NOW THAT President Carter has postponed his tax-reform plan, the administration has a chance to tie it more closely to employment patterns. There has been rising anxiety in Congress over the shift in the country's economic balance away from the Northeast, toward the South and West. The drift of population out of the big aging cities of the Northeast imposes one kind of cost, the rapid expansion of the new cities of the Sun Belt imposes another kind. Neither is small.

Rep. Al Ulmann (D-Ore.), chairman of the House Ways and Means Committee, has begun to think about the possibility of using te tax system to slow down the drift. The idea has been floated before, but in the past it has been shouted down by the boosters among us - the states that were growing fastest and the people there who saw a soaring population as a source of pride and wealth. In this decade, some of those states have changed their minds about the desirability of that kind of growth. Mr. Ullman's Oregon is a leading example. The constituency for a stabilization policy sems to be increasing.

In 1950, 32 out of every 100 American jobs were in the Northeast. By 1975 it was down to 24.5 jobs. In 1950, 24 out of every 100 jobs were in the South; in 1975 it was up to 30. The figures come from a study that the Library of Congress did for Sen. Henry Bellmon (R-Okla.), who was concerned that the Sun Belt-Frost Belt debate was outrunning the data. All quadrants of the country have become richer and more heavily populated in those 25 years, but they grew at very different rates. In those years employment nationwide rose by 31 million jobs. Out of that total, there was a gain of 4.4 million jobs in the Northeast, which here includes Pennsylvania and everything beyond it. The gain in the North Central states was 7.1 million jobs, and in the West 7.6 million. In the South, it was 12.2 million.

But even this kind of statistics understates the reality. Within each of these broad regions the local disparities ar far more extreme. One great illustration of the process is the abandonment of wide stretches of some of the inner cities, most notably in New York's South Bronx.

The hard question ere is not whether to try to stabilize employment where peopel now live - but rather how to do it. In this newpaper's Outlook section last Sunday, James Sundquist of the Brookings Institute suggested an expanded investment credit in the areas of highest unemployment. That might be a beginning. But it's difficult to think that an investment credit alone could determine the location of many plants and payrolls.

Effective intervention would be expensive. But it could hardly be as expensive as tolerating complacently the decay of regional economies in one part of the country, while localities in other parts urgently build new schools, new roads and new hospitals for the people moving n. It would be impossble to stop the long-term changes in American economic geography. But it would be useful to pause for another look at the rules of the game, to ensure that the advantages for each employer do not add up to the disadvantage of the country as a whole.