Monday, Nov 3
Melvin Stephens, Estimating Program Benefits
Shawn Bushway (School of Criminal Justice, University of Albany, State University of New York)
02/08/2010, at noon in room 6050 ISR-Thompson.
co-authors: Hui-shien Tsao and Herb Smith
The sheer size of the prison population- there are now more people serving time in U.S. prisons than there are people serving in the U.S. military ---has recently led demographers to consider the potential impact of prison on demographically relevant factors such as family formation, fertility, death and racial inequality. However, little attention has been paid to the demographic character of the prison population itself. Using a nationally representative sample of inmates conducted in roughly five year intervals since 1974, we present the first detailed exploration of the changing age structure of the state and federal prison populations. After establishing that the sample will support an age-period-cohort analysis, we demonstrate that the age-prison curve has moved dramatically to the right during the last thirty years, with a median increase in age of eight years. We further establish that this is not strictly the result of the aging of the overall U.S. population – the age-specific rates of incarceration also shift to the right by around 5 years. Using a technique by O’Brien and Stockard to separately identify the cohort and period effects, we find that this shift in the age structure is largely a cohort effect. Taken together with fairly constant age-arrest curves during this period, our results cast doubt on recent claims by economists that the increase in prison population is due almost entirely to the increased use of prison sentences, rather than longer prison terms. Future research will attempt to link changes in the age structure in the federal prison population since 1991 to changes in sentencing practice.