ABSTRACT
As Duncan explained in Chapter 3, the arrival of long runs of panel data has radically altered the American perception of poverty (Duncan et al, 1984; Hill, 1985; Bane and Ellwood, 1986; Ellwood, 1988; Ruggles, 1990). As we have seen in the last two chapters, rather than being conceptualised as a static state, with an immutable distinction between the poor and the not poor, poverty is recognised to be dynamic and to occur in finite spells with beginnings and ends. This has opened the way to thinking about the specific causes of poverty and the means by which spells of poverty might be brought to an early end.
