In the 18-month study of the concentrated employment program in Boston, preliminary analysis indicated that program analysis was not possible without first examining the operation of the low-wage labor market. Accordingly, an exploration was made of the dynamic relationship between manpower programs and the economic and social environment in which they operate. Alternative labor market models were compared. The "queue" theory, which assumes a giant "shapeup" in which employers rank employees by their potential net productivity or desirability, was found to be inadequate. Other theories stressed a dual labor market, in which ghetto workers are largely confined to secondary jobs; decision making (weighing wages versus risk and job stability); or the "two queue model" (hiring followed by the more precise job placement stage). One major finding was that high turnover in ghettos is determined mainly by the quality of employment available to the disadvantaged labor force. Two policy alternatives--improving the quality of less preferred employment, and opening preferred employment to the disadvantaged--were suggested. The latter seems preferable because little is known about upgrading less preferred jobs, and because the training arrangements, supervisory services, and economic conditions associated with preferred employment would probably entail less cost. (LY)