Eight years ago, two Stanford political scientists set out to explore a theoretical question about public education. They wanted to know what difference it makes that public school systems are controlled by elected officials.

John E. Chubb and Terry M. Moe deliver their blunt answer in a new book published by the Brookings Institution, "Politics, Markets and America's Public Schools," which is based on an survey of 20,000 principals, teachers and students from 500 schools.

Basically, Chubb and Moe answer that politics makes no good difference in public schooling. They recommend a radical solution: Not only would they let any student choose any public school, but they would also let almost anyone operate a public school, ending the geographic monopolies of school districts.

The researchers concluded that public bureaucracies that control education block the formation of effective schools. They propose "creating a system that is almost entirely beyond the reach of public authority." Chubb and Moe estimate that the constraints imposed by central offices and union contracts cost a high school student the equivalent of a year of study.

Chubb, now a senior fellow at Brookings, said in an interview that school bureaucracy is "not just a problem, it's the problem. It's kind of a good news finding. If bureaucracy is the key problem, that's a problem the public policy-makers can control."

The endorsement of public school choice -- which has been primarily advocated by the Bush administration and conservatives -- has drawn attention partly because the research was financed mostly by Brookings, which retains a reputation for being a liberal think tank. The Education Department and private foundations also supported the study.

"We welcome converts," said Herb Berkowitz, spokesman for the conservative Heritage Foundation. He described the book as "full of all sorts of things the Heritage Foundation has been talking about for years, full of school choice, vouchers and all that good stuff."

One critic of the Brookings study, California Superintendent of Public Instruction Bill Honig, challenged the conclusion that schools cannot improve without radical surgery on bureaucracy. He cited an increase in test scores, equivalent to a year of study, registered by California seniors in the last five years after improvements in curriculum and teacher training.

"Theirs is just all theory," Honig said. "We've got that gain another way. . . . There are other ways of building in incentives {to improve schools} without taking these radical steps."

Chubb and Moe found that a high degree of autonomy prevailed at effective schools, which are described in other education research as those having strong principals, high expectations of students and a clear sense of purpose. The authors also found that school autonomy was likely to be present only at public schools located in small suburbs with homogenous populations. But autonomy is found at private schools anywhere.

"Private schools aren't working because of some magic. . . . They're working because they're focusing on the client {student} and are more effectively organized," Chubb said.

Their proposed overhaul is intended to introduce into public education the market incentives that shape private schools.

States would set minimal standards for public schools -- roughly analogous to the existing requirements for graduation, health and safety and teacher certification imposed on private schools. States would relinquish broad authority over other matters such as instructional hours, textbooks and tenure.

Conceivably, private schools could become public schools by signing a new charter. Other groups could organize new public schools and operate outside the control of existing districts. Each school could set its tuitions.

State and local funds would be rolled into "scholarships" approximating per pupil expenditures, except that states would reduce their contributions for students in rich districts and increase them for poor districts. State and federal funds for disadvantaged students would be added to their scholarships. Families could not supplement their children's scholarships, but an entire district could vote to raise its per pupil contribution.

Then through state-run "choice offices," students would get information on various schools and apply to them. Schools could set their own admissions criteria, "subject only to nondiscrimination requirements." The offices would send the scholarship money directly to the schools.

Chubb acknowledged the political difficulties of achieving such a system, which would crimp the power of local officials, teachers unions and state bureaucrats. But in another sense, he added, it would be easier to curb the government role in schools than to expand it.

"It doesn't require elaborate legislation. While it may seem radical, it's a fairly easy thing to do," Chubb said.