The main thesis of Punishment and Social Structure (1975) is that the relationship between social-structural variables, such as the condition of the labor market, and the changing modalities of punishment over time, challenges the legitimacy of the State's power to punish. Researchers developed this theory by examining the quality and quantity of punishment and focusing on its characteristics in a capitalist, disciplinary, and industrial society. This research leads to the debate over decarceration or the diminishing relative weight of imprisonment in the whole economy of control. Punishing behavior develops in several types of cycles: the secular trend beginning with an upward movement in the 19th century followed by a downward movement in the 20th century; long-term (40-50 year) oscillations determined by the viscosity of institutional and cultural change; and short-term oscillations similar to those characterizing business cycles. The relationship between a changing economy and changing imprisonment rates is demonstrated by the case of Italy from 1896 to 1986, when the significant negative association between business cycle and imprisonment rates was not mediated through the rate of general criminal behavior. The observed association between change in imprisonment rates and change in economic indicators depends on another variable, vocabularies of punitive motive. A grounded labeling theory relates the way in which variable social-structural elements impact upon the variable degree of probability that a certain behavior will be classified as normal or deviant. The author maintains that ongoing deep changes in the nature of the whole arena of social control are connected to a decreasing role of imprisonment. 6 notes, 54 references.