Thread by @samswey

Here’s what we know about police violence, based on the data that’s been collected over the past several years. A thread (1/x).

Despite the US govt’s unwillingness to track police violence, we now have comprehensive data on killings by police via WaPo, mappingpoliceviolence.org & fatalencounters.org. While we don’t have enough data on non-fatal violence, we can learn a lot from the data we *do* have...

First, we know that police have killed people at a similar rate each year as far back as we have good data (~2013). It’s not getting worse but also not getting better. But this conceals important dynamics at the local level that give us clues as to how to stop police violence...

While police violence hasn’t reduced on a nationwide level since 2013, it *has* reduced in cities. It’s also *increasing* in rural/suburban areas, offsetting those reductions. So it looks like some progress is being made by (some) cities. So which cities and how did they do it?

When we break down the data by city, there are massive differences in police violence rates per population. St Louis, OKC and Orlando consistently have the highest rates of police violence. Killings are 3x less frequent in SF or Philly, 4x less in Detroit. mappingpoliceviolence.org/cities

Who’s impacted by this violence? Black and brown people. This chart shows the race of unarmed people killed by police in major cities from 2013-19. Almost everyone was Black or Latinx. Black people are 3x more likely than whites to be killed by police & more likely to be unarmed.

So which of these departments have made the most progress in reducing police violence? When we look data on all police shootings (both fatal and non-fatal) from 2013-19 and examine the change over time, here’s which major city departments have had the biggest reductions:

Among places that reduced police shootings, SF, Philly, Chicago, Denver, San Jose, LA, Phoenix and Baltimore made their use of force policies more restrictive during this period. Many had DOJ investigations. & many reformed enforcement of drugs/low level crime, reducing arrests.

This is consistent with previous research finding departments with more restrictive use of force policies (useofforceproject.org) and DOJ interventions predict reduced police violence. Despite this, the Trump admin has stopped these DOJ investigations. vice.com/en_us/article/…

Importantly, cities that reduced police shootings did so regardless of levels of crime. Crime rates (including violent crime rates) actually went up slightly in major cities during this period - but police shootings declined consistently with a 41% reduction during this time.

There are other factors researchers have identified as impacting police violence rates. More racial segregation predicts more police violence (bu.edu/sph/2018/02/05…). So does receiving military weapons from the feds (journals.sagepub.com/doi/full/10.11…).

On the other hand, having more Black officers is NOT associated with reducing police violence in most studies (some studies suggest more women on the force can help tho). Implicit bias training also hasn’t been shown to make a real difference. But a factor that DOES matter is...

THE ACCOUNTABILITY SYSTEM. Just as some depts have higher violence rates, some officers have higher violence rates too. In Columbus, 6% of officers commit 50% of all use of force. When the system fails to hold them accountable it causes MAJOR problems... theappeal.org/just-6-percent…

See now we can track how police misconduct spreads. And officers in close proximity to officers who have records of misconduct end up being 4x more likely to use force & 5x more likely to shoot someone. We can *contact trace* police violence like a virus. theintercept.com/2018/08/16/chi…

Accountability systems work if they intervene & remove officers BEFORE misconduct spreads/escalates. But only 7% of excessive force complaints are upheld/result in discipline and 1% of killings result in prosecution. So these officers remain on the force. stories.usatodaynetwork.com/data_stories/p…

Here too, data helps us understand how to increase accountability and decrease police violence. Misconduct complaints are more likely to be upheld when depts have body cams (video=key) but much less likely to be upheld in depts with police union contracts. gothamgazette.com/city/9167-init…

This is because police union contracts are where the accountability system is set. Most contracts purge misconduct records, restrict misconduct investigations and help officers overturn discipline and get reinstated after being fired for misconduct. checkthepolice.org

In fact, police depts with union contracts that grant more procedural protections to officers accused of misconduct are:
1) more likely to be reported for misconduct
2) less likely to uphold misconduct complaints
3) more likely to kill unarmed people papers.ssrn.com/sol3/papers.cf…

So the contracts matter, and they need to be re-negotiated to increase levels of accountability and to reduce police violence. But what about the exceedingly low rate of officers being prosecuted? Well, we know more about this topic now than we did 5 years ago. For example...

While only 1% of all killings by police result in officers being charged with a crime, this 1% isn’t evenly distributed. There are a handful of places that routinely charge officers while almost every other place fails to do so. And those places tend to have Black prosecutors.

https://twitter.com/samswey/status/690599927588585473

We still don’t know what tactics these prosecutors are using to secure indictments in these cases, but electing prosecutors (esp. Black prosecutors) who genuinely care about holding police accountable is important. This + video evidence are factors that can make the difference.

So we now know:
1) which depts have the worst police violence rates & which are reducing it
2) a set of interventions/factors associated with reducing this violence and increasing accountability

By continuing to study & scale up what works we can end police violence nationwide.