With its expiration last month, Section 8 joins a long line of government programs that failed to answer one of the nation's most intractable public policy questions:

How do we provide housing for people who cannot pay for it?

Over the past 45 years, the government has tried almost every conceivable variation of subsidy and incentive. All seemed promising at the start, but none proved to be the answer.

In a sense, Section 8 came closer than most. It produced units. Indeed, it produced hundreds of thousands of units. But it also produced two other things -- scandal and envy -- and they proved to be its undoing.

The Section 8 program, following as it did a long line of unsatisfactory programs that had tried to produce housing at minimal cost to the government, grew out of the conviction that there just wasn't any way to do the job cheaply. Housing is expensive to build and troublesome to maintain, the thinking went, and the only way to produce it is to make it very profitable.

And that is what Section 8 did. It worked this way:

First, the Department of Housing and Urban Development calculated the "fair market" rent for various sizes of new apartments in different parts of the country. This figure became the standard for what developers could charge for apartments built under the program.

Then HUD provided subsidies to tenants so that the developer obtained the full "fair market" rent for his units, though no tenant was required to pay more than 25 percent of his income in rent. This had the effect of guaranteeing the developer his rent roll.

Then HUD lent the developer 90 percent of his costs. But since the program allowed a 10 percent profit, and since HUD was willing to accept that paper profit as a down payment, the loan in practice could amount to 100 percent of the cost.

Finally, HUD allowed the developer to sell the tax shelter aspects of his project -- depreciation and the like -- for cash, and at the end of the mortgage term to withdraw his building from the program.

It sold like hotcakes.

It was so popular with developers, in fact, that it wasn't long before charges of bribery and influence-peddling became commonplace. In places like Rhode Island, lists of developers read like a political who's who in the state.

At the same time, the amenity-laden units that were being built drew fire as luxury housing for poor people.

Because of the way the loans and rent subsidies worked, there was no penalty for running up costs. And because the building could later be moved onto the private market -- after only five years in the early days of the program -- builders tended to add luxury features.

The result was publicity about "plush" apartments, and complaints from middle class people who felt their own homes weren't as nice. These complaints translated into the kind of political trouble that eventually brought the program to an end.

But if a lesson learned from earlier programs was that housing cannot be had cheaply, then a lesson learned from Section 8 is that a free-spending program is likely to buy trouble.

Escaping this trap will be difficult. Much of the so-called housing lobby is concerned solely with providing shelter for the poor. Other parties, however, are interested in making money. There's nothing wrong with that; the people who build houses cannot be expected to work for free. But early programs, reflecting congressional concern about profiteering, often reduced the profit potential to the point where many reputable builders weren't interested. If this view returns in the wake of the Section 8 demise, it seems unlikely that whatever program emerges will fare any better than its ancestors.

The administration, in keeping with its philosophy of returning welfare responsibilities to state and local governments, likes the idea of giving block grants to localities. That has the advantage of removing responsibility from the federal government. But localities have widely varying interests in low-income housing -- the interest is plainly zero in many areas -- and such a program is likely to follow those interests to a much greater extent than a federally controlled plan.

The Democrats, on the other hand, tend to favor a strong federal hand, both for the added control and for the added political capital they can derive -- the lead time on housing being such that any fallout is almost certain to be farther down the road than the next election.

So the prospects are not encouraging.

Housing expert Gertrude S. Fish has noted that "it took about 100 years from the time that a housing inspector defined bad housing conditions a social problem (1834) to the time when society voted to provide some units for those unable to attain them for themselves (the Public Housing Law, 1937)."

After 45 more years, the best that can be said is that the provision of low-income housing is now accepted as one of society's responsibilities. Agreement on how to meet that responsibility is not yet in sight.