Skip to contentSkip to navigation

William G. Ouchi: The Thought Leader Interview

A champion of decentralized management shows how better education means competitive advantage, but only if school principals are treated like business unit managers.

(originally published by Booz & Company)

Photograph by Vern Evans
The most pernicious management problem in the world is also the most significant. It’s the problem of designing public school systems that can educate young people for the unprecedented global society that awaits them when they come of age. If we take that problem seriously, then we need to pay attention to the research and thinking of William G. Ouchi. Bill Ouchi, age 62, is the Sanford and Betty Sigoloff Professor in Corporate Renewal at the UCLA Anderson School of Management, and a long-standing expert in organizational design.

Although he has single-mindedly pursued the reform of public school management for 25 years, Professor Ouchi (pronounced “oh-chee”) is still best known for introducing the principles of Japanese management to Western business audiences. His best-selling book Theory Z: How American Business Can Meet the Japanese Challenge (Addison-Wesley, 1981) made him famous as an advocate of such then exotic approaches as team-based management, lifelong career paths, and consensual decision making.

Today, Professor Ouchi is one of the very few writers who can claim substantial influence in both education and business management. His 2003 book, Making Schools Work: A Revolutionary Plan to Get Your Children the Education They Need (Simon & Schuster), is based on in-depth study of the few large urban school systems that made consistent improvements in student test scores. The book identified seven elements common to good school systems: giving school principals local autonomy (as if the schools were business units), giving families a choice about which school they could send their children to, making schools accountable for student results, giving local schools control over their budgets, delegating authority as low as possible in the hierarchy, instilling a “burning focus on student achievement,” and setting up schools as “communities of learners,” where all the teachers figured out solutions together.

Since Making Schools Work was published, its precepts have been adopted by school systems in New York; Chicago; Oakland, Calif.; and Hawaii (where Professor Ouchi was born). In 2005, Richard Riordan, the former Los Angeles mayor who was Governor Arnold Schwarzenegger’s first secretary of education, evoked Professor Ouchi’s work when he wrote, “If you made a list of people’s silver bullets for public education — smaller classes, better pay for teachers, more phonics, longer school years, no social promotions — the concept of changing governance structure would be near the bottom of the list. None of the favorite silver bullets is going to work, though, unless principals are empowered and can in turn empower teachers and parents.”

Though Professor Ouchi’s education research has been limited to North American and Japanese schools, his points seem universal. For example, the issues of centralization help explain the difficulties that European educators have now in dealing with children from immigrant Muslim families. And his working hypothesis for schools — that they cannot succeed without local autonomy — has implications for leaders in any sort of global organization who are trying to make a difference in the world around them.

Professor Ouchi sat down with strategy+business in his office at the University of California at Los Angeles in January 2006.

S+B: How would you describe the state of education in the U.S. today?
It’s dire. Right after World War II, the average American felt that the public schools were excellent and that children would get a better education than their parents had received. Today, the average American gives the schools a C-minus.

S+B: Were schools in fact different then? Or is it just perception that has changed?
Schools were very different, mainly because the students were more homogeneous. Some people will go on from there to say that the difference was cultural. They’ll say that today’s poor children come from uneducated families, where people don’t understand education and don’t care about their children. This is absolutely dead wrong.

The critical difference was this: In any industry, if you have a homogeneous product or service, your costs go down and you can deliver results more effectively. The U.S. had that homogeneous student body in the 1950s. Now, we have a tremendously diverse student body, which we’re jamming into the same old education factory that was set up for homogeneous throughput. We need an organizationally more flexible and more complex solution.

S+B: You mean we need the educational equivalent of a differentiated product line.
That’s exactly right. Moreover, school systems are much bigger. Before World War II, we had 25 million public school students in this country and 127,000 school districts. Today, we have 50 million students and only 16,000 districts. The typical American school district today has 15 times as many students as did the typical district of World War II. No company can grow 15 times in size, not alter its structure, and continue to be effective. But our school districts have not altered their structure. They are organized and run exactly as they were in the 1950s. And they don’t work.

S+B: Why was there that much consolidation?
After Sputnik went up in 1957, there was a hue and cry in the U.S. that Soviet children were better at math and science than American children. This was blamed on having too many small high schools that couldn’t offer advanced courses in physics and math. So we rushed to replace small school districts with big ones.

American businesses did much the same in building gigantic factories that turned out products of horrible quality. At that time, the productive capacity of every economy in Western Europe and Asia had been destroyed in World War II; virtually all manufactured products had to be bought from the U.S. American factories had 20, 30, or 40 acres under one roof. Nobody in their right mind would operate such a factory today. But we are still managing 4,000-student schools.

Failing to LEARN

S+B: What led to your interest in school management?
It started with the 1984 Los Angeles Olympics. This Olympics had been a great success. It made money. The traffic didn’t tie anything up. And it left Los Angeles with a real can-do attitude.

Subsequently, some of the Olympic organizers formed a broad-based civic improvement group called the Los Angeles 2000 Committee. Roy Anderson, who was then the chief executive of Lockheed, was the chairman. I was asked to cochair the committee on the economy. We also had committees on pollution, health care, transportation, housing, jobs, and water and power, but at the end of this enterprise we were unanimous: Fixing the public schools outranked every other issue in determining the future of Los Angeles. All roads led back to K–12 education.

S+B: Why was that so important?
Because if people don’t have a good education, they can’t afford even the most inexpensive housing. They can’t contribute to the cost of their own health care. They will not attract employers to locate in L.A. And on and on.

So 12 of us formed a group called LEARN: the Los Angeles Educational Alliance for Restructuring Now. Bob Wyckoff, then the president of Arco, and Dick Riordan, who would later become the mayor of Los Angeles, co-led it. I was the only academic member; most of the others were business leaders. We met once a week.

Our question was how to fix the L.A. Unified School District. This is the second-largest school district in the U.S., with 740,000 students. Only about 20 percent of the students in that district were performing at an acceptable level. We all felt that the problem was management: The district had become top-heavy, autocratic, bureaucratic, and unresponsive. My feeling was that every school district in America was run in the same top-down, centralized, bureaucratic way. So if managing it differently would make a difference, there was no available evidence, because no one had ever done it differently.

I had spent 35 years studying the management of very large companies, and one of the most consistent principles is decentralization. In a competitive world, you must make decisions in the smallest operating units possible, or you will go out of business. The public school districts, of course, get students and money whether they succeed or fail. This means that they can’t fail in the traditional business sense. They can fail, but only in terms of educating their students.

Surely, we thought, if the board and the superintendent saw their management system as we saw it, then they would naturally move to create more autonomy for their “operating unit” leaders, the school principals. After all, leaders of large companies, once they see the benefits that accrue, naturally give more decision authority to people who manage the operating units.

But after several years of intense effort, the LEARN initiative failed. No one in public education seemed to care about the way that school systems were organized — or even to acknowledge that it mattered. I was particularly disappointed by the response from academics in the education field. Most researchers who study education will tell you that 97 percent of the difference between a high-performing student and a low-performing student is accounted for by the education and income of the parents and by the training of the teacher. They will say that how you manage a school, or a school system, makes no difference. But that is obviously incorrect, and I decided to spend the rest of my career studying how to organize school districts.

S+B: What happened next?
The LEARN group hired a team of researchers to go around the world and look for any successful example of rehabilitation of a large, urban school district. They came back in 1991 with only two. One was the charter school concept — the idea of creating semi-autonomous entrepreneurial schools within a public system. This had only been implemented in England and Canada; it had not yet come to the U.S.

And the other was a radically decentralized school district in Edmonton, in the Canadian province of Alberta. Since 1975, school principals had controlled their school budgets, decided what staff to hire, set their own curriculum, and designed their own schedules. We were persuaded that this was the right approach; we set on Edmonton as the model to study and introduce to Los Angeles.

S+B: Did that mean persuading the school board?
No, neither the superintendent nor the school board were interested in change at that time. We started with the president of the teachers union, Helen Bernstein. Then we went out and enrolled 700 community organizations to become trustees of LEARN. These were groups like the NAACP, the Mexican American Legal Defense and Education Fund, and the Urban League. We went through massive planning sessions with them in which, for example, Bob Wyckoff and I cochaired a workshop on governance with 75 people. Some of these meetings went from 8 a.m. until 6 at night. We revised and redrafted our plan until everyone was satisfied and all 700 community organizations signed on to it.

S+B: Why was that important?
Many businesspeople, especially in the 1980s, tried to get involved, just as we did, in fixing the schools. They did not want to face a labor pool of illiterates, to rely for customers on an uneducated, low-income population, or simply to watch their communities decline. With the best of intentions, they put a lot of their time and their companies’ money into the fray and they came up empty. I think they concluded that they weren’t wanted and didn’t know how to do it right, so their efforts would be better spent elsewhere.

But in retrospect, these efforts had failed because the businesspeople saw the need for change as a logical, rational business design problem. They thought all the schools needed was some good advice. But school change was really a political process; school board members are elected, and public schools are a unit of government. We could not succeed unless we developed a politically undeniable consensus.

In 1993, we marched down to the school board with all of our organizations. We filled the auditorium. We presented the board with a box full of petition signatures. They voted unanimously to adopt our plan. But we didn’t realize that school board members can look you dead in the eye and tell you a stone-cold lie. We later found out that they turned right around and told their staff, “Don’t you dare implement that plan.”

S+B: What was threatening about this plan to this school board?
Any school board is threatened by the idea of granting power to the principals and teachers. That means the central office bureaucrats lose power, and they will not do it voluntarily.

Decentralize or Perish

S+B: Theory Z, which came out in 1981, had made a similar case to corporations, hadn’t it?
Yes. In retrospect, the core of our argument was the simple recognition that human beings have limited computational ability. None of us is smart enough to tell our subordinates how to do their jobs, so everyone must delegate authority down. But you cannot do that effectively unless you also put in place an accountability system that gives you adequate control, which is not easy to do. It takes thought, time, and investment. But the rewards are great.

This is based on research conducted on the American corporation since the mid-1950s by sociologist Peter Blau at the University of Chicago, by the Aston Group in the United Kingdom, by myself, and by many others. We all consistently found, in every industry and country, that there is a limit to the size of an effective organization. If employees essentially perform repetitive tasks, like assembly-line work or telemarketing, the maximum effective size is about 1,500 people. If employees are professionals, like consultants, accountants, doctors, lawyers, or teachers, then the maximum effective size is about 150. Larger sizes produce the pathologies of centralization, rising general and administrative costs, rigidity, and bureaucracy. And the only antidote is local unit autonomy: Decentralize or perish.

S+B: Was there a difference between your message to business leaders in the 1970s and what you would go on to say to the education leaders in the 1990s?
In the domain of management, the message was the same. But schools are not businesses. First, community leaders — business leaders and community group members alike — have no jurisdiction. You can be smart, rich, and energetic, but only the school board, a small group of people elected every few years, has any power to change the situation. Back in the 1970s, there was a widespread movement called school-based management, where committees of parents and teachers and community members diagnosed school problems and recommended solutions. But they had no power to change anything, and by the 1990s, they had all drifted away.

Second, education is the most emotionally charged realm imaginable. People care about their children even more than they care about their money. A child has only one shot at first grade, and if you ruin that year, you’ve made a big dent in his or her life prospects. So you cannot experiment with schools, nor should you be able to. You cannot say, “I’m not sure this is going to work, but I’m going to reorganize your school.” And especially in cities, public schools serve those who have no choice. The city of Detroit, at its peak, had 320,000 public school students. Today, it has 120,000. Everyone who could leave the schools has left them.

Third, issues of race and religion lie at the heart of our public schools. Why would poor or Latino parents trust a wealthy white man from a giant company to actually care about their children? Why should they trust him and take his advice? That’s a reasonable question. So, if you are a businessperson trying to change the schools, you have to commit to be involved long enough for community leaders to get to know you and decide they can trust you. Until that happens, I don’t know how you can possibly be effective.

Attractive School Leadership

S+B: What happened after you realized that the L.A. school board wouldn’t implement your proposal?
Local community support was necessary but not sufficient; we would have to build an ironclad case. So we went back to research — to the study that I ultimately wrote about in Making Schools Work. We hypothesized that local autonomy is imperative in a large school district, just as it is in a large company. By 1994, there were three school districts in North America with local school autonomy: Edmonton, Houston, and Seattle. So I compared them against the three largest U.S. districts, which were New York City, Los Angeles, and Chicago.

We interviewed a 5 percent sample of the school principals in each district; 66 principals in New York City, for example. We went through each one’s budget, and found that in New York, Los Angeles, and Chicago, principals respectively controlled 6.1 percent, 6.7 percent, and 19.3 percent of the money spent in their schools. Then we went to the three decentralized districts. In Seattle, principals controlled 69 percent of the money. In Houston, 80 percent. And in Edmonton, 91.7 percent.

S+B: Just for the sake of comparison, what would a typical corporate business unit manager control?
Of the money spent in that unit, close to 100 percent. I’m not including the money spent by the district’s central office, which constitutes 30 percent of the budget of a large urban school system. Incidentally, no business could survive at that rate.

Then I found that, indeed, the districts with local autonomy showed higher student performance and a smaller achievement gap between white and Asian students on the one hand and Latino and African-American students on the other. In short, if you go to local school autonomy, student performance gets better. I wrote about this in my book, and today, in part because of that book, eight school districts in the U.S. are implementing local school autonomy.

S+B: Only eight.
Well, listen to who they are. Boston, New York City, Oakland, and Chicago are each launching programs. San Francisco and St. Paul have had district-wide local school autonomy for six years; Seattle and Houston for 10. Hawaii is introducing a state law mandating local school autonomy, which I helped draft. And Dick Riordan, who was California’s secretary of education until June 2005, managed to pass a similar law for the state.

The results prove the concept. Boston has only 19 schools in its program, all in hard-core high-poverty areas. They issued a report in January 2006 that shows that those 19 schools far outperformed the other schools in Boston in student performance. In New York City’s “autonomy zone,” as Chancellor Joel Klein calls it, there are 48 schools where the principals control 100 percent of the money. They have been so successful that New York is increasing the number of autonomous schools to 200 in September 2006.

S+B: Why does giving principals control over the budget make such a difference?
We have a research project under way now in which we’re interviewing 527 principals with local autonomy and visiting their schools. We’re focusing on inner-city high schools, which have proven in the past to be the hardest schools to improve. We’re finding that control over the budget gives principals control over three key school decisions: the staffing mixture, curriculum, and schedule.

The most important single indicator of a school’s quality is a metric you’ve never heard of: total student load. It’s the number of classes a teacher teaches times the number of students per class. In New York City, by union contract, a teacher may teach up to five classes, and a class may have up to 32 students, for a total load of 160. English teachers at Bronx Science, one of the greatest high schools in America, have to figure out how to comment on the essays of 160 students. In Los Angeles, the total load is 200; in some districts, it’s as high as 240. Then visit an elite private school, like those where many of your readers send their children. The total load is 55 to 60.

In October 2005, I visited pilot senior high schools in Roxbury and South Boston — hard-core, inner-city neighborhoods. Both had total loads of 53. Each teacher handles two classes of 20 students each and a writing workshop of 13 more students. The teachers meet three times a week and they discuss each student one by one, because they know every student well.

A young woman at an autonomy school in New York said to me, “When I came here, I was always getting into fights. But at this school, everybody knows who you are. You can’t get away with that kind of thing.”

S+B: And why would the total student load drop simply because the principal is more powerful?
First of all, the principal is the key to every school. A strong principal attracts strong teachers. Then, when you give autonomy to the principal, the principal has no choice but to consult with the teachers. And they invariably decide they don’t want a large school of 3,500 students. It’s unmanageable. So they break it down into small schools, each with its own principal, bell schedule, and teaching staff.

Next, they get rid of administrative overhead. New York City has 75,000 credentialed teachers, but only 50,000 of them are in the classroom. When you give principals autonomy, they put their resources into the classroom. Everybody teaches. There’s no registrar; a teacher handles scheduling. There’s no dean of discipline or guidance counselors. They also dump the seven-period, five-day-a-week class schedule, which nobody likes; it produces a tremendously inefficient use of the scarce teacher resource.

Then they innovate. At one school I visited, the math and science teachers set up a combined program. They all went to night school to get credentialed in both disciplines, and then they created a joint course, team-teaching twice a week in a three-hour block. The students got more class time, and the teachers were in class a bit longer, but they each had half as many students as before. What teacher wouldn’t sign up for that?

S+B: Some parents might want to keep to the old system, with more advanced math and science.
That’s why you also need school choice. Every one of these districts has a system where students can attend a variety of schools without a waiver from the central office. If schools are oversubscribed, they run lotteries, giving preference to neighborhood children. This introduces a healthy form of competition; no principal wants a school that’s 30 percent empty.

I visited an elementary school in Seattle where they established a core objective: Every child, including those from non-English-speaking families, must read English well by third grade. They hired 12 reading and math coaches, drawing on retired teachers who could work part-time without benefits. They changed the curriculum as well. The result was a flying squad, moving from classroom to classroom, working with groups of five to seven students at a time. In that school today, black and white children have identical reading scores and 100 percent of them are at or above grade level.

Autonomy and Accountability

S+B: In your book, you name two other key elements: the “burning focus on student achievement” and making everyone accountable for student performance.
They’re very important. You can have a great decentralized system, but if you don’t focus everyone’s attention on student achievement, performance will not go up. This means that principals cannot have tenure. Their continued employment and compensation should depend on whether or not students achieve.

S+B: What about teacher tenure?
To me, teacher tenure is a good thing. American businesspeople tend to blame tenure and teachers unions. That’s a mistake. People don’t go into teaching to make money. Scratch the most cynical, battle-scarred teacher, and the idealist is still there. And that’s what’s going to save us. Teachers support new ideas when they prove out. Richard Stutman, the president of the Boston teachers union, had led opposition to our pilot schools there because he opposes differential teachers’ pay. But he was quoted in Education Week in January 2006, saying, “We are committed to what works well. We see the need for pilot schools to grow and expand.”

S+B: Your views on accountability were linked with U.S. Secretary of Education Rodney Paige, who was superintendent of the Houston school district when you were researching Making Schools Work. Inevitably, you’ve been associated with Secretary Paige’s program “No Child Left Behind,” where schools are judged by the performance of their students on national standardized tests. And that’s put you in the middle of a firestorm.
I think the firestorm has subsided somewhat. People were willing to write off examples of success in Houston and Canada, but now when they see the effect of accountability in Boston, Chicago, and New York City, they have to take it seriously.

The issue at the heart of this is balancing autonomy with accountability. It’s just like in a business. What allows a huge company to give a division general manager autonomy? It’s the fact that you can measure profit and loss, never with perfection, but pretty clearly. Then, if you’re a central office bureaucrat and you order a division manager to do something contrary, he or she can reply, “I’m maximizing profit right now. Are you saying you want me to sacrifice profit?” No bureaucrat can withstand that argument.

So for a school system to succeed, there must be a simple, unassailable measure of performance. In the Edmonton approach, which became our model, there were three legs of the accountability stool. First is the measures of student achievement on standardized tests, which is now legally required in the U.S. under No Child Left Behind. Second, they must stay within budget. In other words, if a school runs a deficit, the principal must repay that money out of his or her own budget in the future. If he runs a surplus, he or she gets to keep and spend it in that school in future years.

And third, there must be some other kind of consumer transparency as well. In Edmonton, every child in grades 1 through 12 fills out a simple, two-page anonymous questionnaire every year. “Is the school safe? Do you like coming here? Are the teachers trying to help you? Do the other children make fun of you? Do you use computers in class?” Teachers, custodians, and other school employees fill out other questionnaires: “Is the school building properly maintained? Are there adequate books and materials? Do you have adequate opportunity to influence the decisions that affect your work? How would you evaluate the leadership provided by your principal?” And parents have theirs: “Is the school welcoming? Is it properly maintained?” Principals fill out questionnaires about the responsiveness and quality of service they get from the central administration and its staff. All of these results are published publicly.

Everything is open. You cannot escape from that kind of accountability. Nobody who is unable to perform can withstand that kind of public scrutiny.

S+B: Should measures like that have been part of No Child Left Behind as well?
I don’t know if America was ready. You first need to have New York and Seattle and some other places try it, and let people see how it works.

S+B: Has your view on accountability changed in the three years since Making Schools Work was published?
In the wake of the Sarbanes-Oxley earthquake in American business, I have come to realize that there are no reporting standards in the education business. When senior executives at Enron, Tyco, or WorldCom filed fraudulent reports to the shareholders, they faced criminal prosecution. If a superintendent and a school board file fraudulent reports about the progress of their district, nothing happens. It’s not a crime. There are no penalties.

Take spending, for example. In every district we study, my team and I always find that school districts spend 30 to 50 percent more per student than they report to the public. The graduation rate in big cities, like Chicago, New York, and Los Angeles, runs about 50 percent, but those cities all reported 88 percent. If a student stops coming to school, they typically send a letter recommending that the student enroll in a GED program; and then they stop counting that individual as a student. They don’t have to follow up and find out if the student received the letter, let alone whether he or she acted on it.

Finally, districts commonly test only 60 or 70 percent of their students, the good test takers, and report that result as their performance. That was the point of the No Child Left Behind law. Now schools are required by law to test 95 percent of all groups — white, black, Latino, Asian Pacific, Native American, English learning, and special-education children — and report all the results to the public. It’s creating a furor, because people for the first time are finding out what’s going on and they are horrified.

Making a Difference

S+B: Those results drive many businesspeople to suggest that schools should be privatized, through vouchers and similar structural solutions. What’s your opinion?
There are 50 million public school children in America. Fewer than 30,000 of them are on vouchers. Why are we even discussing this? We have had two initiatives on the ballot in California to establish a voucher system. The first one had a campaign budget of $50 million and lost 70 to 30. The second one had a campaign budget of $70 million and lost 70 to 30. There is a message here: The public doesn’t trust vouchers. They believe that one way or another, perhaps in a way we cannot presently anticipate, vouchers will produce a vastly more economically segregated schooling system than we have even now.

My guess is that when competing schools, including charter schools, private schools, and religious schools, get a share above 20 percent of the local student population, they will become a genuine competitive threat to the district establishment. And then schools improve. But it doesn’t have to be private competition; I’m all for charter schools. In Seattle, by the way, 47 percent of the children attended private school in 1995. Public schools have taken back eight market share points from them since then.

S+B: What would you suggest to a businessperson who wants to make a genuine difference in education?
If you go on the board of a company for the first time, what’s the first thing you do? You get to know the articles, the bylaws, the P&L, the cash flow, and the balance sheet. You need to do the same with schools. Businesspeople haven’t a clue how much their school district spends per child. They don’t know the graduation rates or the test scores. Instead, they walk into a classroom and say, “I want to help.” How naive.

Even if you ask for the information, it’s hard to get, because the district doesn’t want you to know how badly they are doing. They typically have six, seven, or eight incompatible legacy information systems, which might use the same general ledger account number to refer to completely unrelated activities. Talk to whoever audits your company, and you’ll find they have probably conducted an audit of the local school district. They will tell you the information systems are hopelessly unfixable. The only thing to do is blow them up and start over again. This is very much a size-related phenomenon. The smaller the school system, the less of a problem they tend to have.

S+B: Do you see the same problems outside the United States?
Nobody else seems to have school districts as big. Even metropolitan Tokyo, which has 30 million people, is divided into 28 wards, each with its own school board and school system. The curriculum is mandated at the national level, but the use of the curriculum, the staffing mixture, and the schedule are up to the principal. And no two schools are alike. They’re all different, given that flexibility.

S+B: As China and India enter the developing world, as Europe becomes more urban, will other countries face the same kinds of problems?
Yes. I get questions all the time from people in China, in particular, where the public is growing more demanding of its public schools. Schools in China are typically part of the municipal government and very centralized. But the winds of change in education are blowing in China — and everywhere else.

Reprint No. 06212

Author profile:

Art Kleiner ( is editor-in-chief of strategy+business and author of Who Really Matters: The Core Group Theory of Power, Privilege, and Success (Doubleday, 2003).

Get s+b's award-winning newsletter delivered to your inbox. Sign up No, thanks
Illustration of flying birds delivering information
Get the newsletter

Sign up now to get our top insights on business strategy and management trends, delivered straight to your inbox twice a week.