(FORTUNE Magazine) – IS THERE an economic phenomenon more frustrating than the federal budget deficit? For a decade it has mocked us, defying all efforts to eliminate it -- from the Gramm-Rudman-Hollings Act of 1985, which promised a balanced budget by 1991, to last fall's Budget Enforcement Act, which is supposed to trim the deficit to zero by 1996. Now comes the news that not only is the deficit rising, but in the next fiscal year, which starts October 1, it will set a record of $348 billion, far higher than the deficits that had us so worried in the mid-1980s.

Should we start worrying again? After all, the deficit decade produced none of the catastrophes that many experts predicted. Private investment was not crowded out, inflation didn't reignite, and the dollar didn't collapse. But the U.S. did pay a price. Because of those past deficits, real interest rates were higher and national saving and investment lower than they would otherwise have been, which translated into slower productivity growth and less abundant lives for everyone. If last year's budget act works as advertised, the deficit should soon resume its downward path -- but that is an if of Himalayan proportions. Should the deficit instead stay high, the federal debt's share of GNP will climb to altitudes not seen in decades. The inevitable result: even higher real interest rates and even lower investment. So by all means worry. But don't fret too much about next year's huge numbers. The deficit's size is distorted by the effects of the Gulf war and the S&L bailout (see chart). The war actually reduced the fiscal 1991 deficit by $24 billion, since the U.S. collected more in contributions from its allies than it paid out. In 1992, however, war finances will add an estimated $16.8 billion to the deficit as spending to replace used-up weapons continues while contributions fall. Far larger is the impact of the money spent to shut or sell failed savings and loans. Government bean counters record the expense of making depositors whole only when S&Ls are sold or closed. Because the government has been slow to shed moribund thrifts, those costs in 1991 turned out lower than estimated, which improved the deficit in 1991 and will worsen it in 1992. If all goes well, most big payments to depositors should be finished by the end of 1993. Once that's over, proceeds from additional sales of S&L assets are expected to cut the deficit by tens of billions per year. The 1990-91 recession has also swelled the deficit by lowering tax revenues and boosting spending on unemployment insurance and entitlement programs. To measure the underlying impact of federal spending and tax policy on the economy, economists adjust the deficit to net out the effects of the business cycle. By one such measure, the Congressional Budget Office figures that the deficit, net of the war, the S&L costs, and the recession, rose from $145 billion in 1990 to $175 billion in 1991 -- $107 billion less than the official tally.

Far more troubling than the deficit's level is its trend. After peaking in 1983 at 6.3% of GNP and hovering around the $200 billion mark through 1986, the deficit actually fell during the late 1980s, both in dollar terms and, more sharply, as a percent of GNP. By 1989, net of payments to depositors at failed banks, it was down to $131 billion -- just 2.6% of GNP. In 1990 that benign decline stopped, and the deficit net of deposit insurance and Gulf war spending climbed to $162 billion. This year, the net deficit will be up to $222 billion, or 4% of GNP. Even using the CBO's measure that subtracts the cost of the recession, the 1991 deficit is still more than 3% of GNP. WHAT happens next? That largely depends on whether last year's budget act, which forced George Bush to renege on his ''no new tax'' pledge and outraged Republican conservatives, turns out to be a tiger or a poodle. White House projections show the deficit coming down smartly, to less than 1% of GNP by 1996. The CBO, though a bit more gloomy, also predicts a sharp reduction in five years, to around 2% of GNP. Both projections assume that the tight spending caps in the budget law stay in place. These were imposed on the three main categories of discretionary spending -- defense, domestic, and international -- for each year through 1993. (Spending on so-called mandatory programs, such as food stamps, whose level is determined by the eligibility of the recipients, is not capped because it is not controllable.) In addition, the act ordained that any proposals for new spending must be accompanied by a method of payment -- either new taxes or offsetting cuts in other programs. Called pay as you go, this concept, while familiar to most adults, was news in Washington. So far, it has stifled discussion of any big new spending plans. Exceptions are allowed for emergency measures, but they have been used sparingly (example: humanitarian aid for the Kurds). Says Robert D. Reischauer, director of the CBO: ''There's been a lot of pounding on that door, but so far it has remained fairly tightly closed. There's no question that the budget deal is working.'' But come fiscal 1994 -- the first budget that will be prepared after the 1992 election -- the debate promises to get ugly. In that year and in 1995, the budget act replaces the three individual spending caps with a single cap. This will free Congress to swap defense cuts for domestic spending increases. Those two years are also when the budget caps are supposed to squeeze hardest. Based on the Bush Administration's latest budget request, for example, defense spending in 1995 is already projected to be 22% lower, after inflation, than in 1990. Even with defense at that level, the targets can be met only if other discretionary spending, both domestic and international, falls by nearly 10% in real terms over the two years. Daunting? You bet. Washington hasn't delivered budget cuts that big since the mid-1950s, when America was demobilizing after the Korean war. In the most likely post-election scenario, a reelected George Bush will still be facing a Democrat-controlled Congress. In that case, the two parties might choose to break their bitter stalemate over whether to slash defense even further or trim domestic programs by convening yet another budget summit, one that would push serious deficit cutting still further into the future. Should that happen, the federal debt held by the public will climb past $3 trillion and remain stuck above 50% of GNP for the first time since the 1950s, when America was still growing out of its World War II debts. (The total debt outstanding fell to less than 25% of GNP by 1974.) At those heights, interest payments will account for more than 15% of all federal spending by 1994, up from 7% 20 years ago. As each year's deficit adds to this mountain, the risk rises that those long-feared nightmares -- crowded-out investment, soaring inflation, the Brazilianization of the U.S. economy -- could finally be realized. If, instead, the current budget deal is honored, federal debt should plateau as a percent of GNP, and begin gradually waning after the mid-1990s. But even that welcome rigor won't be enough. Many economists believe that the U.S. must eventually start running budget surpluses, in order to build savings to pay for the baby-boom generation's retirement after 2012. THE NEED for further fiscal discipline is understood in the White House. In July, Budget Director Richard Darman told the Senate Budget Committee: ''The figures, no matter how you present them, present a large enough problem to justify more policy action than we have collectively taken.'' The Administration, in fact, proposed a further $47 billion in spending reductions in the 1991 budget, but it didn't push for them and Congress didn't pay much attention. The right way to cut the deficit is -- as it always has been -- to contain spending, not to raise taxes. Although there will always be a constituency for tax hikes, recent economic history should give pause to the tax-and-spenders: Would we be better off if, instead of running up the deficit over the past two years, we had financed higher government spending with higher taxes? Says Michael Boskin, chairman of the Council of Economic Advisers: ''You have to convince yourself of two things: One is that it wouldn't have hurt the economy, and I think the evidence is that it would have hurt the economy substantially. Second, you have to convince yourself that if you had put more taxes in, they actually would have reduced the deficit, and there wouldn't have been a political dynamic that led to a further expansion of spending.'' Despite last year's budget deal, the war against the deficit is clearly far from over. It won't be won until Congress and the White House stop striking grand bargains that avoid pain in the short run and make only the ''out years,'' as they're known in Washington, the ouch years.

CHART: NOT AVAILABLE CREDIT: SOURCES: COUNCIL OF ECONOMIC ADVISERS: OFFICE OF MANAGEMENT AND BUDGET CAPTION: WILL WHAT WENT UP COME DOWN? Whether that projected steep fall materializes is the $200 billion question.

CHART: NOT AVAILABLE CREDIT: SOURCES: COUNCIL OF ECONOMIC ADVISERS: OFFICE OF MANAGEMENT AND BUDGET CAPTION: UP, UP, AND . . . Paying the interest on this debt now claims 14% of federal spending.