The prospect of balancing the federal budget is now foundering on the mistaken belief that further cuts in nondefense spending must hurt the poor. Although everyone recognizes the importance of decreasing the huge budget deficits that are projected for the next several years, there remains substantial opposition to the necessary reductions in nondefense spending.

Much of that opposition reflects a genuine concern that spending reductions mean cutting the benefits now provided to low-income households. Fortunately, that's not true. Nondefense outlays can be substantially reduced without taking anything away from the poor.

The administration's budget is faulty because its proposed cuts exclude the major programs aimed at middle- and upper-income families. The administration has not suggested any significant changes in the Social Security retirement, disability and Medicare programs that currently account for some 40 percent of nondefense spending. Because these programs provide benefits without regard to a recipient's other income, a very significant portion of these benefits goes to families that are not poor. Substantial savings in government outlays can be achieved by gradually reducing the benefits of these nonpoor families.

It is very important to recognize that the truly poor can be exempted from these benefit cuts without significantly impairing the overall savings that result from the general reduction in benefits.

Consider the Social Security program of benefits for retired workers and surviving dependents. This is the largest government program with expected outlays of nearly $140 billion.

In the decade after 1970, the average benefits per retired worker rose 50 percent when measured in dollars of constant purchasing power. During the same period, average constant dollar gross weekly earnings of private nonagricultural workers were essentially unchanged, rising less than 2 percent over that entire period. Thus the ratio of average retiree benefits to the average earnings of those who pay Social Security taxes rose nearly 50 percent. This increase represents a striking departure from the relatively constant ratio of benefits to earnings since the basic pattern of Social Security benefits was established in the years just after World War II.

The jump in the relative size of Social Security benefits has added more than $40 billion to 1982 government outlays and thus to our current and projected deficits.

I think the growth of Social Security benefits should be gradually slowed until the traditional ratio of benefits to earnings is re-established. Last fall, I proposed to the Joint Economic Committee that the slowdown should be achieved by limiting the inflation indexing of benefits to the excess of inflation over 2 percent. Thus, if inflation is 7 percent, benefits would automatically rise by 5 percent instead of by the 7 percent that is provided in current law. A couple that now receives a monthly benefit of $600 would see those benefits increase by $30 instead of by $42.

There is no reason why the inflation "deductible" should be 2 percent rather than the 3 percent recently suggested by Sens. Pete Domenici and Ernest Hollings or the 4 percent proposed by Sen. Rudy Boschwitz. The basic idea is to have a rate that is small enough to avoid significant personal dislocations from year to year but large enough to have a substantial cumulative effect on aggregate outlays over a period of several years. If an inflation-minus-three rule for adjusting benefits were adopted now, benefits in 1987 would be about 15 percent lower and the deficit that year would be reduced by about $30 billion.

Most of this $30 billion savings would still be achieved if those who are truly poor are exempted and continue to receive the full inflation adjustment. According to the most recent data published by the Social Security Administration, less than 8 percent of the 7.9 million families with a head aged 65 or over who received Social Security benefits in 1978 were classified as poor. In addition, 25 percent of the 7 million individual aged recipients were classified as poor. Since the poor obviously receive lower average benefits than the nonpoor, completely exempting the poor from the indexing change would diminish the cost saving by less than 15 percent. The reduction in total outlays would still be more than $25 billion in 1987.

Exempting the poor in this way would require obtaining information on the combined benefits of husbands and wives and on any other income that they receive. A less accurate but administratively simpler method would be to have full indexing for monthly checks below some amount and reduced indexing for checks above that amount. Completely exempting the lowest 25 percent of monthly benefits would only reduce the total savings by about 12 percent. Instead of a $30 billion reduction in outlays, the savings would be about $26 billion.

Whether low-income recipients should be completely exempted from the reduction in benefits is a separate question that Congress must decide on the basis of its own views about income redistribution. But as a technical economic matter, it is clear that a major savings in Social Security outlays could be achieved without affecting the poor at all.

The same is true about other so-called entitlement programs. For example, requiring Medicare patients to pay a modest share of their hospital charges in the same way that patients with private insurance do would not only reduce the $45 billion Medicare outlay but would also help to control the inflation in hospital costs. Medicare beneiciaries with low total incomes or low monthly social security checks could be exempted from the copayment requirement without significantly diminishing the overall cost saving.

The rapid growth of nondefense spending over the past decade is the primary reason for the vast projected deficits that now threaten our economy. In 1970, federal nondefense spending accounted for 13 percent of the gross national product. This year, such spending will be 18 percent of the GNP. Without this increase in nondefense spending, there would be no deficit today and no deficit in 1987 after the personal and business tax cuts were fully phased in.

It would, of course, be possible to accommodate the increase that has occured in nondefense spending by permitting taxes to rise substantially. Fortunately, Congress voted last year not to let that happen. The 25 percent reduction in personal tax rates between 1981 and 1984 will be just about enough to prevent bracket creep from making taxes in 1984 take a larger share of personal income than they did in 1980.

Unless Congress wants to reverse the decision it made last year and allow taxes to rise substantially, it must find ways to reduce significantly the share of GNP that now goes to nondefense spending. It is important, therefore, to know that major reductions in the broad income-transfer programs can be achieved without hurting the poor. A concern for the poor is not a legitimate excuse for the failure to prune back the overgrowth of government spending.