The economic stimulus program of early 1977 featured a strong dose of what might be termed indirect countercyclical policy. Rather than altering federal expenditures and taxes directly, the stimulus program consisted mainly of three different grant programs for state and local governments: a) countercyclical revenue sharing (CRS); b) public service employment (PSE); c) local public works (LPW). In the parlance of the public finance literature, the first of these grants was an unconditional block grant, the second was a close-ended categorical grant with no local matching for the purpose of stimulating local government employment, and the third was a close-ended categorical grant with no local matching for the purpose of stimulating local government construction. Stimulating, or attempting to stimulate, aggregate spending through state and local governments in this way is fiscal federalism with a vengeance. The federal government is not abdicating its stabilization responsibilities and leaving it up to states and localities to do the job,' but it is placing its own stabilization policy at the mercy of the behavior of state and local governments. There are no restrictions at all on the use of the CRS grantsthey can be spent, used for tax reduction, or used to rebuild financial net worth (asset stocks less outstanding debt), with only the first two uses having any stimulative effect at all. There are restrictions on what can be done with the other two grants, but the well-known displacement phenomenon implies that with these grants it also may be possible for states and localities to frustrate the restrictions and use the grants as they would any other source of revenue sharing. What happens to all three grants then is an empirical issue, and the timing and magnitude of any stabilization impact depends on how the numbers come out. In this paper I briefly describe a model for estimating this stabilization impact and show what it suggests for the three grants. Only the PSE grant will be seen to have any positive short-run impact on aggregate spending at all, and that impact will prove to be both diluted and short-lived, indicating that the general idea of stimulating the economy through state and local governments is probably not a very good one. Plain old permanent federal income tax cuts retain their superiority as a fiscal stabilization device. But even though as stimulation devices the three grant programs leave much to be desired, as policies they may still be valuable, and the paper also suggests how one might do a more complete evaluation of each of the grant programs.

The New York City fiscal crisis as it was played out in the nation's newspapers this year had all the elements of a first class drama. There was first of all the tension-would the city make it through its periodic financial hurdles, would the Ford Administration blink, what would happen if the city defaulted? Then there were the accusations was it the fault of Wagner, Lindsay, Beame, Rockefeller, Ford, the unions, or economic and social forces beyond the city's control? Then the controversy-the issue seemed ideally suited to split deficit spenders from budget balancers, soft-headed liberals from hardheaded accountants, eastern establishment intellectuals from the silent majority. Finally, though it did not capture as much press coverage, the crisis also graphically illustrated several basic issues in the economics of federalism that are now creeping into public finance textbooks-the proper role of local and national governments in stabilizing the economy and redistributing income, whether the federal or the state government has an obligation to protect the financial integrity of local governments, whether public expenditures can be effectively controlled in the short run. This paper discusses the city's fiscal plight in the context of all of these issues. I. How Big are the Deficits?

The U.S. federal budget deficits averaged about 1 percent of GNP in the 1960s, about 2 percent of GNP in the 1970s, and then rose to a peak of 5.5 percent of GNP in the four bad years of fiscal 1983 to fiscal 1986. By fiscal 1989, they had fallen back to 2.9 percent of GNP, and under present legislation they are now projected to drop to about 1.5 percent of GNP by 1994. When deficits were at their peak, before the start of the budget bargain for fiscal 1987 (the first budget to show any sizeable deficit reduction), the country passed the GrammRudman-Hollings (GRH) deficit control law. This law has been revised and amended over time, but it is still operating in roughly its initial form. Superficially, the numbers given above suggest that it must be working pretty well. Moreover, in many ways GRH was written to respond to the standard criticisms of deficit constraint legislation: it did not call for abrupt cuts in deficits but rather gradual declines, it was legislative and flexible rather than constitutional and rigid. Yet just as the country is on the verge of finally getting its deficits under control, the legislation is coming under vicious attack, including even that of one of its sponsors, Senator Hollings. Why should such an apparently successful piece of legislation be so widely scorned? In this paper I take on this question. I first illustrate how GRH was intended to operate with a simple indifference curve model, and use this model to see what kinds of changes should have occurred in budget bargains and spending patterns. Then I look for these changes by comparing numbers from the pre-GRH era (1983-86) with those from the post-GRH era (1987-89). I also identify other sources of improvement in budgetary position from the earlier to the later period. This numerical comparison suggests that the direct effects of GRH played a very minor role in the improvement in the U.S. budgetary position, though it seems likely that difficult to evaluate indirect effects were much more significant.