Baltimore school officials, stymied in their search for ways to reduce truancy now running at between 17 and 28 percent a day, have recommended cutting welfare benefits for families whose children are frequent truants.

It's an interesting idea, proposed for very good reasons. But is it a good idea?

The panel of secondary school officials that reluctantly proposed it obviously thinks it is. "The only way the welfare cycle is going to be broken is if these children get an education," said Boyce Mosley, a member of the panel.

Mosley's intentions are beyond reproach. He means to do what he can to help break the cycle of poverty. He said he didn't see the proposal to have the city's social services department dock the families of truants as a violation of anybody's rights.

"Welfare is a kindness extended by a benevolent government to unfortunate people," Mosley said.

On the other hand, there are those who see the proposal as unfair and probably unconstitutional. They argue that welfare, unlike private charity, is a right and should not be used to reward or punish recipients.

I remain ambivalent, even after applying the touchstone I devised to help me handle just such questions of public policy: Is it something I would propose for people I care about -- my own children, for instance? If not, I won't propose it for anyone else.

If my own children made a habit of skipping class, would I cut their allowances? You'd better believe I would. Well, suppose I had two class-cutting children, one who got an allowance from me, the other who earned his own spending money. If I cut the allowance of the first one, how would I punish the second for the same offense?

That's one of the problems with the Baltimore proposal. It punishes poor children but not those from self-supporting families. Is that fair?

It may also be unfair to assume that hard-pressed parents of welfare families always have effective control over their adolescent children. The technical answer to that one is that a truant child who insisted on jeopardizing the family income could always be declared "beyond parental control."

But severing even a tenuous link between child and family may not be in anybody's best interest.

Still, there is something to be said for any proposal that drives home to both the children and their families the fact that actions have consequences. Indeed, one of my frequent criticisms of the way welfare operates in America is that it tends to reward bad choices and punish sensible ones.

The Baltimore proposal might force the parents of some of the city's highest-risk children to start acting like parents. It might also arm them with one means of enforcing their parental authority.

On the other hand, you have to wonder how academically useful it would be for a child to attend school under such duress. He might learn something just by being there, but would it be enough to justify the degree of intrusion the proposal would demand?

My own notion is that the best way to get young people interested in school is to have them see the benefits of education. Those benefits are hard to see in inner cities, where both high school graduate and dropout often find themselves scrounging for temporary, ladderless jobs, or hustling to survive in the underground economy. If these youngsters could be made to believe, on the basis of real evidence, that there was a payoff for academic diligence, it might not be necessary to coerce them into school attendance.

But wouldn't regular school attendance for any reason be better than truancy? No doubt. But that's assuming that the coercion would work. Greg Conderracci of Baltimore's Associated Catholic Charities doubts that it would. He suspects that the principal result would be a useless tangle of red tape.

"People seize on ideas like this because they think it will be cheap, easy and effective. We don't think it will be any of these."

He may be right. Still, you have to be grateful that the city's school officials recognize the link between ignorance and continued poverty and are ready to move, however desperately, to do something about it.