Back
to Appendix B: Federal and State Income TaxesBy Amy L. Watts
From Education and the Common Good
pp. 47-50, published 2001
Expected incarceration costs for Kentucky men and women at varying levels of educational attainment were used to estimate the potential criminal justice savings associated with education. The expected cost of incarceration was a combination of the likelihood of incarceration, given a Kentuckian’s age, gender and education level, and the average annual cost of imprisonment. Data on prisoners of state facilities in Kentucky were from the Kentucky Department of Corrections, 2000,(1) data on prisoners in state correctional facilities for the United States were from the Survey of Inmates in State and Federal Correctional Facilities, 1997,(2) and Kentucky population data were from the March Supplements to the Current Population Surveys, 1998-2000. These datasets were used together to estimate the share of Kentucky’s adult population in state correctional facilities.
Data limitations hindered our use of regression analysis in estimating the relationship between education and the criminal justice system. A random sample of noninstitutionalized and institutionalized persons was not available for use in estimating this relationship. Therefore, another method utilizing the three datasets described previously was used to estimate the probability a person would reside in a state correctional facility, based on education, age, and gender.
Within each dataset, subpopulations based on age(3) and gender were estimated. Data on educational attainment were not available in the Kentucky inmate dataset. Within each gender and age group, the national incarcerated population and the Kentucky general population are estimated for four educational attainment categories (less than high school graduate, high school graduate or equivalent, some college or two-year degree, and bachelor’s degree or more). To construct these categories for the Kentucky-level inmate dataset, national inmate percentages were applied. The percentage of national inmates in each gender, age, and education category were multiplied by the number of Kentucky inmates in each gender and age category to obtain state-level estimates of educational attainment of Kentucky’s prison population. The number of Kentucky inmates in each demographic category was divided by the number of persons in the general population in each demographic category to obtain the probability that a person, given their gender, age and educational attainment level, would be imprisoned. The present value of a lifetime of costs, from age 19 to 75, was calculated for men and women and discounted at a rate of 3.05 percent, to reflect the “time value of money.”
This analysis excludes federal prisons and local jails, (4) as well as the costs of probation, parole and the court systems. Therefore, this analysis does not include the full cost of the criminal justice system. Federal prisons hold a relatively small share of the total number of prisoners and comprise a relatively small share of the costs of incarceration. While a large number of people are on probation, the cost of probation is relatively small. To the extent better educated individuals tend to be jailed in federal prisons or are put on probation, this analysis may overstate the savings associated with education.
Expected annual costs per person, given each one’s age, gender and educational attainment level, were estimated by multiplying the probability that a person would be incarcerated by the average annual cost of incarceration per inmate in Kentucky. Expected annual costs vary, based solely on the variation in the likelihood that a person will be incarcerated, given each person’s demographic status based on age, gender and educational attainment level.
Since education levels were not available for Kentucky’s state prison population, national data were used to estimate education shares for the state level data. As mentioned previously, the percentage of persons in state prisons for each gender, age and education category at the national level was multiplied by the number of persons in each gender and age category at the state level to obtain estimates for the education categories. The national sample was similar to the Kentucky data in gender and age. In both datasets, 93 percent of the total observations were male and 7 percent female. In addition, the percentages of men in both datasets in the age groups defined in Table C.1 were similar, with over half of the men in both between the ages of 18 and 35, a little over a third between the ages of 36 and 55 and less than 5 percent of both data sets between the ages of 56 and 75. The female percentages break down in a similar manner at both the state and national levels. Frequency distributions of all three data sets used in this section are provided in Table C.1. By construction the education percentages for the state inmate sample are the same as those for the national sample.
To view a list of all chapters in this book, click here. To read the chapters in sequential order, please follow the arrow below.
Back
to Appendix B: Federal and State Income Taxes
Kentucky Department of Corrections, Justice Cabinet, Commonwealth of Kentucky. Current Extract File. Data on inmates correct as of December 11, 2000. Return to text.
U.S. Dept. of Justice, Bureau of Justice Statistics, and U.S. Dept of Justice, Federal Bureau of Prisons. Survey of Inmates in State and Federal Correctional Facilities, 1997 [Computer File]. compiled by the U.S. Dept. of Commerce, Bureau of the Census, ICPSR ed. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [producer and distributor], 2000. Return to text.
Data on age were calculated using 3-year rolling averages. Return to text.
The state data include Class D inmates who are those that are temporarily housed in local jails until space is available in a state prison. Although physically in local jails, funding support comes from the state corrections department, therefore these observations were included in the final analysis. There are 2,251 Class D felons included in this data set, representing approximately 15 percent of the total number of observations. Return to text.