This appendix summarizes and critiques major sources of descriptive statistics used in this report and by many scholars of incarceration.
HISTORICAL AND COMPARATIVE INCARCERATION RATES
The adult incarcerated population is generally counted as the number of people held in jails and prisons. Prison population counts have been reported by the Bureau of Justice Statistics (BJS) in a continuous time series dating back to 1925. Counts of the jail population are available in a continuous time series from 1980, although earlier years are available periodically through special BJS collections and in the U.S. census. The scale of a penal system is usually measured by an incarceration rate that expresses the number incarcerated per 100,000 of the resident population. The annual rates are usually formed with census population counts and intercensus estimates.
The rate of state and federal imprisonment 1925-2012 (Figure 2-1 in Chapter 2) was taken from Maguire (n.d., Table 6.28.2012). Data for jail incarceration 1980-2011 were taken from Maguire (n.d., Table 6.1.2011). Figures for 2012 are from Glaze and Herberman (2013). Data on jail incarceration 1972-1979 were taken from Hindelang et al. (1977, p. 632) and Parisi et al. (1979). Missing years were interpolated. International incarceration rates in Aebi and Delgrande (2013) for European countries and Walmsley (2012) for Australia, Canada, and New Zealand provided data and methodologies for comparative considerations. Figure 2-2 on
international comparisons was created with most recent rates available from International Centre for Prison Studies (2013).
Growth in federal and state prison populations and local jail populations 1972-2010 are for men and women under age 65. These data were compiled by Bryan Sykes, University of Washington, Seattle, and are described below.
Data on counts of the total correctional population and its constituent prison, jail, parole, and probation populations (Figure 2-4 in Chapter 2) were taken from Maguire (n.d., Table 6.1.2011). Data for the 1972-1979 period were taken from the Sourcebook of Criminal Justice Statistics (1982).
Data on state imprisonment rates 2000-2010 are for sentenced prisoners under the jurisdiction of state correctional authorities (Figure 2-4) and were taken from Maguire (n.d., Table 6.28.2011). Data for 1972 are from the Sourcebook of Criminal Justice Statistics (1982, p. 471).
DATA AND METHODS FOR DISAGGREGATED INCARCERATION RATES
Data for incarceration by sex, race, ethnicity, age, and education were constructed by Bryan Sykes and Becky Pettit at the University of Washington, Seattle. Estimates of persons incarcerated by sex, race, age, and education are for the period 1972-2010. Unlike many BJS series, microdata from prison and jail surveys of inmates were used to obtain estimates of the prison and jail populations for non-Hispanic blacks, non-Hispanic whites, and Hispanics. To obtain these estimates, Sykes and Pettit used aggregated data on penal populations from BJS. Aggregated data for the entire time series are available by facility type, not for specific sex, race/ethnicity, age, and education groups. Data on inmate totals come from the Sourcebook of Criminal Justice Statistics Online.1 Data for federal and state inmates 1982-1984 and 1986-1989 were provided by BJS. Jail counts are for the last business day in June; state and federal prison counts are for December 31 of the year.
Following methods outlined in Pettit and Western (2004), Western (2006), and Pettit (2012), microdata from BJS correctional surveys were used to estimate proportions of inmates within sex, racial/ethnic, age, and education groups. The Survey of Inmates of Local Jails (1972, 1978, 1983, 1989, 1996, 2002), the Survey of Inmates of State Correctional Facilities (1974, 1979, 1986, 1991, 1997, 2004), and the Survey of Inmates of Federal Correctional Facilities (1991, 1997, 2004) were used to interpolate
between survey years (within facility type).2 Estimates for state inmates prior to 1974 were assumed to follow the distributions of the 1974 Survey of Inmates of State Correctional Facilities, while estimates for federal inmates prior to 1991 were assumed to follow the distributions of the 1991 Survey of Inmates of Federal Correctional Facilities. Estimates for inmates after the last survey year (2002 Survey of Inmates of Local Jails, 2004 Survey of Inmates of State Correctional Facilities, and 2004 Survey of Inmates of Federal Correctional Facilities) were calculated based on the demographic distributions of respondents in the last survey. These proportions were applied to aggregated population counts of inmates by facility type to obtain the demographic distributions of prison and jail inmates.
The U.S. civilian population was obtained from the weighted March Current Population Survey (CPS) for 1972-2010.3 Because the CPS is a household-based, noninstitutional sample, we constructed a series of civilian incarceration rates in which the weighted population totals from the March CPS were adjusted to include inmate totals from BJS, as outlined in Pettit (2012) and Pettit et al. (2009). This analysis reports civilian incarceration rates of men and women by race and educational attainment for different age intervals. Race is coded as non-Hispanic white, non-Hispanic black, Hispanic, and non-Hispanic other. Educational attainment was measured as less than high school, high school, and some college education. Age was disaggregated into 5-year age groups except for those aged 18-19. However, the grouped table of the analysis collapses these age categories into 18-19, 20-39, and 40-64 to minimize small cell counts and extreme variability for demographic groups in which incarceration is less prevalent.
2For more information on BJS surveys of correctional institutions, see http://www.bjs.gov/index.cfm?ty=dcdetail&iid=274; http://www.bjs.gov/index.cfm?ty=dcdetail&iid=275; and http://www.bjs.gov/index.cfm?ty=dcdetail&iid=273 [August 2013].