Suppose you were told that the president would agree to a budget-balancing plan that exempts over half of total spending from cuts and requires that half the cuts come from defense. Further, suppose you were told that the plan would require cuts in affected areas of 10 to 15 percent over 4 consecutive years, while outlays in unaffected programs such as Social Security would rise steadily at about 7 percent a year.

Given these facts, you might suppose that a deficit reduction plan weighted against defense and in favor of Social Security might be justified by an imbalance between defense and social spending that had grown up over past years.

But consider the facts: since 1967, the year before Lyndon Johnson's guns-and-butter budget proposed a record peacetime deficit of $25 billion, the share of GNP devoted to entitlements -- including Social Security, Medicare and Medicaid -- has risen from 5.9 percent to 11.3 percent in 1985. Meanwhile, the share of GNP devoted to national defense has fallen from 9.2 percent in 1967 to 6.5 percent in 1985. Nondefense discretionary spending that finances government operations, grants to state and local governments, and research and education programs aimed at the young has stayed between 4 and 5 percent of GNP.

A clear pattern has developed. Defense spending is cyclical. It rose during the Vietnam conflict and fell sharply thereafter until 1981 when it reached 5.5 percent of GNP. In contrast, the growth of spending on entitlements has followed a steady, upward trend. During the decade between 1957 and 1967, spending on entitlements rose from 3.9 percent of GNP to 5.9 percent. It rose much more rapidly during the eight years after 1967, reaching 10.6 percent of GNP by 1975 and rising to over 11 percent of GNP in the 1980s.

It i the growth of spending on entitlements and other social programs that has made the guns-and- butter budget option far more expensive. Even Lyndon Johnson's 1968 guns-and-butter budget set outlays at 21 percent of GNP. In 1985 the guns-and- butter option, with defense at 6.5 percent of GNP instead of 1968's 9.8 percent, cost nearly 25 percent of GNP. The difference absent a tax increase has meant deficits running at about 5 percent of GNP.

The president's budget strategy has been to engineer a cyclical buildup in defense spending while shrinking overall spending to about 20 percent of GNP. This plan has not worked given the upward march of spending on economic entitlements, the failure to achieve sustained rapid economic growth and congressional resistance to virtual elimination of discretionary nondefense spending. Ruling out tax increases has disciplined neither Congress nor the president to slow spending growth enough to avoid large deficits that threaten the economic recovery and pile up overwhelming burdens on future generations.

As a measure that largely exempts from cuts the entitlements that constitute the fastest growing portion of spending, Gramm-Rudman is a bizarre piece of legislation for the president to have signed. It overlooks the historical anomaly of the deficit problem. Entitlements as a share of GNP have nearly doubled since the last defense buildup in the late 1960s. If he insists on continued growth in defense while agreeing not to slow the growth of spending on entitlements, the president has in effect ordained a tax increase.

There are three things that must be done to end the deficit dilemma. First, we must end the boom-and-bust cycle in the defense budget and acknowledge that defense spending should grow steadily at a real rate of 2 to 3 percent a year. That range is consistent with the real growth of the economy and consistent with the best estimates of the real growth of defense spending b the U.S.S.R. Second, we need to address the overly generous entitlements programs by omitting every other year the cost-of-living adjustments. A one-year COLA freeze for Social Security and other entitlements would save $80 billion over five years. A two-year freeze would save $185 billion over five years.

Finally, we must make a difficult choice. Without a return to irrational boom/bust defense spending, we must choose to moderate growth of entitlements for the elderly, sharply cut education and research programs for the young, raise taxes or some of each.

A sensible "some of each" formula would include the biannual COLA adjustment for entitlements, 2 to 3 percent real growth in defense spending and a 5 percent consumption tax. The consumption tax would yield about $100 billion annually, enough to put $15 billion toward deficit reduction and still lower tax rates to 30 percent or below.

Low-income taxpayers could be exempted from the consumption tax with tax credits. The low rates would provide an incentive for younger workers to concentrate on productive effort instead of searching for tax loopholes. A consumption tax removes the double tax on savings and helps to raise our low savings rate, thereby slowing our rapid rise in debt. As a levy on spending out of accumulated assets concentrated in the hands of well-to-do and older Americans, a consumption tax represents a transfer from old to young. That is a fair way to help pay for the rapid growth of entitlements of the last 20 years.

Gramm-Rudman will fail to survive in its current form. Its goal of a zero deficit in five years is unnecessarily harsh, and it precludes cuts in the fastest growing portion of spending. If Congress and the president would begin this year's budget cycle by remembering how we got into this deficit mess, they might find constructive ways to compromise. Neither Americans nor the rest of the world will be well served by a repeat of last year's budget fiasco.