Skip to main content

It is lunchtime, it is Wednesday, it is the year 2000, and the principal of Nathan Hale High School is leading 50 juniors on a march up Lake City Way.

Last year, the teachers told their sophomores that if they passed all four sections of the Washington Assessment of Student Learning, the statewide standardized test, they would get a free lunch at Dick's, the burger place 15 minutes' walk up the hill from Nathan Hale.

So many students passed that the walk has become a parade: Six police escorts, a row of traffic cones blocking two lanes of traffic, the marching band leading the way. The juniors hold balloons that say, "I passed the WASL and I'm going to Dick's!" The next morning, the Seattle Times runs a story "Kudos Come in Burger Form."

Eight years ago, when Mr. Eric Benson, the principal, started at Nathan Hale, this would have been unthinkable. For at least a decade, Nathan Hale had the reputation as the worst high school in north Seattle—the only one without a waiting list, the drainage chute for students who couldn't get into, or had gotten kicked out of, more selective public schools. The last time it got a major write-up in the Seattle Times, the story was about a melee in the parking lot that had put a senior in the hospital with stab wounds.

In the early 1990s, the problems at Nathan Hale looked a lot like the ones at other troubled schools in troubled districts around the country. Decades of forced busing had pushed white, affluent kids out to the suburbs, and "gifted" programs had segregated the students left behind.

They came to be known as dropout factories: Large urban high schools—up to 3,000 students each—that weren't just failing to keep kids from falling through the cracks, but were actively opening up new ones. Nathan Hale's 1,000-plus students were divvied up between special-education, standard, and honors tracks, then taught, separately, in classes of more than 30 students each. They attended the same school, but they were in different worlds, the problem kids concentrated and the college-track kids cocooned.

A version of this story originally appeared in the August/September 2017 issue of Pacific Standard. Subscribe now and get eight issues/year or purchase a single copy of the magazine.

A version of this story originally appeared in the August/September 2017 issue of Pacific Standard.

"The good students were being well taken care of because they were in advanced classes," Mr. Benson says. "But overall, we had the worst attendance rates and worst disciplinary action rates in the city."

Soon after he started, though, Mr. Benson began hearing about an idea that could turn Nathan Hale around. It was simple: Split the school into a bunch of smaller ones. Rather than being organized as one large unit, Nathan Hale would become a cluster of "small learning communities": groups of students who shared the same classes, the same teachers, and the same counselors and administrators for all four years.

Results from early pilots had been promising. A school in Oakland started a "graphics academy" within the same building; it sent nearly 100 percent of its seniors to college. Another, in Baltimore, went from second worst to second best in the citywide rankings in just two years after splitting its 2,170 students into five smaller schools.

Mr. Benson decided to adopt the idea and apply it to Nathan Hale. Starting in 1998, freshmen were sorted into clusters of 90-odd students, handpicked to ensure demographic and academic diversity.

The idea was to de-industrialize the school experience, making sure the kids knew, and were known by, the adults around them. High-performing students still completed AP-level work and students with disabilities still received specialized help, but within the same classrooms. Teachers started coordinating lessons: Kids read Lord of the Flies in English and then conducted a mock trial of Jack Merridew in world history. The staff transformed from an assembly line to a pit crew, each academy with its own team of teachers and counselors, each teacher and counselor responsible for meeting the needs of their own cluster of students.

Almost immediately, the results started to show. The dropout rate plummeted from 12 percent to 2 percent. Attendance rose 8 percent. By 2006, Nathan Hale's WASL pass rates were beating the state and the district by 10 points.

The rest of the country took notice. Nathan Hale and the other schools that piloted small learning communities started to get visits from donors and journalists. Mr. Benson won a citywide principals' award, and the district received a $25 million "transformation grant" from the Gates Foundation, part of which would go toward expanding small-schools reforms.

Across the country, the idea of splitting large public schools into small ones went from intriguing to imperative. In 2001, the United States Department of Education stated definitively that small learning communities "positively affect student achievement" and began giving grants to districts that rolled them out. The Gates Foundation followed suit, spending somewhere between $650 million and $2 billion (the numbers are still disputed) between 2000 and 2008 to establish 800 new small schools and break up 1,000 existing ones. Florida passed a law that capped every new high school at 900 students and encouraged schools larger than that to break up.

It was, quietly, one of the country's biggest-ever experiments in education policy.

And then, just as suddenly as it began, it was over.

By 2008, evaluations conducted by the Gates Foundation and the Department of Education showed that, despite what was happening at Nathan Hale and other early adopters, the hundreds of millions spent converting large schools into small schools hadn't made any difference to the outcomes of the students attending them.

The Gates Foundation looked at three different districts and found that the conversion "did not produce a measurable impact on student academic achievement." One school, Denver's Manual High, was profiled in Businessweek ("Bill Gates Gets Schooled") in 2006 and The New Yorker ("Can the students who became a symbol of failed reform be rescued?") in 2007. Both stories found that converting Manual from one school into three—one on each floor—had chased away the high-performing kids and concentrated the challenges of the ones who remained. In 2009, the Department of Education ended its small learning communities grants program. In 2012, the Gates Foundation did too.


I was a student at Nathan Hale from 1996 to 2000, one of the last classes under the old system, before Mr. Benson's reforms kicked in. I remember the fights in the halls, the stabbing in the parking lot my sophomore year, the way we were segregated into "normal" and "high-performing" classes.

For decades, American public education has been in a state of perpetual crisis. Our diagnosis has gone from dropout factories to bad teachers, rigid unions, low expectations, and over-testing. We endlessly debate which idea is going to fix it—which new structure or technology will scale up to match the size of the problem it's intended to solve.

"It's the Starbucks model," says Michael Klonsky, the author of Small Schools: Public School Reform Meets the Ownership Society. "You find a successful coffee shop, then you replicate it. The problem is that education can't be standardized like a latte."

It is easy, in other words, to come up with the next big idea in education reform. It is harder to make it work everywhere at once. Breaking large schools into smaller ones was not the first intervention that promised, then failed, to transform American education. Nor was it the first time the country moved on without seeking to understand how and why that failure occurred.

Nathan Hale was a troubled school when I attended. Then, for a while, it wasn't. These days, it is troubled again, but in new ways and for new reasons. In the fall of 2014 I spent two weeks there attending classes and interviewing dozens of teachers, administrators and district officials to find out why the reforms had worked in my school and didn't anywhere else.

How did small learning communities, an idea that was once so promising, end up so discredited? And what does its rapid rise and fall mean for the future of education reform?

The Implementation


"Please welcome ... genital warts!"

This is Oprah.

Well, this is Matt pretending to be Oprah. It is 11:30 a.m., it is Thursday, it is 2014 and this is Ms. Cindy Jamieson's first-period health class.

Matt, Kelly, and Steve, (whose names, along with those of the other students in this story, have been changed) have been assigned a sexually transmitted disease and they have five minutes to present it to the class. They're doing it as a skit—a talk show. Matt is wearing a 1970s wig, the closest thing to Oprah he could find. Steve is holding a sign with 'Applause' written in Sharpie, walking back and forth like one of those boxing-ring models. He's 6-foot-4, and the sign nearly touches the ceiling.

"Actually, I go by Genitalia Schwartz now." Kelly sits down, flipping her waist-length hair behind the chair so she doesn’t sit on it.

Oprah—Matt—is interviewing Kelly, asking her about the symptoms she causes and how she's treated. Every group had to make a slogan for their disease, and now Steve is holding a poster with a version of a T.I. quote on it: "Where They at Doe? In Your Genitals!"

Nathan Hale, it turns out, still has small learning communities, still splits itself into academies of around 90, and still puts advanced students at desks next to kids with learning disabilities. Teachers and administrators still work like a pit crew, meeting once a week to plan college-application workshops and identify the kids who need The Great Gatsby read out loud to them.

"I haven't broken up a fight in the hallways in years," says Mr. Brian Coon, my old math teacher. He's been here since the mid-1990s, way before the school went through the reforms. He, like all the other teachers I speak with, tells me that splitting the school into small learning communities was the key to transforming it.

But he also tells me that it wasn't actually the restructuring that made the difference. It's how it was done. Before Mr. Benson instituted the small learning communities, he spent years preparing teachers and administrators for the change.

"It had to be bottom-up," Mr. Coon says. Principals rarely last more than a few years. If teachers aren't behind the reforms, he says, they can simply wait them out—keep doing whatever they're doing until the next principal, the next overhaul.

Four years before he implemented the small learning communities, Mr. Benson gave his staff a mandate to design the turnaround themselves. Teachers, tutors, and administrators started meeting at each other's houses on Saturday mornings, butcher paper on the wall.

"There was never any agenda," Mr. Coon says. "It was just, 'How can we make this place better?' Mr. Benson used to say: 'No sacred cows. If you think the school would be better without a principal, that's fine.'"

It wasn't just the slowness and deliberateness of the reforms that made them work. Mr. Benson says changes from above were also decisive. In 1994, the district classified Nathan Hale as a "vanguard school," a designation designed to allow flailing schools to try out drastic changes.

Crucially, this gave Mr. Benson control over his budget. Before 1994, Nathan Hale was allocated a fixed number of staff members—administrators, counselors, tutors, janitors—based on its student headcount. If Mr. Benson needed, say, an extra science teacher or an extra drug-and-alcohol counselor, there was nothing he could do. The only spending he controlled was around $40,000 for supplies.

After 1994, though, Mr. Benson controlled nearly his entire $4 million budget himself. All of a sudden he had the power to adapt his staff to his student body. He immediately started trading in back-office staff—vice principals, deans, secretaries—for young, cheap, hungry teachers right out of graduate school.

This extra control is what allowed Mr. Benson and his teachers to customize the reforms they dreamt up on Saturday mornings. They knew that the first year of high school was a pivotal one, setting the trajectory toward, or away from, graduation and college. So they front-loaded teachers onto the freshmen, assigning extra staff to make sure classes stayed below 25 students. They lengthened periods from 50 minutes to 90, and even gave freshmen a different schedule than the rest of the school (this had the additional benefit, one teacher tells me, of "keeping the senior boys away from the freshmen girls").

But the most important reason the reforms worked here, teachers tell me, wasn't the changes for the students. It was the changes for the staff. Starting in 1998, school began 90 minutes later one day per week—time specifically set aside for teachers in each academy to discuss how their students were performing, to plan common lessons, to keep the butcher-paper spirit going. On the other days, teachers set up "critical friends groups" to sit in on other classes and provide critiques afterward.

And it turns out that the consultation, the administrative enfranchisement, and the backstage collaboration—all this boring but crucial stuff—is exactly what most other schools skipped when the idea of small learning communities scaled up nationwide.

"No one spent any time understanding why the early pilots worked," says Elena Silva, the director of pre-K-12 policy at New America. "They just said, 'We'll make the schools smaller and that will do it.'"

Originally, implementing small learning communities was a means to an end, a way to personalize teaching and homework for each student. As the idea expanded, though, smallness became an end in itself. After Florida passed the law encouraging its large high schools to split up into smaller ones, some closed in the spring as large comprehensives and then reopened in the fall as small learning communities. One district tried to convert 30 large schools in just five years, a pace that made it impossible to prepare students or teachers for the change. A 2008 evaluation of Department of Education grants found a high school in Long Beach, California, where a quarter of the students couldn't even identify their own small learning community, despite being enrolled in it.

The result of this rush was that teachers and administrators were implementing the reforms before they understood how, or why, they were even happening. Some schools assigned teachers to be "principals" of their own academies—suddenly responsible for things like rewriting curricula and managing staff—without any training.

Other schools took the reforms as an invitation to maintain segregation rather than abolish it, tracking low-performing students—disproportionately poor and racial minorities—into communication and arts academies and reserving science, math, and engineering for "gifted"—i.e. white—kids. Students at a school in Yakima, Washington, told a researcher that they had been relegated to the "loser academy."

In fact, only one-quarter of the first set of schools that received federal small learning communities grants even implemented them completely. Most of them skipped major components—things like teacher collaboration and parent involvement. Some schools that tried to take it slow, to roll out the reforms one grade at a time, like Nathan Hale had, lost their Gates grants before they had even converted all four grades.

All of this should have been predictable. Mr. Benson tells me many of the teachers at Nathan Hale resisted the reforms too. When he retired in 2003, he says, they were still tweaking the reforms and defending them against teachers who weren't convinced.

Within a few years, though, it was a moot point. By the time the full teaching staff was on board with the reforms, they were already starting to fall apart.

The Aftermath


"Thou whoreson kidney-faced rabbitsucker!"

Carol and John are standing up, holding lists of adjectives. It is 10:30 a.m., it is Monday, it is 2014, it is the first day of the Romeo and Juliet unit, and Ms. Megan Simmons has instructed her students to shout Shakespearean insults at each other.

"Thou yeasty, dog-hearted catpurse!"

The laughter and applause wake up Sheila, head down in a red beanie. The two girls next to her, sharing a pair of earbuds, barely flinch.

The pandemonium today is routine. Part of the problem is John. He's one of the smartest kids in the class; he gets what Ms. Simmons means the first or second time she says it. By the third or fourth, he's restless.

"It's someone's choice to be offended," he said yesterday during a discussion of the infamous "Auschwitz selfie." It's obvious he doesn't believe this; he's just trolling the class to keep himself awake. "No one should ever do anything differently just because it will offend someone else."

Today, the cluster of girls in the front of the classroom maintain their own conversation nearly the whole 90 minutes of Romeo and Juliet: laughing, dozing, eating, like they're in a room by themselves. Sometimes Sheila stands up and dances, snapping in rhythm to whatever song is playing in her head.

The combination—John disruptive and engaged, the girls disruptive and unengaged—keeps the learning-to-shushing ratio at about one-to-three. Ms. Simmons spends most of each class in full triage mode, trying to cram as much teaching as possible into the two or three minutes of calm between outbursts.

Every kid seems to be in a different classroom. Two students can barely speak English; one of them arrived from East Africa only a month ago. She sits attentively at her desk, nods, looks like she's following along. But whenever she's called on to speak, she goes mute. Another, tiny in a giant Yoda sweatshirt, balls up paper towels and spends most of the class shooting free throws at the recycling bin. Whenever he misses, he gets up, retrieves his wadded-up towel, then returns to his seat and shoots again.

Sixteen years after Nathan Hale split into small learning communities, all of the structures of the reforms are still here. But everything supporting them, everything that made them work, has fallen away.

Budget control was the first to go. In 2008, the district went back to assigning a fixed number of staff members to each principal. At the time I visited, Nathan Hale got one teacher for every 30 students, one tutor for every 45 bilingual kids, an extra assistant principal when total enrollment surpassed 999, and so on.

Dr. Jill Hudson, the current principal at Nathan Hale, came here in 2009, just in time to inherit Benson's reforms but none of his autonomy. The inflexible staffing allocations mean she can't trade in administrative staff for younger, cheaper teachers. She can't customize her spending to her students.

Nathan Hale's inclusion model has attracted special-education kids from all over Seattle, she tells me, and up to 20 percent of the students here are now on an "individual education plan," compared to around 13 percent citywide. The district sends her extra chaperones and tutors to account for the larger population of special-education students, but the formula doesn't take into account cumulative effects. One of the reasons Ms. Simmons' language arts class is such chaos is that it has one student on the autism spectrum, one with attention-deficit/hyperactivity disorder, one with oppositional defiant disorder, and one with post-traumatic stress after fleeing a war zone for the U.S. as a child. All of these challenges manifest, and combine, in ways that make it nearly impossible to teach.

Next to go was the butcher-paper collaboration. In 2012, the district rolled out new teacher evaluations: 22 components ("Establishing a culture for learning," "organizing physical space," "demonstrating flexibility and responsiveness"), each rated from "unsatisfactory" to "innovative."

This has flipped the teacher assessment process from horizontal to vertical. Instead of getting informal, ongoing critiques from their peers—"teaching with the door open," they used to call it—teachers mostly get scheduled, intermittent evaluations by their bosses.

It's part of the district's, and the country's, push for schools to be more "data driven."

While this sounds rigorous and quantitative, teachers tell me the new evaluations ultimately boil down to one person writing words in a box about another. Dr. Hudson doesn't like to give out the "innovative" rating, so teachers cross their fingers for the vice principal, an easier grader, to assess them. Meanwhile, according to the school's 2012 Climate Survey, only 30 percent of teachers said they observe other classes anymore.

But the biggest challenge to maintaining the reforms isn't the administrative changes. It's the sheer number of kids they're being applied to. The entire purpose of small learning communities was to customize the school experience to each student. But as class sizes grow, and the percentage of special-needs kids rises within them, tailored lessons and individual attention become more and more difficult. While I'm here, the school paper, the Sentinel, reports that teachers are being strained by classes of over 30. "We used to have 22 to 25 kids in each class, and it was totally different," Ms. Simmons says.

The ballooning class sizes aren't Nathan Hale's, or even the district's, fault. In 2001, Washington voters passed a referendum that restricted the growth in property tax revenues—not the tax rate, the actual revenues—to 1 percent per year.

Since then, inflation has averaged 1 to 2 percent and the population of Seattle has grown by more than 100,000. That means that, in real terms, school spending shrinks every single year. The district can augment its funding through ballot initiatives, but those are only allowed to fund "enhancements," not basic operations. That's how Nathan Hale got an $86.7 million renovation in 2011—a new lobby, a synthetic-turf football field, an auditorium with opera-house acoustics—but can't, according to the staff, afford to keep the heat on past 3:15 p.m. every day.

It's not just Seattle. This turns out to be at the heart of why small learning communities failed in so many places: Nobody wanted to pay for them.

Screen Shot 2017-08-14 at 3.20.43 PM

"Schools build their programs around the grant, but then the grant runs out," New America's Silva says. She spent the early 2000s helping schools in Maryland and California get small learning communities funding. She says she saw it happen over and over: Districts received grants to turn around their schools, implemented the reforms, then gradually abandoned them when the funding dried up.

"They like the idea," Silva says, "but they don't have the money to keep it. So the ninth-grade small learning community becomes an even looser ninth-grade cohort, which becomes 'Wouldn't it be great if we just paid more attention to the ninth graders?' which becomes 'It never even happened.'"

The funders found the same problem. "In many schools," a 2006 internal Gates Foundation evaluation concluded, "we found that teachers voted to accept a grant or agreed to conversion not because they thought it was a good idea, but because they did not want the school to 'miss out' on some extra money."

In the early 2000s, the Gates Foundation and the Department of Education gave hundreds of millions of dollars to districts to convert to small learning communities. Few districts, however, gave enough funding to their schools to account for all the additional time teachers and administrators had to spend implementing them. High Tech High in San Diego, one of the early Gates Foundation grantees, was already trying to fundraise $1,000 extra per student just two years after it received its grant. Evaluations from across the country found that, due to district budget cuts, schools were laying off teachers who helped design the reforms and replacing them with teachers from traditional schools who didn't know—or didn't like—the small-schools model. Some schools had classes as large as 50 students.

Not surprisingly, the dramatic nature of the changes, combined with their slapdash implementation, turned staff members against them. In schools that received Gates Foundation grants, an average of 30 percent of the teachers quit after the first year. According to the Seattle Weekly, Mountlake Terrace, a few schools north of mine in Seattle, lost 25 percent of its teaching staff the summer after it accepted a Gates grant—when the typical turnover rate is 5 to 10 percent a year.

Silva, the education researcher, points out that the small learning communities reforms were rolled out right after the dot-com bubble burst. Districts across the country were slashing their budgets, laying off teachers, and stretching their administrators thinner and thinner. Not only that, she says, but, at the same time, the country was re-diagnosing the fundamental problem of American schools.

"As a nation," she says, "we became focused on No Child Left Behind and then Race to the Top and then Common Core."

By the mid-2000s, low test scores and poor teacher quality had replaced dropout factories as the looming existential threat to America's children. With the metrics of success shifting out from under them, districts abandoned the goals of smallness and personalization in favor of teacher training and standardized tests.

What really happened to small learning communities, in other words, is not that they failed, but that they drowned.

The Reasons


It is Thursday, it is lunchtime, it is my last day at Nathan Hale, and Dr. Hudson, the principal, is telling me what she would do here if her funding doubled overnight.

"I'd extend the day,” she says. "Some students take more time to learn things. I'd give them that time." She'd add two periods for students who needed extra help, hire full-time drug-and-alcohol counselors, offer paid tutoring jobs to the kids who go home to an empty house every night.

Some of these students, she says, "Their families don't have money, and they want to contribute. We should help them do that."

After two weeks at Nathan Hale, this is the question I'm left with: Every single teacher and administrator here has ideas for improving the school. So why do education reform efforts always ask them to implement somebody else's?

The rise and fall of small learning communities, after all, was not an isolated incident. Since then, education reformers have cycled through fads like yo-yo dieters. First it was teacher quality, then it was STEM programs, then it was "blended learning"—giving teachers souped-up Kindles to track students' progress in real time. These ideas always follow the same trajectory: First they are promising, then they are scaled, then they are too slow, too complicated, or too expensive. Then they are discarded.

I ask Kim Whitworth, the Seattle Public School District representative responsible for Nathan Hale, why the independence given to Mr. Benson was so systematically dismantled, why principals and teachers are no longer allowed to design their own solutions for their own problems.

Whitworth says that, for about a decade, they could. In the late 1990s, the district gave every principal in Seattle total control over their schools and their budgets.

In some places, for some principals, it worked. For most of them, though, the extra independence simply meant a lot of extra work. Designing a school from scratch is a huge administrative task. Principals, many of them former teachers, were overwhelmed by the new responsibilities of budgeting and hiring and planning a curriculum.

Within a few years of instituting the reforms, an independent audit found, the school district was in financial chaos. Principals were spending money without reporting it and overpaying for teachers. High-performing schools were swamped, low-performing schools were languishing, and teachers across the city were stretched thin from all the extra meetings, the loss of planning time, and the start-stop restructurings. In 2005, the district got in a public tiff with its donors after several high school principals reportedly couldn't figure out how to spend their grants and simply left them sitting in their accounts. Shortly afterwards, the district started rolling back the reforms.

The lesson the district learned from this experience—systems must be designed for their worst performers, not their best—is, in fact, the same lesson that the Gates Foundation took from the small learning communities reforms.

"Nobody has developed a reliable framework other than closing schools and reopening them," says Tom Vander Ark, one of the architects of the Gates Foundation's small-schools grants in the early 2000s. Every time the foundation opened a new school, he says, the small learning communities worked. When it tried to convert existing schools, however, it ended up in a briar patch of confused parents, resistant teachers, and aloof districts that took years to untangle. For a donor seeking quantitative progress and scalable models, the slow and difficult and human process of reforming an institution, rather than creating one from scratch, looks like a bad investment.

Whitworth pulls a book off the shelf behind her. It's a copy of John Hattie's Visible Learning, the 2008 meta-study that sparked, in part, the current "data-driven education" trend. Whitworth turns to a list of classroom interventions and the results, down to the decimal point, they will yield from students.

According to the book, for every year that a student receives "micro teaching"—personalized attention—her education will improve by an "effect size" of 0.88. Every year of phonics instruction yields a 0.5 effect. Other items on the list include "outdoor/adventure programs" (0.43), "desegregation" (0.28), and "use of calculators" (0.27).

This is how it works now. For districts that can't afford to train their teachers and principals, for donors looking for transformative turnarounds, this is how they decide which ideas to send into the classrooms. Something with a number behind it looks like a sure thing. Letting schools decide for themselves looks like a gamble. And gambles are a luxury that American education can no longer afford.

The Epilogue


Perhaps the most surprising thing about small learning communities is what happened to them after the country stopped paying attention.

New York City converted some of its schools to small learning communities in 2002. A decade later, the district found that its "small schools of choice" had higher graduation rates and better college-readiness indicators than the ones that stayed big. At the Highline School District, just south of Seattle, my old vice principal, Rick Harwood, tells me he ran a school that looked a lot like Nathan Hale in the late '90s: integrated classes, personalized assignments, kids separated out into clusters. It's worked there for years, he says.

Dozens of districts across the country, it turns out, kept the reforms after the Gates Foundation and Department of Education money dried up. Hundreds of schools are still organized into academies, still strive to personalize education, still pit-crew the performance of each of their students.

Small learning communities, in other words, can succeed. They can also catastrophically fail. The decisive factor isn't the Big Idea itself—splitting up dropout factories floor by floor—but the millions of little ideas that hold it together. When they worked, small schools allowed educators to know their students, to tailor their lessons and their days and their hallways to the kids they were serving. When done poorly, small schools did the opposite. They became just another unreachable expectation, another constraint imposed on teachers and principals and districts that already lived with far too many.

In the decade since the small learning communities experiment inflated and popped nationwide, every successive Big Idea in education reform—charter schools, blended learning, "early college" programs—has ended up in the same place. It works for one school and doesn't for another. It serves one group of students and leaves another behind. Over and over again, studies investigate exciting new models and come to the same pedestrian conclusion: School reforms work when they are implemented by good leaders, empowered teachers, and adequately funded administrators. When they aren't, they don't.

This is as close as I've come to a conclusion from the two weeks I spent at Nathan Hale and the two years I spent afterwards trying to understand what I saw there: Maybe when it comes to education reform, big ideas don't matter. There will never be a structure or a technology or a method that is more powerful than the environment in which it is applied. Big ideas are, at their best or their worst, simply a mirror—an amplified or diminished reflection of the leaders, the institutions, and the people expected to carry them out.

"You have to find out what every single kid needs and get it for them. There's no shortcut that's going to make that easy," says Ms. Sarah Smith, my old history teacher. She was one of the young, hungry graduates Mr. Benson hired in the mid-'90s. Now she works for Rainier Scholars, a non-governmental organization that helps low-income students make it to college. Some of her kids need tutoring, she says. Others need SAT prep, or rides to school, or warm clothes or anger management or anxiety medications. "It's never one thing—it's everything."


It is 11:35 a.m., it is Thursday, it is Ms. Jamieson's second-period health class. She's sitting on a high stool at the front of the room, showing us how to use a dental dam ("In a pinch, Saran Wrap is fine, but it won't be minty fresh") when a voice comes on the PA telling us we're having an earthquake drill. The kids all crouch down under their desks and a minute goes by. Ms. Jamieson tries to get them to sing "If You're Happy and You Know It." The kids ignore her, and then one of them shouts, "Started from the bottom, now we here," and then they are doing Drake, percussion and all.

Suddenly the door opens and Dr. Hudson, the principal, runs in. "Phil! Put your leg under the table! It just got smashed!" She makes a lap through the class, critiquing our form. "Kathryn, your head is crushed!"

She runs out and the room goes quiet and there we are, silent, crouched under our desks, waiting for the door to close behind her so we can start again.

A version of this story originally appeared in the August/September 2017 issue of Pacific Standard.