Last month, the U.S. Department of Justice issued a voluminous report on uses of deadly force by the Philadelphia Police Department. In recent years, there has been a drop in both violent crime and assaults on police officers in the City of Brotherly Love, but officer-involved shootings (OISs) have remained stubbornly high. Amidst media coverage of rising OIS numbers in 2013, the Police Department requested assistance from the DOJ in order to assess the problem.

The new report, authored by George Fachner and Steven Carter, finds there were 394 OISs in Philadelphia between 2007 and 2014, for an average of 49 per year. The suspects were unarmed in 15% of the cases. Fachner and Carter provide a wealth of data regarding the 394 OISs and dozens of recommendations for the Department.

One recommendation is that the “PPD should publish a detailed report on use of force, including deadly force, on an annual basis. The report should be released to the public.”

I’m pleased to say that we are already doing such annual reports here in Milwaukee. How do the numbers compare?

In 2013, the MPD reported 40 uses of force involving a firearm. However, in 26 of the incidents, the firearm was used against a dog; these must be excluded because the Philadelphia report only counts incidents involving humans. That leaves 14 OISs in Milwaukee, in comparison to the annual average of 49 in Philadelphia.

Of course, Philadelphia is a larger city than Milwaukee. Still, even holding population constant, we see a sizable gap between Philadelphia’s 3.2 shootings per 100,000 residents compared with Milwaukee’s 2.3.

Arguably, though, the best denominator is not raw population, but some measure of police activity. For instance, as between two cities of the exact same size, we might expect more shootings in the city with more arrests; after all, each such police-citizen interaction carries with it a certain risk of violence.

Based on Bureau of Justice Statistics data from 2012, the most recent year for which numbers are available, the Philadelphia Police Department made 87,050 arrests, while the Milwaukee Police Department made 51,274, giving the MPD a higher per capita arrest rate than the PPD. Combining these numbers with OIS numbers from 2012, the PPD had 6.7 OISs for every 10,000 arrests, while the MPD had only 1.8.

But perhaps arrests are not the most appropriate basis for comparison. The vast majority of arrests are for low-level nonviolent offenses. Presumably, arrests for violent crime provide a better measure of the dangerousness of the work done by a police department. And, in this regard, the PPD does look rather different than the MPD; the number of violent arrests in Philadelphia (8,482) was more than twice as high as the number in Milwaukee (2,973).

Even correcting for this, however, the OIS gap remains. The PPD had 6.8 OISs per 1000 violent arrests, while the MPD had only 3.0.

These numbers are troubling on their face, but they do not necessarily demonstrate that the PPD has a greater problem with unjustified shootings than other police departments. One would need a careful examination of the facts in each OIS in order to judge whether the officer pulling the trigger acted improperly.

Concerns are heightened, though, by the 15% of PPD shootings in which the subject was unarmed. Here again, we can make a comparison with the MPD; in all of the Milwaukee incidents in 2013, the subject was armed. Indeed, in 79% of the cases, the subject carried a firearm, as compared to only 56% of the Philadelphia shootings. This provides further support for the impression that the Milwaukee officers may be doing a better job of assessing threats accurately and responding proportionately.

In at least one area, though, the OIS numbers in Philadelphia and Milwaukee were quite similar: when officers in either city opened fire, chances were good that someone would be hurt. In the PPD shootings, 23% of the subjects were killed and an additional 47% were injured. The MPD numbers from 2013 were quite similar at 29% killed and 50% injured.

Cross posted at Life Sentences.