Technology won’t save us from the police – Round 2

By Dianne Post

In June 2019, I wrote a blog saying that technology won’t save us from the police. I focused on suggestions from the 1960s and what developed i.e. tazers, body cams, sting rays, facial recognition.  None have lived up to their hype.

As Braunstein and Erickson point out in an article in the South Dakota Law Review the use of technology and scientific investigation has a long history in law enforcement.  Fingerprints were the newest technology in 1902. By 1990, we were looking at hair strands, bite-marks (now discredited), tire tracks, bullet marks and other wonders you can see on CSI.   Sometimes new science trumped old science and someone found guilty under old science was declared innocent under new science. Today we focus on body cameras. In November 2014, 41 of the 100 largest cities had some or all officers wearing body cameras with 400 deployed.  Two years later, that number was 2,800. By 2016, nearly half of law enforcement agencies had them.

But in January 2020, Daniel Lawrence, a researcher at the Urban Institute in D.C. is finding that just buying them is not enough.  It’s how you use them. When are they turned on, who reviews them, do officers look at them before writing their reports, who gets to see them – all of these factors help determine how useful they are. 

In March, Cynthia Lum and Christopher Koper from the George Mason University’s Center for Evidence-Based Crime Policy produced a meta-study of 70 studies published through June 2018. They found that the cameras have not had statistically significant effects on either officer or citizens behaviors or citizens’ views of the police.   Five studies showed officers used less force but eight others showed no difference.

However, the study did find that the police are using the cameras for evidence collection and protection. The cameras did not improve transparency or trust but in fact did the opposite.  The public had restricted access to the video so didn’t trust what was on it. The use of cameras did not improve police performance, accountability, or relationships with citizens. The cameras did reduce the number of complaints against officers – so it benefited them more than the public they allegedly serve.

The study also showed police use of body-worn cameras did not lead to de-policing, also known as a “Ferguson effect,” in which officers pull back from being productive in their duties. Cameras did not appear to discourage police contacts or officer-initiated activities or have any noticeable effect on arrest rates. That hyperventilating by law enforcement has zero substance.

Artificial intelligence is yet another technology deployed in court rooms and by judges to help determine bail, parole sentencing, and risk assessment of a defendant.  As ProPublica and the Washington Post report in January 2020, it is plagued by the same problems as human intelligence – bias. Garbage in – garbage out. I argued this time and time again when I served on the Supreme Court Task Force for Fair Justice – to no avail.  A study done by Abou-Elyounes found that among black defendants 42% were falsely labeled as future criminals versus 22% for white defendants. COMPAS systems likewise did the opposite and flagged white defendants as low risk when they were not. How do you compare the risk of five trespassing or burglary arrests for an African American defendant versus one murder or rape for a white defendant?  The computer cannot do that. But apparently, neither can we.

As the Kerner Commission said in 1967, we need to drastically rethink about crime in society.  Not one of their suggestions has been brought to fruition unless you count the tazer – a non-lethal weapon.  We don’t need another study. We don’t need another commission. We need to do what needed to be done back in 1967 and never was.  Phoenix made a tiny step forward in March 2020 with the citizens review board for the police but we need far broader and deeper ideas to truly integrate everyone into society.