CSAT Prisoners

Methodology

The CSAT - Prisoners tool allows you to examine statistics on inmates under the jurisdiction of both federal and state correctional authorities. The CSAT - Prisoners tool draws from two BJS collections: the National Prisoner Statistics (NPS) program and the National Corrections Reporting Program (NCRP). NPS collects aggregate information from states (e.g., the total year-end population); the NCRP collects offender-level information (e.g., characteristics of each offender in the year-end population). The sum of the offender-level records in NCRP can differ from the aggregate totals in NPS for several reasons.

• NPS background
• NCRP background

Responses to the NPS survey are displayed on the mapping tool and the annual counts of admissions, releases, and year-end population tabular reports.

• Admissions collected in NPS
• Releases collected in NPS
• Year-end population collected in NPS

NCRP data are used to generate tabular reports showing the characteristics of persons admitted to prison, released from prison, and in the year-end prison population.

• Admissions collected in NCRP
• Releases collected in NCRP
• Year-end population collected in NCRP
• Weighted state estimates for the NCRP

NPS background

The National Prisoner Statistics (NPS) data collection began in 1926 in response to a congressional mandate to gather information on persons incarcerated in state and federal prisons. Originally under the auspices of the U.S. Census Bureau, the collection moved to the Bureau of Prisons (BOP) in 1950, and then in 1971 to the National Criminal Justice Information and Statistics Service, the precursor to the Bureau of Justice Statistics. Since 1979, the Census Bureau has been the NPS data collection agent. BJS depends entirely on the voluntary participation of state departments of corrections and the Federal Bureau of Prisons (BOP) for NPS data.

The NPS is administered to 51 respondents. Before 2001, the District of Columbia was also a respondent, but responsibility for housing the District of Columbia's sentenced prisoners was transferred to the BOP, and by year-end 2001 the District of Columbia no longer operated a prison system.

The NPS provides an enumeration of persons in state and federal prisons and collects data on key characteristics of the nation's prison population. NPS has been adapted over time to keep pace with the changing information needs of the public, researchers, and federal, state, and local governments. Three distinct collections comprise the NPS:

1. The NPS-1 is an annual collection known as the "Summary of Sentenced Population Movement." Before 1978, the collection tracked admissions and releases by type and sex during each calendar year, as well as the number of prisoners in custody by sex and sentence length on December 31.
• In 1978, NPS-1 was expanded to include the number of prisoners under the jurisdiction of state and federal adult correctional officials, the number of inmates held in local jails solely to ease overcrowding in prisons, and the race and Hispanic origin by sex of the jurisdiction population at year-end.
• In 1991, NPS-1 was further expanded to include questions on the number of human immunodeficiency virus (HIV) infections and confirmed acquired immune deficiency syndrome (AIDS) cases in the custody population.
• In 1999, the NPS-1 was expanded again to collect data on inmates housed in local facilities without reference to prison overcrowding, inmates held in other state or federal facilities, and inmates held in privately operated facilities. In addition, the race and Hispanic origin questions were combined into a single item to better comport with administrative recordkeeping systems maintained by state departments of corrections. The survey also began collecting more detailed counts of inmate deaths by cause.
2. The NPS-1A was introduced in 1981 to track the rapidly growing prison population. It included counts of the number of inmates under jurisdiction and in custody of prisons, by sentence length and sex, on June 30 of each year. This collection was eliminated in 2010 to reduce burden on data respondents.
3. The NPS-1B was added to provide advance counts of the December 31 (year-end) populations and to obtain data by sex, race, and sentence length. It was expanded in 1982 to include counts of inmates housed in local jails because of prison overcrowding, and again in 1983 to provide measures of housing capacity and numbers of early releases due to prison overcrowding.

NPS-1 and NPS-1B were merged in 2007 under the title NPS-1B "Summary of Sentenced Population Movement." This single collection captures counts by sex of year-end jurisdiction and custody populations by sentence length; number of prisoners held in federal, local, private, and other state facilities; the race and ethnicity of inmates; types of admissions and releases during the calendar year; prison system capacity; and HIV infections and confirmed cases of AIDS. In 2011, two items were added from the former NPS-1A collection, including the year-end custody counts of noncitizens and inmates age 17 or younger.

The NPS datasets used in this online tool draw from the NPS-1 collection until 2007, when the data were consolidated in the NPS-1B. The counts of noncitizens and persons age 17 or younger have been retained from the NPS-1A in the current NPS-1B collection. To present a consistent data series, the CSAT - Prisoners tool includes those data collected after the 1977 expansion of the survey to include jurisdictional counts. Sex-specific data on custody counts from 1925 to 1986 are archived for public use at the University of Michigan's National Archive of Criminal Justice Data (NACJD) https://www.icpsr.umich.edu/web/NACJD/studies/8912/versions/V1.

Additional information about the NPS, including the data collection instrument, is available on the BJS website (https://bjs.ojp.gov/data-collection/national-prisoner-statistics-nps). The full NPS data set is archived at the National Archive of Criminal Justice Data website at https://www.icpsr.umich.edu/web/NACJD/series/886.

NCRP background

NCRP collects offender-level data annually on prison admissions and releases, and post-confinement community supervision. BJS has administered the NCRP since 1983. State departments of corrections and community supervision provide these data, which are used at the federal and state levels to monitor correctional populations and address policy questions related to recidivism, prisoner reentry, and trends in demographic characteristics of the incarcerated and paroled populations. The U.S. Census Bureau served as the NCRP data collection agent for BJS until October 2010, when Abt Associates assumed this position.

Until 1999, three separate files were requested annually: prisoners admitted to prison (Part A), prisoners released from prison (Part B), and persons released from parole (Part C). In 1999, BJS began requesting data on the custody population for each state at year-end (Part D). In 2012, BJS requested that states submit data on persons entering post-confinement community supervision (e.g., parole) programs (Part E) and exiting post-confinement community supervision programs (Part F).

From 1983 to 2009, NCRP data were analyzed and archived each year in three or four year-specific files that corresponded to the file types that states were asked to submit to the Census Bureau. The four files are: Prison Admissions (Part A), Prison Releases (Part B), Parole Exits (Part C), and Prison Custody (Part D). For example, the 2009 NCRP dataset consists of prison admissions occurring in 2009, prison releases occurring in 2009, parole exits occurring in 2009, and prisoners in custody on December 31, 2009.

Starting in 2011, BJS’s data collection agent developed an algorithm to link the prison records (Parts A, B, and D) based on inmate ID numbers, dates of birth, admission, release, offense, and sentencing information in the NCRP data. The majority of states were able to have records linked for the time they submitted data between 2000 and the present. These linked term records represent a single period of imprisonment for an individual offender, allowing analysts to more accurately calculate time served in prison, as well as identify recidivists.

In 2011, BJS began archiving the NCRP data as a single, multi-year Term Record file. A Term Record represents a single period of incarceration for an individual offender. Each year, the archived Term Record file will be replaced by a new Term Record file that incorporates new NCRP data collected and processed during the previous year as well as updates to previously collected data. For those few states with records that could not be linked, BJS will continue to archive the data in the year-specific separate NCRP files (A, B, D). Parole exit records (Part C) will continue to be archived separately, although BJS hopes to eventually link the Part C records with parole entry records (Part E). Because some users of NCRP will undoubtedly want to retain the old archive format, BJS has also archived statistical software that will return the four-file annual format.

In 2012, BJS received OMB clearance to ask states for several additional items of data, including the unique FBI identification number and several questions on previous military service and veteran status. In 2015, BJS received OMB clearance to collect security level at which inmates were held in prison.

For more information on the NCRP, visit the NCRP Resource Guide at the National Archive of Criminal Justice Data.

Year-end population collected in NPS

The NPS distinguishes between prisoners in custody and prisoners under jurisdiction. To have custody of a prisoner, a state or federal prison must hold that inmate in one of its facilities. Jurisdiction over a prisoner means state or federal officials have legal authority over that prisoner, regardless of where the prisoner is incarcerated or supervised. Some states are unable to provide counts that distinguish between custody and jurisdiction.

The NPS jurisdiction counts include persons held in prisons, penitentiaries, correctional facilities, halfway houses, boot camps, farms, training or treatment centers, and hospitals. Counts also include prisoners who are—

• temporarily absent (fewer than 30 days), out to court, or on work release
• housed in privately operated facilities, local jails, or other state or federal facilities
• those serving concurrent sentences for more than one correctional authority.

The NPS custody counts include all inmates held within a state's facilities, including inmates housed for other states. The custody counts exclude inmates held in local jails and in other jurisdictions. With a few exceptions, the final custody counts reported by BJS include inmates held in privately operated facilities.

In the fielding of a current year's NPS survey, respondents are permitted to update the prior counts of prisoners held in custody and under jurisdiction. Statistics on the jurisdiction and sentenced prison populations for the prior year are updated in this data tool. All tables showing data based on jurisdiction counts (including tables of imprisonment rates) contain the most recent data available and will be updated annually.

The NPS has historically included counts of inmates in the combined jail-prison systems of Alaska, Connecticut, Delaware, Hawaii, Rhode Island, and Vermont. The District of Columbia (D.C.) has not operated a prison system since year-end 2001. Felons sentenced under the D.C. criminal code are housed in federal facilities. Jail inmates in D.C. are included in the Annual Survey of Jails. Some previously published prisoner counts and the percentage change in population include D.C. jail inmates for 2001, the last year of collection.

Imprisonment rates are calculated from U.S. resident population figures provided to BJS by the U.S. Census Bureau. The population estimates are generated for January 1 of the year following the year-end NPS data collection.

Admissions collected in NPS

Admissions data include entries to prison through new court commitments; parole violator returns and other conditional release violator returns; transfers from other jurisdictions; returns of prisoners absent without leave (AWOL), with or without a new sentence; escape returns, with or without a new sentence; returns from appeal or bond; and other admissions. For reporting purposes, BJS admission counts exclude transfers from other jurisdictions, AWOL returns, and escape returns in the annual Prisoner bulletins, but these types of entry are included in the data tool’s total admissions count. Between 1979 and 1998, admissions of returned AWOL or escaped prisoners were measured by a single variable; for this data tool, BJS combined AWOL and escape admissions for the years they were measured separately in order to provide a consistent longitudinal measure.

Releases collected in NPS

Releases from prison include unconditional releases (e.g., expirations of sentence or commutations); conditional releases (e.g., probations, supervised mandatory releases, or discretionary paroles); deaths; AWOLs; escapes from confinement; transfers to other jurisdictions; releases to appeal or bond; and other releases. For reporting purposes, BJS release counts exclude AWOLs, escapes, and transfers to other jurisdictions in the annual Prisoner bulletins, but these methods of exit are included in the data tool’s total releases count. Between 1979 and 1998, releases due to AWOL and escape were measured by a single variable; for this data tool, BJS combined AWOL and escape releases for the years they were measured separately in order to provide a consistent longitudinal measure. The total number of releases is a separate variable on the NPS-1B, and while the individual types of release should sum to this total, there are occasions when jurisdictions submit parts that do not sum to the total. In this data tool, the totals for several states have been adjusted to match the sum of the individual release types. The original data as submitted by the states and BOP are available for public download and analysis at the National Archive of Criminal Justice Data website at https://www.icpsr.umich.edu/web/NACJD/series/886.

Year-end population collected in NCRP

Under NCRP, state departments of corrections are asked to submit year-end population data files that contain one record for each sentenced inmate under physical custody, regardless of sentence length or jurisdiction, at year-end.

States are asked to include in the year-end population file—

• sentenced inmates in the state’s prison facilities, regardless of jurisdiction or sentence length
• sentenced inmates under the state’s jurisdiction held in county or local jails
• sentenced inmates under the state’s jurisdiction held in in-state or out-of-state private prisons, including both privately owned facilities and facilities operated by a private entity under contract to the state
• any inmate in the above categories who was temporarily released (less than 30 days) from a facility.

States are asked to exclude from the year-end population file—

• sentenced inmates under the state’s jurisdiction held in federal facilities or another state’s facilities
• unsentenced inmates held in the state’s prison facilities (e.g., civil commitments, inmates awaiting trial)
• inmates who have escaped and are not in custody.

Prison facilities include prisons, penitentiaries, and correctional institutions; boot camps; prison farms; reception, diagnostic, and classification centers; release centers, halfway houses, and road camps; forestry and conservation camps; vocational training facilities; prison hospitals; and drug and alcohol treatment facilities for prisoners. For inmates under home confinement, a private residence is not considered a prison facility.

Admissions collected in NCRP

Under NCRP, state departments of corrections are asked to submit data files that contain one record for each admission of a sentenced inmate to the state’s prison system during the calendar year, regardless of sentence length or jurisdiction.

NCRP defines admissions as including—

• new court commitments
• revocations from probation, parole, or other types of post-confinement community supervision
• transfers from other jurisdictions
• escape or AWOL returns
• returns from appeal or bond.

States are asked to include in the admission file—

• admissions of sentenced inmates to the state’s prison facilities
• admissions of sentenced inmates under the state’s jurisdiction to county or local jails
• admissions of sentenced inmates under the state’s jurisdiction to in-state private prisons, including both privately owned facilities and facilities operated by a private entity under contract to the state.

States are asked to exclude from the admission file—

• admissions of sentenced inmates to one of the state’s prison facilities who are being transferred from another one of the state’s prison facilities
• inmates re-entering a prison facility after a temporary leave of 30 days or less (e.g., for a court appearance, funeral furlough, or medical care)
• admissions of sentenced inmates under the state’s jurisdiction to federal facilities, another state’s facilities, or out-of-state private facilities
• admissions of unsentenced inmates to the state’s prison facilities (e.g., inmates awaiting trial or civil commitments).

Prison facilities include prisons, penitentiaries, and correctional institutions; boot camps; prison farms; reception, diagnostic, and classification centers; release centers, halfway houses, and road camps; forestry and conservation camps; vocational training facilities; prison hospitals; and drug and alcohol treatment facilities for prisoners. For inmates under home confinement, a private residence is not considered a prison facility.

Releases collected in NCRP

Under NCRP, state departments of corrections are asked to submit data files that contain one record for each release of a sentenced inmate from the state’s prison system during the calendar year, regardless of sentence length or jurisdiction.

NCRP defines releases as including—

• conditional releases from prison to parole, probation, or other forms of post-confinement community supervision
• unconditional releases
• releases or transfers to other authorities
• deaths
• releases on appeal or bond if credit for time served is not given while on release
• escapes from custody.

States are asked to include in the release file—

• releases of sentenced inmates from the state’s prison facilities, regardless of jurisdiction or sentence length
• releases of sentenced inmates under the state’s jurisdiction from county or local jails
• releases of sentenced inmates under the state’s jurisdiction from in-state private prisons, including both privately owned facilities and facilities operated by a private entity under contract to the state.

States are asked to exclude from the release file—

• sentenced inmates who are being transferred from one of the state’s facilities to another one of the state’s prison facilities
• temporary releases of sentenced inmates of 30 days or less (e.g., for a court appearance, funeral furlough, or medical care)
• releases of sentenced inmates under the state’s jurisdiction from federal facilities, another state’s facilities, or out-of-state private facilities
• releases of unsentenced inmates from the state’s prison facilities (e.g., inmates awaiting trial, civil commitments).

Prison facilities include prisons, penitentiaries, and correctional institutions; boot camps; prison farms; reception, diagnostic, and classification centers; release centers, halfway houses, and road camps; forestry and conservation camps; vocational training facilities; prison hospitals; and drug and alcohol treatment facilities for prisoners. For inmates under home confinement, a private residence is not considered a prison facility.

Weighted state estimates for the NCRP

BJS employs a ratio adjustment method to weight the individual-level data from the NCRP to the control totals obtained in the NPS data collection. That is, the distribution of each variable is derived from the NCRP data, and then the distribution is scaled to the total value of that variable from the NPS data.

For example, suppose a state reported in the NPS that 10,000 males were admitted to prison in 2016. Via the NCRP, that same state reported 9,500 admissions of males in 2016 and that their education level was distributed as follows:

After weighting the NCRP data to the NPS control totals, CSAT - Prisoners returns the following:

NPS and NCRP totals differ for several reasons, including definitional differences between the two collections (e.g., NCRP requests data on all admissions while NPS requests data only on admissions of offenders with a sentence of more than one year); states compiling the data at different times of the year (e.g., if NPS is reported in February and NCRP is reported in June, then NCRP incorporates three additional months of data updates to the previous year’s records); and the different methodologies employed (e.g., a survey requesting aggregate data that may be compiled from different sources or agency reports versus a data file assembled from records in the agency’s offender management system). The NCRP and NPS data differ in the six states that have unified jail and prison systems (Alaska, Connecticut, Delaware, Hawaii, Rhode Island, and Vermont), due to the substantial number of offenders in these states serving sentences of less than one year. Users are cautioned against comparing offender profiles in these states to other states that only have jurisdiction over offenders serving a sentence of more than one year.