William J. Wilson (1978, 1985, 1986) has hypothesized that the combination of increased spatial concentration and increased inequality of income among blacks has caused adverse behavioral consequences for poor blacks and contributed to the development of an underclass. Lessening segregation and the general rise in black economic well-being in the postwar period enabled middle-income blacks to move out of segregated inner-city neighborhoods. As a result, low-income blacks in these areas now rarely come in contact with middle-class blacks, who had previously influenced social organizations and community institutions, and provided role models of economic and social success. Wilson hypothesizes that poor blacks have changed their labor force and family behaviors because of the social and economic consequences of this selective outmigration. Wilson's hypothesis has both an empirical and a causal component. In this paper we focus primarily on the former. In the first two sections, we review changes in the level and distribution of male earnings, and in the spatial concentration of poverty. The trends for blacks are compared to those for whites. The third section discusses the links between the empirical evidence and the causal component-did these changes lead to behavioral responses that contributed to the development of an underclass?

In the past ten years, discussions of the system and reform have increasingly been concerned with the extent of welfare dependence. However, while the idea that some individuals may be dependent upon the system for support is intuitively clear, the concept of dependence is rather ill-defined. Our aims in this paper are to clarify what is meant by dependence; to discuss the implications for data handling, measurement, and statistical modeling; and to present new results on the level and trend in dependence in the United States.

Discussions about the antipoverty effects of economic growth in the United States have largely been predicated on John Kennedy's metaphor that a rising tide lifts all boats. But the magnitude of these effects has been a subject of debate since the inception of the War on Poverty (see Lowell Gallaway, 1965, and Henry Aaron, 1967). This debate has public policy as well as academic implications-the greater the antipoverty effectiveness of growth, the less the need for special programs or income supplements during economic expansions. Elsewhere, we have shown that increased real income need not be associated with a decline in poverty (see our 1984 and 1985 papers). In fact, poverty rates did not fall from 1982 to 1983, even though real median income increased. And in 1984 the official poverty rate was about the same as it was in 1967, while real median family income was 7.1 percent above its 1967 level.' If a rising tide was lifting all boats, the tide was late in many harbors. In this paper we examine the relationship between macroeconomic conditions and poverty. Section I argues that several factors now limit the effectiveness of growth in reducing poverty. Section II differentiates the effects of secular economic growth from those of cyclical recoveries. The next section presents our interpretation of the data, followed by a brief conclusion. We show that growth had a large antipoverty effect through the early 1970's, but that the more recent experience has been different because growth rates have slowed and inequality has increased.