The National Police Misconduct Statistics and Reporting Project utilizes news media reports of police misconduct to generate statistical information in an effort to approximate how prevalent police misconduct may be in the United States.
As part of this project, reported incidents of misconduct are aggregated into a news feed on Twitter and then added into an off-line database where duplicate entries and updates are removed and remaining unique stories are categorized for statistical information in monthly, quarterly, and yearly reports here on this site. To view data from other months, refer to the Police Misconduct Statistics menu item located on the top menu bar.
While the use of news reports to generate statistical data may seem strange, keep in mind that police departments do not normally release any detailed information about disciplinary matters, and sometimes they don’t release anything at all. The use of court records by themselves would only garner information about misconduct cases that were successfully prosecuted and would miss confidential settlements and cases of misconduct that were not prosecuted but did result in internal disciplinary action.
It should be noted that the use of media reports acts as a filter that limits the number of outwardly questionable allegations of misconduct but may also suffer from under-reporting due to laws that limit the amount of information law enforcement agencies report to the press. Therefore, if anything, the resulting statistics we publish should be considered as a low-end estimate of the current rate of police misconduct in the United States and for any locality we cite.
Also, In order to allow for accurate comparisons between this project’s statistics and the US DOJ/FBI Uniform Crime Reporting (UCR) statistics, it should be noted that this project utilizes the same methodology as the UCR by way of a hierarchical reporting system that only records the most serious allegation when more than one allegation is associated with an alleged incident of misconduct. It should also be noted that both the NPMSRP and UCR report on alleged instances, not convictions.
The following report was generated from data gathered in the months of April 2009 through September 2009. In the last 6 months there were:
2,568 – Alleged victims of reported police misconduct.
2,854 – Law enforcement officers alleged to have engaged in misconduct.
207 – Law enforcement leaders (police chiefs or sheriffs) that were cited in those reports.
215 – Fatalities reported in connection with alleged instances of misconduct.
14.7 – Reported incidents of misconduct tracked per day on average or a report of misconduct every 98 minutes.
1 out of every 116.4 – Estimated number of officers who will be involved in a reported act of misconduct this year.
$128,306,406 – Reported costs in police misconduct related civil litigation, not counting legal fees or court costs.
When examining misconduct reports by type, non-firearm related excessive force complaints were most common at 21.3% (652) of all reports, followed by sexual misconduct complaints at 13.0% (397), and then fraud/theft reports at 9.8% (300).
When examining reports by last reported status, 32.8% had resulted in some sort of adverse outcome for the officers involved (25.7%) or their representative employers (7.1%). 215 (7.0%) officers were disciplined, 138 (4.5%) were fired, and of the 1018 who were criminally charged, 317 were convicted of a criminal offense for a 31.1% conviction rate.
The following statistics only count state, city, and county law enforcement agencies since current federal law enforcement employment rates were not available for calculation. The statistical rates are based on the NPMSRP statistics and employment data provided by the 2008 US DOJ/FBI UCR.
The following density map shows the police misconduct incident rate per general population.
While the next density map shows the police misconduct rate per law enforcement officer.
The average national police misconduct rate is estimated to be 834.69 per 100,000 police officers. In 2008 there were an estimated 712,360 state and local law enforcement officers employed in the US for an average of 1 officer for every 231.5 people.
The following table shows the state misconduct rates ranked from worst to best.
|State||Misconduct rate per 100k officers||State||Incidents per 100k population|
*note: West Virginia state statistics are based on an estimated law enforcement population since they do not provide statistical information to the federal government.
As a form of “parity check” the following table lists law enforcement agencies that are known to publish data on their police misconduct rates along with our estimated police misconduct rates for the same agency in order to make a rough estimate as to how accurate the NPMSRP statistics might be.
|City||ST||Agency Sustained Incidents||Agency Rate||NPMSRP Projected Incidents||NPMSRP Projected Rate|
*Note: Due to the relatively few law enforcement agencies that make misconduct information available the number of comparisons we can make are limited, but do show more potential for under-reporting than over-reporting for our data in comparison to what known agencies self-report. (If you would like to see data for your own location, contact us at email@example.com)
By projecting this month’s NPMSRP totals out to one year, the following comparisons can be made between the reported police misconduct allegation rate and the reported 2008 general crime rate* as published by the FBI and DOJ for 2008 (*please note that both the NPMSRP police misconduct rates and the FBI/DOJ UCR general crime rate statistics are reported incidents, not convictions):
(all assault, excessive force, forcible rape, murder, and domestic assault allegations)
- 1 out of every 268 (372.5 per 100k) police officers will be accused of a violent crime.
- 1 out of every 220 (454.5 per 100k) citizens will be accused of a violent crime.
(all non-negligent manslaughter, murder, and homicide allegations)
- 1 out of every 2,374 (42.1 per 100k) police officers will be accused of homicide
- 1 out of every 18,518 (5.4 per 100k) citizens will be accused of homicide
(all sexual assault, coercive sexual battery, and rape allegations but not including consensual sexual misconduct, exposure, solicitation, or child pornography)
- 1 out of every 846 (118.2 per 100k) police officers will be accused of sexual assault.
- 1 out of every 3,413 (29.3 per 100k) citizens will be accused of sexual assault.
The following comparisons are made between the NPMSRP 6 month statistics projected out to one year and the 2004 US DOJ Bureau of Justice Statistics Criminal Sentencing Statistics:
- 68% of civilians charged are ultimately convicted
- 31% of police officers charged are ultimately convicted
Probation Sentence Rate
- 28% of convicted citizens are sentenced to probation.
- 38% of convicted police officers are sentenced to probation
- 37 months – average civilian prison sentence length
- 14 months – average police officer sentence length
About This Report
Accountability – Incidents involving evidence of police misconduct cover-ups, lack of investigations, allegations of lax disciplinary response to sustained allegations, and other activities that involve accountability policies or processes.
Animal Cruelty – Acts of violence resulting in harm to animals both on and off duty that may include unnecessary shooting incidents, inappropriate training of K9 units, or other such activities.
Assault – Unwarranted violence occurring while off-duty
Brutality – Unwarranted or excessive hysical violence occurring while on-duty
Civil Rights – Violations of general civil liberties that would be ruled unconstitutional yet not covered by other categories. For example, excessive force would be a violation of constitutionally protected rights, but is already covered in the Brutality class. However, complaints of warrantless eavesdropping or illegal disruptions of lawful protests would be deemed civil rights violations.
Sexual Misconduct- Sex related incidents including rape, sexual assault, harassment, coercion, prostitution, sex on duty, incest, and molestation.
Theft – includes robbery, theft, shoplifting, fraud, extortion, and bribery
Shooting – gun-related incidents both on and off-duty, including self-harm
Color of Law – incidents that involve misuse of authority such as bribery or extortion by threat of arrest
Perjury – includes false testimony, dishonesty during investigations, falsified charging papers, and falsified warrants.
Allegation – First stage of a misconduct complaint, can be from victim, witnesses, relatives of the victim, and other sources. Simply an allegation of misconduct.
Investigation – Second stage of a misconduct complaint, can be an internal investigation, criminal investigation, external investigation, or a DOJ/FBI civil rights investigation.
Lawsuits – Civil complaints filed in court, generally requires more evidence than a simple allegation, but still within the realm of allegations.
Charges – Criminal complaints filed in court, generally requires more evidence than a simple allegation, but still within the realm of allegations.
Trials – Criminal trials in court, requires enough evidence to establish probable cause, higher threshold than civil litigation or criminal charges, but still allegations.
Judgments – These are rulings that support a civil litigation complaint but also include settlement agreements that are typically, officially, said to not be admissions of guilt. Should be considered a confirmed case of misconduct.
Disciplinary – Results of investigations that confirm misconduct complaints but do not result in termination of employment.
Firings – Results of investigations that confirm misconduct severe enough to warrant termination of employment.
Convictions – Results of criminal trials that confirm allegations serious enough to warrant criminal charges. These include both rulings and guilty pleas.
Data is gathered from various media outlets by manual searches and review of daily news stories several times a day. There are no sufficient key terms that work well enough to automate this data gathering tasks, the results must be vetted by human intervention.
Confirmed stories about police misconduct that have been vetted to ensure that the story is about a case of misconduct or allegation of misconduct are published to a Twitter-based National Police Misconduct NewsFeed. From there, the stories are copied to a spreadsheet where they can later be sorted and analyzed.
At the first day of the month, data from the previous month is sorted and analyzed in the spreadsheet. All duplicate stories, stories that are informational, stories involving policy, and legislative issues are purged from the spreadsheet. Any items involving a status change about a specific incident are culled so that only the latest status story remains to avoid duplicate data. Only the most serious charge in a series of charges related to a single incident of misconduct are recorded to maintain parity with the national UCR statistical analysis methodology.
After all data has been analyzed it is presented on this site by General, Geographical, Type, and Status datasets.
The data collected and presented here should only be used to provide a very basic and general view of the extent of police misconduct within the US. It is, by no means, an accurate gauge that truly represents the exact extent of police misconduct since it relies on the information voluntarily gathered and/or released to the media, not from information gathered first-hand by independent monitors who investigate complaints of misconduct since no such agency exists nationally.
This information has been gathered here because nobody else is gathering it and the national government has not gathered it for several years. Keep in mind that geographical distribution of misconduct reports can be representative of concentrations of corruption or permissive attitudes towards abusive police policies or can be indications of more open information sharing between police agencies and local media along with departmental efforts to reduce misconduct by actively engaging problematic officers. There is no real way to determine which is the case since there is no independent monitoring and investigation into allegations of police misconduct.
In generally, monthly reports do not provide as accurate a depiction of the overall extent of police misconduct in the US as do quarterly and yearly reports as there is a fair amount of fluctuation between incident types and rates month by month. Therefore, monthly reports should only be considered as the state of police misconduct in that month itself while the longer-term reports paint a more comprehensive and accurate picture of police misconduct in the US.
As always, I appreciate any recommendations, advice, requests, and general comments.