All podcasts / People I (Mostly) Admire / Summary

Arne Duncan Says All Kids Deserve a Chance — and Criminals Deserve a Second One (Update)

2025-07-26 - source - Read full transcript
Steve Levitt (host)Arne Duncan

Key insights

Talent is evenly distributed across neighborhoods; opportunity is not.
Duncan watched kids in his mother's afterschool program who were as smart, talented, and resilient as his classmates at an elite school, living a mile and a half away but with wildly fewer chances. He frames his entire career as chasing this gap rather than any deficit in the kids themselves.
equity-of-opportunity
Giving one class of kids full opportunity and funded college access flipped their outcomes dramatically versus the class just one year ahead.
The I Have a Dream cohort Duncan ran had a 91 percent high school graduation rate, versus 66 percent for the class one year older in the same school, neighborhood, and socioeconomic conditions. The only difference was six years of intensive, funded support.
equity-of-opportunity
Equity, not equality, should be the goal of school design.
Duncan argues schools should give each child what they specifically need - more meals, more time, better teachers, eyeglasses, extracurriculars - rather than identical treatment. He'd also flip the constant/variable relationship of schooling: keep learning outcomes constant and let time in school vary by student need.
equity-of-opportunity
Publicly owning a scandal, rather than covering it up, produced a better outcome than any political strategy could have.
When Levitt brought Duncan statistical evidence that roughly 5 percent of C.P.S. teachers were cheating on standardized tests, Duncan's reaction was to call it a "Godsend," grant full access to retest classrooms, and hold a press conference framing himself as fighting cheating. The resulting headline was "Arne Duncan Puts His Foot Down Against Cheating" instead of a scandal story, and nearly all implicated teachers resigned without contesting it.
data-driven-accountability
Cheating on standardized tests directly harms kids by denying them services they need.
Duncan's argument for zero tolerance on cheating isn't reputational - when teachers mark below-grade-level students as above grade level, those students lose access to interventions like summer school, setting them up for failure later without the support that could have closed the gap.
data-driven-accountability
C.R.E.D.'s intensive violence-intervention model produces a roughly 19-to-1 return on investment compared to the cost of homicide and incarceration.
The program costs about $33,000 per person per year (variable by case), against a $1.4 million average cost per homicide to the city and about $60,000 per year for a Cook County Jail bed. A pro-bono bank consulting study estimated the 19-to-1 R.O.I.; C.R.E.D.-specific data shows about a 58 percent reduction in victimization among participants, studied by Northwestern.
trauma-and-violence-intervention
Reducing gun violence requires working directly with shooters, not just victims or bystanders.
Duncan describes an epiphany that came from years of Chicago conversations: violence reduction only works if you intervene in the crisis space itself, using credible messengers (often formerly incarcerated people) with direct relationships to conflicted groups, rather than programs that avoid engaging the highest-risk individuals.
trauma-and-violence-intervention
The scale of gun violence in his C.P.S. tenure was a defining, unrelenting cost that dwarfed every operational challenge.
Duncan says a child was killed by gun violence roughly every two weeks during his seven-and-a-half years running C.P.S., and that this got harder emotionally over time, not easier - unlike budget, labor, or academic challenges, which he describes as wildly easier by comparison.
trauma-and-violence-intervention
Physical, social, and emotional stability is the foundation the rest of academic achievement rests on.
Duncan traces this directly to his mother's afterschool program, where she fed kids before anything else because "it's hard to concentrate if your stomach's grumbling." He argues rigor (A.P. classes, college prep) is unstable without first addressing bullying, fear, hunger, and trauma.
trauma-and-violence-intervention
Everyone Duncan has worked with through C.R.E.D., including people who have shot others, is redeemable - the difference between who becomes a shooter and who doesn't is largely circumstance, not character.
Duncan contrasts his own upbringing in a household full of books (both parents educators, all three siblings went into education) with a program participant whose earliest memory was his father shooting heroin and who grew up in a household full of guns instead of toys. Duncan explicitly rejects framing outcomes as making one person better or worse than another.
redemption-and-second-chances
A man who killed Duncan's childhood friend now works alongside him leading C.R.E.D.'s alumni program, illustrating what functional forgiveness looks like in practice.
Billy Moore killed Benji Wilson (the nation's top high school basketball recruit and Duncan's friend) at age 16, served 20 years, and was later introduced to Duncan on a peace march. Moore now leads C.R.E.D.'s alumni work; Duncan says he'd "take a bullet for him." After Moore's own son was later murdered, Moore said that if his son's killer entered the program, he would mentor him - because he cannot ask for forgiveness he isn't willing to give.
redemption-and-second-chances
Moral clarity (hero vs. villain, good guy vs. bad guy) breaks down inside the actual work of violence intervention.
Duncan describes the world of street outreach and gang conflict as "the grayest world I've ever been in," where questions like whether the street worker or the policeman is the hero don't have clean answers, and progress requires tolerating that discomfort rather than resolving it into simple categories.
redemption-and-second-chances

Books referenced

Media referenced

Companies

Techniques and frameworks

Summary

This is an encore presentation of Levitt's 2021 conversation with Arne Duncan, a friend, former next-door neighbor, and one of the few guests Levitt introduces as unreservedly admired. The frame for the whole conversation is Duncan's memoir line "Education runs on lies," which becomes a lens for a career built on refusing comfortable dishonesty - about whether the U.S. actually values education, teachers, or children's safety despite what people say publicly.

The interview traces Duncan's arc from growing up alongside kids in his mother's under-resourced but volunteer-run afterschool program on Chicago's South Side, through running the I Have a Dream program (where an entire cohort of kids given full opportunity and funded college access saw a 91 percent graduation rate versus 66 percent for the class just a year ahead of them), to becoming CEO of Chicago Public Schools at 36 and then U.S. Secretary of Education under Obama at 44. A recurring thread is Duncan's approach to bad news: when Levitt brought him statistical evidence of widespread teacher cheating on standardized tests early in his C.P.S. tenure, Duncan treated it as a gift rather than a threat, granted full access to retest classrooms, and got ahead of the story publicly - turning what could have been a career-ending scandal into a credibility win, because he understood that cheating directly harmed kids by denying them services they needed.

The second half of the conversation moves to Duncan's current work running Chicago C.R.E.D., a violence-intervention organization he co-founded with Laurene Powell Jobs after leaving Washington. He lays out its five-pillar model - street outreach by credible messengers, life coaching (often from formerly incarcerated staff), an intensive clinical/trauma team, an education team, and job placement - and makes an unusually blunt economic case for it: at roughly $33,000 per participant per year against a $1.4 million average cost per homicide, outside analysis pegs the return at 19-to-1. He's equally direct about the human cost underlying the work, noting that during his seven and a half years running C.P.S. a child was killed by gun violence roughly every two weeks, and that this got harder to sit with over time, not easier.

The most striking material centers on redemption. Duncan's childhood friend Benji Wilson, the top-ranked high school basketball player in the country, was shot and killed in 1984; his killer, Billy Moore, served 20 years and today works alongside Duncan leading C.R.E.D.'s alumni program. Duncan describes trusting Moore completely, and recounts that after Moore's own son was later murdered, Moore said he would mentor his son's killer if that person entered the program - because he can't ask for forgiveness he isn't willing to extend. Duncan repeatedly resists simple moral categories in this work, describing it as "the grayest world I've ever been in," where circumstance rather than character often explains who ends up shooting and who ends up an educator.

The episode closes on a lighter note about Duncan's basketball obsession (a national three-on-three championship win at nearly 50, and Levitt's account of watching him make dozens of consecutive shots from his backyard court), followed by a short 2025 update in which Duncan reports Chicago C.R.E.D. is having its best year yet, with a real chance of the city dropping below 400 homicides for the first time since 1965.

Notable Quotes

"Talent is evenly distributed. Opportunity's not." - Arne Duncan

"We as adults, we as educators, we as parents, we as leaders, we failed to keep our kids safe." - Arne Duncan

"Hurt people hurt people. But the converse is also true, that healed people help to heal people, help to heal their families, help to heal their community." - Arne Duncan

"This is the grayest world I've ever been in, Steve. Who's a hero? Who's a villain? Who's the good guy? Who's the bad guy?" - Arne Duncan

"He can't ask for forgiveness and redemption if he can't give it." - Arne Duncan, describing Billy Moore