BETA
This is a BETA experience. You may opt-out by clicking here

More From Forbes

Edit Story

What Educators Are Learning From Money Managers

Following
This article is more than 10 years old.

Brownsville Elementary School in Brooklyn is surrounded by neighborhoods with the highest murder rates in New York City. But inside the charter school at 10:30 a.m. everything is tranquil. Dressed in identical uniforms of green polo shirts and khakis, students walk carefully along tape strips in the halls before sitting down at their desks. Except for controlled bursts of excitement, as when a teacher asks kids to yell out their goals for a lesson, classrooms are hushed in concentration. Children perform tasks like organizing papers on their desks and placing pencils next to their books with precision.

The emphasis on polite manners and discipline is straight out of the 1930s. Except for what is going on behind the scenes: From quiz scores to homework and attendance records, every detail of a student's performance at Brownsville Elementary is fed into computer databases where teachers and administrators examine the constantly unfolding record and quickly adjust lesson plans and individual teaching strategies in response. Achievement First, the New Haven, Conn. nonprofit that operates Brownsville Elementary and 16 other schools in Connecticut and New York, is more like an information-driven company than an old-fashioned school district. "We're obsessed with using data on an ongoing basis," says Douglas McCurry, Achievement First's co-chief executive and a frequent presence in school halls. "Schools are fundamentally undermanaged."

American education is, as always, in a state of crisis. In the past four decades spending per pupil (adjusted for inflation) has gone up 2.6 times, but sat scores have not budged. Despite the $661 billion a year this country puts into public K--12 education, we are churning out a nation of mediocre graduates ill equipped to meet global competitors. Thousands of teachers are being laid off. Central Falls, R.I. fired all of its high school teachers (half will be hired back); in Kansas City, Mo. half the schools are closing. Reformist politicians in Florida, Colorado, Washington, D.C. and New Jersey are confronting teachers' unions and the sacred rights of tenure and rising compensation.

Away from the angriest national debates, however, a quiet revolution in American public education is occurring at organizations around the country like Achievement First (see sidebar stories listed below). Most were launched by idealistic liberals with dreams of social equality. But with annual budgets exceeding $50 million, sophisticated computer systems and hundreds of employees, they are starting to resemble corporations--tracking and responding to minute changes and putting resources to efficient and innovative uses. The question is whether these strategies can be writ large, like Wal-Mart , to work in thousands of schools with millions of students nationwide. There are plenty of doubters.

One believer in data-driven management is Joel Klein, chancellor of the 1.1 million-student New York City Department of Education. Klein takes a portfolio-theory approach to education reform, meaning he wants a selection of large, professional organizations to choose from when he sets up a new charter school to operate outside the district's maze of union contracts and bureaucratic rules. Like a pension sponsor looking to put assets with winning portfolio managers, Klein and his crew want to pick school operators (or traditional school principals) that can reliably move poor students ahead. The goal is to close the so-called "Scarsdale gap" between wealthy suburbs and urban schools in student test performance. Eliminating the gap may be impossible; narrowing it is not.

Charters, says Klein, are a "core part of our portfolio strategy," which includes working with traditional public schools as well. "Our view is those that are good we want expanded and those that are doing a poor job we want closed down." As an assistant attorney general in the Clinton Administration and a longtime corporate lawyer before that, Klein says, "I learned two things: competition and accountability."

Charter operators in New York City, including Achievement First, Knowledge Is Power Program, Green Dot and Victory Schools, now run 125 schools with 40,000 students. They tend to pay teachers more than union scale (albeit for longer hours and more school days per year) and collect a stipend of roughly $12,300 a year for each student they enroll, close to what the district spends on traditional public schools (not counting occupancy and police costs). In the aggregate, they are getting results. Last year charter-school students exceeded the New York City average by eight percentage points in reading and seven in math, and outperformed students from their own districts by at least three times as much. These kids were not cherry-picked; nearly all the students were selected by lottery.

A 2009 study by Stanford University researcher Caroline Hoxby showed that the New York City charter schools achieved this with students who were 63% African-American versus 34% districtwide, and whose household incomes were 37% lower. Hoxby compared applicants who won the lottery for a slot in a charter school to those who didn't. She found the charter students had closed 86% of the Scarsdale gap by the end of eighth grade, while the noncharter pupils remained just as far behind. Perhaps reflecting the motivation that led them to apply for the lottery, all students outperformed their peers in poor districts. (There are no such controlled experiments nationally.)

Frustration with the persistent gulf between rich and poor students drove Dacia Toll, a Rhodes scholar and Yale Law School graduate, to found her first charter school in New Haven in 1998. Now 38, Toll launched Amistad Academy--named after the Spanish ship hijacked by slaves and sailed to the U.S. in 1839--on the theory that social justice begins with better education. Achievement First has a $60 million budget and 17 schools with 4,500 students (300 of them at Brownsville Elementary), making it the equivalent of a good-size school district. The company spends less than 10% of its budget on central administrative costs, compared with 15% to 25% at most urban school districts, which tend to have a heavy load of patronage jobs. The savings get spent at the school level. Teachers receive higher salaries--and help. Brownsville Elementary backs 23 teachers up with five administrators, including deans and "coaches." "We don't think you can coach and manage more than six people effectively," co-chief McCurry says.

Sidebars:

Homecoming Day For a Reformer

Think College

Fire Bad Teachers

Tech and Teaching


Like other successful charter school operators, Achievement First focuses on lagging students, because once they fall behind it is extremely difficult to bring them back up to speed. Children in kindergarten through second grade are given a one-on-one reading comprehension test every six weeks and graded on a scale of 1 to 12. If the entire class struggles with a concept, McCurry says, "we can go back and reteach that." But if individual students fall behind, the school pulls them out into separate groups for intensive instruction on their individual weak points. The extra lessons can be delivered on a computer or during a lunchtime tutoring session; the important thing is that teachers and administrators are constantly watching and adjusting their methods as test results come in.

In one sunny classroom at Brownsville Elementary eight second graders sit in a circle of desks. One teacher fires questions at them from in front of a whiteboard while a second, younger teacher watches and takes notes. These students are the poorest-performing readers in their class, so Achievement First has put them in a separate group with Samantha Hooper, one of the school's most experienced and highest-paid teachers.

Other classes at Brownsville Elementary run 28 to 30 students apiece, enough to drive a suburban parent into a nervous rage. But national studies show no relationship between class size and student performance. (In fact, the Stanford study found a positive correlation between class size at charter schools and academic performance.) By expanding classes, organizations like Achievement First can add a junior teacher to the classroom to help out and learn under a more experienced colleague. The goal is to achieve better results without spending any more than traditional public schools, which in New York City is $16,700 per pupil, including occupancy and police costs. "We spend more on curriculum development, leadership development, and we try to spend a lot less on purely bureaucratic roles," says Max Polaner, a former McKinsey & Co. consultant and member of the Polaner jelly family who serves as Achievement First's $130,000-a-year chief financial officer.

The organization's $34 million New York operation will break even this year, Polaner says, except for $200,000 in startup costs for two new schools. Ninety-six percent of its students passed the 2009 state achievement tests, compared with 82% statewide and 71% in the districts where they're situated. In Connecticut, where the state pays Achievement First 75% of what it gives urban school districts, the group runs the second- and third-best schools for African-American student performance; its Bridgeport middle school led the state last year in student improvement.

These results mean something to portfolio managers beyond school chancellors: to capitalists like Bill Gates, whose foundation has poured $5 billion into public education, including $2 billion in scholarships. Chicago billionaire Jay Pritzker has donated $17 million to the cause, driven partly by research showing that improvements in education yield positive economic returns via reduced crime and higher productivity. "I've seen thousands of business plans and invested in over 50 companies, and just scattershot spending money doesn't work," says Pritzker. "It's testing, watching it, seeding a little more--all those steps tell you whether you should put the big dollars behind it." Cisco Systems , Google and billionaire Michael Dell's foundation are backing the School of One, for example, an experiment in New York City public schools where students receive their lesson plans on airport-style overhead monitors in the morning according to how they performed on a quiz the afternoon before.

Now used in math classes in a Manhattan middle school, the School of One approach allows teachers and administrators to choose among computer instruction, traditional classes or even remote tutoring to keep students on track. In a trial run last year students scored 42% to 70% higher on math tests after participating in the program, says Joel Rose, a former middle-school teacher who runs the New York City Department of Education experiment. If it works, "and that's still a big if," Rose cautions, "our model is quite scalable because all the analytics can be done centrally with cloud computing."

At Wireless Generation in Brooklyn software engineers are working with Achievement First to build a commercial version of the software that the charter operator uses to monitor student and teacher performance. Operating out of stylish offices in the shadow of the Manhattan Bridge in Brooklyn, the 350-employee firm will take in $65 million in revenue this year and is growing at a 20% annual rate. It has contracts with schools in all 50 states--including school districts in Chicago and Washington, D.C. and the entire state of Indiana--and has compiled 3 terabytes of data it uses to refine teaching methods for subjects like reading and math.

"Education is in a revolution of sophisticated analysis of the data set," says Larry Berger, Wireless Generation's founder and chief executive. When a student stumbles, "alarm bells go off early in kindergarten, and they go off with the precision that comes from tracking performance every two weeks." A Yale graduate and Rhodes scholar, Berger started the company in 2000, when "the belief was we would put the student in front of the computer and the computer would do all the work." That idea has shifted, he says, with companies like Wireless Generation functioning more like the electronic helpers that work behind the scenes in the medical industry, planning schedules and warning doctors against dangerous drug interactions.

For $12 to $25 per student per year, Wireless Generation provides software that lets teachers regularly evaluate students for reading proficiency and math skills. Using Palm handhelds or Web browsers, teachers can input results of, say, a kindergarten reading comprehension test that requires students to pronounce three-letter nonsense words. The software can differentiate causes of failure, distinguishing between students who are too slow and those who make errors; it can also flag kids who don't blend letter sounds together. Then it prompts the teacher to group children at similar developmental stages together and provides proven instructional techniques for their particular problems. Instead of being a cudgel to discipline teachers whose students fail to make the grade, software is becoming an efficiency tool to make them more effective. "It's the same as Google," says Berger. "Optimize small things about the user experience to get big results."

Sidebars:

Homecoming Day For a Reformer

Think College

Fire Bad Teachers

Tech and Teaching


Can best practices be replicated across America? "It's pretty much the same recipe you would have had putting a school together in 1910--an execution-based model," says Frederick Hess , a scholar at the American Enterprise Institute and author of Education Unbound (ASCD, 2010). He has doubts. "It's bumping up against natural limits." Whereas Wal-Mart can achieve superior results with the workforce it finds in any region, Achievement First and its peers rely on young, inspired teachers coming out of training programs like Teach for America.

Achievement First hires its teachers under one-year contracts and pays them 15% to 20% more than their union counterparts, although the staff works 10-hour days and a longer school year. (The most experienced teachers can earn $85,000 or more, and McCurry and Toll each earned $154,000 in 2008, the most recent year for which the organization's financials are available.) Roughly 10% to 15% of its teachers quit each year; another 5% or so are fired for poor performance, compared with 9% attrition and 4.4% dismissal rates for public schools.

McCurry says the teachers at the oldest school, Amistad, are among the most experienced and highest paid in the organization, suggesting he doesn't have trouble retaining talent. But Hess thinks the talent gap will hit charter operators before long, leaving technology like that offered by the School of One as the only way to spread the learning revolution across the country.

"Scaling up without losing quality is a huge issue for charter operators," says Christopher Williams, senior program officer with the $35 billion Bill & Melinda Gates Foundation. Searching for better results from the nation's existing stock of public-school teachers, the foundation is investing in high-tech means of collecting data on pedagogical methods. One $45 million program has mounted panoramic video cameras in 3,000 classrooms in six states. Outside experts monitor teachers and students four times a year and match their observations with test data to discern what works best. "One problem is nobody has performance metrics everybody can agree upon," says Williams.

Chancellor Klein is determinedly upbeat. "When people said, 'You can scale this to 20 or 30 schools but no further'--well, we're at 100 [charter schools]. The people opposing it are the unions, not because it can't scale but because it can."

Education pays. James Heckman, a University of Chicago economist and 2009 winner of the Nobel Prize, has spent much of his career studying this point. Heckman is a realist about the challenge that public schools confront. "Most of the gap in test scores is there at age 5, before they enter kindergarten," he says. But when schools do make up for lost ground, they get a giant payoff. A high school diploma, Heckman and others have shown, is worth $11,600 a year in incremental salary.

That income differential represents more than just the "sheepskin effect" of a diploma; some of it comes from the innate difference in intelligence of high school grads versus dropouts. But Heckman has concluded that graduate equivalency degrees confer almost no benefit, because ged holders, who are presumably equally intelligent, tend to be "wise guys" who lack attentiveness and discipline. In addition to learning concrete skills in math and computers, high school grads acquire other skills, such as the habit of showing up on time.

Heckman goes on to estimate society's interest in increasing the number of high school graduates at $1,500 to $3,000 a year per student, mainly reflecting the lower rates of incarceration and jail costs of graduates. (He ignores the actual cost of theft, because the goods remain in the economy.)

It's lamentable how many defective products the U.S. education industry sends out of its $660 billion factory. But it's encouraging to see that there are ways to boost the output.



Sidebars:

Homecoming Day For a Reformer

Think College

Fire Bad Teachers

Tech and Teaching



Special Offer: Free Trial Issue of Forbes