Executive Summary
OFFICIAL DATA with ALPRP analysis and proposed responses
ADOC reports 21,125 residents in-house at June 30, 2026 and 2,397 security staff. The verified June capacity comparison is 174.4% occupancy: 21,125 residents against 12,115 designed places. ABPP granted parole in 20.3% of hearings through August 2026. Programs, safety, financial flows and parole each retain their own reporting periods; these records inform ALPRP’s design for safer custody and supported reentry.
OFFICIAL DATA · Quarter ending June 30, 2026, PDF p. 4,5
OFFICIAL DATA · June 2026 Monthly Statistical Report, PDF p. 3
OFFICIAL DATA · August 2026 Monthly Statistical Report, PDF p. 11
In-house population
21,125
ADOC · June 30, 2026
Occupancy
174.4%
June 2026 · in-house ÷ design capacity
Security headcount
2,397
June 2026 · all shifts and classifications
Parole grant rate
20.3%
560 grants ÷ 2,752 hearings · YTD Aug
ALPRP ANALYSIS · Key Findings
Population exceeds architectural capacity
The June report places 9,010 more residents in-house than the system’s designed capacity. System totals cannot establish the safety of an individual dorm or the number of officers on duty.
+9,010
June 2026
Program access needs an assessed denominator
The June quarterly report records 564 unduplicated basic education students and 1,122 career and technical students. Enrollment is not the same as completion, and these totals do not establish the share of assessed need served.
Parole decisions and releases measure different events
ABPP reports hearing decisions; ADOC counts actual releases. A parole grant does not necessarily produce a release in the same reporting month.
Prison Population vs. Designed Capacity
June 2025–June 2026 · month-end counts from one report vintage
OFFICIAL DATA · June 2026 Monthly Statistical Report, PDF p. 3 · Printed p. 2. Jurisdictional population includes people outside ADOC facilities; occupancy uses in-house population only.
Race / Ethnicity
In-house population · June 30, 2026 · 21,125 residents
OFFICIAL DATA · Quarter ending June 30, 2026, PDF p. 4 · Percentages calculated from source counts. Commitment classification does not establish present behavior or individualized risk.
Commitment Classification
In-house population · June 30, 2026 · 21,125 residents
OFFICIAL DATA · Quarter ending June 30, 2026, PDF p. 4 · Percentages calculated from source counts. Commitment classification does not establish present behavior or individualized risk.
Release Types
June 2026 only · 791 jurisdictional releases
OFFICIAL DATA · June 2026 Monthly Statistical Report, PDF p. 3
Admission Types
June 2026 only · 495 jurisdictional admissions
OFFICIAL DATA · June 2026 Monthly Statistical Report, PDF p. 3
ALPRP PROPOSAL · Response
ALPRP connects Revive, Rehabilitate, Rebuild, Restore and Release through individualized assessment, safe classification, treatment, education, lawful release preparation and measurable accountability. Pilot 100 tests that design with documented safety, service access and reentry outcomes. Technology scales communication, records and monitoring while officers, clinicians and authorized agencies retain responsibility for decisions.
Sources, definitions and ALPRP evidence standards
June 30, 2026; fiscal YTD Oct 2025–Jun 2026 · Printed pp. 1–4, 9, 11 · source reviewed 2026-10-08
June 30, 2026; Q3 Apr–Jun 2026 · Printed pp. 4–8, 57 · source reviewed 2026-10-07
FY2026 YTD Oct 2025–Aug 2026; supervision Sep 3, 2026 · Printed pp. 2, 10–11 · source reviewed 2026-10-07
FY2027 enacted; FY2026 budgeted comparison · PDF p. 4, row 78 · source reviewed 2026-10-08
OFFICIAL DATA · INDEPENDENT RESEARCH · RESIDENT/FAMILY-REPORTED · ALPRP ANALYSIS · ALPRP PROPOSAL · ILLUSTRATIVE PROTOTYPE. Evidence categories distinguish agency records, external findings, reported concerns, interpretations, proposed services and demonstrations.
Missing or N/R values are not zero. Counts retain their population, reporting period and unit. Derived percentages state their denominators. All published data in this version is maintained in one reviewed dataset; AI extraction creates private candidates for review.
Download the source-reviewed data and definitions →