chicago_police_complaints — Chicago police complaints, use of force and awards (1967-2023)

Description

Networks of the officers of the Chicago Police Department (CPD) and the complaints of misconduct against them, the reports of their use of force, and their awards, from the Citizens Police Data Project (CPDP) of the Invisible Institute, which obtained the records through Freedom of Information Act (FOIA) requests and litigation, and matched the officers across the files of several FOIA responses to unique ids. The entry contains three bipartite networks, each with all 37,545 officers of the matched profiles (also those without complaints, reports or awards):

complaints: the nodes are officers and complaints (complaint register, CR, numbers), and an edge connects a complaint with each officer accused in it. Complaints are filed by civilians or by CPD employees, and investigated by the CPD's Bureau of Internal Affairs (BIA) or civilian review agencies (OPS, IPRA, COPA). Officers accused together in complaints form the co-accusal networks used in studies of police misconduct. The network contains all 234,971 complaints (also the 89,776 complaints whose accused officers are unknown), with 263,315 accusations, from 1967 to 2023 (most complaints from 2000 onward). Four accused officers are missing from the profiles; they are included as nodes without attributes (has_record false).
use_of_force: the nodes are officers and incidents in which they used force, from the Tactical Response Reports (TRR) of 2004 to 2023 that officers file after using force. An edge connects an officer with an incident for each report of the officer on it (one per subject: an officer and an incident can be connected by several edges). The incidents are the CPD event numbers of the reports; the 15,252 reports without event number are incidents of their own. There are 65,816 incidents and 98,055 reports; the 809 reports without officer are left out (557 incidents have no reports with officers, and are included without edges).
awards: the nodes are officers and awards (e.g. honorable mentions, commendations, attendance and crime reduction awards), and an edge connects an award with each officer who received it (or for whom it was requested): 113,775 awards and 807,173 officer awards, from 1967 to 2020.

Left out are the investigators and the police witnesses of the complaints (not matched to officer ids), the complainants, victims and civilian witnesses; of the use of force, the reports that are only in the secondary files of the TRR (without officer), the subject records, the details of the weapon discharges, and the chain of review of the reports (by other officers, not matched to officer ids); and the salaries and unit histories. The officers are named, as in the public records.

The networks have the following properties:

(Node) is_complaint, is_incident or is_award: whether the node is a complaint, incident or award (otherwise an officer).
(Node) uid (officers): the officer's id in the data of the Invisible Institute (-1 for the other nodes).
(Node) has_record (officers): whether the officer is in the file of profiles.
(Node) officer_id (officers): the id of the officer on the CPDP site (https://cpdp.co/officer/<officer_id>; -1 if unknown).
(Node) first_name, last_name, middle_initial, middle_initial2 and suffix_name (officers): the name of the officer.
(Node) appointed_date, resignation_date, org_hire_date and start_date (officers): the dates of appointment, resignation, hire (by any police organization) and start in the CPD, as given (M/D/YY, with two-digit years).
(Node) birth_year (officers): the year of birth (-1 if unknown).
(Node) gender, race (officers): the gender and race of the officer.
(Node) current_star (officers): the last star (badge) number.
(Node) cleaned_rank (officers): the standardized rank.
(Node) current_rank and current_unit (officers): the last rank and unit.
(Node) active (officers): whether the officer is active (1, or 0).
(Node) profile_count (officers): the number of datasets in which the officer appears.
(Node) The properties of the officers are the most common values over the files, as determined by the Invisible Institute.
(Node) cr_id (complaints, in complaints): the complaint register number (-1 for officers).
(Node) complaint_date, closed_date and incident_date (complaints): the dates of filing, closure and incident (Unix time; -1 if unknown).
(Node) incident_time (complaints): the time of the incident (as given).
(Node) days_to_closure (complaints): the number of days to closure.
(Node) address (complaints): the block address of the incident (number and street).
(Node) beat (complaints): the police beat.
(Node) city (complaints): the city.
(Node) location_code (complaints): the type of location.
(Node) old_complaint_address (complaints): the address in the older format.
(Node) latitude and longitude (complaints): the geocoded block address (NaN if unknown).
(Node) current_status (complaints): the current status of the complaint.
(Node) investigating_agency (complaints): the last agency investigating the complaint.
(Node) investigating_agencies (complaints): all agencies investigating the complaint.
(Node) complainant_type (complaints): CIVILIAN or CPD EMPLOYEE.
(Node) filename (complaints): the last FOIA response with the complaint.
(Edge) complaint_category and complaint_code (complaints): the category of the allegation against the officer (e.g. Excessive Force).
(Edge) complaint_categories and complaint_codes (complaints): all categories and codes listed for the complaint in all files.
(Edge) disciplined (complaints): whether the complaint resulted in discipline.
(Edge) final_finding (complaints): the final finding (e.g. SU sustained, NS not sustained, UN unfounded, EX exonerated, NAF no affidavit, AC administratively closed).
(Edge) final_outcome and final_outcome_desc (complaints): the final outcome (e.g. Reprimand, 5 Day Suspension) and its description.
(Edge) recc_finding and recc_outcome (complaints): the recommended finding and outcome.
(Edge) days (complaints): the recommended days of suspension (NaN if none).
(Edge) cv (complaints): the version of the complaint files from which the row comes.
(Node) event_no (incidents, in use_of_force): the event number (-1 if none).
(Node) incident_date and incident_time (incidents): the date (Unix time of the day) and time (as given) of the incident.
(Node) block, direction and street (incidents): the block address.
(Node) beat (incidents): the police beat.
(Node) latitude and longitude (incidents): the geocoded block address (NaN if unknown).
(Node) indoor_or_outdoor, lighting_condition, weather_condition, location and location_recode (incidents): the type of location.
(Node) The properties of the incidents are the values of the first report on the incident; the reports of different officers sometimes differ, and their own values are edge properties.
(Edge) trr_id (use_of_force, the reports): the id of the report.
(Edge) date and time (use_of_force): the date (Unix time of the day) and time.
(Edge) block, direction, street, beat, latitude, longitude, indoor_or_outdoor, lighting_condition, weather_condition, location, location_recode (use_of_force): the place, as for the incidents.
(Edge) trr_status (use_of_force): the status of the report.
(Edge) firearm_used and taser (use_of_force): whether the officer used a firearm or a taser (1 or 0).
(Edge) notify_DET_division, notify_OEMC, notify_OP_command and notify_district_sergeant (use_of_force): whether the detective division, the emergency management office, the OP command or the district sergeant were notified (as given; empty if not).
(Edge) number_of_officers_using_firearm, number_of_weapons_discharged and total_number_of_shots (use_of_force): the numbers of officers using a firearm, of weapons discharged and of shots (-1 if not given).
(Edge) party_fired_first (use_of_force): the party that fired first.
(Edge) officer_assigned_beat, officer_on_duty, officer_in_uniform, officer_injured, officer_rank, officer_unit_id and officer_unit_detail_id (use_of_force): the officer at the time of the incident (-1 if unknown).
(Edge) subject_armed, subject_injured, subject_alleged_injury, subject_age, subject_birth_year, subject_gender and subject_race (use_of_force): the subject of the force.
(Edge) member_actions and subject_actions (use_of_force): the actions of the officer (e.g. VERBAL COMMANDS, TAKE DOWN/EMERGENCY HANDCUFFING, TASER) and of the subject (e.g. PULLED AWAY), each listed once.
(Edge) max_action_category (use_of_force): the highest level of the officer's actions, from 0 (member presence) to 6 (firearm; NaN if not given, as in the older files).
(Edge) subject_weapons (use_of_force): the types of weapons of the subject (e.g. HANDS/FISTS).
(Edge) cv (use_of_force): the version of the files from which the report comes.
(Node) award_id (awards, in awards): the id of the award.
(Node) award_type, award_start_date and award_end_date (awards): the type of the award and its start and end dates (Unix time; -1 if unknown), the values of the first record of the award.
(Edge) current_award_status (awards): the status of the award (FINAL, DELETED, DENIED, PENDING APPROVAL or PRELIMINARY), as recorded for the officer.
(Edge) award_type (awards): the type of the award, as recorded for the officer.
(Edge) award_request_date and ceremony_date (awards): the dates of the request and of the ceremony (Unix time; -1 if unknown), as recorded for the officer.
(Edge) requester_full_name (awards): the name of the officer who requested the award (REDACTED for some), as recorded for the officer.

Tags

Social Offline Affiliation Bipartite Unweighted Multigraph Timestamps Metadata

Added

2026-10-31

Citation

Invisible Institute, "Citizens Police Data Project: Chicago police data." GitHub repository (2025) github.com
L. Zhao, A. V. Papachristos, "Network Position and Police Who Shoot." The Annals of the American Academy of Political and Social Science 687, 89-112 (2020) 10.1177/0002716219901171 Sci-Hub

Upstream URL OK

https://github.com/invinst/chicago-police-data
https://cpdp.co/

Networks

Tip: click on a column header to sort the list, or hover over it for a description.

Column legend

$\left<k\right>$: Mean degree
$\sigma_k$: Standard deviation of the degrees
$\lambda_h$: Largest eigenvalue of the Hashimoto (non-backtracking) matrix
$\tau$: Random walk mixing time
$r$: Degree assortativity
$c$: Global clustering coefficient
$\oslash$: (Pseudo-) diameter
$S$: Relative size of the largest component
NPs: Node properties
EPs: Edge properties

Name	Nodes	Edges	$\left<k\right>$	$\sigma_k$	$\lambda_h$	$\tau$	$r$	$\oslash$	$S$	Kind	Mode	NPs	EPs	gt	GraphML	GML	csv
complaints	272,520	263,315	1.93	4.63	14.75	206.75	-0.31	28	0.60	Undirected	Bipartite	is_complaint uid has_record officer_id first_name last_name middle_initial middle_initial2 suffix_name appointed_date resignation_date org_hire_date start_date gender race cleaned_rank current_rank current_unit birth_year current_star active profile_count cr_id complaint_date closed_date incident_date days_to_closure address incident_time beat city location_code old_complaint_address current_status investigating_agency investigating_agencies complainant_type filename latitude longitude	complaint_category complaint_code complaint_categories complaint_codes final_finding final_outcome final_outcome_desc recc_finding recc_outcome disciplined days cv	11.2 MiB	15.7 MiB	15.8 MiB	17.3 MiB
use_of_force	103,361	98,055	1.90	3.61	5.28	452.81	-0.35	32	0.69	Undirected	Bipartite	is_incident uid has_record officer_id first_name last_name middle_initial middle_initial2 suffix_name appointed_date resignation_date org_hire_date start_date gender race cleaned_rank current_rank current_unit birth_year current_star active profile_count event_no incident_date incident_time block direction street beat latitude longitude indoor_or_outdoor lighting_condition weather_condition location location_recode	trr_id date time block direction street beat latitude longitude indoor_or_outdoor lighting_condition weather_condition location location_recode trr_status firearm_used notify_DET_division notify_OEMC notify_OP_command notify_district_sergeant number_of_officers_using_firearm number_of_weapons_discharged total_number_of_shots party_fired_first taser officer_assigned_beat officer_on_duty officer_in_uniform officer_injured officer_rank officer_unit_id officer_unit_detail_id subject_armed subject_injured subject_alleged_injury subject_age subject_birth_year subject_gender subject_race cv member_actions max_action_category subject_actions subject_weapons	6.3 MiB	9.2 MiB	9.1 MiB	10.2 MiB
awards	151,320	807,173	10.67	83.93	236.72	22.54	-0.07	10	0.95	Undirected	Bipartite	is_award uid has_record officer_id first_name last_name middle_initial middle_initial2 suffix_name appointed_date resignation_date org_hire_date start_date gender race cleaned_rank current_rank current_unit birth_year current_star active profile_count award_id award_type award_start_date award_end_date	current_award_status award_type requester_full_name award_request_date ceremony_date	7.1 MiB	12.3 MiB	11.7 MiB	14.6 MiB

Ridiculograms^*

^* These are automatically generated force-directed visualizations, and can be quite meaningless for networks both big and small. They should not be taken seriously as sources of scientific insight. See here for a discussion.