Tuesday, September 29, 2020
Home Video How Do Algorithms Predict Criminal Behavior?

How Do Algorithms Predict Criminal Behavior?

Host Myles Bess takes a deeper look at how algorithms are being used in the criminal justice system — and whether not computer programs make better predictions about criminal behavior than humans.

SUBSCRIBE to Above the Noise: [https://www.youtube.com/abovethenoise?sub_confirmation=1]

ABOVE THE NOISE is a show that cuts through the hype and takes a deeper look at the research behind controversial and trending topics in the news. Hosted by Myles Bess and Shirin Ghaffary.

*NEW VIDEOS EVERY OTHER WEDNESDAY*

It’s no big secret that the United States has a prison problem. We lock up people at higher rates than any other nation, and there are huge racial disparities in who we lock up. According to a study from The Sentencing Project, in state prisons, African Americans are incarcerated 5 times more than whites. There are lots of reasons for why we may see this racial disparities, including
law enforcement practices, crime rates, and punitive sentencing policies. Keeping so many people in prison is really expensive– it costs about $80 billion dollars a year– and it contributes to racial inequalities in America. As a result, there’s a big push among both Democrats and Republicans to reform our prison system. And one popular strategy many people advocate for as part of this reform effort are risk assessment tools. The tools use data to predict whether a person will commit a future crime. This video explores how these tools work and some of the controversy surrounding their use.

What do you think does a better job of predicting someone’s risk of committing a future crime — data-driven algorithms or human judgement?

Check out some of these tools:

Sample of COMPAS Tool (Obtained by ProPublica):

http://www.northpointeinc.com/downloads/research/DCJS_OPCA_COMPAS_Probation_Validity.pdf (Appendix A: page 32)

Indiana Risk Assessment System Tools: http://www.pretrial.org/download/risk-assessment/Indiana%20Risk%20Assessment%20System%20(April%202010).pdf

Missouri Tool:
http://www.mosac.mo.gov/page.jsp?id=45500

Virginia: (Appendix D, page 132)
http://www.vcsc.virginia.gov/risk_off_rpt.pdf

In-depth article on this topic:
https://ww2.kqed.org/lowdown/2017/06/01/can-algorithms-predict-criminal-behavior/

Sources:
Risk Assessment in Criminal Sentencing https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2662082#

Risk Assessment Instruments Validated and Implemented in Correctional Settings in the United States: An Empirical Guide
https://csgjusticecenter.org/wp-content/uploads/2014/07/Risk-Assessment-Instruments-Validated-and-Implemented-in-Correctional-Settings-in-the-United-States.pdf

Risk Needs Assessment in the Criminal Justice System
https://fas.org/sgp/crs/misc/R44087.pdf

Bureau of Justice Statistics
https://www.bjs.gov/index.cfm?ty=pbdetail&iid=5177

Incarceration gap widens between whites and blacks

Incarceration gap widens between whites and blacks

Marijuana Arrests by the Numbers
https://www.aclu.org/gallery/marijuana-arrests-numbers

Mass Incarceration: The Whole Pie
https://www.prisonpolicy.org/reports/pie2016.html

Follow KQED:
KQED: http://www.kqed.org/
Facebook: https://www.facebook.com/KQED/
Twitter: https://twitter.com/KQED?lang=en

Teachers follow KQED Learning
KQED Learning: https://ww2.kqed.org/learning/
Facebook: https://www.facebook.com/KQEDLearning/
Twitter: https://twitter.com/KQEDedspace?lang=en

About KQED
KQED, an NPR and PBS affiliate in San Francisco, CA, serves Northern California and beyond with a public-supported alternative to commercial TV, Radio and web media.

Funding for Above the Noise is provided in part by S.D. Bechtel, Jr. Foundation, David Bulfer and Kelly Pope, Horace W. Goldsmith Foundation, The Dirk and Charlene Kabcenell Foundation, The Koret Foundation, Gordon and Betty Moore Foundation, Smart Family Foundation, The Vadasz Family Foundation and the members of KQED.