THE URBAN AFFAIRS task force advising Ronald Reagan has suggested replacing the current housing programs for the poor with a system of vouchers to help them pay their housing costs. That is an idea that has been around since the Nixon-Ford days and, like most housing programs, has had a checkered career. But there is this to be said for the idea: a lot is known about it, thanks to a series of housing allowance experiments begun in the 1970s.

What do those early experiments show? First, if the main purpose of housing programs for the poor is to improve the quality and availability of housing for them, then the results of the experiments must be seen as discouraging. When participants were given allowances to pay for any housing they chose, some of them did improve their housing. But most simply stayed where they were and used the benefits to cover housing costs they had been paying previously from their own pockets. The freed-up money was then used to buy other things the family needed.

This is not a bad result in itself -- although other experiments show that the same result, including the same average improvement in housing, can be produced with much less administrative cost by simply giving the poor money. But it is surely not what housing program advocates usually have in mind. They want the money to be used to buy better housing, not other things.

Other plans tested tried to circumvent this problem by limiting benefits to families who were able to find housing that met certain standards. But establishing standards and monitoring dwellings to make sure they qualified turned out to be an administrative hassle. And anyway, most of the eligible families weren't able to find better housing, even with the vouchers, especially minority families and those with the biggest families and the worst current housing. In fact, most of the benefits ended up going to families already in relatively good housing.

Still, there were some pluses. The program did not seem to cause landlords to raise their rents, as had been expected. And the program also seemed to have an important effect on encouraging families in aging but still closely knit neighborhoods to keep up and improve their homes. The fact that the program covered not only renters, but homeowners as well, made it comparatively effective in reaching older people.

In judging housing programs for the poor, the fact is that one is choosing among a set of poor alternatives. No one would claim that any of the variety of approaches that have been tried -- public housing, subsidies to private builders, grants for rehabilitation, loans for low-income house buyers, to name a few -- has been a roaring success. As a swap for some of the less efficient, more heavily red-taped of these programs, a modest program of housing allowances might conceivably not be such a bad idea. It wouldn't build any new houses for the poor, but it might at least provide some of them with the chance to buy better housing. And, if it were fairly distributed among all the poor -- not just those lucky enough to find good housing -- it could prove the best substitute available for a more adequate system of income support for the poor. s