(FORTUNE Magazine) – The idea sprang fully formed from Chris Whittle's mind about a decade ago, and it was a stunner: transform public education in America with a chain of 1,000 or more for-profit, privately run grammar schools and high schools, all crisply efficient and brilliantly designed, and all paid for with tax dollars and school vouchers.

Entrepreneur though he is, this was no get-rich-quick scheme on Whittle's part. Rather, the venture that came to be called Edison Schools--which currently runs 150 learning institutions in 22 states across the country--was a strange combination of idealism and opportunism, propelled by the idea that business could succeed where government had failed. Alas, instead of the bold assault on bloated, corrupt, and failing school bureaucracies that Whittle envisioned, it has turned into a long, slow, bloody war, like Napoleon reeling through Russia, with few tangible successes, enormous controversy and resentment, and hundreds of millions of dollars in losses since the company opened its first school in 1995.

Whittle, 55, who calls his CEO job "grinding," miscalculated how tricky and expensive it was to run a school. He expanded too far, too fast. He badly underestimated the suspicion and opposition he would arouse among parents, school boards, and school staff. As Edison's forecasts of huge leaps in academic achievement have failed to occur, a wave of unhappy clients has begun to terminate contracts with the company before expiration. Edison stock, which hit nearly $37 a few years ago, now trades at less than $2. Though Edison posted record revenues of $525 million last year, it also had record losses of $86 million, and it carries a heavy $160 million debt load. "Whittle was never willing to be a niche player," says Dan Quinn, an analyst at Morningstar. "He wanted to be king of the world. Now it's all come collapsing around them. I don't know where they're going to get the money to pay off their debt."

We can all agree that it would be great if Edison were a success. After all, few things in this country need improvement more than our public schools. But Edison has been in business for seven years now, and the verdict is clear: It doesn't work.

Education, in some form, has been in Chris Whittle's blood from the beginning. His father, a doctor in rural Tennessee, was president of the local school board. While Whittle was at the University of Tennessee, he started a series of guides and magazines aimed at college students, which became the basis of one of his companies, Whittle Communications. At 32, he and a friend bought the venerable but ailing magazine Esquire. Later he launched a controversial ad-supported classroom television network called Channel One. As he visited innumerable schools to sell them on the network, Whittle became increasingly dismayed by what he saw: self-serving unions, uninspired administrators, ill-prepared students. All this could be fixed, he decided, with for-profit companies running public education.

When he made speeches about for-profit education before various conferences of state governors in the early '90s, many implored him to take over their troubled public schools. His gubernatorial fans--largely Republican--included Lamar Alexander of Tennessee, Tom Ridge of Pennsylvania, and George W. Bush of Texas. (William Weld, former governor of Massachusetts, later joined Edison's board.) Venture capitalists and investors became enchanted with the idea, too, lured by presentations from Merrill Lynch, an early Whittle backer, that spoke of an inefficient secondary-education industry that spent $1 billion a day, only half--half!--of which was spent directly on education. The rest, it was implied, was pork and featherbedding, easily trimmed by sharp-eyed, reform-minded businesspeople.

Backed by $45 million of early financing, Whittle pulled together a raft of academics and education experts in 1992 and spent three years researching the best ways to teach and run schools. The plan that emerged was ambitious. Lots of teacher training, longer school days and school years, enormous attention to reading. Foreign languages, art, dance. Home computers for every kid. Big schools reorganized into smaller "houses," with the same cluster of kids and teachers together for years. Lots and lots of benchmark testing designed to let schools know which kids needed help and which teachers were having trouble. Incredibly, it wouldn't cost the school district any extra money, Whittle's team figured. Edison would accept the same per-pupil allotment that the district was already spending, and profit from the efficiencies it would impose.

Whittle's first mistake was to count on the nascent school voucher movement of the early 1990s. Vouchers--where parents get a few thousand dollars from the government to spend at any school--were never popular. And Whittle's plan turned out to be extraordinarily difficult to sell. Middle-class suburbanites, it turned out, were almost always satisfied with their local schools and didn't want a for-profit company educating their kids. Teachers and teachers' unions objected to the longer hours, lower pay, and lack of tenure that Edison generally proposed. Existing administrators and support staff, who faced wholesale firings, were predictably unenthusiastic. Whittle "didn't realize how hard a sale it would be," says Henry Levin, head of the National Center for the Study of Privatization of Education at Columbia University. "They landed maybe one or two schools for every 20 they romanced. It was incredibly naive."

Whittle's second mistake was to position Edison, which went public in late 1999, as a growth company. To get big fast, Edison started making deals that have turned out to be foolishly expensive. The company signed several contracts that charged so little that Edison was practically paying school districts for the right to teach. Worse, the deals often allowed local boards to terminate agreements at any time for any reason--which could cause the company to forfeit much of the millions in startup costs it pours into each new school. If Edison wanted to get the contract to run a budding charter school, it often had to lend millions of dollars to the founding school board to construct or renovate each school building. Accordingly, Edison shelled out $200 million to dozens of charter schools and guaranteed loans worth another $20 million. In Las Vegas a few years ago, Edison was so sure that philanthropists would donate $10 million to establish its schools there that it guaranteed the contributions to local officials. The philanthropists didn't cough up, and Edison was on the hook.

Now Edison is desperately short of cash, with $80 million in charter school loans still outstanding. "The company needs to refinance those loans if it wants to keep growing beyond 2004," says Brandon Dobell, an analyst at Credit Suisse First Boston. "And its stock will have a hard time gaining ground until that happens." With Edison's track record, it's unlikely the company will be able to refinance at anything close to market rates.

Thanks to resistance from schools in middle-class areas, Edison's eventual customers wound up being vastly different from those Whittle had first imagined. The takers were charter schools started from scratch within existing school districts, and desperately troubled inner-city schools with rock-bottom test scores and little money. Says Adam Tucker, a spokesman for Edison: "They were the ones most willing to take the chance, who felt they had nothing to lose." But even troubled school systems were ambivalent about Edison. School boards were reluctant to give up control, unions were often vigorously opposed, and parents feared Edison "profiteers" would cut corners at their kids' expense.

In fact, Edison was spending ridiculous sums on its students. For example, at Boston Renaissance, a grammar school that Edison helped establish, the company gave home computers to all 650 students in 1995. The giveaway proved a fiasco. Not only did the computers cost nearly $1 million, but many parents--poor and poorly educated--didn't know how to set them up. Edison's tiny technical staff was swamped trying to install, maintain, and repair them. Nevertheless, Edison continued for years to provide home computers to most students in its other schools. Whittle concedes that he was slow to abandon the idea. "We couldn't afford any early failures," he explains, so "our schools were oversupported and overengineered at the beginning." He adds, "The costs of classroom technology--purchases and support and networking--are close to the entire losses of the company." That's more than $338 million to date.

Other losses are more recent. "I told Chris Whittle he wasn't going to make any money here," says Thomas Persing, a gruff ex-Marine who heads the school district in Chester, Pa., a deeply distressed town just outside Philadelphia. "But he said that if Edison could take Chester--urban, troubled, black--and show academic success, he could use it as a marketing tool and really make big money." Persing figures Edison lost $14 million in Chester in 2001, the first year it was there. Edison says its startup costs were closer to $11 million.

What about all those efficiencies Edison had promised to find? It seemed to have a hard time locating them. When managers did attempt to cut costs, they sometimes made embarrassing, even dangerous, blunders. For example, in Philadelphia last October, Edison eliminated all the so-called nonteaching assistants who monitor students in the hallways. Disaster ensued: Fights and fires broke out, and police had to be called into one school to restore order until staffers were rehired. As for the company's central office in New York, Whittle acknowledges that its expenses--originally expected to run about 7% of revenues--have been considerably higher than planned: 23% of revenues two years ago and about 15% last year. "We are still working," he says, "to drive them down." Last July he vowed to slash the central office's then-300-person staff in a "reengineering." So far he's cut 45.

As if Whittle didn't have enough problems, last May the SEC declared that Edison was confusingly inflating revenues by including salaries paid directly by Edison-managed public schools to teachers. Edison argued that doing so didn't violate Generally Accepted Accounting Principles. After spending millions to defend itself, the company backed down. It agreed to restate revenues, lowering them by a total of $77.6 million over the past four years.

You'd be a lot more willing to overlook Edison's financial results if its educational results were outstanding. The company insists that its students' improvements on standardized tests are substantial and significant--an average of five to six percentage points a year--compared with one to two percentage points at most big urban schools. But to many, that difference just isn't dramatic enough. In Boston's Renaissance school, "the superior academic performance that the board sought never materialized," says Dudley Blodget, the school's COO, who nevertheless takes pains to praise Edison's extraordinary effort. "We were moving a lot of children from low achievement out of that category, and doing that more than most schools. But we were not moving children up to a higher level as much as we'd like." Renaissance terminated its contract with Edison, effective next June--three years early.

The Dallas school board announced last August that it was booting Edison prematurely too. The problem there wasn't declining test scores; they were up at six of the seven schools Edison runs. It wasn't that the schools were shabby either. A tour of Medrano Elementary near downtown reveals a sparkling place with friendly, uniformed students walking quietly, single file, hands behind their backs. The school board was unhappy because, in standardized tests, kids with similar backgrounds at other schools in the area did even better than kids at Edison. "I really wanted Edison to succeed," says Ken Zornes, president of the school board, who notes that Edison schools cost the district 10% more than other schools because it has longer hours. "If they proved their concept, it would have been worth it. But we did the numbers."

The biggest blow for the company this year came in April. Whittle had spent years negotiating with Pennsylvania governors Tom Ridge and Mark Schweiker, both staunch Edison supporters, and state education officials about taking over the entire 264-school Philadelphia district. But Philly mayor John Street wasn't so keen, threatening to barricade himself in a school to protest a deal with Edison there. Edison wound up getting a contract to manage just 20 schools, a fraction of what it had been hoping for. Investor confidence collapsed, sending the stock as low as 15 cents a few months later. Meanwhile, cash-strapped Edison had to scramble to raise money at a whopping 12% interest rate to buy supplies for those schools. (The Philly public-schools superintendent held up payments to the company until October, when he extracted a promise that the district could keep the textbooks and supplies if Edison went belly up.)

Perhaps the simplest reason Edison doesn't work, though, is that for-profit education just isn't a very good business. Though a dozen or more competitors have sprung up in recent years--most much smaller than Edison--about the only profitable ones are $15.6 million Nobel Learning Communities, which runs mostly private schools, and National Heritage Academies, which runs 28 quasi-religious schools. For-profit schools are "low margin no matter how you slice it," says Steven Wilson, who started Advantage Schools in 1996. Economies of scale are almost nonexistent, says Columbia's Levin: "Schools have very high variable costs, and as school systems get larger than 6,000 students they actually become less and less efficient." Edison, ten times bigger than the optimum-sized district and with schools scattered thousands of miles from its New York headquarters, has especially high costs, he says. Then there's the quality of the education. For-profit schools have to be orders of magnitude better than their public-school rivals in order to overcome the political opposition that confronts them. And so far that hasn't happened.

No one doubts that Whittle is passionate about his educational mission. "The changing of human institutions," he says, "happens to be one of the most important ideas." The team he's assembled seems to be equally committed. "I have a chart I stare at in bed at night," says Richard Barth, Edison's senior vice president in Philadelphia. "There's an elementary school in the city with 1,200 students, where only 3.5% are proficient in math. I will change this or die trying."

At the same time, no one doubts that Whittle would have loved to get rich from Edison. His holdings include a mansion in the Hamptons and a townhouse on Manhattan's Upper East Side. His chic Italian wife is a member of Fiat's wealthy Agnelli clan. And one of his former companies had two jet pilots on the payroll. Whittle is a complicated fellow, says a former colleague. "Is he a schemer or an idealist? I think he is both. Chris is a dreamer with a highly developed business sense: seriously interested in education and seriously horrified by what he has seen of public schools."

Edison's decline has left Whittle, who holds more than 14% of the company's shares, financially soaked. "I own only one stock," he says with a sigh. Last summer he was forced to put his huge Hamptons house on the market, for $45 million. "That was painful," he says. "It's where my children grew up." It remains unsold. Meanwhile, the company Whittle founded continues to bleed, prompting some analysts to predict that if things don't pick up dramatically next year, Edison won't be around much longer--at least in its present form.

But you won't hear that from Whittle. "Things are much better than you might guess," he insists. He points to growing revenues and growing enrollments: Edison currently educates 80,000 students--6,000 more than last year. He talks of scaling back the company's growth to save startup costs at new schools. He predicts that a new federal law, No Child Left Behind, which requires school districts with failing test scores three years in a row to change management, will boost demand for the company's services. Edison will show a profit in the fourth quarter ending next June, Whittle insists. "There are 15,000 school systems out there," he muses. "If we get 1%, that will make us a FORTUNE 500 company." Chris Whittle is clearly one die-hard dreamer.

FEEDBACK boreilly@fortunemail.com