class: left, middle, inverse background-image: url(https://www.unomaha.edu/university-communications/downloadables/campus-icon-the-o/uno-icon-color.png) background-position: 95% 89% background-size: 10% # Police Use of Deadly Force ## What We Know and What We Need to Know </br> </br> [Justin Nix](https://jnix.netlify.app) Distinguished Associate Professor School of Criminology and Criminal Justice <br> <br> <br> .white[Curious People Series] .white[University of Nebraska Omaha] .white[April 12, 2022] --- class: middle, center, inverse # What Do We Know? ??? **Disclaimer**: I'm going to keep this conversation pretty high-level. - Much of the work on this topic focuses on a single jurisdiction, or a small number of them - Which OTOH is great because policing predominantly takes place at the local level - But OTOH we can't be sure how well such findings generalize --- class: top ## A lot more than we did before Ferguson <img src="ferguson.png" width="65%" style="display: block; margin: auto;" /> <p style="text-align: center; color: gray">.small[Image by [Jamelle Bouie](https://www.flickr.com/photos/jbouie/) on [Flickr](https://flic.kr/p/oA5tcz), [CC BY 2.0](https://creativecommons.org/licenses/by/2.0/)]</p> -- (Thanks to [WAPO](https://github.com/washingtonpost/data-police-shootings), [Fatal Encounters](https://fatalencounters.org/), and [Mapping Police Violence](https://mappingpoliceviolence.squarespace.com/)!) ??? Reporting "justifiable homicides" to the FBI is voluntary. We always knew underreporting was likely an issue, but since 2014, we've come to learn that the FBI was under counting the true number by about 50%. Fortunately journalists and private citizens have stepped up and provided the data But users beware! Each uses different inclusion criteria. Read their methodology before you go using them! I've reviewed multiple papers in the last year that saw key findings change after I pointed out the need to clean and/or consider the inclusion criteria. --- class: top ## On average, police kill 3 people per day -- <img src="ois_per_day.png" width="95%" style="display: block; margin: auto;" /> ??? Each year, WAPO consistently tracks about 1,000 fatal police shootings (range ~950 to ~1050) - Remarkable consistency - Over this period we've seen violence rise and fall - Debate about a "war on cops" and widespread de-policing after Ferguson and Floyd - A global pandemic - **And throughout it all, 3 people per day.** - The daily number of shootings bounces around between 0 and 9 - Nothing really jumps off the page here - No noticeable COVID or Floyd effect --- class: top ## There are racial disparities in terms of who police kill -- <div class="figure" style="text-align: center"> <img src="ois_by_race.png" alt="Source: WAPO 2015-20" width="90%" /> <p class="caption">Source: WAPO 2015-20</p> </div> ??? - Black people make up about 25% of police killings despite being just ~12% of the US population - Black men face about a 1 in 1000 chance of being killed by police over the life course - About 2.5x as likely as white men to be killed by police over the life course --- class: top ## Police seem to kill at higher rates in the western US -- <div class="figure" style="text-align: center"> <img src="schwartz_jahn_1.png" alt="Source: Schwartz & Jahn (2020)" width="90%" /> <p class="caption">Source: Schwartz & Jahn (2020)</p> </div> ??? This is from Gabe Schwartz and Jacquelyn Jahn in PLOS ONE. - They mapped police killings in all 382 MSAs from 2013 to 2017 using data from Fatal Encounters - This map shows each MSA's rate, color coded by quintiles - A clear pattern is visible here --- class: top ## ...And racial disparities vary a lot by region, too <div class="figure" style="text-align: center"> <img src="schwartz_jahn_2.png" alt="Source: Schwartz & Jahn (2020)" width="90%" /> <p class="caption">Source: Schwartz & Jahn (2020)</p> </div> ??? But check out how racial disparities differ by region. - This map shows the Black-white incidence rate ratios, again color coded by quintile - All of the IRRs are >1, suggesting disparity. - Black-white disparities way more pronounced in the Northeast --- class: top ## ...And racial disparities vary a lot by region, too <div class="figure" style="text-align: center"> <img src="schwartz_jahn_3.png" alt="Source: Schwartz & Jahn (2020)" width="90%" /> <p class="caption">Source: Schwartz & Jahn (2020)</p> </div> ??? This one shows Latino-white IRRs - Not all are >1 as in the Black-white IRRs - And the largest disparity is 1.5 (whereas for Black-white it was 6.5) - The highest Latino-White disparity is less than the lowest Black-White disparity. Here, we see greater disparity in the southwest, Colorado, but also in some places in the northeast. --- class: top ## Those killed are overwhelmingly men -- <div class="figure" style="text-align: center"> <img src="ois_by_sex.png" alt="Source: WAPO 2015-21" width="90%" /> <p class="caption">Source: WAPO 2015-21</p> </div> ??? Edwards et al: Lifetime risk for men is 1 in 2,000; for women its 1 in 33,000 - Women make up about 51% of the population, but just 5% of people killed by police. --- class: top ## The average decedent is nearly 40 -- <div class="figure" style="text-align: center"> <img src="ois_by_age.png" alt="Source: WAPO 2015-20" width="90%" /> <p class="caption">Source: WAPO 2015-20</p> </div> ??? For young men between the ages of 25 and 29, police killings are a "leading cause of death" - 6th, behind accidents (including overdoses), suicides, homicides, heart disease, and cancer But note that this distribution doesn't taper off until about age 40, whereas the "age-crime curve" tapers off quickly around age 25. --- class: top ## Most decedents were armed and/or attacking officers -- <div class="figure" style="text-align: center"> <img src="ois_by_weapon.png" alt="Source: WAPO 2015-20" width="90%" /> <p class="caption">Source: WAPO 2015-20</p> </div> ??? Almost 60% of those fatally shot were in possession of a firearm - Another 35% had a knife, blunt object, or some other object - And about 7% were unarmed Of course a simple "armed/unarmed" dichotomy doesn't tell the whole story. --- class: top ## Most decedents were armed and/or attacking officers <div class="figure" style="text-align: center"> <img src="ois_by_threat.png" alt="Source: WAPO 2015-20" width="90%" /> <p class="caption">Source: WAPO 2015-20</p> </div> ??? WAPO also includes a field that indicates whether a person posed an **imminent threat** - Imminent - firing a gun - attacking with a weapon other than a gun - pointing/brandishing a gun - Other - brandishing a knife - refusing to drop a non-gun weapon - driving a vehicle at a person - moving quickly/lunging at an officer without a weapon - furtive movements - fleeing - accidental shootings We could bicker over how some of these were coded. But a lot of them are pretty clear. In any event: these data suggest about 1 in 3 people fatally shot by police didn't pose an **imminent** threat. --- class: top ## Most decedents were armed and/or attacking officers <div class="figure" style="text-align: center"> <img src="ois_by_weaponXthreat.png" alt="Source: WAPO 2015-20" width="90%" /> <p class="caption">Source: WAPO 2015-20</p> </div> ??? Now, look what happens when we cross-tab *weapon* and *threat*. - 20% of those who had a gun apparently weren't pointing or shooting it. - 52% of those with some other weapon apparently weren't attacking anyone with it. On the flipside, note how 43% of those who were unarmed apparently still posed an imminent threat. Maybe going for the officer's gun, for example. - [One guy](https://www.tampabay.com/news/publicsafety/crime/hillsborough-state-attorney-clears-tampa-police-officer-in-18-year-olds/2262431/) was attempting to drown both an officer and his k9. - These sort of incidents are where BWCs can be very useful. Otherwise its the officer's word against the dead person's word. --- class: top ## About 1 in 4 decedents display signs of mental illness -- <div class="figure" style="text-align: center"> <img src="ois_by_mental_illness.png" alt="Source: WAPO 2015-20" width="90%" /> <p class="caption">Source: WAPO 2015-20</p> </div> ??? A lot has been made of the fact that 1 in 4 fatal shootings involved a person who displayed signs of mental illness. Here I've plotted the % of shootings where WAPO indicates the person had a mental illness each year, and there does appear to be an ever-so-slight downward trend. Two points are worth considering: - [Lum et al. (2021)](https://journals.sagepub.com/doi/10.1177/10986111211035002) recently analyzed millions of 911 calls to nine agencies and found that calls for mental health concerns made up about 1.3% of all calls on average. They conclude that getting police out of the business of responding to mental health crises won't do much to reduce the footprint of policing. Add to this how extremely rare fatal shootings are, and that 50% of the mentally ill decedents had a gun...and I'm not as optimistic that we'll see big reductions in these specific shootings any time soon. - I've also worked extensively with new "official" datasets (some of which I'll discuss momentarily), and they report a much lower percentage of shootings involve mentally ill persons. I suspect agencies use a more conservative definition than the news media, and that the truth might be somewhere in the middle. --- class: middle, center, inverse # What We Need Know š¤ --- class: top ## How often do police **use deadly force**? -- > *The true frequency of police decisions to employ firearms as a means of deadly force...can best be determined by considering woundings and off-target shots as only fortuitous variations of fatal shootings. All are of a kind.* <div style="text-align: right"> - Jim Fyfe (1978) </div> ??? Jim Fyfe was a pioneer in the study of police use of deadly force - He defined "deadly force" as *force that kills or is likely to kill*, noting that *it does not always kill* Any time a police officer discharges their firearm toward a person, they're using force that is likely to kill. Think of the Jacob Blake shooting in Kenosha, Wisconsin. Blake didn't die, but this was obviously a use of deadly force and it resulted in significant blowback. After Ferguson, everyone wanted to know how often police kill. I wish that after Kenosha, there was the same push to understand how often police shoot but don't kill. Some of Fyfe's early work higlighted noteworthy differences across agencies in terms of their "fatality rates." - Now much of that variation could just be random noise. But there could also be more to the story. - Consider those maps I showed earlier from the Schwartz and Jahn study. They only had fatality data. As did the Edwards paper that calculated the lifetime risks of being killed by police. - Now ask yourself, how much of those differences in police killing rates (or the disparities therein) are explained by officers' decisions to use deadly force, as opposed to, say, how close the nearest hospital happens to be, or what departmental policy is on administering aid, and if it is safe to administer aid? - **Note that these have different policy implications** -- So our **best datasets** only provide us with a non-random sample of the outcome we seek to understand <div class="figure" style="text-align: center"> <img src="puzzle.png" alt="Image by Mike Sweeney from Pixabay" width="30%" /> <p class="caption">Image by Mike Sweeney from Pixabay</p> </div> ??? **A non-random sample** This is critical because if the data we have are biased in some way (e.g., certain people or places more likely to show up in the data, or be ommitted from the data), it can cloud our vision on this. I've likened it to working on a puzzle where you don't have the box to know what the picture is, and you've only got about 35% of the pieces. You'd probably get a vague sense of what the picture is, but you could also have huge holes in what you see. **Researchers/academics in the audience**, have you ever tried to publish a paper with a 35% response rate? Unless you fell over yourself explaining why your data was still useful (and maybe regardless), you probably had at least one reviewer tell you this was a serious problem with your study. --- class: top ### Fatal shootings ā a random sample of deadly force incidents -- <div class="figure" style="text-align: center"> <img src="vice_fatality_rates.png" alt="Source: VICE News" width="90%" /> <p class="caption">Source: VICE News</p> </div> ??? Around 2016, VICE News published shooting data from 47/50 largest jurisdictions in the US. *All* shootings, not just those resulting in death. - Like Fyfe's earlier work, one of the first things I noticed was the variation in fatality rates. - Most agencies hovered in the 30-50% range, but you can see here that Boston's was 71% and at the other extreme, St. Louis was 17%. - **Transparency shenanigans**: The Detroit Police Department said it would take up to 3,120 business days and cost at least $77,532 to retrieve records that other departments made available online for free. --- class: top ### Fatal shootings ā a random sample of deadly force incidents <div class="figure" style="text-align: center"> <img src="vice_boston_atl.png" alt="Source: VICE News" width="90%" /> <p class="caption">Source: VICE News</p> </div> ??? Let's do a thought exercise. Imagine you pull data from one of the three sources I mentioned earlier - WAPO, FE, MPV - which only track fatalities. And you want to compare agencies to see how they stack up in terms of police killings. - So according to VICE, you'd see that Boston and Atlanta each fatally shot 10 people. So you'd conclude they're on par with each other. - **But actually,** Atlanta used deadly force on 33 other occasions, compared to Boston's 4. The truth is, Atlanta PD used deadly force over 3x as much as Boston during this period. --- class: top ### Fatal shootings ā a random sample of deadly force incidents <div class="figure" style="text-align: center"> <img src="vice_vegas_stlouis.png" alt="Source: VICE News" width="90%" /> <p class="caption">Source: VICE News</p> </div> ??? From the same data, you might see that Las Vegas had 47 fatal shootings to St. Louis' 20. And you'd conclude that LVMPD kills people over twice as often as SLMPD. - **But actually**, their total shooting numbers were nearly identical. And granted, I'm just talking about frequencies here - we've not yet gotten to the need to calculate meaningful *rates*. That's coming. --- class: top ### Fatal shootings ā a random sample of deadly force incidents <div class="figure" style="text-align: center"> <img src="pone_vbar.png" alt="Source: Nix & Shjarback (2021)" width="90%" /> <p class="caption">Source: Nix & Shjarback (2021)</p> </div> ??? Last year, my colleague John Shjarback (who just became a dad!) and I published a paper that leveraged several years of data from four states. Three of these states had recently passed legislation requiring all local agencies to submit data on fatal AND injurious police shootings. Florida's data were compiled for a story by the Tampa Bay Times. So note in this case we still don't have information on incidents where police shot but missed people. This first bar chart just shows the raw number of people killed or injured by police gunfire in each state during the study period. The first takeaway is that in just these 4 states, **there were 1322 police shootings that resulted in injury but not death**. --- class: top ### Fatal shootings ā a random sample of deadly force incidents <div class="figure" style="text-align: center"> <img src="pone_dot.png" alt="Source: Nix & Shjarback (2021)" width="90%" /> <p class="caption">Source: Nix & Shjarback (2021)</p> </div> ??? A follow-up question is: are there factors that increase (or decrease) a person's odds of death upon being shot by a police officer? This dot plot suggests the answer is yes. Take note of each state's symbol. On the y-axis, we've got a list of several individual/ecological factors, and on the x-axis, we've got the mortality rate that runs from 0 to 100%. - In all 4 states, Black people shot by police were *less likely* to die than White, Hispanic, or Asian people. - In every state but Texas, men were more likely to die than women. - Pretty consistently, we see that the mortality rate increases with age. - In all 4 states, people armed with a deadly weapon were quite a bit more likely to die than people who were unarmed. - No clear patterns emerged with respect to trauma care availability or the "urbanicity" of the county. --- class: top ### Fatal shootings ā a random sample of deadly force incidents <div class="figure" style="text-align: center"> <img src="pone_race_margins.png" alt="Source: Nix & Shjarback (2021)" width="90%" /> <p class="caption">Source: Nix & Shjarback (2021)</p> </div> ??? John and I went on to run regression models that estimate the odds of death (conditional on being shot) while controlling for these same factors. This margins plot shows each group's probability of death upon being shot by police. - So while controlling for the presence of a weapon, age, gender, access to trauma care, etc., we find that Black people in the pooled sample were about 7% less likely to die when they were shot. - I also like that this plot shows how much more uncertainty there is for, e.g., Asian or Native American People. Because the number of such people shot each year is so much smaller in comparison than the number of White, Black, and Hispanic people, it's much harder to discern meaningful variation from noise. - E.G., for "Other" people, our 95% confidence interval runs from about 54% (which would be on par with White and Hispanic) to about 74% (which would be much higher than the other groups). --- class: top ### Fatal shootings ā a random sample of deadly force incidents <div class="figure" style="text-align: center"> <img src="pone_age_margins.png" alt="Source: Nix & Shjarback (2021)" width="90%" /> <p class="caption">Source: Nix & Shjarback (2021)</p> </div> ??? And here we've plotted the probability of death for each of four different age groups. This suggest a rather clear positive correlation between age and the likelihood of death upon being shot by police. On average, as we get older, our bodies become less effective at responding to trauma. --- class: top ### Fatal shootings ā a random sample of deadly force incidents <div class="figure" style="text-align: center"> <img src="hou_fig10.png" alt="Source: Hou (2022, p. 122)" width="95%" /> <p class="caption">Source: Hou (2022, p. 122)</p> </div> ??? Just a couple months ago, Yuchen Hou (from John Jay) published his dissertation, which involved scraping the Gun Violence Archive for all fatal and nonfatal police shootings in the US in 2015. In just that one year, he discovered 733 injurious shootings and 246 that resulted in no injuries (probably a significant undercount). Here you can see that in the Pacific, Mountain, and West South Central census divisions, fatal shootings were more prevalent than nonfatal shootings. But the opposite is true in the other 6 divisions. --- class: top ### Fatal shootings ā a random sample of deadly force incidents <div class="figure" style="text-align: center"> <img src="hou_fig11.png" alt="Source: Hou (2022, p. 123)" width="70%" /> <p class="caption">Source: Hou (2022, p. 123)</p> </div> ??? In this figure, he sorts states by: - The total number of shootings (top) - The rate of shootings per 100K (middle) - And the difference between the fatal and nonfatal shooting rates (bottom) - Recall my study with John where we had data from TX, CA, FL, and CO. Most of these states appear on the left side of the chart. How well might our findings generalize to, say, New Mexico or Kentucky? - The honest answer is we don't know. --- class: top ## Those tricky counterfactuals ??? OK, hopefully I've convinced you that we need better data on nonfatal police shootings. To be clear, some agencies provide this information on their websites or in annual reports. Or on request. But I think it'd be really useful if we had more comprehensive data, such as what we have for fatal shootings. That said, knowing all the details of when police shoot only tells us so much. To really understand some of the variation I've discussed thus far, we need to know the counterfactuals. And this is way trickier. -- <img src="iceburg.jpg" width="75%" style="display: block; margin: auto;" /> ??? Imagine you're on a boat and you can see an iceberg up ahead in the water. You've got no idea how big it *really* is because you can't see what lurks beneath the surface. But just because you can't see it doesn't mean you can simply ignore it! --- class: top ## Those tricky counterfactuals <img src="iceburg2.jpg" width="75%" style="display: block; margin: auto;" /> ??? - Researchers and other users of data on police use of deadly force have to be mindful of the iceberg problem. - **What's the right denominator?** - In other words, *who was exposed to the risk of being shot or killed by the police*? And *how much*? - It's not the population. But nor is it crime or arrest data. - It is well-documented that police are more likely to initiate contact with nonwhite people (e.g., stop or arrest). And in order to be killed by police, you have to first come into contact with police. - At the same time, we can't focus only on who gets stopped, because that erases any bias that might have gone into the decision to make contact with individuals of various groups. --- class: top ## Those tricky counterfactuals <img src="denominators_538.PNG" width="85%" style="display: block; margin: auto;" /> ??? Here I'm sharing an image from an excellent story in FiveThirtyEight a while back, which illustrates how the iceburg problem can mislead us. The story is linked at the end of my presentation, which I'll make available on my website. - **Key takeaway:** Comparing use of force rates only among those who were stopped can obscure bias. - My sense in studying deadly force over the years is that it's useful to analyze the various "decision points" separately - E.G., the decision to stop is quite different from the decision to use deadly force, so I find it useful to examine whether there are disparities in use of force, conditional on who was stopped. Finding no disparity doesn't mean there **isn't** disparity - in fact it probably just occurs earlier in the "decision tree." - And again, this is important information for reforms meant to reduce the use of deadly force. It would suggest that best way to reduce disparities in police killings would be to reduce stops. Then the key question is whether/to what extent there's a public safety tradeoff. --- class: top ## More about the officers involved ??? Another thing we need to know more about is who the officers involved in these shootings are. -- Just identifying them has proven difficult ??? Agencies often withhold the identity of officers - citing officer safety concerns or that investigations are ongoing. - WAPO tried to start tracking down information about the officers who shoot in 2016, but quickly found this much more difficult than tracking down information about the decedents. And ultimately, AFAIK, they abandoned the effort. At minimum, they haven't shared that information publicly. -- But beyond identifying them, we need to tease out *why* we see the variation that we do <img src="thinker.png" width="30%" style="display: block; margin: auto;" /> <p style="text-align: center; color: gray">.small[Image by [Stephen Carlile](https://www.flickr.com/photos/stephencarlile/) on [Flickr](https://flic.kr/p/K5Mdx), [CC BY-ND 2.0](https://creativecommons.org/licenses/by-nd/2.0/)]</p> ??? Two points to keep in mind here. š --- class: top ### Beware **data density bias!** (see Chalfin & Kaplan, 2021) ??? The first is "data density bias." (h/t to Aaron Chalfin & Jacob Kaplan). - By definition, when some outcome (deadly force, complaints) is relatively small compared to the number of officers, it will be true **by definition** that a small share of the officers will account for a large share of the outcome. - Imagine an agency with 1000 officers that has five deadly force incidents in a given year. Even if there were five separate officers involved (which is by definition **not clustered**), it will be true that 0.5% of the agency is responsible for 100% of the deadly force incidents. This sounds scary but it's just a statistical artifact of studying a rare outcome. -- .pull-left[ <img src="tweet.png" width="100%" style="display: block; margin: auto;" /> ] .pull-right[ <img src="tweet2.png" width="100%" style="display: block; margin: auto;" /> <img src="wapo_misconduct.png" width="80%" style="display: block; margin: auto;" /> ] ??? Here I've shown two tweets by an individual with a huge platform and social media reach, who suggests we simply lay off the officers who are involved in the most use of force incidents. - This is not a sound recommendation based on these data IMO And in the lower right corner, a recent story from WAPO. Over 11 years, 5 officers were involved in misconduct that resulted in 101 or more cash settlements each. One officer racked up 143. - I don't mean to suggest we shouldn't pay attention to outliers. Of course we should. We just shouldn't leap to conclusions, especially given the possibility of data density bias. - What we need is more context... --- class: top ### The counterfactuals matter here, too -- Officers vary in the rate at which they [stop/investigate people](https://doi.org/10.1126/science.abd8694) <div class="figure" style="text-align: center"> <img src="ba_science.png" alt="Source: Ba et al. (2021)" width="75%" /> <p class="caption">Source: Ba et al. (2021)</p> </div> ??? This figure is from a recent study of the Chicago Police Department in Science Magazine by Bocar Ba and colleagues. - Here the authors are comparing officers who work the same beats on the same shifts on the same days of the same months - As close to "apples to apples" as you can get - And they're comparing the rate of stops, arrests, and uses of force per 100 shifts - **Takeaway**: Compared to white officers, Black officers make fewer stops and arrests, and use force less often. Same is true of Hispanic officers, albeit the disparity is smaller. - Women stop at the same rate, but make fewer arrests and use force less often. Granted, this is just Chicago, and what I said earlier about generalizing still applies here. But at least in Chicago, when analyzing how often certain officers use force, it's necessary to account for **differences in how often they expose themselves to the opportunity to use force**. -- As well as in their ["thresholds"](https://doi.org/10.1177%2F0002716219896553) and ability to [talk to people](https://doi.org/10.1177%2F0002716219887366) ??? My friend Natalie Todak has done extensive fieldwork on de-escalation. Not surprisingly, she found considerable variation in officers' ability to communicate with people, to empathize with their situations, and so on. And officers were keen to these differences as well - there were certain individuals within the agency she studied who were recognized by their peers as being especially skilled communicators. --- class: top ## What happened **before** the "final frame?" -- <img src="reel.png" width="75%" style="display: block; margin: auto;" /> <p style="text-align: center; color: gray">.small[Image by [Steve Snodgrass](https://www.flickr.com/photos/stevensnodgrass/) on [Flickr](https://flic.kr/p/nJfVwE), [CC BY 2.0](https://creativecommons.org/licenses/by/2.0/)]</p> ??? Before moving on to some reform ideas, I just want to mention that with all this new data, we have to remember not to obsess over the "final frame" to the point that we ignore important details during the lead up to the final frame. Again, BWCs are very helpful here. Questions to consider include: - Was this a case of officer-initiated jeopardy? - Could the officer have waited for backup? Or waited the citizen out? - Could the officer have safely attempted less-lethal options? Did he? My point is, these 1000 fatal shootings per year come in many different forms, and there's so much more to consider than simple questions like "was the person who was shot armed or unarmed at the moment he was shot." --- class: middle, center, inverse # What Are Some Promising Reforms? ??? OK. Moving on. I'm not suggesting this list is exhaustive, or fool-proof. It's just a short list of ideas I believe are promising based on what I know. --- class: top ## š±š£ We need better data! <img src="cowbell.jpg" width="60%" style="display: block; margin: auto;" /> ??? Hopefully I can get off this soapbox at this point in my presentation. Better data can provide us with a better understanding of the issue. As long as we understand how to use it in a meaningful way. Again, just 10 short years ago we didn't know nearly as much as we know now. --- class: top ## Maybe some of what Omaha's doing? <div class="figure" style="text-align: center"> <img src="opd_shootings.png" alt="Source: OPD & Omaha World Herald" width="90%" /> <p class="caption">Source: OPD & Omaha World Herald</p> </div> ??? Better data can reveal which agencies are performing well so that we can learn from them. Here in Omaha, there has been a sharp decline in police shootings since 2010. From 2010-11, there were 20 shootings. In the last 3 years, AFAIK, there have been 5. **Why?** Chief Todd Schmaderer points to BWCs, better training, mental health co-responder program, hiring more veterans from other agencies, and a culture shift. Tough to say as we've not done any sort of rigorous analysis that I'm aware of, but the truth is it probably is some combination of factors like these. There probably isn't some silver bullet reform - it'll take more work than that. --- class: top ## Document when officers point their guns -- </br> .pull-left[ <div class="figure" style="text-align: center"> <img src="jennings.PNG" alt="Source: Jennings & Rubado (2017)" width="100%" /> <p class="caption">Source: Jennings & Rubado (2017)</p> </div> ] .pull-right[ <div class="figure" style="text-align: center"> <img src="shjarback.jpg" alt="Source: Shjarback et al. (2021)" width="100%" /> <p class="caption">Source: Shjarback et al. (2021)</p> </div> ] ??? According to a 2016 estimate, only about 54% of all agencies require written documentation when officers display their firearms. - When officers point their guns at a person, that's a use of force. And I believe that all uses of force should be tracked, just from a moral perspective. - But it turns out we have good evidence that this reporting requirement is associated with lower rates of deadly force. - On the left, a national study of 1100 agencies which found that this reporting requirement is associated with lower police-involved death rates, while controlling for inequality, racial composition, urbanicity, and community policing training, among others. - On the right, a study by some of my colleagues who found that since the Dallas Police Department implemented this policy circa 2013, shootings have declined significantly. And moreover, they uncovered no evidence that cops were more at risk because of this policy. - So to me this seems like an easy one. Require officers to document this type of force. --- class: top ## Consider de-escalation training programs -- <div class="figure" style="text-align: center"> <img src="icat.png" alt="Source: Engel et al. (2022)" width="90%" /> <p class="caption">Source: Engel et al. (2022)</p> </div> ??? Much has been made of de-escalation training and we're finally accumulating a body of evidence to go on. Here's a figure from an evaluation of Louisville's "ICAT" training, which suggests that as the training was rolled out, LMPD's use of force rates declined significantly. - Researchers found a 28% reduction in use of force incidents, - A 26% reduction in citizen injuries - And a 36% reduction in officer injuries --- class: top ## Consider de-escalation training programs <div class="figure" style="text-align: center"> <img src="t3_training.png" alt="Source: McLean et al. (2020)" width="80%" /> <p class="caption">Source: McLean et al. (2020)</p> </div> ??? However, more research is needed before agencies go all in on this. In Tucson and Fayetteville, my colleagues didn't find the same noteworthy reductions in use of force incidents. This was true among both low and high-dose treatment conditions. So my recommendation is for communities to **consider** this sort of training. They're not cheap nor are they guaranteed to generate the reductions that Louisville experienced. --- class: top ## Cut down on shootings of people who aren't brandishing or shooting a gun ??? My last recommendation is the one that I'm sure will generate the most pushback from police. But I'll say it anyway. In 2017 Frank Zimring published a book called "When Police Kill." It's a fantastic read and I highly recommend it. On page 87 he makes the following observation: -- > *Knives and blunt objects and personal attacks do not threaten the lives of police officers on either side of the Atlantic. Yet they produce lethal responses from American police hundreds of times a year. Why is this?* <div style="text-align: right"> - Zimring (2017: 87) </div> <div class="figure" style="text-align: center"> <img src="leoka.png" alt="Source: FBI LEOKA" width="52%" /> <p class="caption">Source: FBI LEOKA</p> </div> ??? I've paired his observation with LEOKA data showing officers killed each year going back to 2021, and how they were killed. - It's true in the US that officers are very rarely killed by anything other than guns. - Zimring goes as far as to sugggest that knives aren't "deadly weapons" when used on the police. - That's maybe pushing it a bit... - Probably can't *forbid* shooting people with knives or blunt objects, as Zimring argues... - Consider **Kisela v. Hughes** and the **Ma'Khia Bryant** shootings - But the alternative shouldn't be to do nothing - We know that police shoot and kill about 1000 people every year. - If we set out to reduce killings by 10%, where would we start? - All I'm saying is that on average, it's probably harder to reduce killings of people armed with guns (~600) than everyone else (~400). --- class: top ## Cut down on shootings of people who aren't brandishing or shooting a gun <div class="figure" style="text-align: center"> <img src="chitwood.png" alt="Source: Montgomery et al. (2017)" width="45%" /> <p class="caption">Source: Montgomery et al. (2017)</p> </div> ??? To do so would require a culture shift, as well as a change in how we train officers to respond to persons armed with knives or blunt instruments. I've pasted here a quote from a police chief in Florida, who recounted the time a guy almost bit off his finger. Though he might have been legally justified to shoot, he instead chose to strike the man until he let go. In doing so he broke his left hand in four places. - With this example, all I'm saying is I suspect many cops might have handled the situation differently. --- class: middle, center, inverse # In conclusion... --- class: middle ## It will be difficult to reduce police shootings significantly without **drastic societal changes** -- ## That said: 1,000 fatal shootings per year doesn't have to be normal -- ## The data we have, imperfect as it is, provides us with several good starting points --- class: top, center background-image: url(https://www.unomaha.edu/university-communications/downloadables/campus-icon-the-o/uno-icon-color.png) background-position: 50% 59% background-size: 10% # Thank you! ## Questions? Justin Nix *School of Criminology and Criminal Justice* *University of Nebraska Omaha* </br> </br> </br> <svg viewBox="0 0 512 512" style="height:1em;position:relative;display:inline-block;top:.1em;" xmlns="http://www.w3.org/2000/svg"> <path d="M459.37 151.716c.325 4.548.325 9.097.325 13.645 0 138.72-105.583 298.558-298.558 298.558-59.452 0-114.68-17.219-161.137-47.106 8.447.974 16.568 1.299 25.34 1.299 49.055 0 94.213-16.568 130.274-44.832-46.132-.975-84.792-31.188-98.112-72.772 6.498.974 12.995 1.624 19.818 1.624 9.421 0 18.843-1.3 27.614-3.573-48.081-9.747-84.143-51.98-84.143-102.985v-1.299c13.969 7.797 30.214 12.67 47.431 13.319-28.264-18.843-46.781-51.005-46.781-87.391 0-19.492 5.197-37.36 14.294-52.954 51.655 63.675 129.3 105.258 216.365 109.807-1.624-7.797-2.599-15.918-2.599-24.04 0-57.828 46.782-104.934 104.934-104.934 30.213 0 57.502 12.67 76.67 33.137 23.715-4.548 46.456-13.32 66.599-25.34-7.798 24.366-24.366 44.833-46.132 57.827 21.117-2.273 41.584-8.122 60.426-16.243-14.292 20.791-32.161 39.308-52.628 54.253z"></path></svg> [jnixy](https://twitter.com/jnixy) | <svg viewBox="0 0 512 512" style="height:1em;position:relative;display:inline-block;top:.1em;" xmlns="http://www.w3.org/2000/svg"> <path d="M464 64H48C21.49 64 0 85.49 0 112v288c0 26.51 21.49 48 48 48h416c26.51 0 48-21.49 48-48V112c0-26.51-21.49-48-48-48zm0 48v40.805c-22.422 18.259-58.168 46.651-134.587 106.49-16.841 13.247-50.201 45.072-73.413 44.701-23.208.375-56.579-31.459-73.413-44.701C106.18 199.465 70.425 171.067 48 152.805V112h416zM48 400V214.398c22.914 18.251 55.409 43.862 104.938 82.646 21.857 17.205 60.134 55.186 103.062 54.955 42.717.231 80.509-37.199 103.053-54.947 49.528-38.783 82.032-64.401 104.947-82.653V400H48z"></path></svg> [jnix@unomaha.edu](mailto:jnix@unomaha.edu) | <svg viewBox="0 0 496 512" style="height:1em;position:relative;display:inline-block;top:.1em;" xmlns="http://www.w3.org/2000/svg"> <path d="M336.5 160C322 70.7 287.8 8 248 8s-74 62.7-88.5 152h177zM152 256c0 22.2 1.2 43.5 3.3 64h185.3c2.1-20.5 3.3-41.8 3.3-64s-1.2-43.5-3.3-64H155.3c-2.1 20.5-3.3 41.8-3.3 64zm324.7-96c-28.6-67.9-86.5-120.4-158-141.6 24.4 33.8 41.2 84.7 50 141.6h108zM177.2 18.4C105.8 39.6 47.8 92.1 19.3 160h108c8.7-56.9 25.5-107.8 49.9-141.6zM487.4 192H372.7c2.1 21 3.3 42.5 3.3 64s-1.2 43-3.3 64h114.6c5.5-20.5 8.6-41.8 8.6-64s-3.1-43.5-8.5-64zM120 256c0-21.5 1.2-43 3.3-64H8.6C3.2 212.5 0 233.8 0 256s3.2 43.5 8.6 64h114.6c-2-21-3.2-42.5-3.2-64zm39.5 96c14.5 89.3 48.7 152 88.5 152s74-62.7 88.5-152h-177zm159.3 141.6c71.4-21.2 129.4-73.7 158-141.6h-108c-8.8 56.9-25.6 107.8-50 141.6zM19.3 352c28.6 67.9 86.5 120.4 158 141.6-24.4-33.8-41.2-84.7-50-141.6h-108z"></path></svg> [jnix.netlify.app](https://jnix.netlify.app) </br> <svg viewBox="0 0 496 512" style="height:1em;position:relative;display:inline-block;top:.1em;" xmlns="http://www.w3.org/2000/svg"> <path d="M165.9 397.4c0 2-2.3 3.6-5.2 3.6-3.3.3-5.6-1.3-5.6-3.6 0-2 2.3-3.6 5.2-3.6 3-.3 5.6 1.3 5.6 3.6zm-31.1-4.5c-.7 2 1.3 4.3 4.3 4.9 2.6 1 5.6 0 6.2-2s-1.3-4.3-4.3-5.2c-2.6-.7-5.5.3-6.2 2.3zm44.2-1.7c-2.9.7-4.9 2.6-4.6 4.9.3 2 2.9 3.3 5.9 2.6 2.9-.7 4.9-2.6 4.6-4.6-.3-1.9-3-3.2-5.9-2.9zM244.8 8C106.1 8 0 113.3 0 252c0 110.9 69.8 205.8 169.5 239.2 12.8 2.3 17.3-5.6 17.3-12.1 0-6.2-.3-40.4-.3-61.4 0 0-70 15-84.7-29.8 0 0-11.4-29.1-27.8-36.6 0 0-22.9-15.7 1.6-15.4 0 0 24.9 2 38.6 25.8 21.9 38.6 58.6 27.5 72.9 20.9 2.3-16 8.8-27.1 16-33.7-55.9-6.2-112.3-14.3-112.3-110.5 0-27.5 7.6-41.3 23.6-58.9-2.6-6.5-11.1-33.3 2.6-67.9 20.9-6.5 69 27 69 27 20-5.6 41.5-8.5 62.8-8.5s42.8 2.9 62.8 8.5c0 0 48.1-33.6 69-27 13.7 34.7 5.2 61.4 2.6 67.9 16 17.7 25.8 31.5 25.8 58.9 0 96.5-58.9 104.2-114.8 110.5 9.2 7.9 17 22.9 17 46.4 0 33.7-.3 75.4-.3 83.6 0 6.5 4.6 14.4 17.3 12.1C428.2 457.8 496 362.9 496 252 496 113.3 383.5 8 244.8 8zM97.2 352.9c-1.3 1-1 3.3.7 5.2 1.6 1.6 3.9 2.3 5.2 1 1.3-1 1-3.3-.7-5.2-1.6-1.6-3.9-2.3-5.2-1zm-10.8-8.1c-.7 1.3.3 2.9 2.3 3.9 1.6 1 3.6.7 4.3-.7.7-1.3-.3-2.9-2.3-3.9-2-.6-3.6-.3-4.3.7zm32.4 35.6c-1.6 1.3-1 4.3 1.3 6.2 2.3 2.3 5.2 2.6 6.5 1 1.3-1.3.7-4.3-1.3-6.2-2.2-2.3-5.2-2.6-6.5-1zm-11.4-14.7c-1.6 1-1.6 3.6 0 5.9 1.6 2.3 4.3 3.3 5.6 2.3 1.6-1.3 1.6-3.9 0-6.2-1.4-2.3-4-3.3-5.6-2z"></path></svg> [Replication Code](https://github.com/jnixy/replication-materials/blob/master/nix_curious_ppl_talk/curious_ppl_code.do) --- class: top, references-font # References/Additional Readings Ba, B. A., Knox, D., Mummolo, J., & Rivera, R. (2021). [The role of officer race and gender in police-civilian interactions in Chicago](https://doi.org/10.1126/science.abd8694). *Science*, 371(6530), 696-702. Bronner, L. & Mithani, J. (2020, June). [Why statistics don't capture the full extent of the systemic bias in policing](https://fivethirtyeight.com/features/why-statistics-dont-capture-the-full-extent-of-the-systemic-bias-in-policing/). *FiveThirtyEight*. Chalfin, A., & Kaplan, J. (2021). [How many complaints against police officers can be abated by incapacitating a few ābad apples?ā](https://doi.org/10.1111/1745-9133.12542) *Criminology & Public Policy*, 20(2), 351-370. Conley, A. (2020, July 27). [Use of force rare among Omaha police; Decrease in shootings 'impressive,' expert says](https://omaha.com/news/local/crime-and-courts/use-of-force-rare-among-omaha-police-decrease-in-shootings-impressive-expert-says/article_31f8b976-5dba-558e-89b1-61998e619e24.html). *Omaha World-Herald* (accessed 7 April 2022). Edwards, F., Lee, H., & Esposito, M. (2019). [Risk of being killed by police use of force in the United States by age, raceāethnicity, and sex](https://doi.org/10.1073/pnas.1821204116). *Proceedings of the National Academy of Sciences*, 116(34), 16793-16798. Engel, R. S., Isaza, G. T., Motz, R. T., McManus, H. D., & Corsaro, N. (2021). [De-Escalation Training Receptivity and First-Line Police Supervision: Findings from the Louisville Metro Police Study](https://doi.org/10.1177%2F10986111211049834). *Police Quarterly*. (Advance online publication). --- class: top, references-font # References/Additional Readings Engel, R. S., Corsaro, N., Isaza, G. T., & McManus, H. D. (2022). [Assessing the impact of deāescalation training on police behavior: Reducing police use of force in the Louisville, KY Metro Police Department](https://doi.org/10.1111/1745-9133.12574). *Criminology & Public Policy* (Advance online publication). Federal Bureau of Investigation (2019). [*Table 28: Law Enforcement Officers Feloniously Killed by Type of Weapon, 2010-2019*](https://ucr.fbi.gov/leoka/2019/tables/table-28.xls). Fyfe, J. J. (1978). *Shots fired: An analysis of New York City police firearms discharges*. (Unpublished doctoral dissertation). State University of New York at Albany. Hemenway, D., Azrael, D., Conner, A., & Miller, M. (2019). [Variation in rates of fatal police shootings across US states: the role of firearm availability](https://link.springer.com/article/10.1007/s11524-018-0313-z). *Journal of Urban Health*, 96(1), 63-73. Hou, Y. (2022). [Fatal and Non-Fatal Police Shootings in the United States, 2015: An Examination of Open-Source Data](https://www.proquest.com/docview/2632097653) (Doctoral dissertation, City University of New York). Jennings, J. T., & Rubado, M. E. (2017). [Preventing the use of deadly force: The relationship between police agency policies and rates of officerāinvolved gun deaths](https://doi.org/10.1111/puar.12738). *Public Administration Review*, 77(2), 217-226. --- class: top, references-font # References/Additional Readings McLean, K., Wolfe, S. E., Rojek, J., Alpert, G. P., & Smith, M. R. (2020). [Randomized controlled trial of social interaction police training](https://doi.org/10.1111/1745-9133.12506). *Criminology & Public Policy*, 19(3), 805-832. Montgomery, B., Humburg, C., & Herndon, M. (2017, April). [Why cops shoot: An unprecedented review of Florida police shootings reveals how fear and bias breed confusion, how order quickly dissolves into chaos, and ways to avert the violence](https://projects.tampabay.com/projects/2017/investigations/florida-police-shootings/why-cops-shoot/). *Tampa Bay Times*. Nix, J., & Shjarback, J. A. (2021). [Factors associated with police shooting mortality: A focus on race and a plea for more comprehensive data](https://doi.org/10.1371/journal.pone.0259024). *PLOS ONE*, 16(11), e0259024. Nix, J. (2020). [On the challenges associated with the study of police use of deadly force in the United States: A response to Schwartz & Jahn](https://doi.org/10.1371/journal.pone.0236158). *PLOS ONE*, 15(7), e0236158. Riddell, J. R., & Worrall, J. L. (2021). [Predicting firearm and CEW displays as police officers' response to resistance](https://doi.org/10.1016/j.jcrimjus.2020.101775). *Journal of Criminal Justice*, 72, 101775. Ridgeway, G. (2020). [The role of individual officer characteristics in police shootings](https://doi.org/10.1177%2F0002716219896553). *The ANNALS of the American Academy of Political and Social Science*, 687(1), 58-66. --- class: top, references-font # References/Additional Readings Schwartz, G. L., & Jahn, J. L. (2020). [Mapping fatal police violence across US metropolitan areas: Overall rates and racial/ethnic inequities, 2013-2017](https://doi.org/10.1371/journal.pone.0229686). *PLOS ONE*, 15(6), e0229686. Shjarback, J. A., White, M. D., & Bishopp, S. A. (2021). [Can police shootings be reduced by requiring officers to document when they point firearms at citizens?](http://dx.doi.org/10.1136/injuryprev-2020-043932) *Injury Prevention*, 27(6), 508-513. Tregle, B., Nix, J., & Alpert, G. P. (2019). [Disparity does not mean bias: Making sense of observed racial disparities in fatal officer-involved shootings with multiple benchmarks](https://doi.org/10.1080/0735648X.2018.1547269). *Journal of Crime and Justice*, 42(1), 18-31. VerBruggen, R. (2022, March). [Fatal police shootings and race: A review of the evidence and suggestions for future research](https://www.manhattan-institute.org/verbruggen-fatal-police-shootings). *The Manhattan Institute*. VICE News (2017, December). [Get data on nonfatal and fatal police shootings in the 50 largest U.S. police departments](https://www.vice.com/en/article/a3jjpa/nonfatal-police-shootings-data). --- class: top, references-font # References/Additional Readings Wheeler, A. P., Phillips, S. W., Worrall, J. L., & Bishopp, S. A. (2017). [What factors influence an officerās decision to shoot? The promise and limitations of using public data](https://doi.org/10.1177%2F1525107118759900). *Justice Research and Policy*, 18(1), 48-76. Wolfe, S., Rojek, J., McLean, K., & Alpert, G. (2020). [Social interaction training to reduce police use of force](https://doi.org/10.1177%2F0002716219887366). *The ANNALS of the American Academy of Political and Social Science*, 687(1), 124-145. Worrall, J. L., Bishopp, S. A., Zinser, S. C., Wheeler, A. P., & Phillips, S. W. (2018). [Exploring bias in police shooting decisions with real shoot/donāt shoot cases](https://doi.org/10.1177%2F0011128718756038). *Crime & Delinquency*, 64(9), 1171-1192. Zimring, F. (2017). *When police kill*. Cambridge, MA: Harvard University Press. - See my short review [here](http://transformativestudies.org/wp-content/uploads/10.3798tia.1937-0237.1730.pdf) <!-- ```{css, echo=FALSE} --> <!-- @media print { --> <!-- .has-continuation { --> <!-- display: block; --> <!-- } --> <!-- } --> <!-- ``` --> <style> p.caption { font-size: 0.5em; color: gray; } </style>