The school board and the Washington Teachers Union have agreed to a contract provision that would deny sixth-year raises to teachers rated merely "satisfactory" by their principals.

Apparently they think that the prospect of missing out on a $700-a-year pay boost will transform mediocre teachers into outstanding ones and, like magic, improve classroom instruction.

They're trying to bell the wrong cat. If they want to improve the schools, the place to begin is with the principals. I give you a rule of thumb: if the principal is mediocre, the school will be mediocre--even if the faculty includes one or two outstanding teachers. If the principal is an outstanding educator, the school will be outstanding--even if the teachers are only average.

Actually, there are two pieces to this back- door merit pay proposal: equity and incentive. The first, which I have no trouble accepting, is that outstanding teachers--like outstanding practitioners of any discipline--are deserving of financial rewards. This is the idea behind the merit pay proposals that have been made, and defeated, across the country for as far back as I can remember. Teachers, and especially teachers' unions, don't like merit pay--not because they are interested in protecting mediocrity but because they fear such a system would open the door to favoritism and politics.

The local proposal--part of a contract package teachers have until the end of the month to accept or reject--seeks to come at the equity idea another way: by withholding pay raises from teachers not rated "better than satisfactory." (Under the old standard, the raises would be withheld only from those teachers rated less than satisfactory.)

The incentive piece is a ghost. It assumes that the financial threat is enough to induce mediocre teachers to outstanding performance --or else weed them out of the system. It will do neither. Incompetent teachers aren't incompetent because they wish to be but only because they don't know how not to be. The lure of $700 won't show them how. And if they could earn more in another field, they wouldn't be teaching in the first place. They won't leave; they'll stay in the system, incompetent and embittered.

You can see how the thing would work by looking at any private sector pursuit--newspapering, for instance, where merit pay is a matter of course. The contracts hammered out between reporters and their employers are base- pay agreements. It is accepted that a newspaper has the right to decide which reporters do their work uncommonly well, and to reward them commensurately. But if the base pay is high enough to attract them in the first place, they are unlikely to leave for failure to win merit increases. The reporter who either cannot or will not do outstanding work isn't likely to improve because of the prospect of merit pay.

Outstanding newspapers almost always are the newspapers with editors who know the business and know how to get the most out of their reporters, by some combination of instruction, encouragement and pressure. Schools, I suspect, work pretty much the same way.

School principals, like newspaper editors, set the standards. The crucial difference, which few school boards have been willing to address, is that editors have it in their power to get rid of incompetent reporters--and they will do so if their own future depends on it.

If it were left to me, I would install an idea proposed a year and half ago by school board member Eugene Kinlow: hire principals on fixed-term contracts--say five years at the outside--with the understanding that the performance of their pupils would determine whether the contracts would be renewed. Along with that burden would go a substantial grant of authority, including a significant say in deciding which teachers to hire, retain and reward for outstanding success.

Under the Kinlow notion, it would not be necessary for a principal to declare a teacher incompetent to get rid of her: only that she wasn't working out in her present assignment.

As a matter of fact, the private schools already operate pretty much on the Kinlow idea. Private schools have a lot of advantages over public schools, of course, including the right to select their students. But their teachers as a rule are no more gifted or better trained than their public-school counterparts. The crucial difference is that their principals know they will have to produce results or face unemployment.