AI summary by TwinMind
146. Is There a Fair Way to Divide Us? | People I (Mostly) Admire
Freakonomics Radio Network
1h 5m 32s
Introduction and Background
- Moon Duchin: Cornell mathematics professor, specialist in abstract geometry
- Discovered practical application of geometric research to political gerrymandering
- Works with state commissions and courts to transform redistricting processes
- Key insight: "we don't have a baseline we don't know what normal districting looks like"
- Math provides "better and better methods for sampling from that huge Unthinkable Wilderness of plans"
Path to Gerrymandering Research
- Teaching undergraduate "mathematics of social choice and voting" course in 2016
- Course for non-majors fulfilling math requirement
- "Sneaky deep" despite appearing easy
- Wanted to connect theory to Trump primary season events
- Discovered redistricting literature wasn't "well solved" as expected
- Gradual realization that geometric toolkit could help with redistricting problems
- Geometry relevant due to:
- People unevenly distributed over territory
- Political subgroups with varying concentration patterns
Gerrymandering Definition and History
- Definition: "abusive redistricting" - using line-drawing power to advance agenda
- Broadest consensus: controlling district lines to get more seats for one type of representative
- Etymology: Elbridge Gerry (pronounced "Gary"), Massachusetts governor, 1810 census
- Approved state senate map with controversial North Shore district
- Political cartoon showed district as "Gerry's salamander" with wings, teeth, claws
- Became "gerrymander" (though salamanders don't have wings)
- Historical context: 1810 district looks "just fine" by modern standards
- Whole towns, no precinct cuts
- Criticism focused on "skipping over some population to grab other population"
Redistricting Rules and Constraints
- Only two universal US rules:
- Population balance (equal-sized districts)
- Racial fairness (Voting Rights Act 1965, 14th Amendment equal protection)
- Additional four common rules ("big six"):
- Contiguity: districts must be connected pieces
- Compactness: districts should have "nice shapes"
- Respect for political boundaries: keep counties, cities, towns whole
- Communities of interest: shared historical/cultural/economic interests
- Notable absence: no strict partisan fairness requirements
- Only Ohio has constitutional proportionality requirement (40% votes → 40% seats)
- Ignored by partisan actors in recent cycle despite constitutional mandate
Mathematical Complexity of Redistricting
- Justice Alito's underestimate: "maybe 100, maybe 25, probably thousands" of possible plans
- Actual scale: "probably Google (10^100) with 100 zeros"
- Problem characteristics:
- Enormous possibility space (more than particles in galaxy)
- No good recursive structure
- Can't break big problems into easier smaller ones
- 1960s prediction that computers would search all plans proved false
- Need for representative sampling methods like polling
Markov Chain Monte Carlo Approach
- Duchin's research group developed new sampling method
- Markov chain concept: start with districting plan, make incremental changes
- Traditional approach: move one precinct at a time (inefficient, like taking one card off deck top)
- Duchin's innovation: "riffle shuffle" equivalent
- Fuse two neighboring districts
- Redraw boundary in totally new way
- Fundamental reconfiguration vs. small particle changes
- Mathematical control over long-term behavior and probability distributions
Practical Applications
Outlier Analysis
- Generate ensemble of random redistricting plans
- Compare proposed plan to distribution
- Extreme outliers suggest gerrymandering intent
- Courts find this "pretty persuasive"
Pennsylvania Case Study
- Republican legislature plan looked typical of "blind" (neutral) plans
- Geographic reality: Philadelphia corner location, Pittsburgh opposite side
- Blind process favored Republicans due to Democratic concentration
- Distinction: "blind isn't always fair"
- Argument: free and fair elections clause should require better than blind process
- "Drunken monkey can eventually type Hamlet" - why outsource redistricting?
- Amanda Holt (piano teacher) drew plans without partisan data
- Plans fell in middle of random distribution
- Legislature strategically adopted Holt's plan as PR move
Visual vs. Mathematical Analysis
- Traditional focus on "crazy looking districts" (goofy kicking Donald Duck)
- Pennsylvania's 7th district: connected only by hospital width, seafood restaurant width
- Duchin's key insight: "reasonable looking" districts can still be completely unfair
- Shape constraints less limiting than initially expected
- Legislators under pressure for nice shapes can gerrymander just as effectively
Segregation Paradox
- Conventional wisdom: Democratic disadvantage due to urban clustering ("huddle for warmth")
- Duchin challenges confusion between density and inefficiency
- Research finding: clustered minority populations get better representation
- Example: 41% minority uniformly spread = 41% in every district = lose everything
- Massachusetts case: most uniform political geography in US
- Republicans consistently ~30% statewide
- Same percentage at all geographic levels
- Result: all-Democratic delegation regardless of line-drawing
- 1992: last Republican elected to Congress from Massachusetts
Proportional Representation Discussion
- Most world countries use party list voting (~50% of countries)
- Germany: mixed member system (single districts + proportional adjustment)
- US geographic representation benefits:
- Direct voting for people vs. parties
- Local interests representation (shrimpers, loggers, etc.)
- Federalist Papers: "competing local jealousies"
- Downside of party list: party control over diversity
- Many countries need "gentleman's agreements" for diverse representation
Voting System Reforms
- Multi-member districts with ranked choice voting
- Alaska's "final four" system example
- Portland Oregon reform: 4 districts electing 3 members each = 12 city council
- Multi-member systems provide "big head start" toward proportional representation
- Duchin's modeling shows effectiveness for minority representation
Business Model Discussion
- Levitt's suggestion: sliding scale consulting
- $500K for fair plan
- $5M for skewed but reasonable-looking plan
- Duchin's actual involvement:
- Michigan independent commission (communities of interest testimony)
- Arizona commission (competitiveness guidelines)
- Preference for "better processes" over just "catching wrongdoing"
Freedom to Vote Act Contribution
- Duchin contributed to partisan gerrymandering language
- Test: proportional representation 3 out of 4 recent elections
- Research with Gabe Schonbach: blind plans often pass proportionality test
- Even gerrymandering hotspots (North Carolina, Wisconsin, Pennsylvania) had adequate supply of passing plans
- Dave Wasserman's 538 Atlas of Redistricting: hand-drawn plans passed tests
- Evidence: proportional past performance predicts proportional future performance
Rush Limbaugh Stories
Duchin's Story (2003)
- University of Chicago grad student
- Helped organize gender-neutral bathrooms
- Administration cooperative: "sure, how many and where"
- Chicago Sun-Times quote: "just more choices for people"
- Limbaugh's week back from prescription drug rehab = huge audience
- Limbaugh's suggestions: "buckets for idiots like Moon Duchin"
- Closing: "good night Moon" - "chef's kiss just beautiful"
Levitt's Story (1995)
- Paper on prison population impact on crime rates
- Magazine headline: "Prisons Reduce Crime"
- Limbaugh: "Real Genius Economist at Harvard" discovered "brilliant new insight"
- Mock follow-up paper: "Nighttime Causes Darkness"
- Both wear criticism as "badge of honor"
Academic Diversity Discussion
- Levitt's concern: talented people have many options
- Question: are we doing harm by recruiting to academia vs. other high-impact fields?
- Duchin's response:
- Fewer undergrads want academic careers (COVID, tech boom effects)
- Would be "patronizing" to redirect people's stated preferences
- Need fair evaluation processes, not just better recruitment
- Unhealthy dynamics in hiring committees between "diversity minded" and "what about white guys" factions
Math and Social Justice Integration
- Harvard double major: math and women's studies
- Interest in Science, Technology, and Society (STS)
- Philosophy class disappointment: "brains and vats" thought experiments
- Women's studies as way to "think about what science does while talking about people"
- Redistricting work brings "interest in people power and science right into mainstream"
- Considers herself "extraordinarily lucky" for this convergence
Future of Academia
- Levitt's experience: real world "poisoned" him to academics
- Duchin's response: still excited by campus intellectual life
- Loves "different kinds of intellectual projects that people bring into collision"
- Finds academic environment "tremendously exciting"
- Appreciates both research and education aspects of professor role
