Data Dashboard FAQs and Methodology
What is the national corrections data dashboard?
The national corrections data dashboard presents eight core indicators that quantify the various points at which people are involved in state corrections and supervision systems. These eight measures are (1) prison admissions, (2) prison population, (3) probation admissions, (4) probation population, (5) probation revocations, (6) post-release supervision population, (7) releases to post-release supervision, and (8) post-release supervision revocations.
What are the state data dashboards?
The state data dashboards are broad overviews of the available data from corrections agencies and jails in all 50 states, using the most up-to-date publicly available data to show the latest trends across key indicators. The purpose of these pages is (A) to consolidate data from several sources to provide policymakers with timely information on criminal justice trends at the state and local levels and (B) to identify and highlight agencies that are reporting this important data publicly. The dashboards currently include prison, community supervision, and jail data. When possible, additional data points will be added to the dashboards. This inventory will help the Justice Counts coalition of state and local leaders develop a set of criminal justice metrics that can be implemented across all 50 states.
Where does the data come from?
These dashboards build on data collected, analyzed, and publicly released by state and local agencies, the U.S. Department of Justice’s Bureau of Justice Statistics (BJS), and the Vera Institute of Justice. The Justice Counts team recognizes and sincerely appreciates the efforts of the aforementioned entities as it seeks to consolidate and leverage their work to create a resource for policymakers, agency leaders, and the public in all 50 states.
The corrections data in these dashboards comes from publicly available records that criminal justice agencies within the state are already producing. The Justice Counts team has sourced these records and manually collected the data from state public safety agencies’ websites and from data that certain state corrections agencies have provided to us directly. You can find the sources used at the bottom of a given state’s data dashboard; some data points on the page will also have an (i) icon that contains the link to the original public data source.
The jails data in these dashboards is powered by data compiled by the Vera Institute of Justice (Vera Jails Survey) and BJS (Annual Survey of Jails). Data from Vera is continuously integrated from January 2020 onwards, while BJS data is pulled in manually as reports are published with the most recent report showing 2018 data. Note that this leads to a gap in data in 2019, as BJS data is forthcoming.
If you represent a state’s department of corrections (or similar government agency) and would like to contribute additional reports to the dashboard, please contact us at email@example.com.
How were the data points featured on the dashboards chosen?
The Justice Counts state data dashboards aim to identify a common set of data points that provide a useful picture of the state of different parts of the criminal justice system.
The corrections data points chosen here are meant to represent how people move through the corrections system, including prison and supervision (probation and parole) populations, admissions, and releases. The jails data points represent jail populations and incarceration rates statewide and in the counties that report data. The data points do not represent every piece of data that a state collects or reports. When possible, additional data points will be added to the dashboards.
Are the data points on the site standardized?
The data points shown on this site are not standardized. The data is taken directly from state reports without transformation (other than occasionally aggregating reports from adjacent time periods). As such, please take caution before comparing data between two different states; you can find more information about a given state’s data by clicking on the sources available at the bottom of each dashboard page.
If you have questions about definitions, please contact us at firstname.lastname@example.org.
Why is Justice Counts calculating individual and cumulative jail “confinement rates” for reporting counties?
These calculations account for varying county population sizes and differing reporting timelines. Each county’s confinement rate is the size of its jail population, divided by the county’s population, x 100,000. The State (Reporting Counties) Confinement Rate is not an average of the confinement rates for each reporting county; rather, it is a cumulative confinement rate for all reporting counties, as shown below:
This method weights each county’s jail confinement rate by its size. Using this method, a small county with a particularly high confinement rate (or vice versa) has less of a disproportionate impact on the state rate as compared to a strict average of each county’s confinement rate.
How often will the data dashboards be updated?
We will update the state data dashboards on an approximately monthly cadence.
What happens if a state updates data that has already been published?
Sometimes a state will report preliminary data and subsequent reports will show slightly different numbers after an agency finalizes its counts. If a state updates its historical data, we will aim to update its dashboard accordingly in a reasonable amount of time. Refer to the state data sources listed on the dashboard for the most up-to-date data from that state. If your state typically reports data preliminarily and finalizes it at a later date, please let us know by contacting email@example.com.
Where can I find demographic breakdowns of the data points available?
While we do not currently publish demographic data on the state data dashboards, many states report these breakdowns in the original sources. You can find the sources consulted at the bottom of a given state’s data dashboard; some data points on the page will also have an (i) icon that contains the link to the original public data source.
Why is some of the data missing or out of date?
The Justice Counts state data dashboards serve two purposes: (1) to show aggregate, state-level data that’s more timely, less disjointed, and as useful as possible to policymakers, practitioners, and the public; and (2) to serve as an inventory of what data is and—just as importantly—is not available. In order to accomplish this, we have attempted to collect as much data as is available publicly, up to the most recent report.
Justice Counts publishes the most recent data available and, when available, historical trends for the past decade. Each state data dashboard indicates (1) the dates for which each indicator is being reported, and (2) the date of the most recent update to the state data dashboard. States that publish data online daily but do not publicly archive historical data may find that some of their data is missing from the dashboard. We are building the capacity to gather this daily data regularly.
If data seems to be out of date, there are likely no more recent reports. If you believe we are missing recent data that is already available publicly, please contact us at firstname.lastname@example.org.
Why is the data presented on the dashboard reported at a different frequency than what is in the original source?
In order to preserve consistency, we have chosen to report data as either monthly or annual statistics. You can use the control at the top of the page to switch the frequency of the data from “Monthly” to “Annual.”
For counted events like admissions or releases, we will aggregate data at whichever cadence it is available in the reports up to either monthly or annual statistics; for example, prison admissions data reported weekly might be aggregated up to a monthly statistic when the “Monthly” setting is on, or monthly revocation counts might be aggregated up to an annualized statistic when the “Annual” setting is on.
If you have any questions about what data is used in the dashboard, please contact us at email@example.com.
Why does some corrections data appear on a state’s monthly data tab, but not on the corresponding annual tab?
The data aggregation range control allows users to switch between showing only monthly aggregated data and only annually aggregated data. Monthly and annual time periods were defined based on underlying data; note that not all data within a given state aligned with those time periods for aggregation. For this reason, some data points will appear on a state’s monthly data tab, but will not appear on the same state’s corresponding annual tab. For example, if a state’s annual time period is defined as December through December, data for a measure that is reported the next month (in January) will not be included on the annual tab.
Who gathered the data for this website?
With guidance from the CSG Justice Center and feedback from state corrections and supervision agencies across the country, Recidiviz gathered the data for the Justice Counts data dashboards. Recidiviz also designed the interface for the dashboards and will continually gather and organize the data.
This dashboard contains data for select prison and community supervision measures. The definitions of these measures are deliberately broad, as the data available for each state varies significantly. If a state provided explicit guidance on the data to be collected for a particular measure, this was taken into account and would generally take precedence over the data collection team’s judgement. For each measure in the dashboard, the exact categories covered by the data are shown in the tooltip to clarify what is included for that particular state.
The admissions data covers people who started serving a term of incarceration or supervision in that state during the given time period.
New Prison Admissions
New admissions cover admissions to prison due to a new sentence imposed by a court.
New Probation Admissions
New admissions cover admissions to probation due to a new sentence imposed by a court.
Revocations cover any supervision violation that resulted in the individual being admitted to prison. This is likely different from the total number of revocations that occurred in that state. The dashboard breaks revocations down by the type of revocation (new crime or technical violation) and the type of supervision the individual was revoked from (probation or post-release).
The population data covers people serving a term of incarceration or supervision in that state. In most cases, the population reported is from a single day during the given time period; however, the average daily population for the period may be shown if a single day population was not available. In most cases, the “(Grand) Total” population was collected wherever indicated.
The post-release supervision population includes people on any type of post-release supervision, including parole.
In most cases, the prison population includes people who have been sentenced and are incarcerated in a state. In some states with unified systems, this number may also include the pretrial population. For those states, the “Jails” tab will be disabled.
The probation population typically includes people sentenced directly to supervision by a court. This number may sometimes include populations such as “community corrections” if the distinction between this population and the “post-release supervision” population is clear.
The total releases data covers all releases from prison with or without supervision in the state during the given time period. This may be referred to as “exits,” “terminations,” or other terms in states’ statistical reports.
To Community (Without Supervision)
Releases to the community cover people who are released without supervision after fully serving their sentence. This number typically corresponds with “discharge” or “straight release” in states’ statistical reports as a subcategory of total releases.
To Post-Release Supervision
Releases to post-release supervision include people that are released onto some type of post-release supervision, excluding probation. This number typically corresponds with releases “to parole” or other post-release community supervision, as explained in state reports as a subcategory of total releases.
Measurement Type: Point in Time
Data displayed represents the given metric on the day it was reported.
Measurement Type: Average
Data displayed represents a measurement that is an average over the reporting period.
Time Aggregation Methodology
Picking Time Periods
For monthly data, a time period is created for every month, from the first of the month to the last day of the month, inclusive. This is done for simplicity as most sources report data that aligns with these periods. Some sources report data that does not align (e.g., weekly admissions data may cover days from two different months), but these are omitted.
Annual periods are less consistent, as some sources report data for the calendar year while others report for the fiscal year, even within the same state. For each state, an annual time period is defined based on the underlying data. Data with time periods longer than a month are prioritized as those will not be included in the monthly data. For instance, June will be used as the boundary month if there is admissions data reported by fiscal year ending in June, even if there is population data reported annually in December that then will not be included.
Because monthly and annual time periods needed to be defined based on underlying data and not all data within a given state aligned with those time periods for aggregation, some data points will appear on a state’s monthly data tab, but will not appear on a state’s corresponding annual tab.
Even when data is omitted from the flow chart, it will still be visible in the graphs section of each page, which contains time-series visualizations of all the data available for the state.
For a given time period, data is selected and aggregated in the following manner:
For population metrics, the most recent Point in Time measurement for the time period is used so long as it falls in the last month of the period. If that does not exist, an Average measurement will be used so long as it covers the entire period exactly. If neither exist, no data point will be produced for that metric and period.
For population flow metrics (admissions, revocations, releases), the underlying data must cover the entire period exactly. Data covering continuous non-overlapping periods can be summed to produce a point that covers the desired time period. For example, if a source reports quarterly admissions, those admissions numbers may be summed to produce an annual data point ending in March, June, September, or December.
This dashboard is powered by data compiled by the Vera Institute of Justice (Vera Jails Survey) and the Bureau of Justice Statistics (BJS) (Annual Survey of Jails). Data from Vera is continuously integrated from January 2020 onwards, while BJS data is pulled in manually as reports are published, with the most recent report showing 2018 data. Note that this leads to a gap in data in 2019, as BJS data is forthcoming.
County population data is pulled from U.S. Census datasets. Population counts are interpolated for years that population data is not reported.
Jail Confinement Rate (per 100k)
The jail confinement rate is defined as the rate at which individuals in a given jurisdiction are confined in jail at any given time. This is meant to give a sense of the prevalence of jail confinement in a given state using the data available.
Justice Counts calculates the confinement rate based on the populations provided by the source in order to ensure methodological consistency. For a given moment in time, this can be approximately calculated as the number of people in jail, divided by the total population at that time, x 100,000. Our statewide calculation (in both the Key Insights and charts) is based on the collection of counties actually reporting data rather than an aggregated statewide population; as such, we use the sum of both the reported county jail population and the last known (or interpolated) county population to calculate the confinement rates.
When calculating a change over time (i.e., the percent change from one year before), we find the month closest to one year in the past and calculate the confinement rates based on the collection of counties reporting at those two points.
This is the percentage of counties that were reporting their jail population as of the most recent date that data was reported for any county.
Jail Population (represented counties only)
The jail population indicates the number of people incarcerated in a county jail or other pretrial facility.
This is calculated using only the counties that report data on a regular basis; as such, the population reported here will typically undercount the actual jail population.