, Cohen traversed the country, interviewing educators, researchers and policymakers and observing tutoring sessions in seven states and the District of Columbia. 

Now the vice president of policy for the education group , Cohen shares her findings in a new book, out today from Harvard Education Press: .

She explores 鈥渢he accidental experiment鈥� that took place across American schools starting in 2020, as researchers figured out the principles of what was originally called 鈥渉igh-dosage tutoring鈥� but has come to be known as 鈥渉igh-impact tutoring.鈥� 

Its four pillars, according to Stanford鈥檚 : 

1. It must take place at least three days a week.
2. Sessions last at least 30 minutes.
3. Sessions are with a consistent tutor.
4. There are no more than four students working in a group. 

The moment couldn鈥檛 have been more tailor-made for such a comprehensive intervention. In the course of just a few months, federal aid to K鈥�12 schools more than tripled, with districts slated to get at least 90% of the new funding. Federal rules eventually dictated that they reserve at least 20% of the largest pot of money to treat pandemic-related learning loss. Tutoring, Cohen writes, 鈥渜uickly became the watchword of how learning loss should be addressed.鈥�

Cohen interviewed everyone from Stanford scholar Susanna Loeb, whose research helped lay the groundwork for the movement, to Katreena Shelby, a Washington, D.C., middle school principal who somehow found a way to get a tutor for every student in her school.

Ahead of the book鈥檚 publication, Cohen spoke to 蜜桃影视鈥檚 Greg Toppo about her findings and her belief that, despite the bleakness of the past few years, educators 鈥渨ant to do good things for kids, and they’re willing to try new things.鈥�

Their conversation has been edited for length and clarity. 

I want to start with a kind of impertinent question: I believe it was former U.S. Education Secretary Bill Bennett who said that many schools serve up what he called a “14-egg omelet.” Have you heard of this?

No, but I like where it’s going.

When what they’re doing doesn’t work, they just do more of the same. I’m guessing you would say that high-impact tutoring does not resemble one of Bennett’s lousy omelets. Are schools truly doing something different?

It’s, of course, impossible to answer universally for every school and every tutoring program. And there have been tutoring programs that haven’t been super additive. But at this point, the schools that have implemented high-impact or high-dosage tutoring within the definition of what that is 鈥� and to the gold standard that the evidence suggests 鈥� are offering something different. Whether that’s home fries on the side of the omelet or a salad, you can choose, but it’s something else.

You write that a couple of places have done better jobs than others. New Mexico, for instance, seems to have made a few missteps. What’s the difference between places where tutoring is working and where it’s not?

Where tutoring works the best is where it is a strategy in service of a broader goal. Sometimes in education we make the mistake of thinking the thing is the goal, and tutoring isn’t the goal. I don’t want people to do tutoring just to do tutoring. I care if kids are learning in school, and so the places that are doing a great job with tutoring, first of all, are doing tutoring in service of the goal of improving learning, and that means it’s often connected to lots of other pieces around instruction, curriculum and all sorts of other things. One is being strategic. Two is recognizing that to do this kind of program well requires a lot of effort on the implementation side, and being willing to put in the resources necessary. Literally assigning someone at a district or at a school a role of high-impact tutoring manager 鈥� who a significant part, if not all, of their job for some period of time is making sure this program is working 鈥� is another hallmark of places that have had success as well.

When you were in Louisiana, you looked at this Teach for America Ignite program, and you mention that it’s become a strong pipeline for TFA Fellows and, by extension, teachers. Should we look at tutoring as a pipeline for teaching?

I think so. We have an evergreen population of college students, even if fewer than we used to. We’re always going to have some amount of college students. And what’s generally true about those young adults is that a lot of them are looking for ways to make some money, and a lot of them are not sure what they really want to do with their lives. So one of the interesting things 鈥� and the TFA program highlights this 鈥� is that when you create opportunities for young people to be involved in education, as a tutor, for example, they start thinking, “Oh, maybe this is a career that I would want to do.”

I like to joke that teacher unions have done such a great PR job that they’ve actually convinced people that they shouldn’t want to be teachers. They’ve convinced the American public that teachers don’t get paid enough and aren’t respected. And if you look at parent polls, more than 50% of parents in this country say they to become teachers.

But what we’ve learned from some of the tutoring with college students is that when you actually give them a positive framework to enter the education space and interact with young people in this way, they start thinking about it. It’s not just the TFA program 鈥� I would say also the in charter schools in New York and New Jersey, that also has had partnerships in D.C. and other places. Similarly, they’re using college grads through the AmeriCorps program. A lot of those young people end up sticking around and becoming teachers.

At a school in D.C., you met Delilah, who you say could easily pass for a high school student, but she’s doing this great job leading students on a lesson about Homer鈥檚 Odyssey. It made me think that tutoring could blur the boundaries between who is an effective teacher 鈥� and how we find them. Do you have any thoughts on that?

I don’t know about 鈥渂lur,鈥� but it certainly broadens how we might think about who can play effective roles in the learning of young people. And we see that in a few places. This isn’t in the book, but in Chattanooga, Tenn., they had a that started during COVID where they actually hired high school students to tutor elementary school students. And those high schoolers, I believe, were getting school credit, and were getting paid. I spoke with this young woman, and she would literally walk down the hill from her high school to the elementary school, where she worked as a tutor and got real-world experience. She said she felt like she was treated like one of the staff at the school, and it was an incredibly positive experience. She is now graduating high school a year early and enrolling at the University of Tennessee-Knoxville to become a teacher, and she’s the first person in her family to go to college. 

The other thing that I did write about is the way that education schools are rethinking the role of tutoring in teacher prep. We have all these college kids or young adults that we might want to expose to education. But then what about those who already think they want to work in education? The dean of the ed school of Bowling Green State University, which is the biggest teacher prep program in Ohio, has always been committed to giving kids as much field work and experience as possible, because she says, “I want to make sure before I send these students as graduates into classrooms, that that’s really where they want to be. How many different kinds of opportunities can we give people who think they want to be teachers to actually play teacher-like roles?” And so they’ve really leaned into tutoring. They think that the experience of me, Liz, trying to really just help Greg master how to read or how to do third-grade math is going to help me in the classroom, but also gives me more touch points to make sure this is really what I want to do. 

Another way to think about that: A principal in Alexandria, Va., told me, “The one thing I’m always looking for is how do I get my kids more time? More time learning. How do we give our kids more time?” And it wasn’t just him that I heard this from. This is a repeated theme that school leaders and teachers feel: Tutoring helps them add time. Time on task, quality learning time. And time is often the most precious resource we have in education, and that is how a lot of folks are thinking about this.

One of the things you say is that if tutoring is woven into a school culture, the relationship that the student has with the tutor can be this “fulcrum that changes the student’s trajectory.” You’re imagining that tutoring could really transform schools at a very basic level, that the student-tutor relationship is transformative for a lot of kids.

That’s right. What made this story so powerful was the power of the relationships. To me, the big takeaway is that young people are really hungry for meaningful adult relationships in ways beyond what even the best classroom teacher can possibly give to a full classroom of kids. Even when I interviewed some of those TFA college tutors, the thing they would tell me that surprised them about their experience was that kids were willing to open up to them even after just building a relationship on a Zoom call and doing tutoring. And I don’t know if it’s because after the pandemic there had been so much disconnect and isolation that people were hungry for a reconnect, or if it’s just a truism of human nature that we like to have relationships with other humans.

There’s something really powerful about bringing more people in to interact with young people in education, in an educational setting, in a variety of ways. And that’s why, even though generally I’m pretty bullish on tech 鈥� I don’t write in the book at all about AI because the stuff’s being built too rapidly 鈥� while tech can inform and empower, what’s happened, at least in the last five years, is really a story about human relationships, and it’s worth telling in a time when people feel more separate.

Near the end of the book, you talk about one way to make tutoring work on a large scale, something called outcomes-based contracting. Would you like to talk about that?

I wrote a whole chapter about contracting, and tried to make it so you wouldn’t fall asleep while you read it. Partly why I dedicated so much space to it is because I actually think that we spend a lot of money on education in this country 鈥� we really do 鈥� and we don’t often get a lot for it. And so it’s interesting that we have this model now. Tutoring is the perfect case study to do an outcomes-based contract, because we have potentially clear outcomes that we’re trying to measure: We want kids to grow a certain amount, and then we can actually link the money to what we’re getting from it. 

Especially now that federal COVID funds are gone, district and state budgets are tightening. I hope we don’t throw the success of tutoring that we’ve had to the wayside and instead think about how do we continue helping it deliver on its promise? And so if you can measure it and then pay only for getting the results that you want, that seems worthwhile, and something that we probably haven’t spent enough time exploring.

Speaking of ESSER funds, that’s a lot of money that’s basically gone. You mention AmeriCorps as well 鈥� AmeriCorps is either. Going forward, where can schools turn if they want to fund these sorts of things? What’s out there that is not at so much risk?

First of all, some districts are using their Title I funds. Now, those Title I funds might have been used for something else, and so you have to maybe make some tough choices 鈥� and I’m not going to say you should definitely do tutoring. I’m saying you should look at the evidence: What are you getting out of whatever it was you were doing? If you’re already doing tutoring and it’s going well, I’d rather a district keep it and give up something else that’s not working as well.

Ector County, Texas, has kept their tutoring program going to some extent, using Title I funds. Some other districts have done some similar work, even as districts like Guilford County, N.C., are having to scale back. But they are repurposing existing Title I funds, often to do this. One reason it’s really important to continue making the case for tutoring鈥檚 impact is that you can convince state legislatures, in some places at least, to fund tutoring. Louisiana put , both for last school year and this current year, into high-impact tutoring. And the funny thing about Louisiana is I didn’t even end up writing about it because it was happening so quickly last year while I was trying to finish the book.

I was like, “Wow, it’s a lot of money. Is this really going to happen?” And this year, 2025-2026, Louisiana is tutoring something like 240,000 kids using $30 million from their state budget, and I think some other district funds too, in a pretty effective model tied to their Science of Reading and their math work. And they have funded a lot of other pieces too, around curriculum, teacher professional development and instructional coaches. So for them, tutoring is that exact thing I said earlier about being a strategy within their broader goal of how to overhaul core instruction 鈥� and the state’s put in real money for it.

Connecticut passed to continue some high-impact tutoring work. But then in other states, we aren’t seeing that. Where to look for money? Can you convince your state legislatures to support tutoring because it works? Some places are able to do that.

And also some city budgets: The mayor in D.C. has . And the mayor in Nashville has into tutoring. 

At the end of the book, you lay out these three truisms from your reporting: “1. Public schools are hungry for new ideas that work. 2. Tutoring works. 3. Nothing is perfect.” It sounds like you’re a bit impatient here, and just want us to sort of get on with it. 

I do! Every single day you have kids showing up to school, and those kids either want to learn or it’s our job to help them want to learn, and we need to figure out the tools to do that. If you look, for instance, at continued problems with chronic absenteeism, we flipped a switch during the pandemic, and we thought we could just flip it back on.  That’s not what’s happened. So I believe we have to continue the sense of urgency that we had in 2021 and 2022, because there are kids every day in our schools. But the other thing I really want people to know is that in all of these places I went, people want to do good things for kids, and they’re willing to try new things and implement new programs and make big changes.

That’s not the reputation that K-12 public education has overall. And I want people to believe that that is part of the story of public education in the United States in 2025. I want us to get on with it, because it’s what people want to do. So let’s just do the thing.

The open-source tool, which researchers say other educators can recreate and integrate into their tutoring systems, made the human tutors slightly more effective. And the weakest tutors became nearly as effective as their more highly-rated peers, according to a study . 

The tool, dubbed Tutor CoPilot, prompts tutors to think more deeply about their interactions with students, offering different ways to explain concepts to those who get a problem wrong. It also suggests hints or different questions to ask.

The new study offers a middle ground in what鈥檚 become a polarized debate between supporters and detractors of AI tutoring. It鈥檚 also the first randomized controlled trial 鈥� the gold standard in research 鈥� to examine a human-AI system in live tutoring. In all, about 1,000 students got help from about 900 tutors, and students who worked with AI-assisted tutors were four percentage points more likely to master the topic after a given session than those in a control group whose tutors didn鈥檛 work with AI.

Students working with lower-rated tutors saw their performance jump more than twice as much, by nine percentage points. In all, their pass rate went from 56% to 65%, nearly matching the 66% pass rate for students with higher-rated tutors.

The cost to run it: Just $20 per student per year 鈥� an estimate of what it costs Stanford to maintain accounts on Open AI鈥檚 GPT-4 large language model.

The study didn鈥檛 probe students鈥� overall math skills or directly tie the tutoring results to standardized test scores, but Rose E. Wang, the project’s lead researcher, said higher pass rates on the post-tutoring 鈥渕ini tests鈥� correlate strongly with better results on end-of-year tests like state math assessments.听

Wang said the study鈥檚 key insight was looking at reasoning patterns that good teachers engage in and translating them into 鈥渦nder the hood鈥� instructions that tutors can use to help students think more deeply and solve problems themselves.听

鈥淚f you prompt ChatGPT, ‘Hey, help me solve this problem,’ it will typically just give away the answer, which is not at all what we had seen teachers do when we were showing them real examples of struggling students,鈥� she said.

Essentially, the researchers prompted GPT-4 to behave like an experienced teacher and generate hints, explanations and questions for tutors to try out on students. By querying the AI, Wang said, tutors have 鈥渞eal-time鈥� access to helpful strategies that move students forward.

鈥滱t any time when I’m struggling as a tutor, I can request help,鈥� Wang said.

She said the system as tested is 鈥渘ot perfect鈥� and doesn鈥檛 yet emulate the work of experienced teachers. While tutors generally found it helpful 鈥� particularly its ability to provide 鈥渨ell-phrased explanations,鈥� clarify difficult topics and break down complex concepts on the spot 鈥� in a few cases, tutors said the tool鈥檚 suggestions didn鈥檛 align with students鈥� grade levels. 

A common complaint among tutors was that Tutor CoPilot鈥檚 responses were sometimes 鈥渢oo smart,鈥� requiring them to simplify and adapt for clarity.

鈥淏ut it is much better than what would have otherwise been there,鈥� Wang said, 鈥渨hich was nothing.鈥�

Researchers analyzed more than half a million messages generated during sessions, finding that tutors who had access to the AI tool were more likely to ask helpful questions and less eager to simply give students answers, two practices aligned with high-quality teaching.

Amanda Bickerstaff, co-founder and CEO of , said she was pleased to see a well-designed study on the topic focused on economically disadvantaged students, minority students, and English language learners.  

She also noted the benefits to low-rated tutors, saying other industries like consulting are already using generative AI to close skills gaps. As the technology advances, Bickerstaff said, most of its benefit will be in tasks like problem solving and explanations. 

Susanna Loeb, executive director of Stanford鈥檚 National Student Support Accelerator and one of the report鈥檚 authors, said the idea of using AI to augment tutors鈥� talents, not replace them, seems a smart use of the technology for the time being. 鈥淲ho knows? Maybe AI will get better,鈥� she said. 鈥淲e just don’t think it’s quite there yet.鈥�

At the moment, there are lots of essential jobs in fields like tutoring, health care and the like where practitioners 鈥渉aven’t had years of education 鈥� and they don’t go to regular professional development,鈥� she said. This approach, which offers a simple interface and immediate feedback, could be useful in those situations. 

鈥�The big dream,鈥� said Wang, 鈥渋s to be able to enhance the human.鈥�

Benjamin Riley, a frequent AI-in-education skeptic who leads the AI-focused think tank and writes a on the topic, applauded the study’s rigorous design, an approach he said prompts 鈥渆ffortful thinking on the part of the student.鈥�

鈥淚f you are an inexperienced or less-effective tutor, having something that reminds you of these practices 鈥� and then you actually employ those actions with your students 鈥� that’s good,鈥� he said. 鈥淚f this holds up in other use cases, then I think you’ve got some real potential here.鈥�

Riley sounded a note of caution about the tool鈥檚 actual cost. It may cost Stanford just $20 per student to run the AI, but he noted that tutors received up to three weeks of training to use it. 鈥淚 don’t think you can exclude those costs from the analysis. And from what I can tell, this was based on a pretty thoughtful approach to the training.鈥�

He also said students鈥� modest overall math gains raises the question, beyond the efficacy of the AI, of whether a large tutoring intervention like this has 鈥渕eaningful impacts鈥� on student learning. 

Similarly, Dan Meyer, who writes a on education and technology and co-hosts a on teaching math, noted that the gains 鈥渄on’t seem massive, but they’re positive and at fairly low cost.鈥�

He said the Stanford developers 鈥渟eem to understand the ways tutors work and the demands on their time and attention.鈥� The new tool, he said, seems to save them from spending a lot of effort to get useful feedback and suggestions for students.

Stanford鈥檚 Loeb said the AI鈥檚 best use is determining what a student knows and needs to know. But people are better at caring, motivating and engaging 鈥� and celebrating successes. 鈥淎ll people who have been tutors know that that is a key part about what makes tutoring effective. And this kind of approach allows both to happen.鈥�

It was the third attempt in two years.

With the school year winding down, some districts never even got word the program existed. Those that participated quickly scrambled to cram it into their schedules.

鈥淭he timing wasn’t optimal,鈥� said Matt Monta帽o, superintendent of the Bernalillo Public Schools, north of Albuquerque, and one of just five districts out of the state鈥檚 89 to sign up. Staff members, he said, were 鈥渁 little bit less than enthusiastic鈥� about the interruption.

The late rollout was only the most recent snag in the state鈥檚 troubled effort to spend millions in federal relief funds for tutoring before the deadline to use the money hits next month.

The first attempt 鈥� with an on-demand, virtual provider 鈥� met with a meager response from families. A second try never got off the ground because of a contract mishap the state still won鈥檛 fully explain. And the delayed start on the third effort means only a fraction of the students slated for tutoring got it. State officials estimate that between 2,000 and 3,000 students received the extra help 鈥� far less than the 8,000 they were hoping to reach.

“Clearly, it was not the best,” Amanda DeBell, New Mexico鈥檚 deputy education secretary, said of the condensed program. But in July, the legislature pumped new life into the effort, providing $8.5 million for high-dosage tutoring this fall. The state also plans to use what鈥檚 left of the $4 million in federal relief funds that they鈥檇 hoped to spend last school year to support math tutoring for middle school students.

Data shows New Mexico students still have a lot of ground to make up to combat pandemic learning loss. The state in fourth grade math and reading in the most recent iteration of the National Assessment of Educational Progress.

The experience underscores the difficulty of pulling off a statewide tutoring effort 鈥� even one backed by convincing research and millions of dollars in federal relief funds.

At a May tutoring conference at Stanford University, Education Secretary Arsenio Romero spoke candidly about the state鈥檚 false starts. 

鈥淪ometimes we as educators are our own worst enemies,鈥� he said. 鈥淲e go through year-long cycles before we 鈥� make changes. You need to be able to pivot.鈥� 

鈥楢ll the way to the living room鈥�

Especially when the needs are so great. 

On state tests, less than a quarter of New Mexico students meet math standards and just 38% score proficient in English language arts. The state also continues to operate under to improve education for English learners and low-income, special education and Native American students.

In late 2022, Gov. Michelle Lujan Grisham announced the state had signed a with Paper, a virtual, on-demand tutoring company. The promised to offer students in high-poverty elementary and middle schools 鈥� those hit hardest by school closures 鈥� up to 20 hours of free tutoring.

But the state abruptly terminated the contract less than three months later. The model expected families to sign up for help on nights and weekends, which research shows who are furthest behind. Those students might not know the right questions to ask a tutor, and technical glitches associated with online programs tend to frustrate both kids and parents who are already discouraged.

鈥淭his service is not providing the results in terms of engagement, support or delivery of service to the state鈥檚 students,鈥� Mariana Padilla, then-interim secretary of education, wrote to the company.

Monta帽o in Bernalillo doesn鈥檛 think any students in his district signed up for the program. 鈥淒eployment from the state level all the way to the living room of families is a hugely difficult process,鈥� he said.

Paper officials cited multiple reasons for the rocky rollout. The program launched just as students returned from holiday break in January 2023, and the state didn鈥檛 give the company enough time to get buy-in from families and schools, said spokeswoman Ava Paydar.

Re-envisioning tutoring 

Romero, appointed secretary by Lujan Grisham in March 2023, faced the immediate challenge of finding a more-effective tutoring provider.

鈥淚t really 鈥� allowed us to re-envision what we wanted tutoring to look like,鈥� he said at the Stanford conference. 

Three months after it canceled Paper鈥檚 contract, the state education department for vendors who could offer a high-impact model, either in person or virtually. The virtual classes that predominated during the height of the pandemic set students back academically by months, even years. But research shows that live instruction from a tutor working remotely can produce positive results if schools schedule sessions during the school day and offer the same consistent and frequent support as an in-person tutor.

The state chose three providers, who were slated to begin serving students last August. But officials abruptly canceled that program before it got started because of a protest from another vendor that wasn鈥檛 chosen. The department declined to explain the nature of the dispute, and Romero said the education department never finalized contracts with the three providers.

Some education advocates grew impatient as they watched the school year go by without a program in place. 

鈥淲e failed to offer consistent access to quality, high-impact tutoring,鈥� said Amanda Aragon, executive director of NewMexicoKidsCAN, part of a national network of education policy and advocacy groups. She called the spring effort 鈥渋n no way sufficient.鈥�

While New Mexico may have faced more obstacles than most, other states trying to provide tutoring to thousands of students have weathered similar ordeals.

New Jersey to get funding to districts to hire tutors, and Virginia initially got a from districts when Gov. Glenn Youngkin announced his new All in VA plan, which includes high-impact tutoring in third through eighth grades. In Louisiana, some vendors passed on participating in a program that pays for one-to-one sessions 鈥� about half what providers normally charge. 

鈥淎ny state that was ambitious enough to take on large-scale implementation of tutoring has experienced growing pains,鈥� said Nakia Towns, chief operating officer of Accelerate, a funding tutoring programs and research. Many have struggled to find high-quality vendors and convince districts to participate. 

With the new state funding, New Mexico is trying something different. The state will provide the money, but districts will issue their own contracts and have flexibility to hire teachers or choose the outside vendors they want. 

District efforts

One reason New Mexico leaders ultimately changed course is that they saw that districts had succeeded in blending tutoring into the school day.

Ten Las Cruces schools participated in a program this past school year with , a virtual model led by credentialed educators. Students who were a grade level or more behind gained roughly twice as much learning as those who didn鈥檛 get tutoring, leading the district to invite the provider back this fall, said co-founder Rahul Kalita.

Romero visited one of the district鈥檚 schools in October and saw Spanish-speaking students practicing their English skills with a bilingual tutor while also getting math support.

Kalita attributed some of the state鈥檚 prior difficulties to a lack of 鈥渟teady leadership鈥� at the top. Romero is New Mexico鈥檚 third education secretary since 2019.

“Funding is critical, but it’s just the first step,鈥� he said.

Further evidence on in-school tutoring comes from on a virtual model that has helped prepare over 500 New Mexico middle school students for high school algebra. The program, continuing this fall, is used in large districts like Chicago, Miami-Dade and Fulton County, Georgia. In New Mexico, the effort includes 19 districts, many of them small and isolated, like Tatum Municipal Schools. 

Located about 15 miles from the Texas border, the rural district had just 26 seventh graders last school year. All of them received tutoring, and over half met or exceeded goals by the spring. That鈥檚 a small improvement over their scores from sixth grade, said Superintendent Robin Fulce, but he considers that progress significant because of the 鈥渂ig jump鈥� in rigorous material in seventh grade.

The program has convinced Fulce that students can form tight relationships even with tutors they meet online. 

Recently, two of those tutors passed through town for a visit.

鈥淭hey brought doughnuts and every kid in that seventh grade went over and hugged them. 鈥淚t was a very good experience,鈥� Fulce said. To him, the state鈥檚 multiple tutoring efforts reinforced that offering services outside the school day doesn鈥檛 benefit 鈥渒ids who need it the most.鈥�

The results, Romero said, influenced the state鈥檚 decision to shift gears and make 鈥渄ecisions based on research and data.鈥�

Monta帽o, the Bernalillo superintendent, estimated that about 800 students in his district received services 鈥� roughly half those he felt should have gotten the support. But he doesn鈥檛 consider it a wasted effort.

鈥淚t was too good of an opportunity for us not to take advantage鈥� of it, he said. 

Focused on schools in Washington, D.C., the preliminary results show middle school students attended an additional three days and those in the elementary grades improved their attendance by two days when they received tutoring during regular school hours.  

But high-impact tutoring 鈥攄efined as at least 90 minutes a week with the same tutor, spread over multiple sessions 鈥� had the greatest impact on students who missed 30% or more of the prior school year. Their attendance improved by at least five days, according to the study from the National Student Support Accelerator, a Stanford University-based center that conducts tutoring research. 

Susanna Loeb, who leads the center, called the data 鈥渢he first evidence of a strong causal link between tutoring specifically and attendance.鈥� 

Hedy Chang, executive director of the nonprofit Attendance Works, said it makes sense that students come to school more often when they鈥檙e keeping up in class and getting good grades. 

鈥淧art of why kids don鈥檛 show up is because they don鈥檛 feel successful in school,鈥� she said. Forming a connection with a tutor over several weeks or months can also make students more motivated to attend, she added. 鈥淚 do think it’s an impact of high-dosage tutoring, not necessarily just tutoring.鈥�

The early findings, which will be expanded in a future paper, reinforce the benefits of offering high-impact tutoring during the school day. The extra instructional time helps schools address two of their biggest post-pandemic problems 鈥� learning loss and chronic absenteeism, the researchers said. The White House has urged districts to not only target remaining federal relief funds toward those areas, but explore ways to sustain those efforts when they dry up. 

Districts that continue tutoring programs will likely keep 鈥渟tudent achievement top of mind,鈥� Loeb said, 鈥渨ith greater engagement 鈥� including increased attendance 鈥� as another outcome they hope to see.鈥�

also demonstrated how to successfully integrate tutoring sessions into the school day. The state education agency, which has spent $35 million on the program, funds staff members in charge of rearranging the schedule to accommodate the sessions and track data on student participation.听

鈥淭hey took that off the plate of the principal,鈥� Christina Grant, D.C.鈥檚 state superintendent, said at a January conference hosted by Accelerate, an organization that works to scale high-dosage tutoring. She added that working with researchers like those from Stanford can help districts communicate the impact of federal relief funds. Without those partnerships, she said, 鈥渨e would look back three years later and not be able to tell the authentic story around what happened to $35 million.鈥�

The district, which had a chronic absenteeism rate of last school year, began its tutoring program in 2021. Officials awarded grants to a variety of providers, including , which focuses on high school math and teacher preparation program.

Sousa Middle School, in southeast D.C., works with George Washington University鈥檚 , which pays college students interested in STEM or education to work as tutors.

鈥淢y challenge, when this program first began, was getting students to come and not look at it as a form of punishment,鈥� said Sharon Fitzgerald, Sousa鈥檚 tutoring manager. Now students who have 鈥済raduated鈥� out of the program ask why they can鈥檛 come back. 

Students responded well, she said, because it鈥檚 a 鈥渂reak away from seeing their regular teachers every day鈥� and because they look up to the college students. The tutors, she added, also have a clever way of giving students a taste of how much more they鈥檒l learn during their next meeting and if they attend class everyday.

鈥淚t was what the tutors left them with in the last session that encouraged them to come to school,鈥� Fitzgerald said.

The results are likely to spark more interest in how tutoring and attendance initiatives can work in tandem.

鈥淲e have not intentionally used tutors as a way to address attendance. I can imagine that it could help if part of their work focused on that,鈥� said A.J. Gutierrez, co-founder of Saga Education. 鈥淚 see potential.鈥�

Chang, with Attendance Works, said the results are 鈥渙n the right track,鈥� but don鈥檛 go far enough. During the , several states still had chronic absenteeism rates over 30%, including Alaska, New Mexico and Oregon.

Tutoring doesn鈥檛 address all of the barriers that keep students from attending school, like health conditions or bullying, she said. But tutors could refer students to school attendance teams when those concerns surface.

鈥淲hat more could we get,鈥� she asked 鈥渋f tutoring was tied to a bigger strategy, a more comprehensive approach?鈥� 鈥�

According to from Stanford University, about 60 to 70 percent of high school students surveyed in the fall of 2023 have engaged in cheating behavior 鈥� the same number prior to the debut of ChatGPT in the fall of 2022.

鈥淚 thought that we would see higher numbers in the fall so it was a little surprising to me,鈥� said Denise Pope, a senior lecturer at Stanford鈥檚 Graduate School of Education who surveyed students across 40 high schools through an she co-founded.

Victor Lee, an associate professor at Stanford鈥檚 Graduate School of Education who helped oversee the research with Pope, said high school students are 鈥渦nderwhelmed鈥� by AI chatbot tools.

鈥淚t just sounds very sterile and vanilla to them,鈥� Lee said. 鈥淭hey may have heard about it, but the media a lot of kids are using are quite different than the ones adults and working professionals are attuned to.鈥�

A conducted by the in the fall of 2023 found nearly one-third of students aged 13 to 17 have never heard of ChatGPT and another 44 percent have only heard 鈥渁 little鈥� about it. 

From those who were familiar with ChatGPT, the vast majority 鈥� about 81 percent 鈥� said they had not used it to help with school work.

鈥淢any teens are using a variety of technology鈥but] among those who鈥檝e heard at least a little about ChatGPT, shares of them still aren鈥檛 sure how they feel about it,鈥� said Colleen McClain, a research associate at the Pew Research Center.

Here are four things to know about the effects AI chatbot tools have had on high school cheating:

1. High school students who weren鈥檛 cheating before aren鈥檛 cheating now.

According to the , surveys of more than 70,000 high schools from 2002 to 2015 found about 64 percent of students cheated on a test 鈥� a similar outcome to Stanford鈥檚 findings after the rise of AI chatbot tools.

Pope said what surprises educators and parents the most is how common cheating has been.

鈥淲e know from our research that when students do cheat, it鈥檚 typically for reasons that have very little to do with their access to technology,鈥� Pope told .

鈥淲hen a student is less engaged, when they feel like they don’t belong or are not respected or valued in their community, when they鈥檙e stressed and highly sleep deprived 鈥� these are things that tend to correlate with cheating,鈥� Pope said. 

Lee said this number will 鈥渃onsistently stay there unless schools engage in certain steps to be thoughtful about what climate they’re creating that motivates cheating.鈥�

This includes tapping into the topics students are already interested in and developing useful skills based on how they naturally enjoy learning.

鈥淎 lot of the time, the AI students encounter is via Snapchat because they have a chatbot built into it,鈥� Lee said. 鈥淎nd students aren鈥檛 turning to Google as their primary search, they turn to YouTube鈥or] video-based searches rather than text-based.鈥�

2. ChatGPT awareness is higher among White, wealthier and older students.

Pew found about 72 percent of white students had at least some knowledge of ChatGPT compared to 56 percent of Black students.

In addition, more than 75 percent of students in households with an annual income of $75,000 or more had some knowledge of ChatGPT compared to 41 percent of students in households with annual incomes under $30,000.

McClain pointed to the 鈥渄igital divide鈥� as an explanation for Pew鈥檚 survey findings.

鈥淭he pattern here is quite striking,鈥� McClain said. 鈥淚t certainly speaks to the fact that not every teen is equally likely to have heard about these tools and used them.鈥�

She added how awareness of ChatGPT was seen more in older students 鈥� particularly those in 11th and 12th grade.

鈥淓ven among those who heard at least a little about ChatGPT鈥young] teens may still be figuring out how they feel about it,鈥� McClain said.

3. High school students have adopted a 鈥済ood faith鈥� approach to AI chatbot tools.

Pew found only 20 percent of students aged 13 to 17 said ChatGPT was acceptable to write essays compared to 57 percent who said it was not.

But, nearly 70 percent said it was acceptable to research new topics compared to 13 percent who said it was not.

The Stanford researchers found similar outcomes.

At four high schools surveyed this fall 2023, about 9 to 16 percent of students used AI chatbot tools to write essays and about 55 to 77 percent used it to generate an idea for a paper, project or assignment.

鈥淭he vast majority don鈥檛 want AI to do all the work for them so they鈥檙e coming into this with sort of a good faith effort,鈥� Lee said.

鈥淲hen I鈥檝e had conversations with educators, they sort of breathe a sigh of relief and think 鈥榦h okay let鈥檚 think about some of the cool things we could do鈥� and that鈥檚 exciting,鈥� Lee added.

4. Prohibiting AI chatbot tools won鈥檛 solve the systemic issues of why students cheat.

For Pope, finding comfort around AI chatbot tools starts with educators and parents including their students into the conversation.

鈥淚f you’re going to come up with a classroom or home policy, you want to have the students present, speaking up and telling you what they think will be the most useful and appropriate uses of AI,鈥� Pope said.

Lee said addressing AI chatbot tool usage in high schools is just the 鈥渢ip of a much larger iceberg.鈥�

鈥淧art of why we get concerned is because students feel pretty disenfranchised from the boring assignments, tedious homework and essays in these weird written formats that they don鈥檛 feel will provide them any long term need or use,鈥� Lee said.

鈥淚 don鈥檛 see us as saying AI is the best thing since sliced bread, but I also don鈥檛 think of us as saying AI is going to destroy humanity,鈥� Lee added.

Without a foundational understanding of the relationship between numbers and quantities, Gaffney says, some students begin school 鈥渂elieving that math is almost a foreign language where they need to memorize answers to equations they鈥檙e seeing on the board.鈥�

But to her, memorization is the wrong approach. Instead, she believes games and activities can help students recognize that the numbers in their lessons exist in real life. Once they make those connections, Gaffney thinks students will gain the understanding and confidence necessary to 鈥渇igure out鈥� basic math problems.

This way of thinking is what led her to help launch Heart Math Tutoring in 2013, which trains volunteer tutors to use toys and games to help students comprehend math.

Heart Math works with first through fifth graders in Charlotte, North Carolina, but has recently expanded to other nearby districts, including adding three South Carolina schools to its roster. This year, Heart is working with 26 schools in Charlotte.

Earlier this year, Heart received a $250,000 grant from the tutoring nonprofit Accelerate to expand their program. Accelerate previously told 蜜桃影视 that Heart Math was selected for its use of volunteers, which can be a way to expand tutoring access. 

Gaffney said 鈥渢he Charlotte community has been extremely generous to fund the program鈥� and donated 鈥渙ver $5 million鈥� to public schools through Heart Math Tutoring.

To sign up, schools need to identify at least 50 students who are in need of math intervention. Heart Math Tutoring offers a year-long program, which costs about $75,000 per school to deliver, according to Gaffney. But Gaffney said that through agreements with partners, who offset those costs, some schools pay only 5% to 15 % of that amount.

When schools sign up for the year, Heart Math sends a dedicated staff member to their buildings four days a week for two hours a day. The program also brings a team of volunteers, who each come once a week for one of those hours. Heart鈥檚 staff member supervises and oversees all tutoring sessions, while tutors work one-on-one with students. Each student gets 30 minutes with a tutor twice per week.

Prior to starting the interventions, there is a month-long period of onboarding where the Heart Math employee assesses students鈥� math skills to figure out where they need help. Then, tutoring lasts for eight months. Afterwards, Heart Math reports back data and progress to schools. 

Since officially launching 10 years ago, Heart Math Tutoring has served over 5,000 students. The program鈥檚 states that 97% have met its target for academic growth. According to Heart鈥檚 from September, last school year Heart served 1,072 students across 28 schools. Of those, 96% 鈥渟howed growth on pre/post assessments.鈥�

Charlotte-Mecklenburg Schools is one of Heart Math Tutoring鈥檚 oldest clients and has been working with the group for 13 years, prior to Heart establishing its own nonprofit in 2013. The district鈥檚 math specialists helped design the pilot version of Heart Math Tutoring in 2010, according to executive director of communications Susan Vernon-Devlin.

Vernon-Devlin said the 26 Charlotte-Mecklenburg schools that use Heart Math Tutoring 鈥渧alue both the academics and the one-on-one relationships that tutors build with our students.鈥�

鈥淥ur schools consistently report positively about Heart Math Tutoring鈥檚 work with our students,鈥� she wrote in an email. 鈥淢any teachers report that students return to class more confident, ready to help their peers, answer questions, and explain their thinking in math.鈥�

Gaffney says Heart Math is 鈥渃hanging students’ beliefs about math, going from thinking that it’s something that has to be memorized to it’s something that they can figure out.鈥�

From basic counting to multiplication and division, Heart Math puts students in charge of their own learning, Gaffney said. 

鈥淲e train our tutors instructionally to make sure the student is the one doing the work, meaning the student is the one touching the materials, the student is the one doing most of the talking,鈥� Gaffney said.

This year is Elizabeth Darden鈥檚 fourth volunteering for Heart. As a communications professional, Darden doesn鈥檛 consider herself a math expert, but she said that the training from Heart Math and the guidance from its on-site staff prepared her to get the job done. 

鈥淸Heart Math] makes it really easy for tutors because I mean, I think I’m decent at math, but definitely not in the math field or anything like that,鈥� Darden said. 

Logan Henderson, who has volunteered with Heart for 10 years and is also a 鈥渇inancial supporter,鈥� agreed that 鈥測ou don鈥檛 have to be a math expert鈥� to be successful.

Darden said most of the activities she does during tutoring are 鈥渇un, interactive games,鈥� which she says excites her students.

鈥淚t’s kind of a treat when a student gets called for Heart,鈥� Darden said. 鈥淣one of them see it as something negative or embarrassing, and they always jump right up and come in to play.鈥�

Darden said the games build confidence in her students and help them visualize math concepts. Two years ago, one of her students started 鈥渞eally struggling鈥� and fell behind his grade level, but by the end, he won an award in his class for 鈥渕ost improved.鈥�

Henderson, who works in finance, said that Heart鈥檚 approach to teaching math is different from how he remembered learning math when he was in school, where the focus was on memorizing math facts like multiplication tables.

鈥淲hen I was a kid, it seemed like the focus was on memorization,鈥� Henderson said. 鈥淭here’s been a recognition, I think, that, 鈥楬ey, there’s visual learners,鈥� right, some learn better visually.鈥�

Gaffney said that some of Heart鈥檚 older students are 鈥渢wo-three years behind in elementary math.鈥� These students often 鈥渓ack a concrete understanding鈥� of 鈥渨hat numerals really stand for,鈥� which is why making visual connections is important..

鈥淏y the time they’ve gotten to us, someone has already told them before many times that three plus four equals seven, but when you show them a pile of seven cubes, and you cover up three cubes and they’re seeing four, they don’t know how many are hiding,鈥� Gaffney said. 鈥淪o telling them that three plus four is seven again is not going to fix it.鈥�

To gain that understanding, students need activities that can help them visualize and comprehend math problems, including counting and arithmetic with objects, according to Gaffney. 

鈥淭he important part is that they鈥檙e connecting numerals, abstract numerals, to concrete, hands-on things in the world,鈥� Gaffney said.

The approach is supported by research. 

Dionne Cross Francis, a professor of education at the University of North Carolina at Chapel Hill who specializes in math education, said most students enter kindergarten with 鈥渟ome concept of quantity鈥� and 鈥渁 desire to quantify things,鈥� but that there is wide variation in their levels of understanding. If these problems are left unaddressed, students can have difficulty learning other concepts, as well, she said. 

Cross Francis believes Heart Math has the right idea by emphasizing concrete activities early.

鈥淩esearch would suggest that that’s where you start with children: that they actually have to be able to see one item and label it with the number-word one, and then continue to add additional items and label them with the appropriate number-word,鈥� Cross Francis said. 鈥淲e want to engage kids in really rich worthwhile experiences where they’re developing concepts from engagement with activities.鈥�

Jo Boaler, an education professor at Stanford University and one of the author鈥檚 of California鈥檚 new math education framework, said that for too many students, learning math is a 鈥渃ompletely abstract experience.鈥� 

Boaler thinks students should 鈥渃onnect more with numbers in the world鈥� to build understanding, which she says is supported by research.

鈥淲hat separates the high achievers is that they are able to look at numbers in different ways, break them apart, see numbers inside numbers,鈥� Boaler said. 鈥淭he low-achieving students are just trying to remember memorized facts.鈥�

Disclosure: The Bill & Melinda Gates Foundation and the Overdeck Family Foundation provide financial support to Accelerate and 蜜桃影视.

The release, from the National Center for Education Statistics, shows that with nearly 50 million students, enrollment was still 2% less than 2019 figures. Only Idaho and North Dakota saw enrollment increase about 2% over that time period, while multiple states, including California, Mississippi and New York saw declines of at least 5%.

The data confirms earlier state-level figures and pointing to the loss of students from traditional school districts and a shift toward private schools, homeschooling and newer models, like microschools and hybrid programs. Those trends added to large-scale declines in the number of school-age children that predate the COVID era. All combined, experts say, most districts shouldn鈥檛 expect to see growth anytime soon, and many have already announced school closures.

鈥淭his national demographic decline has first-order implications for whether many schools can reasonably expect enrollment to rebound,鈥� said Thomas Dee, an economist at Stanford University who tracks pandemic-era enrollment data. As , he added, 鈥淲e are seeing its implications for schools 鈥� enduring enrollment loss and the corresponding pressure to close schools and layoff staff.鈥�

In December, NCES released data showing a small uptick in over the same time period that public schools saw their largest declines. Later this year, officials will release newer data on students attending private schools as well as those who are homeschooled. But even those figures could leave some questions unanswered. 

鈥淲e have reasons to expect to see an increase in the number of students being home educated, even though a lot of home education may be unreported,鈥� said Sofoklis Goulas, a researcher at the Brookings Institution who has also analyzed enrollment data. 

Despite enrollment remaining relatively flat between 2021 and 2022 鈥� less than a 1% increase 鈥� there was still a lot of variation at the state level. Louisiana saw the most growth, with a 5% increase, while several states, including California, Colorado, Illinois and New York, saw declines of at least 4%.

First graders in Florida鈥檚 Broward County schools who participated in the program, called , saw more substantial gains in reading fluency than those who didn鈥檛 receive the support, according to the study. They were also 9 percentage points less likely to be considered at risk on a district literacy test.

Chapter One, which combines one-on-one tutoring with computer-based activities, also costs a fraction of other programs 鈥� about $500 annually per student 鈥� compared with programs that . That aspect could make it easier for districts to continue providing students with that support once federal relief funds expire later this year. 

The program鈥檚 鈥渟hort burst鈥� model 鈥渓everages all the knowledge that we have about what works to help children learn to read,鈥� said Susanna Loeb, who leads the National Student Support Accelerator at Stanford University, a leading tutoring research center. 

Chapter One curriculum materials are based on solid reading research, Loeb said. The fact that students work with the same tutor all school year and that the format fits young children鈥檚 short attention spans are also strong features of the program, she added.

鈥淏y embedding a well-supported tutor in the classroom, they are giving students the personalized, relationship-driven instructional approach that really seems to work,鈥� she said. 

As districts continue to look for ways to boost reading scores and curb pandemic losses, the findings strengthen the case for incorporating tutoring into the regular school day. The model demonstrates that even brief contact with a trained tutor who focuses on specific phonics skills can help struggling readers reach grade-level goals. The findings, based on data from over 800 students in 13 Broward schools, build on promising results from the 2021-22 school year. About half the students, then in kindergarten, were randomly assigned to receive tutoring while the rest got business-as-usual instruction. 

In the 2022-23 school year, first graders who continued in the program were more likely than their peers to develop basic reading skills, such as accurately decoding short consonant-vowel-consonant words and those with a silent e on the end. The researchers will continue to follow the students through third grade. 

The program is considered a 鈥減ush-in鈥� model because the tutors, called early literacy interventionists, meet with students during the school day, generally at a table in the back of the classroom. Using a tablet computer, they lead students through short, scripted lessons where students need help 鈥� on specific letter sounds, blends or sight words, for example 鈥� and calculate the number of correct words read per minute. 

Children making adequate progress might only see their tutor a couple times a week, while those who are far behind receive daily sessions. Later in the day, students spend another 15-20 minutes practicing with the same Chapter One software. 

鈥淚t kind of runs itself,鈥� said Ingrid Rosales, a literacy coach at Orange Brook Elementary, a Title I school in Hollywood, Florida. 鈥淚t doesn鈥檛 really interfere with our instruction.鈥�

Convincing school districts

That鈥檚 one key to the success of Chapter One 鈥� now serving over 20,000 students in 21 districts across 14 states, said founder Seth Weinberger. Tutors run their sessions while the teacher is either leading the whole class in a lesson or as students work in small groups. Over the course of the day, a tutor might meet with at least 25 students across multiple classrooms.

鈥淢ore important than the affordability is how nondisruptive our program is,鈥� Weinberger said. 鈥淭hat made it much easier for us to convince school districts to allow us to do high-impact tutoring during the school day.鈥� 

Even effective high-dosage programs often run a full 30 minutes a few times a week, making them . And that when programs operate after school, participation lags due to a lack of transportation and scheduling conflicts.

鈥淥nce you move the sessions to after school, you really lose the consistency of it,鈥� Weinberger said. 

Weinberger founded the Chicago-based program 30 years ago and developed the software to give students practice on basic reading skills. But the pandemic fueled greater demand. The nonprofit, which also works with schools in Canada and the United Kingdom, had 100 employees two years ago and now has 500. 

The original version relied on teachers to blend the mini lessons into their instruction, an approach that wasn鈥檛 always successful. Shortly before the pandemic, Weinberger鈥檚 team substantially changed the model by having former teachers lead the sessions. 

鈥淚t was incredibly successful,鈥� he said. But they also found that it could take two years of 鈥渋ntensive work鈥� to help a student from a low-income family become a strong reader. 鈥淲e said, 鈥楾here’s no substitute for this, but we’re going to have to create our own corps of tutors.鈥欌��

鈥楾hat time adds up鈥�

Other successful tutoring providers have reached the same conclusion 鈥� that expecting teachers to also manage tutoring sessions is unrealistic.

鈥淚deally, this type of work shouldn鈥檛 fall on their backs,鈥� said A.J. Gutierrez, co-founder of Saga Education, which trains recent college graduates, retirees and career-changers to tutor middle and high school students in math. A 2021 study showed that the sessions, also offered during the school day, helped students score higher on tests, get better grades and pass classes at higher rates. 

In Broward, teachers were initially skeptical about having tutors in their classrooms. Tutors weren鈥檛 part of the school staff, and teachers were unsure how easily students would transition from group instruction to their individual sessions.

But it only takes a few weeks for students to pick up the routine. When their session is over, they go tap the next student to let them know it鈥檚 their turn, said Hensley Philogene, who tutors at three Broward schools. After graduating from Florida International University, he joined the program because he was impressed with its track record.

One English-learner鈥檚 progress especially stands out to him. Their sessions had a 鈥渞ocky start鈥� because the student didn鈥檛 know any English and Philogene didn鈥檛 know Spanish.

鈥淵ou think to yourself, 鈥楬ow will I be able to connect with this student?鈥欌�� he said. But by March, the boy was 鈥渁ble to not only read the words, but [read them] fluently.鈥� 鈥�

Teachers were also more supportive once they began to see the impact.

Forty-five percent of students who received the tutoring were considered at risk for reading problems, compared to 54% of those who did not receive the intervention. And even though higher-scoring students didn鈥檛 remain in the study sample through first grade, the Chapter One group still made stronger gains than those without the program.
鈥淭hose five minutes can be very meaningful,鈥� Rosales said. 鈥淪eeing the same person every day 鈥� over time, that time adds up.鈥�

Not only is the program 鈥� called 鈥� targeted to students who to learn remotely during the pandemic, but the study was conducted by experts who typically advocate for in-person tutoring.

鈥淚 was nicely surprised,鈥� said Susanna Loeb, a Stanford University education researcher and leader of the , which has been tracking efforts to expand high-dosage tutoring. 鈥淭he trick is to get [tutoring] to as many students as we possibly can. Being able to do it virtually could really help in the scaling and expansion of this kind of intensive, individualized attention that many students need.鈥�

The evaluation, conducted in 12 Texas elementary schools as part of the Uplift Education charter network, found that over 1,000 K-2 students in the program scored higher on literacy tests than students without the extra support. The results translated into 26 extra days of learning in letter sounds for kindergartners and 55 extra days on decoding for first graders with a one-on-one tutor. Second graders did not benefit as much from the intervention.

While the virtual program was still less effective than in-person tutoring, the model could be a breakthrough for schools in rural areas and those that have struggled to recruit tutors, Loeb said. Districts’ pandemic recovery efforts have sometimes fallen short because they can鈥檛 find trained educators or volunteers to do the job. And and others has found that only a fraction of students who need extra help take advantage of on-demand virtual tutoring programs. 

OnYourMark Education, a nonprofit, is a contrast to the virtual models that researchers like Loeb have long criticized. It鈥檚 offered four times a week during the school day. The tutors, which include college students, retired educators and those who have worked for other virtual tutoring companies, receive training in the science of reading.

鈥淲e’ve put a stake in the ground that our focus as an organization is to really support students to become proficient readers by the time they reach third and fourth grade,鈥� said Mindy Sjoblom, a former Teach for America middle school teacher and principal who founded OnYourMark in 2021. 

But when the program started with Uplift as a pilot, she wasn鈥檛 sure if the tutors would be able to form strong relationships with young children remotely. 

鈥淲e had to get the timing right,鈥� she said. The 30 minute-blocks they started with didn鈥檛 work well. 鈥淗onestly, that was too long to expect a 5-year-old to sit and attend to anything, not to mention be in front of a screen.鈥�

Twenty minutes, she said, has proven to be the 鈥渟weet spot,鈥� allowing tutors to have informal chats with students 鈥� about what they had for dinner last night, for example, or how their basketball game went 鈥� before diving into a solid 15 minutes of work on decoding and fluency. 

OnYourMark now works with 22 schools in seven states, and Sjoblom said she expects to add more students before the end of this school year. Last fall, Accelerate, an organization funding effective tutoring programs, $250,000 to support the research effort. The organization is also a semifinalist for the , a $1 million award that recognizes successful education providers.

鈥楢 great option鈥�

Loeb鈥檚 team used two common assessments to evaluate the impact of the program 鈥� Dynamic Indicators of Basic Literacy Skills, or DIBELS, and MAP Reading Fluency from NWEA, a testing and research organization.

Kindergartners randomly assigned to OnYourMark recognized 3.5 more letter sounds per minute than students who didn鈥檛 receive tutoring. First graders鈥� mastery of sounds and decoding skills also improved.

Loeb said while the one鈥搕o-one model is clearly stronger, the program is still effective when students work in pairs with a tutor. 

鈥淭his is a great option when staffing is hard,鈥� she said, alleviating the need for tutors to commute and get acclimated to a school. 

The results among second graders were not significant. Sjoblom sees a few reasons for the disappointing outcomes. First, last year鈥檚 second graders were in kindergarten during the 2020-21 school year, when many schools were closed for the pandemic. They didn鈥檛 master a lot of the foundational skills that most kids get in kindergarten and first grade.

Older students struggling to read, she added, get embarrassed and have a harder time staying engaged with tutors remotely.

But Loeb said to get such results from a startup is still impressive. Yasmin Bhatia, the CEO of Uplift, added that future research will focus on the specific skills tutors should focus on with second and third graders.

OnYourMark, she said, has met the network鈥檚 needs in a few ways. First, it鈥檚 hard to find tutoring companies even willing to work with younger students. Most, she said, focus on the 鈥渢ested grades鈥� 鈥� third and higher. School leaders, she added, are 鈥減utting their best talent in those upper level grade levels.鈥�

Uplift, she added, serves a high-poverty population that typically would be unable to afford a private tutor. And when the network offered at-home virtual or afterschool tutoring, participation was inconsistent. Bhatia called OnYourMark 鈥渁nother way to support parents鈥� and ensure young readers are getting the extra help they need.

鈥淲e view it as such a high priority,鈥� she said, 鈥渢hat we made it a part of the school day.鈥�

Disclosure: Overdeck Family Foundation provides support to OnYourMark Education and 蜜桃影视.

Kindergarten counts continue to dwindle in many states 鈥� evidence of falling birth rates and an ever-growing array of options luring parents away from traditional public schools. Experts fear those trends, as well as a and the looming cut-off of federal relief funds, amount to a perfect storm for U.S. education.

The $190 billion in pandemic aid that was provided to schools allowed many districts to temporarily salve the loss of funds tied to falling enrollment and to staff and programs. Those funds dry up in 17 months. As budget deficits grow and housing costs drive families out of urban areas, education leaders are staring down a host of unpalatable options, from half-empty buildings to staff.

鈥淚鈥檓 not a pro-school closure guy. That鈥檚 the worst part of school reform,鈥� said Brian Eschbacher, an enrollment consultant and a former Denver Public Schools official. 鈥淏ut if anyone was holding out hope for a bounce back, we have put that to rest.”

The Parkrose School District, outside Portland, Oregon, is one of many grappling with a budget shortfall.

鈥淲e have some decisions to make in the next few months,鈥� said Sonja McKenzie, a board member in the district, where enrollment has fallen 12% since 2018. Now leaders might have to slash positions for special education assistants. Talk of layoffs is also surfacing in , and .

McKenzie went door-to-door last fall asking voters to approve a tax levy to fund 22 positions, reminding them that the district, where nearly 30% of students are Hispanic, heeded their call to hire bilingual family liaisons. Voters .

Some families, she said, have been 鈥減riced out鈥� of the area, heading east to Gresham or across the Columbia River to Vancouver, Washington, where they can find more affordable housing. Those areas, McKenzie said, have 鈥渂enefited from our challenges.鈥� 

Desperation and aspiration

蜜桃影视鈥檚 enrollment analysis is based on figures from 41 states provided exclusively by Burbio, a data company, and additional data from the nation鈥檚 20 largest school systems.

Since last year, enrollment has declined 2.5% in Chicago, 2.4% in Houston and 2% in Nevada’s Clark County, while New York and Los Angeles saw drops of just under 2%. The Hillsborough County district in Florida, which includes Tampa, and the Gwinnett County School District, near Atlanta, are the only two large districts where enrollment now exceeds pre-pandemic levels.

Large district enrollment trends from 2018-19 to 2022-23

The graphic below shows enrollment trends for the nation鈥檚 20 largest school districts. Divided by region, the breakdowns include changes in overall enrollment as well as in kindergarten. (Click here if you’re having trouble viewing the chart)

In California, which has seen a whopping 5% drop in its student population since 2020, the enrollment decline has slowed, according to . But the downward slope in birth rates and exodus of parents from high-priced areas has left district and charter leaders with limited options.

Summit Public Schools in California鈥檚 Bay Area 鈥� a well-established charter network that spawned an online learning platform still used by 300 schools nationwide 鈥� will at the end of this school year. 

Following a community and in Oakland, the local school board decided in January not to close several schools. Now, amid an ongoing , the board is reconsidering whether to because of enrollment decline.

鈥淭here is always this quality and convenience tension,鈥� said Lakisha Young, CEO of Oakland Reach, a parent advocacy organization. 鈥淓veryone wants a school in their neighborhood that they can walk their kids to.鈥�

But she called the emotional debate over closing schools a distraction from more important issues 鈥� namely that a majority of students aren鈥檛 . A third of families in the city , and some have moved further inland to Antioch or southeast to the Central Valley. 

鈥淚f people have the opportunity to move to other places that are slower and quieter and safer, they are going to do that,鈥� she said. 鈥淭hese decisions are not just made out of desperation, they are also out of aspiration.鈥�

鈥榊ou just come here鈥�

Some of those same aspirations are fueling a Republican push to give unhappy parents more options. Twelve states now offer education savings accounts, which allow families to use public funds to pay the costs of private school or homeschooling. Despite pushback from such programs take funding away from public schools and lack accountability, similar legislation has been introduced in several more states, including , and .

鈥淭his pandemic was the perfect incubation event that really caused homeschooling to thrive,鈥� said Bob Templeton, another enrollment consultant with , a housing market research company. 鈥淲e鈥檙e seeing this dramatic change in how we educate kids.鈥�

In Texas, where the legislature is currently , existing options like charters and homeschooling have contributed to a decline in what Templeton calls the 鈥渃apture rate鈥� 鈥� the percentage of children from a particular community attending their local public school. 

鈥淚f they鈥檙e down 200 kids in kindergarten and it doesn鈥檛 return, then in five to seven years, that district is going to be down several thousand kids,鈥� Templeton said. 鈥淵ou need to get ready to close schools.鈥�

Statewide enrollment shifts since 2021-22

*Click the circle next to state to see districts with the greatest enrollment gain, greatest enrollment loss and % change for state鈥檚 largest district. (Click here if you’re having trouble viewing the chart)

He consults for districts surrounding some of the state鈥檚 large urban systems and used to be able to reliably calculate that 100 new homes would result in 50 more students. Not anymore. 

He also monitors between districts. One school system he works with, Pflugerville, near Austin, took in 584 students from other systems this year. But almost 5,400 transferred out to both charters and other districts. Leaders have put off closing schools for now, which Templeton said just 鈥渒icks the can down the road.鈥� 

He and Eschbacher advise districts to stay competitive by designing school models that parents want. In some cases, that鈥檚 paying off. 

The San Antonio Independent School District has had success with a 2017 state law that provides incentives to partner with charters and nonprofit organizations to run schools. 

Rebecca McMains decided to enroll her daughter in one of them, Lamar Elementary, after considering close to 10 public, private and charter schools in the area. Because her daughter has disabilities and an 鈥渆laborate鈥� special education plan, the choice wasn鈥檛 easy.

鈥淚 knew I was going to be heard at Lamar. They are very parent-focused,鈥� said McMains. She said staff members respond to her texts and don鈥檛 push back when she has a request, like having a nurse accompany her daughter on a field trip to NASA. 鈥淚鈥檓 now being thanked for my advocacy.鈥�

But some parents have found their local public schools loath to accommodate the needs of those they are used to seeing as a captive audience.

Jana Wilcox Lavin, a Las Vegas mom, runs Opportunity 180, a nonprofit that supports school choice and formerly led a that converted low-performing schools into charters. Nonetheless, she was willing to consider her Clark County neighborhood school for her daughter, who starts kindergarten in 2024.

When she called the local school to ask for a tour, officials turned her down, citing concerns about student privacy. She turned to a district administrator, who said she could visit the building but not observe classrooms. Spokesman Tod Story said that while no policy prohibits parents from visiting schools, officials 鈥渆rr on the side of caution to protect our students.鈥�

 Lavin said she just wanted to make a well-informed choice.

鈥淲hen I asked how I should assess if the zoned school was a good fit,鈥� she said, 鈥淚 was told, 鈥榃e are your neighborhood school. You just come here.鈥� 鈥�

An 鈥榓bsolute asteroid鈥� 

That鈥檚 less true than ever before. The options available to families have expanded so rapidly that researchers are struggling to keep up.

Counts of how many students are homeschooled are and private school enrollment figures can be a year or two behind. That鈥檚 one reason Thomas Dee, a Stanford University education professor who tracks enrollment trends, was unable to account for of students who left public schools. 

That uncertainty makes it hard to tell whether the American school system is experiencing temporary chaos or a more permanent sea change.

Nat Malkus, the deputy director of education policy at the conservative American Enterprise Institute, called the pandemic an 鈥渁bsolute asteroid鈥� of a disruptive event. Still, he doesn鈥檛 expect ESAs or other emerging models to cause as much damage to the public education system as predict.

鈥淚t鈥檚 hard to overestimate the incumbent鈥檚 strength,鈥� he said.

That鈥檚 the case in Florida, where enrollment grew 1.3% this year and the Hillsborough district expects to keep building schools for years to come to accommodate growth. 

In states with declining numbers, like Oregon, district leaders are more wary. School choice hope to get an ESA initiative on the ballot next year, but McKenzie, the Parkrose board member, is concerned such a program would hobble district schools that are already strapped for cash.

鈥淚 can understand a parent may feel like they have a better option,鈥� she said.鈥淏ut it creates a divisive system of who has the resources and who doesn鈥檛. Less resources for the classroom impacts the whole community.鈥�

It was the year that Dr. Anthony Fauci, America鈥檚 most prominent public health authority, declared that the country was 鈥�,鈥� as COVID case rates plummeted from their Omicron highs. By the fall, President Biden was with that sentiment, noting that most people had laid down their masks and returned to something like normal. 

And around the possibility of winter surges in American schools, the most visible hallmarks of the COVID era have at last receded. The lurching progression from in-person to virtual classes is over, following an explosion of school exposures last winter. Mask mandates, social distancing, and endless disinfectant wipes are also predominantly a thing of the past, with virtually all children approved to receive vaccines. 

But in terms of the pandemic鈥檚 impact on education, it鈥檚 still only the end of the beginning. With each month, new findings emerge revealing more about what remote instruction did to learning and how families reacted. The potentially lifelong shadow the virus has cast over K-12 students 鈥� from how babies develop speech to what today鈥檚 adolescents will earn decades from now 鈥� is largely mysterious. 

Previous editions of this list have covered the wider world of education policy and research: issues like school financing, choice, accountability, and testing. This year, 蜜桃影视 is focusing exclusively on the lessons of the COVID era 鈥� one that is now passing from the scene 鈥� and the questions that remain in its wake.

Here, laid out in charts, maps, and tables, are 14 discoveries that changed how we think about schools in 2022.

The scope of learning loss

By the end of last year, a steady trickle of research had already begun to reveal the harm wrought by prolonged school closures and the transition to virtual instruction. But this fall brought the most definitive evidence yet of the scale of learning lost over more than two years of COVID-disrupted schooling: fresh testing data from the National Assessment of Educational Progress, sometimes called the Nation鈥檚 Report Card, pointing to severe declines in core subjects. 

The unprecedented drop in math scores, which fell by an average of eight points for eighth graders and five points for fourth graders, was especially disturbing. But reversals in literacy were also notable, with sizable increases in the number of students testing below even the 鈥渂asic鈥� level of reading proficiency. What鈥檚 more, the results affirmed dismal findings from NAEP鈥檚 鈥淟ong-Term Trends鈥� test 鈥� an earlier version of the exam that has been administered since the early 1970s 鈥� showing that the pandemic set back nine-year-olds鈥� performance in math and reading to levels last seen two decades ago. 

鈥淲e鈥檙e seeing a lot of that very long-term progress completely erased over the course of a couple of years,鈥� said Dan Goldhaber, a University of Washington professor, of the long-term results.

As many experts warned, additional research has also made clear that the academic damage of COVID was not shared equally. NWEA, the nonprofit testing group whose MAP exam has proven an invaluable assessment tool throughout the pandemic, released a study in November indicating that already-wide achievement gaps in elementary classrooms have grown between 5 and 10 percent in the last few years. Those disparities grew, NWEA analysts specified, because of slumping achievement among struggling students. 

College entrance exams contributed yet another dispiriting perspective, with average scores on the ACT slipping below 20 for the first time since the presidency of George H.W. Bush. Only about one in twelve test-takers from low-income families met standards of college readiness across all of the test鈥檚 four subjects.

In 2022, researchers, educators, and the public discovered the full extent of what COVID did to K-12 learning. 2023 will provide a test of how quickly that learning can be restored 鈥� and how seriously we are approaching the problem.

The geography of remote learning

Multiple studies have identified a strong association between academic backsliding and time spent in remote learning. And while different states and districts switched back to in-person instruction at different speeds, a disturbing commonality emerged: The least-advantaged kids were usually the slowest to return to the classroom.

co-authored by experts at NWEA, the CALDER Center at the American Institutes for Research, and Harvard鈥檚 Center for Education Policy Research used data from over 2 million students to show that 鈥� whether in states that reopened schools relatively quickly, like Florida, or those that stayed remote much longer, like Virginia 鈥� schools serving the highest proportions of low-income students spent the most weeks remote during the 2020鈥�21 academic year. Notably, however, the socioeconomic gaps in exposure to virtual teaching were much larger among the group of predominantly blue states that tended to reopen more hesitantly. In those states, high-poverty schools spent more than two additional months in Zoom classrooms than low-poverty schools. 

Harvard economist and study co-author Thomas Kane observed that the greater prevalence of remote learning among poor students, who are already less likely to succeed academically than their better-off peers, could be an additional driver of achievement gaps for years to come. In an interview with 蜜桃影视, Kane said that the academic recovery interventions planned by school districts were 鈥渘owhere near enough鈥� to compensate for COVID鈥檚 toll.

鈥淏ased on what I鈥檓 seeing, most districts are going to find that students are still lagging far behind when they take their state tests in May 2023,鈥� Kane said.

But was the public convinced by the reams of detailed and well-intentioned research on the results of online learning? Public polling suggests that the answer is ambiguous. At least 鈥� albeit one conducted before much of the research on learning loss was released 鈥� indicated that Americans prioritized curbing the pandemic鈥檚 spread over keeping schools open.

Poorer districts lost the most

Few doubt that some amount of learning loss is linked to the hasty and unplanned adoption of remote instruction. How much is still ambiguous, however. released in October 鈥� devised by Harvard鈥檚 Kane and the eminent Stanford sociologist Sean Reardon, among others 鈥� leveraged a combination of state test scores and federal NAEP results to deliver a granular, district-by-district overview of the pandemic鈥檚 academic impact.

While the researchers found that academic performance in predominantly in-person districts held up much better than mostly remote districts within the same state, they also stipulated that school closures were not 鈥渢he primary factor driving achievement losses鈥�; some states that spent much of the pandemic open as usual, such as Maine, sustained far greater score declines than those that saw widespread closures, such as California. And beyond the question of remote-versus-in-person, it is clear that districts with greater concentrations of poor students experienced the worst academic effects over the last few years.

In districts where 70 percent or more of students qualified for free or reduced-price lunch, average math performance fell by 0.66 grade levels. By contrast, in districts where fewer than 39 percent of students qualified for free lunch, only 0.45 grade levels of math achievement were lost. Above all, the ultra-local look at test scores showed a startling amount of variation in how different school districts experienced the same event; in reading, almost 15 percent of all students were enrolled in districts where achievement actually grew during the pandemic.

Enrollment fell as families fled 

The pandemic left an impact on schools far beyond its blow to student achievement. Due to a combination of public dissatisfaction, increased mobility, and economic upheaval, families withdrew from their public schools in unprecedented numbers 鈥� as many as 1.5 million during the 2020鈥�21 school year, or about 3 percent of all public K-12 enrollment, according to a 2021 report from NCES.

Further scholarly investigation has unearthed the important role that learning modality played in that flight. According to a comprehensive report from the conservative American Enterprise Institute, the districts that spent the most time remote throughout the first pandemic school year lost at least 500,000 more students than they would have if they had stayed open during that time. And in the period that followed, fewer students returned than did to districts where campuses mostly operated in-person. 

The findings suggested that widespread loss of students was not just 鈥減andemic-related; it was pandemic-response related,鈥� Nat Malkus, AEI鈥檚 deputy director of education policy, told 蜜桃影视鈥檚 Linda Jacobson. 

Meanwhile, enrollment trends detected this spring by the data company Burbio showed that major urban districts continued losing students through the 2021鈥�22 school year. Only a handful of states examined by the organization during that time saw an enrollment increase of more than 1 percent compared with the previous year.

The youngest weren鈥檛 spared

While we鈥檝e gained a better empirical understanding of how K-12 students鈥� lives and learning trajectories were altered by COVID, it will be years before we fully grasp the ways in which the youngest Americans were affected. But a provocative study of child development and language acquisition has already given cause for alarm.

Using LENA 鈥渢alk pedometers鈥� 鈥� a that measures the number of spoken interactions occurring in the vicinity of young children, as well as their own vocalizations 鈥� researchers at Brown discovered that babies born after July 2020 produced fewer vocalizations and demonstrated slower verbal growth than comparable children born before 2019. The younger group of babies also experienced slower growth of white matter 鈥� subcortical nerve fibers that facilitate communication between different regions of the brain 鈥� perhaps the result of hearing fewer words spoken and engaging less often with their caregivers. 

If the cognitive development of young learners was slowed by the extraordinary social isolation imposed by daycare closures and lockdowns of public spaces, it will produce unavoidable consequences for schools in the next decade.

Old before their time

Even as social and intellectual growth was apparently slowed for some infants and babies, psychologists warn that the compounded stress of the last few years may have harmfully accelerated the maturation process for older kids.

A slew of surveys highlight newly elevated levels of student stress, the product of public health worries, economic anxiety, and even domestic abuse. But a recently published offers proof that those factors actually changed the neurobiology of some adolescents. Examining MRIs of 128 matched subjects 鈥� half measured before and half after the pandemic began 鈥� a team of psychologists found that the group assessed after COVID demonstrated higher 鈥渂rain age鈥� than their chronological age and experienced faster growth in the amygdala and hippocampus, areas of the brain that regulate fear, stress, and memory.

Such sped-up aging has historically been seen in cases of household trauma and neglect, and its consequences can include decreased capacity across a range of intellectual functions. Follow-up scans are already planned to assess whether the process has been remediated.

Teachers under strain

Adults in schools have shown their own signs of exhaustion. In a survey of nearly 4,000 K-12 teachers and principals conducted by the RAND Corporation, about one-third said they intended to quit their jobs, a significantly higher proportion than it found during the chaotic pandemic months of early 2021. 

That figure almost certainly doesn鈥檛 betoken a future exodus from the profession; educators have historically been much more likely to say they intend to leave than to ultimately act on those plans. But it could mean that large numbers will stay in their jobs past the point of burnout, their effectiveness permanently dimmed. On average, the poll found that the teachers and principals were more than twice as likely to report experiencing frequent, job-related stress than other workers.

Teachers were also twice as likely as comparable adults to say they were not 鈥渃oping well鈥� with their stress. While the most commonly cited contributing factor was the task of addressing learning loss, some school employees also complained of staff shortages and the difficulty of managing their own childcare responsibilities. 

Social shuffle

It shouldn鈥檛 come as any surprise that young adults鈥� personal relationships, no less than their academic prospects, were fundamentally changed by months spent away from their peers. 

In some ways, those changes were positive: According to a June poll released by Pew, 45 percent of American kids between the ages of 13 and 17 said they felt closer to their parents after two years of disrupted schooling. But sizable minorities also reported feeling less close to friends, classmates, teachers, and extended family, a web of social connections that might have proven vital during a lengthy period of difficulty. 

Somewhat surprisingly for a survey administered over two years after the emergence of COVID, nearly 20 percent of the teen respondents said they had not attended classes exclusively in-person during the spring of 2022 (a time of somewhat elevated virus case rates). About two-thirds said they would prefer a return to entirely in-person schooling in the future.

Future earnings endangered

The downstream consequences of thwarted or deferred academic success are destined to include financial disadvantages; after all, today鈥檚 underserved pupils are tomorrow鈥檚 underprepared workers. But until the fall release of NAEP, it was difficult to produce a broadly shared measure of American students鈥� stifled progress. 

With the arrival of those scores, Harvard economist Kane 鈥� him again 鈥� and Dartmouth professor Douglas O. Staiger immediately calculated a projection of how much potential income could be lost due to diminished math learning among eighth-graders since 2020. Based on the historical correlation between math gains on NAEP and professional earnings growth, the figure they reached was astounding: $900 billion of future earnings, if the declines in learning were to remain permanent for all students in the United States.

鈥淲hen there are improvements in scores, those kids coming out of school are going to have better outcomes later in life,鈥� Staiger told 蜜桃影视. 鈥淎nd we can infer from this recent decline that all the cohorts in school now are going to do a bit worse than we expected.鈥� 

The paper was one of a series of analyses focusing specifically on the drop in math knowledge, which appears to have been particularly significant. But the extended disruption to literacy instruction left a substantial mark as well, particularly among students at the beginning of their reading careers. Amplify, a curriculum provider, released data this fall showing that 4 percent fewer second graders and 8 percent fewer first graders are reaching grade-level reading goals than in 2019; meanwhile, almost one-third of third graders were assessed as needing 鈥渋ntensive intervention.鈥�

Those bleak findings echo the results of Curriculum Associates鈥� i-Ready assessment, which revealed that the percentage of elementary students reading below grade level grew between 2021 and 2022. That subgroup of students, sometimes called the 鈥�COVID cohort,鈥� is running out of time to get back on track.

Costs of recovery

The havoc inflicted by the pandemic is now an inescapable fact for schools, families, and public authorities to deal with. But what鈥檚 it going to take to surmount the considerable educational challenges and get kids back on track?

The federal government has allocated roughly $190 billion in relief funding to states for that purpose. But , that amount won鈥檛 be sufficient to get the job done. The true cost, they say, will fall somewhere between $325 billion and $930 billion, huge sums that include not only the pedagogical resources to restore lost learning opportunities from the last several years, but also the out-of-school interventions that power so much of the academic growth that goes on inside classrooms. 

There is no indication that anywhere near that level of funding 鈥� or even any further money at all 鈥� is coming. In the meantime, school districts are only required to spend 20 percent of their federal aid on learning recovery. 

Latino students take a hit

Children of all backgrounds were bruised by the effects of shuttered schools, but among them, Latino students are notable for having recently enjoyed sustained academic momentum. As their share of the national student body has increased to nearly 30 percent, they have also seen rising achievement scores and post-secondary outcomes compared with their parents’ and grandparents’ generations.

COVID put that progress on pause, according to from the advocacy organization UnidosUS. After leaping from 71 percent to 82 percent over the last decade, the on-time high school graduation rate for Latino students fell slightly in 2021. Worse still, the rate of college enrollment for Latino freshmen shrunk by 7.8 percent between the spring of 2020 and 2021. That figure bounced back somewhat over the next academic year 鈥� along with rates of college-going for most Americans 鈥� but still fell below the pre-pandemic norm.

The particular stumbles experienced by Latino kids have explanations that both precede the pandemic and are directly linked to it, the report found. Long before 2020, Latino households were less likely to report having a computer or high-speed broadband in the home. Meanwhile, Latino students were disproportionately likely to be enrolled in low-income schools, which were themselves more likely to stay remote longer during the pandemic.

Explosion of absenteeism

Along with the surge of full-on disenrollment from schools, a shocking number of K-12 students spent the last few years missing day after day of instruction. Just how many days of absence is difficult to know precisely, however, because of ambiguities in the way attendance figures were collected during the COVID era.

An released this fall indicated that over 10 million students were chronically absent (i.e., missing over 10 percent of the school year) in 2020鈥�21. That would be an increase of more than 25 percent relative to the pre-pandemic norm, but from Johns Hopkins University and the nonprofit group Attendance Works, it is also very likely a serious underestimate. Because of challenges in knowing which students 鈥渁ttended鈥� all of their virtual lessons (versus simply logging into Zoom and then logging off, for instance), statewide absence counts in the NCES figures sometimes vary widely from district-level reporting.

Based on the early release of more detailed 2021鈥�22 figures from California, Connecticut, Ohio, and Virginia, the authors wrote, it is reasonable to predict that as many as 16 million kids were chronically absent last school year, a doubling of the pre-pandemic number. 

The teacher exodus that wasn鈥檛

Were American schools plagued with teacher absences this year, or not? It was a question that captivated news sources, but also divided education experts, because it contained an even thornier question within it: If the supply of teachers remains mostly steady, but demand for them spikes, are they truly at a deficit?

In spite of widespread fears that veteran teachers were quitting in huge numbers as a reaction to the pandemic, no mass departure ever took place, according to a paper by Brown economist Matt Kraft. Turnover actually fell slightly in the summer of 2020 and stayed within the typical annual range the next year. But weak hiring during the first few months of the pandemic may have contributed to higher-than-usual vacancy rates, perhaps triggered by fears of Great Recession-style budget cuts that never materialized.

In fact, a windfall of federal cash followed instead, leading districts to add new jobs in late 2020 and 2021, and the resultant hiring spree has indeed made candidates for teaching positions hard to find. But even that phenomenon isn鈥檛 true everywhere, since numbers differ widely across state lines. According to a paper released this summer, Mississippi鈥檚 rate of vacancies per 10,000 students is more than 68 times higher than that of Utah. 

Hopeful signs

As the long legacy of COVID grew clearer, research in 2022 gave the education world plenty of reasons to worry. But it has also contributed some hopeful signs of renewed progress in schools. 

The good omens aren鈥檛 popping up everywhere, but some are to be found in state-level testing, which has resumed around the country after being suspended for at least the first pandemic year. According to Tennessee鈥檚 state exams, the number of students meeting or beating grade-level reading standards rose from 29 percent in 2020鈥�21 to over 36 percent in 2021鈥�22. In all, more than three-quarters of the state鈥檚 school districts reported reading scores higher than were seen in the pre-pandemic period. 

鈥淲e are seeing this broadly across the state, and across district types 鈥� urban, rural and suburban,鈥� Tennessee Education Commissioner Penny Schwinn told 蜜桃影视鈥檚 Beth Hawkins. 鈥淲e are really, really proud of what our districts have done.鈥�

Several other Southern states have begun to make their turnaround, with Mississippi a particular standout. This of 2021鈥�22 testing data showed average scores in math, English, and science nearing or exceeding 2019 levels, while performance on the U.S. history exam skyrocketed compared with 2020鈥�21 (the first in which it had been given). Just as notably, 鈥� a state-mandated test that students must pass to progress to the fourth grade 鈥� fell by only .6 percentage points between 2019 and 2022. 

The by Stanford University鈥檚 Center for Research on Education Outcomes, focuses on pre-pandemic performance, looking at the 2017-18 and 2018-19 school years.

Compared to an average student in the state, students in San Antonio overall showed weaker learning gains in reading, but similar gains in math during that time. 

The study found that in 2017-18, charter school students received the equivalent of 10 more days of instruction in reading, but five fewer days of instruction in math, compared to their peers elsewhere in Texas. 

The following year, however, charter school students outshone others, getting the equivalent of 51 more days of instruction in reading and 46 in math, compared to peers statewide. 

鈥淐harter performance seems to show (an) increasing upward trend鈥� over the two years of the study, said researcher Won Lee. He noted that, as far back as 2014, data shows consistent growth for San Antonio charter schools, in both reading and math. 鈥淪o that’s encouraging, I think.鈥�

Brian Whitley, spokesman for the Texas Charter Schools Association, attributed a portion of the performance boost to the diversity of educational opportunities offered by聽San Antonio charters, such as schools focused on STEM, classical education and rich liberal arts curricula. He also attributed some of the performance gains to charters鈥� longevity in the state, where the first charter law went into effect in 1995.

As a result, he said, 鈥渃harter schools get increasingly good at zeroing in on what different kinds of students need, and refining their curriculum, refining their instructional methods.鈥� 

The district did not respond to several requests for comment on the new findings, but Nora J. Walsh, who leads the George W. Brackenridge Foundation, welcomed them: 鈥淎s a foundation committed to supporting charter schools throughout San Antonio,鈥� she wrote in an email, 鈥渨e were thrilled to see that the data in this study supports what we knew to be true, that charter schools offer a valuable option for students and their families looking for an education beyond the one designated by their zip code.鈥�

A nationwide look at school performance

The new study is part of an ongoing CREDO series examining school performance , first published in 2019. Since then, CREDO has taken on examinations of achievement in another five cities.

Lee said it鈥檚 difficult to generalize overall charter performance from the San Antonio findings, due to heterogeneity of schools and students, among other issues, across the 11 cities. But he noted that CREDO is working on a that covers 25 states and may be able to zero in on the overall effect of charter schools on achievement when it鈥檚 published early next year. 

As in several other cities, CREDO researchers in San Antonio examined achievement in so-called Innovation Schools, district-managed public schools with strategic plans that allow waivers to specific district policies, state statutes, and collective bargaining agreements. These schools were introduced in the city in the 2018-19 school year, so researchers had no data for them from 2017-18.

Lee likened these schools to open-enrollment charter schools that are also accessible to students outside of the district. Priority admission, however, is given to students within district boundaries. 

The data suggest that the sector has shown a few growing pains: Students in San Antonio Innovation Schools in 2018-19 showed weaker growth in reading than the state average, but similar growth in math. 

The researchers found that Innovation Schools students receiving the equivalent of six fewer days of instruction in reading and 13 fewer days in math, compared to peers statewide. 

Progress across demographic groups

Whitley, of the charter schools association, said San Antonio has seen heavy growth in the charter sector over the past few years. In the last school year, he said, the city鈥檚 metro area, which comprises several districts, boasted about 48,800 students, or just under 11 percent, enrolled in charters.

Between the 2018-19 and the 2021-22 school years, Whitley said, charter enrollment rose by about 11,000 students, or nearly 30 percent. 

He said about two-thirds of charter schools in the metro area had a wait list. Statewide, there are about 58,000 students on Texas charter school waitlists, he noted. 

He said San Antonio is also representative of large cities nationwide in terms of student demographics: Charters there enroll a larger share of students of color, low-income students and English language learners than traditional district schools. 

And these students are making some of the most noticeable gains, the CREDO data show: Black students in San Antonio charter schools got the equivalent of 42 more days of instruction in reading and 73 in math compared statewide. 

Likewise with Hispanic charter school students, who showed benefits equaling 49 extra days of instruction in reading and 36 in math. Students in poverty had similar findings: Low-income charter school students got the equivalent of 56 more days in reading and 48 more days in math.

Whitley noted that the new results are similar to CREDO results that researchers found over the past few years in Houston and Austin. 鈥淭hey’re all sort of showing the same thing when it comes to charter school students achieving all these additional days of learning, especially 鈥raditionally disadvantaged student subgroups.鈥�

What鈥檚 perhaps most notable: Charter schools offered eye-popping advantages for English language learners. In reading, these students received the equivalent of 95 extra days, more than half the typical school year.

For disabled students, the findings on charter schools were similarly promising, with charter students received the equivalent of 87 extra days in reading and 68 in math. Innovation School students actually received the equivalent of 16 fewer days in reading, but 55 more days in math. 

, president & CEO of the National Alliance for Public Charter Schools, welcomed the results, saying the gains are due to 鈥渢he flexibility and autonomy offered by public charter schools. The charter school model empowers teachers to provide innovative, high-quality instruction and gives them the autonomy to design a classroom that fits their students’ needs.鈥� 

Rees noted charters鈥� strong philanthropic support in San Antonio, adding, 鈥淐harter schools are led by dynamic individuals who have the flexibility to create a school culture that responds to the needs of the community and fosters student performance and parent satisfaction.鈥�

, released by Stanford University鈥檚 Center for Research on Education Outcomes (CREDO), focused on pre-pandemic performance, looking at the 2017-18 and 2018-19 school years.

It found that in the 2018-2019 school year, charter school students learned the equivalent of 64 more days of instruction in reading and 116 days in math, compared to their district school peers. Black charter school students had even bigger gains, with 86 more days in reading and 144 days in math relative to Black students in district schools. 

In a statement, Indianapolis Superintendent said the study 鈥減rovides another piece of critical data in our relentless mission for all schools to be better.鈥�

The findings reinforce the district鈥檚 belief that diving into data about academic performance at all schools helps educators 鈥渂uild on what works, and fix where we aren鈥檛 delivering for students,鈥� she said.

The findings showed that Black charter school students in Indianapolis had more growth in math than the average Black student statewide; they showed similar growth in reading. Similarly, Black students in Innovation Network Schools saw growth on par with peers statewide. are a group of 20 public schools in the city that enjoy complete, charter-like autonomy over academics and operations. While seven are actually charter schools, the remaining 13 are either new schools, strong district schools whose staffs are trying something new, or struggling schools that have been 鈥渞estarted鈥� with outside partners.

But Black students in traditional district schools performed worse than the typical Indiana student in both reading and math.

The new study, part of an ongoing CREDO series examining school performance in 10 cities, follows a 2019 study finding that growth in both reading and math was weaker than state averages in 2015-16 and 2016-17.

When researchers compared student performance citywide for the current study, they found that students at charters and Innovation Network Schools outperformed district peers across subgroups: Black charter school students saw stronger growth than district students in both reading and math, and Hispanic students in charter schools and Innovation Network Schools showed similar gains.

So did low-income students at charter schools. Similarly, English Language Learners in city charter and Innovation Network Schools saw better gains than district students.

Brandon Brown, CEO of , an Indianapolis nonprofit that has launched 41 schools, said one key to the charter sector鈥檚 success in the city is that the vast majority are locally grown, with leaders 鈥渨ho know Indianapolis.鈥� Most of those leaders, he said, are also people of color who directly reflect the racial backgrounds of students. 

The sector鈥檚 performance is 鈥渁 direct result of schools that are created and sustained relative to what our community wants and needs. And I do think that that’s pretty unique, when you look across much of the work that’s happening nationally.鈥�

Brown also noted that local officials look favorably upon charters 鈥� the mayor鈥檚 office is the largest authorizer in the city 鈥� and don鈥檛 see their growth as 鈥渁 zero sum game.鈥�

鈥淭here’s nowhere in the country where the school district and charters work as closely together,鈥� he said. 

Darius Sawyers, principal of Paramount Englewood, a 5th-8^th-grade school that鈥檚 part of the Paramount Schools of Excellence network, said the sector鈥檚 small scale allows him to collaborate regularly with other charter leaders, in a kind of ongoing principals鈥� consultancy. 鈥淲e’re talking best practices, we’re talking data, we’re talking, 鈥榃hat are you doing to move the data or move the needle?鈥欌��

He said being part of a local network has advantages. 鈥淓verybody’s right here,鈥� he said. 鈥淚f not in the same building, a block or two away.鈥�

Austin Hauser, director of Academic Accountability at Herron High School, said his small network of three Herron Classical Schools is 鈥渁bsolutely homegrown,鈥� founded by a local teacher with more than 30 years of experience. 鈥淚t was started really as a neighborhood movement.鈥�

Being homegrown, he said, allows teachers and administrators 鈥渢o focus on exactly what we need in our community 鈥�. We’re not worried about who we should be in Chicago or in Cincinnati or wherever the network may be located. We are in Indianapolis.鈥�

It鈥檚 a question that has been asked in Columbine, Newtown, Parkland, and elsewhere. And as the number of tragic episodes has climbed in recent decades, it has increasingly drawn the attention of experts studying the effects of trauma on students鈥� wellbeing. Spanning a variety of settings and drawing from the insights of diverse academic disciplines, their work points to substantial emotional damage trailing students who live through school shootings. The hopes of these children 鈥� measured in academic, professional, and psychological terms 鈥� are meaningfully diminished, along with the health of their families.

鈥淎 growing body of research finds that the costs of gun violence in American schools extend beyond the death toll,鈥� said Maya Rossin-Slater, a professor of health policy at the Stanford University School of Medicine who has carefully observed the aftermath of previous Texas shootings. 鈥淭he hundreds of thousands of children and educators who experience and survive these tragedies are likely to carry scars for years and decades to come.鈥�

Rossin-Slater is the co-author of looking at the survivors of 33 school shootings in Texas between 1995 and 2016, including those with or without fatalities. Using administrative data from the Texas Education Agency, and measuring the academic participation of individual survivors against students from a control group of demographically similar schools, the research team detected obvious short-term consequences from shootings: Affected students were more likely to be absent and chronically absent, and over 100 percent more likely to repeat a grade (though this probability rose from a relatively low baseline).

The authors next examined college enrollment and workforce records of students at eight Texas high schools that saw shootings between 1998 and 2006, comparing the trends of students enrolled both before and during the shootings against same-age students at control schools. Tenth and eleventh graders who lived through shootings became 3.7 percent less likely to graduate, 9.5 percent less likely to enroll in college, and 15.3 percent less likely to obtain a bachelor鈥檚 degree. Freshmen, sophomores, and juniors who experienced shootings were more likely to be unemployed between ages 24 and 26; those working by that age earned, on average, $2,350 less in annual wages than their peers, which implies a $115,000 reduction in lifetime earnings.

Evidence of those long-term ramifications can also be found in other recent studies. A analyzed the impact of violence on a broader sample of individuals who were between the ages of 11 and 17 when a school shooting occurred in their home county. Tracking responses to the CDC鈥檚 Behavioral Risk Factors Surveillance System (a nationwide survey querying the health of Americans in their 20s and early 30s) the authors found that girls who lived in the vicinity of school shootings tended to report a host of risky behaviors in adulthood, from increased drinking to driving without a seatbelt.

Boys also demonstrated clear effects 鈥� including a substantial uptick in smoking and the number of days they described themselves as receiving insufficient rest 鈥� and were generally less likely to say they were in excellent or very good health. Similar to the findings of the Texas paper, the authors found that girls in counties where school shootings occurred were less likely to be employed in early adulthood, while boys later earned less than their peers from other counties. Both boys and girls were less likely to be obese in later life, and more likely to be underweight. 

More evidence emerges from a study of the 2011 terrorist attack at Ut酶ya, Norway, the deadliest mass killing perpretrated by a single individual in modern history. at a summer camp, the majority under the age of 20; one poll showed that one in four of the country鈥檚 residents knew someone touched by the event.

The study, conducted by a team of mostly Norwegian researchers, used academic and medical records to pair children who lived through the attack with similarly aged peers who attended different schools, then divided their findings according to different age groupings. In all, they found that relatively young survivors (either 14 or 15 years old) scored vastly lower on standardized tests, while older survivors (between the ages of 15 and 18) were 20 percentage points less likely to complete high school. Relative to the average for the control group, exposed children of all ages made 60 percent more medical visits and received psychiatric diagnoses nearly five times more frequently in the period immediately following the killings.

Families of the survivors weren鈥檛 spared. Siblings also scored lower on state tests by roughly .2 standard deviations (a commonly used measurement illustrating the difference in any population from the statistical mean); a drop of that magnitude is much larger than most effects in education research. Parents were much more likely to visit a doctor, receive a mental health diagnosis, and take sick leave from work (28 percent more likely, in the case of mothers) after the Ut酶ya attack.

Study co-author Prashant Bharadwaj 鈥� a professor of economics at the University of California, San Diego 鈥� wrote in an email that the 鈥渂ig lesson鈥� to be taken from the study was that direct exposure to mass killings can cause enormous ripples even in a Scandinavian setting, where social policy and access to free health care is more generous than in the United States. 

鈥淣orway is a setting with incredible social safety nets: state-provided medical care, high-quality medical care, generous family leave policies, sickness leave, etc.,鈥� Bharadwaj said. 鈥淓ven within this context, the fact that we find large impacts on mental health for children and sickness absences from work for mothers suggests that in contexts like the U.S., where access to medical care and quality of social safety nets are weaker, the impacts can be much more severe.鈥� 

The medical toll on American children is on display in another study conducted by Stanford鈥檚 Rossin-Slater, who measured the impact of 44 school shootings between January 2008 and April 2013. Using information from the IQVIA Xponent panel, which tracks practitioner-level data on medical prescriptions, Rossin-Slater and her colleagues discovered a startling phenomenon: The monthly number of antidepressant medications prescribed to people aged 20 and under increased by over one-fifth in counties that saw school shootings with at least one fatality. The effect continued even three years after the murders occurred. 

Some variety did exist in the effects, however 鈥� the spike in antidepressant use was somewhat smaller in areas with higher concentrations of psychologists and social workers, who can offer behavioral treatment outside pharmacological intervention. Rossin-Slater said that this caveat made a case for providing more mental health resources to communities that lack them.

鈥淎s we mourn the horrific losses of children and teachers in Uvalde and in many other towns across America, we must ensure that our society provides lasting support and resources to the many survivors who are likely to continue to suffer. This need is especially critical in rural and lower-income areas, such as Uvalde, Texas, which tend to have limited access to mental health professionals and other supports.鈥�

A year after the nation鈥檚 schools experienced a historic decline in enrollment, new data shows that many urban districts are still losing students, and those that rebounded this year typically haven鈥檛 returned to pre-pandemic levels.

Whether families withdrew to enroll their children in online charters, school them at home or fled to far-flung suburbs with more affordable housing, the pandemic has triggered population shifts that could change the composition of U.S. school districts for years to come.

Data from Burbio, a company that tracks COVID-related education trends, offers the first look at the degree to which states and districts have recovered from a punishing year of lockdown and remote learning. Out of 40 states and the District of Columbia, few have seen more than a 1% increase compared to 2020-21, when some states experienced declines as high as 5%.

Flat enrollment this year 鈥渕eans those kids did not come back,鈥� said Thomas Dee, an education professor at Stanford University. 鈥淧arents were making these enrollment decisions last summer. There was still a great deal of uncertainty. Parents wanted stability for their kids.鈥�

shows that last year鈥檚 losses were concentrated in the early grades. Those who opted not to enroll their young children in public schools last year, or found an in-person option somewhere else, might never return for middle or high school, Dee said. 

While enrollment in many of the nation鈥檚 urban districts was already shrinking before the pandemic, school closures and economic upheaval forced many families to make decisions they might have put off otherwise.

Barring further pandemic disruptions, student population trends will likely return to their pre-COVID pace, Dee said, but added, 鈥淭he effects of the sharp, recent enrollment declines may be long-lived. The fiscal consequences will remain for some while.鈥�

New York experienced the sharpest decline, a 2% drop 鈥� more than 48,000 students 鈥� since last year. That鈥檚 on top of the previous year鈥檚 3% decline. Enrollment in Florida saw the biggest bounce at 4%, or more than 111,000 additional students 鈥� a reflection of higher birth rates, job growth and fewer COVID restrictions under Republican Gov. Ron DeSantis, experts say. 

Of the 10 largest districts in the nation, only Florida鈥檚 Orange and Hillsborough counties, home to Orlando and Tampa respectively, saw enrollment surpass pre-pandemic figures.

鈥淔lorida was continuing to grow when other states came to a plateau,鈥� said Susan MacManus, a political scientist from the University of South Florida. 鈥淭hings were open and you could still work.”

State data offers a glimpse of what will likely be further enrollment growth in Arizona, Florida and Utah 鈥� states with more affordable housing, growing tech sectors and outdoor living that became an important draw during COVID. At the same time, fewer people are moving to the Northeast from other states and countries, citing . 

District-level figures 鈥� provided exclusively by Burbio to 蜜桃影视 鈥� offer a richer picture of what happened to students after the pandemic began. The data, combined with state-level reports and interviews with district officials and parents, shows many urban districts lost students to growing exurbs. And some districts with no population growth added thousands of students in virtual schools.

Districts with enrollment loss could face tough decisions about layoffs and school closures in the near future. Meanwhile, smaller districts that are rapidly gaining students are struggling to hire staff and preserve the kind of close-knit environment that drew many parents in the first place.

鈥淭he pandemic kind of accelerated some of those pre-existing trends,鈥� said Alex Spurrier, an associate partner at Bellwether Education Partners, a think tank. While school closures forced many parents to look for other options, housing and rental prices were also pushing families out of major metro regions. 鈥淎ll you have to do is go to Zillow and see the year-over-year changes,鈥� he said.

In December 2019, Tanner and Miranda McCutchan relocated from northern California to Boise, Idaho 鈥� one of 10 metro areas that saw the most growth between 2020 and 2021, according to recent . That leaves two fewer children who will enter California鈥檚 schools in the coming years. Miranda stays home with 4-year-old Paige, who attends a Montessori preschool, and 18-month-old Emery, while her husband runs a glass company. 

鈥淲e couldn鈥檛 afford a house where we lived,鈥� she said. 鈥淚t was keep renting or move somewhere we could buy a place.鈥�

The fiscal cliff & 鈥楢rmageddon鈥�

In California, Burbio collected data only from Los Angeles, Oakland and San Diego. All three saw declines, due in part to California鈥檚 high-priced . With the state鈥檚 school-age population expected to keep over the next decade, district leaders are bracing for a to their budgets.

The Oakland Unified Public Schools offers a preview of what other districts with declining enrollment and birth rates will soon confront 鈥� the painful and unpopular decision to close schools. In February, the district, which saw a 5.6% enrollment decline compared to last year, decided it would over the next two years. Four others will merge or reduce grade levels.

In the Granite School District, near Salt Lake City, enrollment fell 2.4%, down to 60,371 this year, even though the state鈥檚 overall enrollment is up. 

The district has seen a decline in birth rates and an increase in families fleeing to 鈥渃heaper areas to build larger homes within [Salt Lake County],鈥� said Benjamin Horsley, chief of staff for the district, adding officials anticipate 鈥渓eveling out around 55,000 students.鈥� The district has already closed three schools and expects to shutter 10 to 14 more in the next five to seven years. 

Districts experiencing similar losses should have been making those tough calls before the pandemic, said Marguerite Roza, director of the Edunomics Lab at Georgetown University.

鈥淔ederal [relief] money is delaying it a year or two, and the fact that state budgets are healthy is delaying it a year or two,鈥� she said about closing schools. Roza advises a network of over 40 urban districts nationwide, the majority of which are shrinking. 鈥淔ederal money will run out, and enrollment for some of them isn’t isn’t going to come back. These cost factors are going to just slam down on people.鈥�

Los Angeles Unified, for example, saw a 5.9% decline this year and is expected to by fall of 2023. Superintendent Alberto Carvalho said Monday that he鈥檚 not yet considering closing schools, but added that at the end of his first 100 days 鈥� in about two more months 鈥� he will discuss 鈥渢echnical corrections鈥� and 鈥渂elt tightening鈥� measures to respond to the loss of students. 

He agrees with Roza about the dangers of the approaching fiscal cliff, and didn鈥檛 mince words about what would happen to the district if it didn鈥檛 turn enrollment trends around by the time federal relief funds dry up in 2024. 鈥淎rmageddon,鈥� he said. Then he added, 鈥淚t鈥檚 going to be a hurricane of massive proportions.鈥� 

The student population in the Clark County School District, which includes Las Vegas, began dropping about five years ago. Superintendent Jesus Jara attributes much of the decline to the growth of charter schools. 

鈥淭he anti-charter discussion 鈥� that was in the 鈥�90s. They鈥檙e not going away,鈥� Jara said. 鈥淭he discussion is how are we more flexible and how we are more agile for our communities.鈥� 

Despite declining enrollment, the district needs to build and renovate 33 schools to better serve its current population, he said. That includes breaking up some large, 4,000-student high schools to offer more 鈥渂outique鈥� and career-focused programs to compete with charters.

鈥楬as not slowed down鈥�

Districts with falling enrollment are strategizing how to keep the students they have. But accelerated growth comes with its own challenges, Roza said, putting pressure on leaders to act fast, especially if they need to recruit staff amid a nationwide hiring shortage. Schools might be 鈥渄igging deeper and deeper into applicant pools鈥� and not necessarily choosing the best candidates, she said. 

Liberty Hill Independent School District, northwest of Austin, Texas, didn鈥檛 lose students during the pandemic. Enrollment, at 5,539 last year, is now over 6,800 鈥� a 23% percent leap. It鈥檚 a bedroom community that just got its first H-E-B, a 鈥渂ig box鈥� grocery store, and is conveniently located near a toll road with easy access to Apple鈥檚 new complex near Austin. 

During the pandemic, the community 鈥渁ctually saw a 40 percent rise in residential home builds, and it has not slowed down,鈥� said Superintendent Steven Snell. The district has eight schools now and will open a ninth next year. 

Parents value the district鈥檚 small-town atmosphere and the sense that educators know their families well, he said 鈥� connections that could be hard to maintain as the district adds 1,000 students a year. Meanwhile, the district has raised salaries for substitutes because of shortages, and there鈥檚 a scarcity of available bus drivers, custodians and cafeteria workers.

鈥淲hen you have a salary that is causing you to live paycheck to paycheck, you鈥檙e going to jump ship for a little more money to survive,鈥� Snell said.

Many of the enrollment swings this year reflect the success of online programs in meeting the needs of families for consistency amid the pandemic鈥檚 many disruptions.

For some virtual charters, the enrollment spike was temporary. Oklahoma鈥檚 Epic One on One, an online program, had 17,106 students in 2019-20. Enrollment roughly doubled last year and is now down to 23,156, according to state data.  

鈥淢any parents decided to enroll their student in Epic once the pandemic hit, but it appears that trend has slowed with this year’s enrollment numbers,鈥� said Carrie Burkhart, a spokeswoman for the Oklahoma Department of Education.

But whether parents are concerned about COVID or found that online school better suits their children, virtual programs remain in high demand. 

South Carolina鈥檚 enrollment has increased almost 2%, due in part to 鈥渟kyrocketing enrollment in virtual charters,鈥� said Ryan Brown, spokesman for the state鈥檚 education department.

The student population in the Huntsville Independent School District, about an hour north of Houston, shot up 40% this year because it operates the Texas Online Preparatory School. And in Colorado, Harrison School District 2, near Colorado Springs, began a partnership with The Vanguard School, a virtual program and one of three charter systems affiliated with the district.

鈥淢any might see it as a public school district versus charter battle,鈥� said Harrison Superintendent Wendy Birhanzel. 鈥淲e believe this makes us stronger and responds to the needs of the community.鈥�

Homeschooling trends

While Burbio data offers an incomplete picture of where lost students have gone, others have been trying to fill in the missing pieces. The Census Bureau鈥檚 Household Pulse Survey showed that homeschooling jumped from about 5% of households to the fall after the pandemic began. By the start of this school year, it had settled back down to about 7%, according to August 2021 data.

Others have left for more established private schools. Michelle Walker, an Oregon mother who became an advocate for school reopening last year, withdrew her daughter from the Canby Public Schools, near Portland. She secured a spot 鈥� and financial aid 鈥� at a private school for fourth-grader MacKenzie. She also took out a loan and received money from family to help cover tuition.

鈥淚 drive 80 miles roundtrip every day to make sure she goes to a good school,鈥� she said. 鈥淚t would take a lot for me to put her back in public schools.鈥�

shows many other parents are following suit. According to Burbio, most districts in Multnomah County, which includes Portland, and nearby Clackamas County have seen enrollment declines this year. 

Some district leaders are still hoping to lure back students they鈥檝e lost. The Bernalillo Public Schools, north of Albuquerque, serves families in Pueblo and Hispanic communities, including many in multi-family households concerned about COVID risk. 

The district was the last in the state to lift its mask mandate. Superintendent Matt Monta帽o said he鈥檚 encouraged that enrollment, while still below pre-pandemic figures, has picked up slightly since last year. 

The district鈥檚 pre-K program, with 190 students at three schools, earned a five-star rating from the state education department 鈥� an accomplishment Monta帽o hopes will help recruit new students.

鈥淥nce we get them in our doors,鈥� he said, 鈥渢here’s no reason why they should leave us.鈥�

How will this chaotic school year have affected them?

Recent findings from the University of Pennsylvania warn that over the next three decades, our recent COVID-related U.S. school closures, as well as the shift to virtual schooling, could massively impact our national gross domestic product (GDP), putting a huge dent in future workers鈥� earning potential.

Get stories like this delivered straight to your inbox. Sign up for 蜜桃影视 Newsletter

The damage from all that reduced schooling could hurt productivity and shrink the U.S. economy 3.6 percent by 2050, economists say. The results will be even worse for workers鈥� personal earnings.

The new estimates come from the nonpartisan Penn Wharton Budget Model, an initiative that examines public policy through an economic lens. The suggests an expensive remedy: extend the school year.

Adding just one month of summer school, they say, won鈥檛 be cheap: about $75 billion, likely financed through the federal government taking on more debt. But they note that the $1.9 trillion COVID relief bill passed in March provides to K-12 public education, with about $22 billion already earmarked for summer school, extended school days, an extended school year, after-school programs and 鈥渙ther enrichment.鈥�

Spending billions on extending the school year, the economists say, could help mitigate learning losses, shrinking GDP loss about half a percentage point, from 3.6 to 3.1 percent. That smaller GDP reduction would produce a gain of $1.2 trillion over the next three decades, equal to about $16 for every dollar spent on more summer school.

That affordance 鈥済ives the kids a few extra hours for them that they wouldn鈥檛 have gotten otherwise,鈥� said Penn鈥檚 , the brief鈥檚 lead author. 鈥淲e include that in their productivity, in their learning capacity. And then, once they come to adulthood, that gets incorporated in their ability in the labor market 鈥� that is, the wages they were going to make.

Her team estimated the relative effectiveness of virtual versus in-person schooling using previous studies that looked at the math skills of students enrolled in the different learning modes.

They also estimated how much a year of learning loss corresponds to later productivity. Looking broadly at and federal , they estimated that graduates鈥� future labor income shrank by 10 percent for those who missed a year of middle school or high school. For those who missed a year of elementary school, it was even worse, 13 percent.

Recently approved federal aid could actually make the Penn prescription happen, in at least a few places. The American Rescue Plan includes for K-12 schools with high levels of low-income students. Districts must spend at least 20 percent addressing learning loss. States, which can hold on to 10 percent of the money the federal government gives, must also spend at least 5 percent of it on learning loss, and at least 1 percent on summer learning.

Thomas Dee, an economist and Professor at Stanford University, said he鈥檚 glad researchers are conducting analyses like this, but said the Penn analysis 鈥渟eems to embed the assumption that an extension to the school year will have the same effects as a school does on average.鈥� That may not be a valid assumption, he said.

Dee said the framing of the policy choices 鈥渟eems to preclude other options鈥� like tutoring and conventional summer learning programs. Extending the school year could also be difficult, since it requires schools to restructure curricula and figure out staffing, among other challenges. And it ignores well-researched summer and tutoring programs that are proven to support students, he said.

Dee and two colleagues last year looked at a long-established that serves low-income middle school students 鈥渁nd features unusual academic breadth,鈥� as well as a social emotional curriculum. The researchers found that participating in the program led to fewer unexcused absences, lower chronic absenteeism and suspensions, and a modest gain in reading scores.

Acosta, the lead author of the Penn analysis, said the 鈥渨hat-if鈥� of extending summer school presented an interesting exercise. Normally, economists would ask what happened if a group of students got more education. 鈥淚n this case, because of the whole COVID environment, it鈥檚 the opposite,鈥� she said. 鈥淲e want to know what happens if you give up one year of your education.鈥�

Acosta鈥檚 team tweaked the formula to account for the benefits of virtual schooling, which were enjoyed more by some students than for others. Then they 鈥渇ast-forwarded鈥� nearly 30 years and compared projected wages to what could have been.

鈥淧eople are making less money 鈥� that means that they are less productive,鈥� Acosta said. 鈥淎nd in an aggregate model, where we want to see what happens to the full economy, it鈥檚 as if we said, 鈥榃ell, you didn鈥檛 have enough education. You don鈥檛 know how to perform some of the tasks, or something was disturbed in your learning process.鈥欌��

The team actually proposed a series of interventions, from extending both the 2021-22 school year and the 2022-23 school year, just extending the 2021-22 school year, or narrowly targeting the aid to offer a longer school year just to 鈥渆conomically disadvantaged鈥� students nationwide. That more focused aid would reduce the cost to $25.6 billion, though the benefit would be slightly smaller.

But it could benefit individual students powerfully: The economists project that today鈥檚 low-income middle- or high school students could earn, on average, 8.2 percent less in 2050 because of the closure. Today鈥檚 low-income elementary school students could earn 10.9 percent less.

Though the results are surprising, Acosta said, she hasn鈥檛 had any peers in academia challenge the figures so far 鈥渂ecause they鈥檝e been fairly in line with what the literature has said鈥� about education, productivity, and the experiences of students the past year during the pandemic. 鈥淚t鈥檚 an entire generation that lost almost a full school year.鈥�