Anyone know where I can find a good dataset on police misconducts? I'd like something similar to what RAND current has out on its policemisconduct.org site, but it looks like its pretty much a Twitter bot scraper. IE I'm pretty sure the same event is tracked at odd times throughout its disposition.

Related datasets I'm interested in would observe expenditures by police by jurisdiction/agency etc. I have seen the recent Federal Report on equipment purchases with federal grant monies; it is broken out by type of equipment and from which granting process. Is this info available somewhere besides PDF form (or how best to parse a PDF with, say, R or Python)?

Lastly, the ex-post observations of the payouts made by cities to citizens who have won misconduct suits against officers would also be of some use, certainly as a measure of "success" in the right contexts.

  • Regarding getting data from a PDF, see opendata.stackexchange.com/questions/4489/… Commented Feb 20, 2015 at 16:29
  • theres a local police force near me that published their internal affairs data annually for a few years. i never really looked at it, but if you're interested i'll find it
    – albert
    Commented Feb 22, 2015 at 0:55

3 Answers 3


Unfortunately, from PoliceMisconduct.net's own FAQ:

Why do this?

Simply because nobody else does. Only a small fraction of the 17,000 law enforcement agencies actually track their own misconduct in a semi-public manner, and even when they do, the data they provide is generic and does not specify what misconduct occurred, who did it, and what the end result was.

This FiveThirtyEight.com article by Reuben Fischer-Baum goes into more detail about the scarcity of this data:

As is the case with police shooting statistics, comprehensive numbers on accusations of police misconduct are hard to come by. There is no national reporting requirement for such accusations; in fact, many places have laws to purposefully keep the details of misconduct investigations out of the public eye.

For the article, Fischer-Baum settles for the NPMRP/PoliceMisconduct.net data.

  • Thanks. What do you think might be the best way to "de-dupe" that dataset if it really is picking up each "status update" of each misconduct? It looks like any time a misconduct hits Twitter, their bot scrapes it and row binds - but that process doesn't seem like its uniquely identifying the events...
    – d8aninja
    Commented Feb 23, 2015 at 4:30

http://www.fatalencounters.org/ Not specifically about misconduct, but attempts to be a database of all fatal encounters will U.S. law enforcement.

  • I have been here, and spoken with the admin about his great site. (One wonders what great "good" Periscopic's new .io site would do with the data he's got.) I think (perhaps as the AG gets his act together on a data-driven solution to police and oversight) recent outrage will propel us Data geeks to spend a little of our opportunity costs on aggregating our little piles into some insight. I hear some calling this notion "Big Data" these days :D
    – d8aninja
    Commented Feb 24, 2015 at 16:13

USA Today published a data-story about police misconduct (archived copy)

Starting with lists of officers who lost their law enforcement certification in 44 states, we are making those records available here.

CSV data: https://gofile.io/?c=bEbPGv

Data snippet (from 30k+ records)

"Name:","State:","Agency:","Year decertified:",""
"Gilberto Abad","Georgia","Atlanta Police","2004","https://www.usatoday.com/documents/5977556-Georgia-Decertifications/"
"Daniel G Abair","Georgia","Lee County Sheriff","2012","https://www.usatoday.com/documents/5977556-Georgia-Decertifications/"

And each record contains a URL to the pdf/txt documents:



URL of documents are derived from the URL in the CSV above

(my source)

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.