Statistics on incarceration are notoriously slippery. Prisoner counts from any given source may or may not include inmates housed in private correctional facilities, inmates under the jurisdiction of one state that are housed in another state, state prison inmates being held in local jails, and so on. For this project, we have used the same definition of "prisoner" used by the federal Bureau of Justice Statistics: an inmate under the jurisdiction of state or federal authorities sentenced to more than one year. (In almost all states, inmates sentenced to less than one year are held in city and county jails.)

Much of the statistical research for this project was performed by the Justice Policy Institute, a nonprofit criminal-justice research organization based in Washington, D.C., and San Francisco.

Increasing Imprisonment
Prisons vs. Education
More Drug Offenders
Racial Inequality

Increasing Imprisonment
Incarceration rates: The number of prisoners sentenced to more than one year per 100,000 residents, as recorded by the federal Bureau of Justice Statistics (BJS). All figures are as of December 31, except 1999 and 2000, which are as of June 30.

Sources: Sourcebook of Criminal Justice Statistics, 1999; Prison and Jail Inmates at Midyear 2000.

Prisoners: The total number of inmates sentenced to more than one year, as recorded by BJS. All figures are as of December 31, except 1999 and 2000, which are as of June 30.

Sources: National Prisoner Statistics data series; Prison and Jail Inmates at Midyear 1999; Prison and Jail Inmates at Midyear 2000.

Notes:
1. Massachusetts holds several thousand inmates serving more than one year in local jails. BJS's incarceration rates include these inmates in all years, but its prisoner totals leave them out for several years. To adjust the counts for those years, (1997, 1998 and 2000), we multiplied BJS's incarceration rate by Massachussetts' general population to derive a total prisoner figure.

2. For some states, the BJS-reported incarceration rates and prisoner totals in this section differ slightly from those in the Racial Inequality section. Data for race and ethnicity were mostly collected on a different day and by a different agency, the Criminal Justice Institute.

Prisons vs. Education
All dollar amounts are adjusted for inflation using the Consumer Price Index and are expressed in constant 1999 dollars. Spending figures shown are state general fund appropriations to higher education and to state corrections, divided by the state's population.

Sources:

Populations: US Bureau of the Census, state population estimates, 1980-1989, 1990-1999, and 2000 Census.

Higher Education spending: Fiscal years 1980-1981, 1985-1986, 1990-1991, and 1995-1996 are from the federal Department of Education's Digest of Education Statistics. 1999-2000 figures are from the National Association of State Budget Officers' State Expenditure Report, June 2000.

Prison spending: fiscal years 1980-1981, 1985-1986, 1990-1991, and 1995-1996 ,Bureau of Justice Statistics, Sourcebook of Criminal Justice Statistics and Justice Expenditures and Employment Abstracts. Fiscal year 1999-2000, National Association of State Budget Officers (NASBO), State Expenditure Report, June 2000.

Full 1999-2000 figures for Washington, D.C. and Colorado were not included in the NASBO report. These were provided directly by the relevant state agencies.

Notes:
1. Seven jurisdictions (Alaska, Connecticut, Delaware, Washington, D.C., Hawaii, Rhode Island, and Vermont) operate unified jail/prison systems. Prison spending for these jurisdictions is adjusted downward for each year by the ratio of (a) number of prisoners serving sentences of one or more years to (b) the total number of inmates in the combined jail/prison system reported in the Sourcebook of Criminal Justice Statistics for that same year.

More Drug Offenders
All prisoners whose primary offense was a drug offense, as reported to MotherJones.com and Justice Policy Institute researchers by each state's Department of Corrections. Every state and Washington, D.C. were contacted. Those with no information noted here either do not collect such statistics or did not respond to repeated requests to provide it. Reported drug-offender numbers were divided by BJS-reported total prisoner numbers to derive the percentage of drug offenders.

Racial Inequality
The Bureau of Justice Statistics last collected state-by-state racial and ethnic information on prisoners in 1998. For 2000, raw numbers of prisoners by race and ethnicity as reported by states were obtained from the Criminal Justice Institute, a nonprofit correctional research organization based in New Jersey. In the seven jurisdictions that operate combined jail/prison systems, inmates serving less than one year were factored out by multiplying CJI's totals by the percentage of inmates serving more than one year in those states, as recorded by BJS.

For states that count Hispanic prisoners but do not separate them from other racial categories (since Hispanics are not considered a racial group), Justice Policy Institute (JPI) statisticians subtracted the total number of Hispanic prisoners from the white and black prisoner totals according to the proportion of Hispanics reported in the 2000 Census to have identified themselves as "white" or "black" in that state.

For the few states that do not count Hispanics at all, JPI calculated the average incarceration rate for each race in bordering states, then multiplied that by the ratio of the original state's overall incarceration rate to the average of those of the bordering states. The resulting incarceration rates by race and ethnicity are then multiplied by the number of each racial and ethnic group in the state's general population to derive a total estimated number of Hispanic prisoners.

The small numbers of "other race" and "unknown race" prisoners in all states were prorated among the five major racial/ethnic groups.

For states in which the resulting calculated numbers differed significantly from those reported in BJS's 1998 survey, numbers on race and ethnicity were obtained directly from state corrections departments, where possible.

General populations by race/ethnicity in each state are taken from the US Census Bureau's 2000 Census. The very small numbers listed as "other race" and "two or more races" were prorated among the five major race/ethnic categories. The prisoner totals above are then divided by the state totals to derive incarceration rates by race and ethnicity.