Nonsampling error can be attributed to many sources, such as nonresponse, differences in interpretation of questions, recall difficulties, and processing errors. Among inmates, the number of first admissions may be slightly overestimated because of underreporting of criminal histories. The full extent of nonsampling error is unknown.
3. No comparable inmate survey was conducted prior 1974 to enable the calculation of first incarceration rates prior to this date. First incarceration rates in 1970 and earlier were estimated to be 80% of each age-specific rate of first incarceration in 1974.
If first incarceration rates in 1970 and earlier had averaged 90% of each age-specific first incarceration rate in 1974, the estimated number of adults alive in 2001 who had ever gone to prison would have been 1.9% higher (5,723,000). Alternatively, if earlier first incarceration rates had averaged 70% of 1974 rates, the estimated number of persons ever to have been incarcerated in 2001 would have been-1.9% lower (5,513,000).
4. Mortality rate schedules for prisoners were not available. Compared with the general population, mortality rates for prisoners were estimated to be 20% higher for adults under age 65, and the same for those age 65 and older. The estimate was based on the lower overall educational attainment of prisoners, and longitudinal studies documenting the relationship between mortality and educational attainment.
If mortality rates for adults ever incarcerated, under age 65 were instead 40% higher than that of the general population, the estimated prevalence of ever having gone to prison in 2001 would be 5,567,000 (-0.9% less). If mortality rates for prisoners had been equal to that of the general population, the estimated prevalence would have been 0.9% higher in 2001 (5,670,000).
5. Comparable mortality rates prior to 1976 were not available. However, there is minimal effect of declining mortality rates since 1900 on the estimated number of persons ever incarcerated. Prevalence rates are only affected to the extent that there may have been a different decline in mortality among those ever incarcerated (the numerator) compared with all surviving members of a birth cohort (the denominator). Furthermore, prevalence rates were applied to estimates of the U.S. resident population (which fully reflect declines in mortality).
6. Age-specific incarceration rates do not incorporate a forecast of future rates of imprisonment, which may be affected by changes in criminal behavior, law enforcement, and in sentencing policies. Consequently, the lifetime likelihood of incarceration, 2001, and projected prevalence rates for 2002 and beyond may be different.
A fuller description of the methodological techniques used in preparing this report is available upon request from the author.
The Bureau of Justice Statistics is the statistical agency of the U.S. Department of Justice. Lawrence A. Greenfeld is director. BJS Special Reports address a specific topic in depth from one or many data sets that cover many topics.