The total jail population (and the population of people who have been detained for over 500 days) has fluctuated significantly in the time surrounding the COVID pandemic as seen in the yearly averages of daily population:
| Year | Average Daily Jail Population | Average Daily Jail Population Detained Over 500 Days | |
|---|---|---|---|
| 0 | 2019 | 2,686 | 528 |
| 1 | 2020 | 2,337 | 514 |
| 2 | 2021 | 2,281 | 643 |
| 3 | 2022 | 2,533 | 647 |
| 4 | 2023 | 2,471 | 659 |
The "Average Daily Jail Population" represents the mean of the total daily populations on the first day of each month between July 1, 2019 and April 1, 2023 grouped by year. These totals include the Shelby County Jail (aka "201 Poplar") which is the pretrial detention facility for men over the age of 18 in addition to "Jail East" which holds juveniles who are waiting to be bound over and women. The "Average Daily Jail Population Detained Over 500 Days" averages the total number of people on the first day of each month whose current booking is greater than or equal to 500 days.
For more granularity, the below figure shows the total population of these facilities on the first of each month. Note that these totals count everyone who "touched" the jail on that day. In other words, this number includes people who got booked and the released in an hour and also people who have been held for a long time including that day. By hovering over the figure, you can see the count and percentage of that population attributed to each race.
The below figure shows the number of people who were in jail for at least 500 days on each date. Not only has this population grown since COVID, but the percentage of that population attributed to black people is also slowly increasing from 91.4% in August 2019 to 93.6% in April 2023.
The graphs in this section use the data contained in the below figure. Each point in the plot represents a single jail booking (i.e. single stay in jail) such that the person was in jail on June 13, 2019 (the day we started recording jail data) or entered thereafter. The points are split/colored by race, and you can hover over any point to see more details about that booking (e.g. arrest date, highest offenses, etc...)
Overall, the average length of stay has been slowly increasing as seen in the below table:
| Release Year | Average Days in Jail | |
|---|---|---|
| 0 | 2019 | 32.1 |
| 1 | 2020 | 32.8 |
| 2 | 2021 | 33.9 |
| 3 | 2022 | 33.5 |
The "Average Days in Jail" column gives us the average length of a booking that exited during the "Release Year".
The disproportionality of long bookings for black people from the above plot is clear, but to split out how much each contributes to the average length of stay in jail, we split out those numbers by race.
Note that only BLACK and WHITE are included in this plot. The relatively low numbers of bookings for HISPANIC, ASIAN, and OTHER people cause the averages to fluctuate significantly, so we exclude them to make the graph tidier.
To see how charge level affects length of stay, we split up bookings according to the maximum charge level on the booking. Note that this does not differentiate between if a booking has one Felony A or a hundred. Severity decreases in alphabetical order (i.e. Felony A is more severe than Felony B) with the exception of Felony M which is reserved for murder and represents the "highest" offense level.
Finally, to illustrate how we use our jail in relation to violent and non-violent crimes, the following table shows the percentage of arrests that were for violent offenses:
| Arrest Year | Violent Arrest % | Arrests | |
|---|---|---|---|
| 0 | 2019 | 14.7 | 27,909 |
| 1 | 2020 | 19.8 | 22,122 |
| 2 | 2021 | 20.2 | 22,545 |
| 3 | 2022 | 19.4 | 24,180 |
| 4 | 2023 | 18.4 | 6,229 |
Note that during COVID and shortly after, we began utilizing our jail less for non-violent offenses, but in 2022 and 2023, we can infer that we are beginning to prioritize using the jail to detain more people for non-violent offenses. Here is a chart to show how a violent offense on a booking affects the length of stay:
| Type | Release Year | Average Length of Jail Stay | |
|---|---|---|---|
| 0 | Non-Violent | 2019 | 16.2 |
| 2 | Non-Violent | 2020 | 15.1 |
| 4 | Non-Violent | 2021 | 14.2 |
| 6 | Non-Violent | 2022 | 14.7 |
| 1 | Violent | 2019 | 131.1 |
| 3 | Violent | 2020 | 110.8 |
| 5 | Violent | 2021 | 120.4 |
| 7 | Violent | 2022 | 119.1 |
The jail population fluctuates based on the difference between arrests and releases. In other words, if more people are arrested than released, the population will grow and vice versa. To better understand if the change in jail population is a result of policies that change arrest or release practices, the following figure plots the number of arrests and releases per month together for comparison. Note that releases consistently outnumber bookings after June 2022, but arrests dramatically decrease in the wake of the Tyre Nichols murder in January 2023:
Or in chart form for the last year:
| Month | Arrests | Releases | Difference | |
|---|---|---|---|---|
| 96 | 2022-03 | 2,160 | 2,217 | -57 |
| 97 | 2022-04 | 2,043 | 2,074 | -31 |
| 98 | 2022-05 | 1,951 | 1,929 | 22 |
| 99 | 2022-06 | 2,152 | 2,184 | -32 |
| 100 | 2022-07 | 2,065 | 2,119 | -54 |
| 101 | 2022-08 | 2,080 | 2,184 | -104 |
| 102 | 2022-09 | 1,986 | 2,044 | -58 |
| 103 | 2022-10 | 1,932 | 2,017 | -85 |
| 104 | 2022-11 | 1,885 | 1,893 | -8 |
| 105 | 2022-12 | 1,629 | 1,902 | -273 |
| 106 | 2023-01 | 1,831 | 1,994 | -163 |
| 107 | 2023-02 | 1,303 | 1,531 | -228 |
| 108 | 2023-03 | 1,647 | 1,901 | -254 |
There are multiple reasons that a person may be released, so to break down what that looks like over time, we plot release type counts over time:
"CASE_DISPOSED" is used if a person is released after their case is dismissed or if they are sentenced to probation/diversion. "SENTENCED" is used when a person is sentenced to incarceration and is therefore (most likely) being transferred to a post-conviction facility.