The Census Bureau's recent cautious challenge to the ingrained idea of counting America's poor on the sole basis of cash income ignores several more serious flaws that come from legislative mixups between real people and statistics. There are, in fact, two "official" gauges of poverty in this country. They are both shockingly out of date.

Using 1980 prices, a nonfarm family of four was labeled "poor" by the conventional "poverty index" if its total annual cash income before Social Security deductions was less than $8,450. This was true whether the family lived in a high-cost area such as Washington or in Louisiana, where rent, food and energy expenses were significantly lower. Yet some federal assistance statutes use a different criterion for determining need, adjusting the minimum income level for geographic variations; and, according to the latter technique (the Bureau of Labor Statistics' lower living standard), such a family was considered "poor" in an urban metropolitan area whenever its income fell below $14,044.

Which is correct? Are there about 24 million poor Americans, or is the number closer to 45 million? Either estimate considers only cash income, and the Census Bureau's latest observation is that taking account of these items might limit the total to around 14 million.

As the high priests of welfare economics trade arcane imprecations over the matter, the average citizen should recognize that neither the "poverty line" nor the lower living standard (LLS) is all that it is made out to be.

Essentially, the poverty index is drawn from an idea of the 1950s that total household expenditures averaged about three times what it cost to provide a subsistence diet for its members. Now, however, the ratio is probably closer to 5 to 1 nationwide. And even an across-the-board adjustment to this effect ignores regional discrepancies.

The LLS does distinguish one area from another, but it is nearly as anachronistic as the poverty index in another respect. All of its fine tuning is based on the "market basket" of goods and services that was typical in 1960-61--when a good many home furnaces still burned coal, when personal transportation requirements were quite different, and before a steep rise in the number of working women changed the whole connotation of a household's "market basket." We live in a new world, and it is uncertain how well a 20-year-old measuring stick can move across the time warp to tell us who today's "poor" are.

Could the much-quoted consumer price index help? Unfortunately, not much. People who cite the CPI habitually overlook the warning that it does not even purport to measure "cost of living," but only inflation. In its case, the base reflects people's habits a "mere" decade ago. As U.S. energy consumption shows, however, it takes far less time for many to modify their use of a price- volatile commodity.

Indeed, energy has probably been the most upsetting single factor in any effort to use old statistical tools to enumerate the "truly needy." At least until free market policies clear away the remnants of artificial pricing, the occupants of two identical row houses in the District may face wildly different degrees of difficulty in stretching identical incomes to cover the operating costs of an old oil burner and a new gas furnace. Does that mean one family is poor and its neighbor not?

One part of a possible solution to the poverty dilemma is simply to recognize its true complexity. Another is to make evaluations of need--and to devise appropriate responses--as close to the individual as possible--here is one of "the new federalism's" strong cards. A specific recommendation is to reevaluate the shortsighted decision to riddle the Bureau of Labor Statistics with reductions-in-force just when BLS is nearing completion of an up-to-date system to measure the true cost of living.

To some extent, "poor" is a relative term --even a subjective one. In that sense, the poor will always be with us. However, a genuine effort to guarantee some aid-of- last-resort to "the truly needy" is both humanitarian and pragmatic. It cannot succeed unless we know who those people are.