Every other week, it seems, another social scientist trots out a new list of figures to challenge or support the contention of Charles Murray ("Losing Ground") that the welfare system has increased poverty in America. The debate over which set of numbers is more accurate, or more accurately interprets reality, tends to obscure a point of fundamental agreement among conservatives and liberals and "neos" of both camps: that there really is a welfare dilemma.

The dilemma is this: It is unacceptable, particularly in a rich country, to allow anyone to live in hunger and squalor for want of money. But unearned payments designed to alleviate hunger and squalor have the perverse effect of making entry- level jobs unattractive. To discourage poor people from taking entry-level jobs is to condemn them to dependency and its attendant pathology (including the erosion of the work ethic), even while alleviating their hunger.

Recent years have seen a number of attempts to resolve the dilemma: work-incentive programs, negative income tax, workfare, and any number of other ideas have been broached and found wanting.

Leonard M. Greene, president of the Institute for Socioeconomic Studies, states the problem clearly enough. "Income transfers are not inherently wrong. Trouble arises from the form of the transfers used. By our well-meaning approach, if a welfare recipient takes a job, his family may lose its public housing and medical benefits in addition to suffering a reduction in welfare payments. The consequent mathematics of the situation are irresistible. We have destroyed the motivation to work and trapped people into a dependency as insidious as poverty itself."

He has also proposed an attractive way out: the Graduated Income Supplement, a single national income supplement that would go to all adults, with a smaller amount going to all children. The supplement would be taxable income, which means that it would result in extra income only for those poor enough to need it. Those too poor to pay taxes would get the benefit in cash.

Greene's proposal sounds a good deal like Richard Nixon's negative income tax, except that since the GIS would be universal, without any eligibility criteria, there would be no need for a welfare system to distribute it. The program would be administered by the IRS, "allowing dismantling of today's cumbersome and amazingly expensive welfare bureaucracy."

Under the proposal, a poor family could earn an additional $4,000 a year and be taxed at a maximum marginal rate of 14 percent. Under present law, the tax on the $4,000 might be above 100 percent, considering loss of eligibility for public housing and health benefits.

Greene believes his proposal would ameliorate another dilemma that is attracting a growing amount of attention: the fact that illegal aliens have more incentive for economic progress than native-born Americans.

"In fact, we have a dual economy. Native-born Americans receive many protections and benefits. Working conditions are generally good and fringe benefits, such as pensions and health care, substantial. At the same time, illegal aliens work in an environment little changed since the 19th century. Sweatshops abound. Families . . . do piecework, toiling around the clock at wages that can be less than $1 per hour.

"Yet, ironically, the sytem appears to be serving the illegal aliens better than America's underclass. Historically, waves of immigrants have worked their way up the American economic ladder, and many recent immigrants seem to be doing the same. The subminimum-wage jobs that they have attained offer opportunity -- a first step up the ladder that for many has already led to prosperity and success."

Greene, obviously, does not advocate a reinstitution of 19th century sweatshops. His point is that it is counterproductive and wrong that a poor person willing to work hard at unpleasant jobs should be subjected to a financial penalty -- which, unfortunately, is how the present system works.