Two generally unpopular fiscal trends have coexisted in the 1980s. The first is the regressive change in the distribution of after-tax incomes, extensively documented by both public (CBO, 1987a; 1988b) and private authors (Robert Haveman, 1988, and Sheldon Danziger et al., 1989). The second is the rise in budget deficits and concomitant fall in net national saving rates. The two trends are usually viewed as independent, even though it should be obvious that policy measures to correct one could make the other worse-if low incomes were raised by tax cuts or transfer increases, budget deficits would rise; deficits could be closed by cutting back on programs benefiting poor people. In this paper we explore this interaction. Every year. the Congressional Budget Office produces a volume giving about 130 ways to reduce future federal deficits (1988a). Many of these options involve changes in exhaustive spending programs where it is conceptually and empirically difficult to estimate the incidence of gains and losses by income group. But many involve tax or entitlement changes where it is not so difficult to estimate the incidence of these gains and losses. We have estimated this incidence for 22 of the CBO's deficit reduction options, and find two packages: One with 7 options quite unfavorable to the rich (the hit-the-rich package) that reduces the 1990 budget deficit by $48 billion, nearly 1 percent of that year's GNP; another with 6 options quite unfavorable to the poor (hit-the-poor) that also reduces the 1990 budget deficit the same amount. We compare these two packages with a relatively neutral broad-based value-added tax (VAT) that would also reduce the budget deficit the same amount, and with historical trends in the distribution of after-tax-transfer incomes. This welter of information gives an indication of the likely future tradeoffs between deficit reduction and income redistribution, and how any expected changes compare to changes in the recent past. Our basic result is that even though there is a noticeable impact on the distribution of projected 1990 aftertax incomes between the hit-the-rich and hit-the-poor packages, and between these and the broad-based VAT, these differences are very modest compared to the time-series changes in the after-tax income distribution that have already taken place between 1977 and 1985. Deficit reduction policy can be used to ameliorate only a small share of the adverse distributional changes that have already occurred. The preliminary suggestion is also that earlier policy changes may play a minor role in explaining recent regressive movements in the income distribution, but this matter will be explored in more depth in subsequent work.

Interpretation of the distribution of occupational success depends crucially on an understanding of the process of mobility. If the holders of low-paying jobs this year are just as likely as anyone else to hold good jobs next year, then the lower tail of the earnings distribution is not a matter of great social concern. Further, if the distribution of success is primarily determined by a random process, the prospects for a significant reduction in the inequality of earnings are unfavorable. On the other hand, high mobility in the short run may conceal systematic stratification of the labor force-it may be that the same individuals spend most of their time in the worst jobs. If so, luck and random events may play a small role in the distribution of well being, inequality may be a major issue, and the elimination of stratification may hold out the promise of an important reduction in inequality. Simple tabulation of distributions of success and of mobility among categories of success cannot answer the basic question of the role of stratification. In this paper, we develop a model and statistical method for studying mobility and relating it to characteristics upon which stratification may be based. Our study rests on annual interviews conducted from 1966 through 1969 with about 3,000 white males born between 1942 and 1952.1 We find that mobility is closely related to personal characteristics, especially intelligence and education. Men with high IQs and extensive education are more likely to move to high-paying jobs and are more likely to remain in them. Nevertheless, random events dominate the distribution of occupational success in the short run. Our results strongly support the finding of Christopher Jencks and his collaborators that redistribution of the determinants of earnings would do almost nothing to reduce the observed inequality of annual earnings.