A Five-Year Budget Package

The Budget Reconciliation Act of 1990 raised the Gramm-Rudman deficit targets from $64 billion in FY91, $28 billion in FY92, and zero in FY93 to $327 billion in FY91, $317 billion in FY92, $236 billion in FY93, $102 billion in FY94, and $83 billion in FY95. Table A shows how the administration's baseline forecast grew from barely $100 billion at the beginning of the year to $367 billion at the end. Of that difference, $103 billion, the change from January to September forecasts on line 5, was due to the change in economic projections on lines 1 and 2; a 4 percent difference in economic growth in FY90–91 is a lot of money. As the CBO figure shows, some of this misestimate was optimism, and some was really news.

Taking social security's surplus out of the calculations increased the deficit by $74 billion (see lines 6 and 7)—a change in accounting only.

A big real change was the $90 billion rise in expected costs of bailing out insolvent thrift institutions. Deposit insurance, however, was a onetime (we hope) megadisaster, and so its costs would go away on their own. In fact, as the government moved from buying thrifts to selling their assets, its cash flow would move from very negative to positive. Table B, showing how the deficit was supposed to fall under the final plan, shows that between FY91 and FY94 the shift in projected deposit insurance flows would reduce the deficit by $150 billion!

This table is based on figures put together by the Republican staff of the Senate Budget Committee, and it should be no surprise that the Democratic staff at House Budget measure the package somewhat differently. They claimed only $137 billion in extra revenue, and $99 billion cut from entitlements. The major difference was whether to treat tax

credits for the poor as a spending increase (Republican) or a tax cut (Democratic).^[6]

Table B also excludes, because no such thing exists, a reliable breakdown between defense and domestic discretionary savings. Since these accounts are annually appropriated, the savings were not enacted in advance. Instead, the Budget Reconciliation Act created caps on spending that would be enforced by automatic sequesters, no matter what the deficit. For the first three years, FY91–93, there are separate caps on defense, domestic, and international affairs spending. If Congress enacts more than the cap in either category, then there will be an immediate sequester within that category . In FY94 and FY95 the categories are combined, so in theory Congress could cut defense more and domestic less, or vice versa.

The table's discretionary savings line is the difference between the caps and a very slippery baseline. The negotiators excluded the costs of a number of immediate emergencies, such as the Iraqi invasion of Kuwait. The law allowed technical adjustments, such as in scoring a new credit budgeting system, which will add another few billion dollars a year to the domestic base. The baseline was also rigged, in a way generally seen as a deal between Director Darman and the appropriators, to hide a domestic increase. In early December House Appropriations staff estimated that increase at about $40 billion in BA and $20 billion in outlays total. Since the BA and outlays don't fit, and the numbers are tighter in FY94–95, we guess the spending gain is closer to $30 billion overall.^[7]

The big discretionary cuts therefore all come from defense. They were made possible by the collapse of the Warsaw Pact military alliance in 1989, symbolized by the dismantling of the Berlin Wall. Only that good luck makes the package look larger than our suggestion; without the defense cuts there are $49 billion in policy savings in FY92, the second year of the deal. The defense windfall makes up for lower growth and the long-term interest costs from the deposit insurance expense over the course of the agreement. Due to social security accounting and strange economic estimates, official figures were at best confusing.^[8] Yet in talking to budget experts around Washington, and in our own rough calculations, we found a consensus that the real figure, including a goodly amount of bad economic news, was a unified budget deficit of $75 billion, about 1 percent of GNP, in 1995.^[9]

On its face, the package was a bad deal for Republicans: by their scorekeeping, less than 20 percent of the savings came in domestic spending. It was perhaps made worse—and for Democrats clearly improved—by its distributional tilt. Although the tax package included regressive excises, it is progressive overall. The increase in the Earned

Income Tax Credit helped the poor. A total of $67.1 billion was raised from the well-to-do by increasing the alternative minimum tax, limiting itemized deductions for taxpayers with adjusted gross income over $100,000, phasing out the tax benefit from personal exemptions, replacing the "bubble" with a flat 31 percent top rate, and increasing the cap on wages taxable for Medicare from $53,400 to $125,000.^[10] Ways and Means Committee Majority staff estimated that, as a result, the package would: raise the income of the poorest fifth of the population by 1.5 percent; lower the income of the richest 1 percent by 2.2 percent; and lower the income of those in between by around 1 percent.^[11] These may not seem like big changes, but the Democrats seemed proud and the administration had fought long and hard for a different distribution.

Last but not least, the agreement in essence abolished the deficit sequester. It did so by requiring, in FY92–93, and allowing, in FY94–95, the president to adjust the targets for changes in economic and technical assumptions when issuing the budget each year. Then the deficit sequester would be calculated on the basis of those same economic assumptions, even if they were wrong! Since the targets and estimates are based on the same economy, the economy can't cause a sequester. And laws to raise spending or lower revenues are inhibited by both the points of order that had worked well for a few years and a new layer of targeted sequesters. Tax cuts or entitlement increases must be paid for as done; otherwise a "PayGo" sequester lies against certain entitlement programs. Increases in a program that raise an appropriated category above its cap will trigger an across-the-board cut within that category.

A lot can go wrong. The deposit insurance mess, or costs from the war and subsequent relief efforts in Iraq, could increase interest costs in FY95. Yet even with such losses, the deal should get the deficit to a reasonable level by then, and decisions to bust the agreement will require a consensus among the president and Congress (including 60 percent of the Senate). The odds of those sides agreeing on how to bust the deal significantly seem, to us, quite slim. It might happen if a united government is elected in 1992—but, then, the people will have voted for it, and we could hardly object.