Since school-level autonomy seems to be so important for effective school organization and performance, how is it that some schools have autonomy, but most do not?

To aid us in figuring out how America's schools might be given more autonomy, we investigated why some schools already enjoy it. Much as we concluded when thinking about how schools could be led to organize effectively, we decided that school autonomy was probably not a virtue that would come to schools just because researchers or reformers thought it was a good idea. Rather, it seemed that autonomy stood a better chance of being increased if the forces that reduced it were understood and then attacked. Thus, we examined a number of factors that we suspected would influence the degree of autonomy that a school would experience. The results support two generalizations, one about public schools, the other about private.

Public schools are given relatively high levels of autonomy only under very special conditions. All things being equal, public schools will fall at least two quartiles below private schools in autonomy from external control. To enjoy the kind of autonomy that the private school receives on average, the public school must exist in the most favorable of circumstances. To be permitted to control its own destiny, the public school must be located outside of a large city in a suburban school system. Its students must be making significant gains in achievement, and its parents must be in close contact with the school. In other words, when the public school is performing well, is being monitored by parents, and is not part of a large administrative system, it will be given relatively great control over its policies, programs, and personnel.

Unfortunately – and predictably – the public schools that now enjoy autonomy are not the ones that are most in need of improvement. And the inner city public schools that most desperately require improvement are the ones that have so little of the autonomy they arguably need. It may even be that urban public schools are caught in a vicious cycle of deteriorating performance, increasing control, and eroding organizational effectiveness. Under political pressure to do something about city schools that are failing, school boards, superintendents, and administra­tors tend to take the only actions that they can. They offer schools more money, if it is available, but then crack-down on underachievement with tougher rules and regulations governing how teachers must teach and what students must "learn." But crackdowns are seldom carried out deftly. And any intervention that responds clumsily to the real needs of teachers and students may undermine school organization rather than build it up.

Private schools, even in urban systems with high percentages of poor students, generally do not face these troubling pressures. Private schools, almost regardless of their circumstances, tend to be free from excessive central controls by admin­istrators, boards, and unions. The main reason appears to be market competition. In a process much the reverse of the one in public schools, where political pressure leads to an increase in central control, competitive pressures lead to an increase in autonomy in private schools. To stay in business private schools must satisfy parents, and satisfy them more than the public schools or alternative private schools. Private schools are therefore forced to organize themselves in ways that above all else respond to the demands of parents. One thing this clearly means is that private schools must vest a lot of control over vital school decisions – about personnel and curriculum, for example – at the school level where the wishes of parents can be more clearly perceived and accommodated. Strong external control is incompatible with the imperative that private schools either satisfy parents or lose them to other schools. In contrast, strong central control fits public schools very nicely. Public schools need not satisfy parents first; indeed they must ensure that parents are not satisfied at the expense of other legitimate groups such as unions, administrators, and various special interests. Policymaking is therefore taken out of the public schools themselves where parents would have a political edge.

Because public schools are ruled by politics, and private schools by markets, public schools may be at a decided disadvantage in developing effective organi­zations and promoting student achievement. Private schools, without the benefit of any reform at all, are encouraged by competitive forces to operate autonomously and to organize effectively. And indeed, the private schools in our study have more of the attributes of organizational effectiveness than public schools, regardless of the quality of their students. Public schools, however, are usually not granted the autonomy that they need to organize effectively – political forces discourage this – and must therefore be periodically reformed from the outside.