September 14, 2015

Freakonomics: Preventing Crimes for Pennies on the Dollar

7/10 microphones, Listen to the Podcast

Problem: Violence in Chicago

• Chicago has 3x higher homicide vs. NYC
• Chicago has very active/violent gangs vs. NY that does not
  • More than 100,000 gang members vs. 20,000 gang members

The Usual Approach: Lots of cash to mentor/advise and try reduce violence

• A massive project was launched in 2009 with mentoring, etc. that cost $27k per student
  • Levitt was asked to model how to best the resources at most at risk students
    • Predictive modelling done looking at past shootings to predict those at highest risk
      • Best predictor: Male
      • Other predictors: Been in Juvenile Detention, HS Dropout, African American (but not once you control for other factors)
    • Able to find a group that was at 20X risk vs. average student (HOWEVER, the average risk was so low that very shootings are avoided even by selected these students
  • Also evaluated efficacy of the program: Results so almost no effect…
  • Confirms conventional wisdom is that there is not much you can do to help once they are already teenagers (too late to rewire kids). Is this true or an excuse?

New Approach: Cognitive Behavioral Therapy (CBT) Type Strategy

• (CBT) Cognitive Behavioral Therapy: Goal to change behaviors without dwelling on underlying reasons that is done with talk therapy
  • Think differently about behaviors that have become almost automatic and be able to respond differently and more thoughtfully to stress
  • Has been shown to work in a variety of ways, but hasn’t been focused on dropout and crime
• Source theory: Daniel Kahneman on System 1 and System 2 processing, Nobel prize winner in Behavioral Economics (Thinking Fast and Slow)
  • System 1 is automatic/subconscious
  • System 2 is when we consciously think and make decisions
  • Goal to shift certain behaviors that are maladaptive from System 1 to System 2 to improve results and decrease violence
• Youth Guidance: Non-profit trying to help in Chicago, led by Tony Divatorio
  • Evaluated by Crime Lab of Chicago as a randomized and controlled study (random and control by lottery)
• Goal is to teach kids to slow down thinking and reduce automatic behavior
  • Kids in Chicago might have learned automatic responses that might be more confrontational when challenged (this is probably reasonable when being robbed, but not appropriate in school)
• Techniques:
  • Fist Exercise: 12 kids invited to sit in circle, told to find a partner, pick a letter per person
    • A’s make a fist
    • B has 30 sec. to open A’s fist
    • NO RULES, GO! = start beating each other up
    • Then switch and do it again = Chaos and Fun
    • Stop to talk about it… Nobody ever asked the partner to open first
      • Ask why they know it won’t work?
    • Lecture: A powerful man becomes an exceptional man when they know how to Ask for what they want
  • High School Day: Role Playing
    • Difference between Savage and Warrior Energy
      • Savage is a normal emotion
      • Warrior energy is taking Savage Energy and choosing to use it positively
      • Help learn and talk about what works, what is difficult and then they open up about the personal issues
    • DO NOT tell them have to behave
      • Dumb to tell them not to fight, just learn how to think a little more
    • Results: 2700 boys from 7-10^th grad (1/3 had been arrested before study)
      • 44% decline in violent crime arrests
      • 36% decline in other arrests
      • Increase in school engagement rates likely to increase graduation rates by 7-22% percent
      • Almost too good to be true… doubted that it was true
      • Repeated with a second study and found 30% decline in overall arrests
    • Is it just Youth Guidance/Tony, or can it work with other people?
    • 3^rd experiment without Youth Guidance involved at a Youth Detention Center (after arrest, before jail)
      • Results: 20% Decline
    • Costs: CBT is relatively low cost, just change what you’re doing for 30 mins a day and train people
      • $1,100 per person vs. $28 with previous program (very cost effective)
    • City of Chicago and Rahm Emannuel are supportive of the approach and studies that rigorously look at the results
      • Other cities should take note and be more willing to work with researchers to evaluate how they spend their money and do studies
    • Where the Effects Lasting?
      • Some short term evidence, first study the violent crime effect faded away in second year, but schooling engagement continued
      • Detention Center Study: Reduced return rates for 12-18 months
    • Even short term only benefits to Violence are hugely beneficial
      • Violence also peaks with age in teens, and fade when older so one key year is important
    • Next Week: CBT with Cash incentives in Liberia