I did not plan to do open-ended interviewing and participant observations: I happened into it, in the course of my graduate education, in two ways: in my thesis research and in my work with civil rights and anti-poverty groups. My thesis topic was the effect of automation upon the skill composition of manufacturing jobs. At the time, there was considerable controversy about whether technological change was increasing or decreasing the skill requirements of jobs. An important group of analysts in the debates about national economic policy were arguing that skill requirements were increasing, leaving a residue of workers who had, or might have, been employed by the old techniques but who were unqualified for the new jobs. This growing residue was supposed to be a barrier to reductions in the level of unemployment.