THERE IS MORE than a little plausibility to the plan for school vouchers that the Reagan administration has now devised. The federal government provides special aid to children in low-income neighborhoods whose academic progress is below average. The administration proposes offering a voucher representing the cash value of that aid -- about $600 a year on average -- to the parents of such children to help pay tuition at any other school, public or private, that they might prefer.

The arguments in favor have to do with both fairness and reform. It is said that a voucher plan would give the poor more power over the education of their children, put poor families on a more nearly equal footing with the better-off in this respect by giving them the means to choose. Proponents believe that the plan would also deliver a healthy shock to the public schools, by depriving them of their near-monopoly status in poor neighborhoods and forcing them to compete. The theory and politics are both enticing to a program-cutting free-enterprise administration that has frequently been on the defensive on educational matters over the last five years. Vouchers put the president on high ground.

But there is another side to the idea, if one assumes that many families would take advantage of it. It would constitute an important transfer of resources. Badly needed money, already in limited supply, would be syphoned away from the public schools in the most difficult neighborhoods in America. As a practical matter this money would find its way instead to church-related schools, for only churches are sufficiently generous to bring tuitions within the reach of low-income families with vouchers of only $600 a year to spend. Even public schools charge more than that. For an out-of-state elementary ol student, for example, Prince George's County charges $3,000 a year. The Supreme Court earlier this year struck down a long- standing practice under which federal funds for the poor were shared with parochial schools. Vouchers would be a broad new conduit for this aid.

It is true that some parents would be helped to buy their way out of bad schools under such a system. It is also true that in some circumstances a child will be far better off in a private than a public school. Private schools have the advantage of being able to exclude all those children who do not fit, who are difficult -- the disruptive, the unusually slow, those with special problems. Public schools are required by law to accept all students. That is their burden, but their virtue and claim on resources as well. It is the public schools that most poor children will continue to attend.

The after-inflation value of aid for poor children has already been allowed to decline in recent years. It is hard to make the case that a further chipping-away at this aid would be of benefit to the majority of poor children in whose name vouchers are proposed. Some form of public aid to private schools has long been an administration goal. It is not clear that such aid on any large scale is either constitutional or, in this society, wise. But in any case the existing program for the poor ought not be the means for achieving it.