- General Overview
- The Central Argument
- Libertarian paternalism: steer people toward better choices while forbidding nothing and coercing no one.
- Nudge defined: any choice-architecture feature that alters behavior without banning options or changing incentives.
- Humans not Econs: real people rely on fallible rules of thumb; textbook rational agents do not exist.
- Choice architecture is unavoidable: someone always arranges the options, so arrange them well.
- As judged by themselves: influence is legitimate only if choosers would endorse the outcome.
- Real third way: neither mandates nor laissez-faire, but freedom-preserving gentle guidance.
- How Humans Think
- Two systems: fast Automatic intuition and deliberate Reflective reasoning compete in every judgment.
- Predictable biases: anchoring, availability, and representativeness breed systematic, confident errors.
- Misread randomness: people see patterns and "hot hands" where only chance operates.
- Optimism and loss aversion: unrealistic confidence and a two-to-one sting from losses distort choice.
- Status quo bias: inertia keeps people in current arrangements, giving defaults enormous power.
- Framing: identical facts framed as losses or gains produce opposite reactions, even in experts.
- Temptation and the Herd
- Planner versus Doer: a far-sighted self battles a myopic self for control of behavior.
- Hot-cold gap: cold states underestimate how arousal will distort future choices.
- Commitment devices: Ulysses contracts, bets, and penalties let the Planner bind the Doer.
- Mindless choosing: people eat and spend whatever is placed in front of them.
- Social contagion: obesity, smoking, and fads spread through networks; conformity is powerful.
- Pluralistic ignorance: groups cling to norms nobody privately approves until a small shock dislodges them.
- When Nudges Are Needed
- Hard, rare, feedback-poor: nudges help most where decisions are difficult, infrequent, and slow to teach.
- Weak feedback: choices with delayed or hidden consequences prevent genuine learning.
- Translation problem: options that cannot be mapped to imaginable experience overwhelm choosers.
- Markets mixed: competition usually protects consumers, but firms also profit from human frailties.
- The Choice Architect's Toolkit
- Defaults: sensible opt-out settings outperform required choice for the absent-minded.
- Expect error: forgiving design, forcing functions, and postcompletion checks prevent mistakes.
- Give feedback: timely, clear signals let people correct course; warnings alone are ignored.
- Structure complex choices: tiered menus, collaborative filtering, and smart ordering tame overload.
- RECAP: Record, Evaluate, Compare Alternative Prices makes hidden fees transparent.
- NUDGES mnemonic: incentives, mappings, defaults, feedback, error, complex choice.
- Money: Saving and Investing
- Save More Tomorrow: automatic increases tied to pay raises turn inertia and loss aversion into wealth.
- Automatic enrollment: switching the default to opt-out pushes participation from 20 percent to over 90.
- Naïve investing: investors chase past returns, diversify poorly, and overhold employer stock.
- Credit markets: complex mortgages and loans exploit unsophisticated borrowers; disclosure must be clear.
- Social Security smorgasbord: too many funds paralyze choosers; a good default serves most.
- Health, Environment, and Freedom
- Medicare Part D: a maze of plans traps seniors; intelligent assignment shows a better default.
- Organ donation: changing the default from opt-in to presumed consent sharply raises donors.
- Saving the planet: better incentives, feedback, and disclosure curb pollution and energy use.
- School choice: simplified information and simpler forms let disadvantaged families choose well.
- Privatizing marriage: separating church blessing from state license expands freedom for all.
- Objections and Conclusion
- A dozen nudges: the same tools apply across charity, self-control, and everyday decisions.
- Slippery slope answered: opt-out safeguards and case-by-case judgment contain the risk of coercion.
- Publicity principle: nudges must be transparent and defensible, never covert manipulation.
- Asymmetric paternalism: help the least sophisticated while imposing near-zero costs on the savvy.
- Government populated by Humans: officials err too, so choice architecture beats prohibition.
- The Central Argument
- Deep Dive
- Nudge
- Humans and Econs
- Two species: real "Humans" rely on rules of thumb; the perfectly rational "Econ" of theory does not exist.
- Biases and blunders: heuristics that usually work breed predictable, systematic errors in judgment.
- Resisting temptation: present bias defeats willpower, so commitment devices outperform good intentions.
- Following the herd: conformity and social pressure quietly determine choices we think are our own.
- When a nudge is needed: payoffs are largest where choices are hard, rare, and slow to give feedback.
- Choice architecture is unavoidable: someone always arranges the menu, so arrange it well.
- Money
- Save More Tomorrow: tying automatic savings increases to pay raises turns inertia into retirement wealth.
- Naïve investing: investors chase past returns and diversify poorly; sensible defaults perform better.
- Credit markets: borrowers misread compounding and teaser rates, so disclosure must be clear and timely.
- Privatizing Social Security: a smorgasbord of funds overwhelms; one well-chosen default serves most people.
- Health
- Prescription drug plans: a mountain of options traps the elderly; good defaults and framing rescue them.
- Organ donations: opt-in consent wastes lives; presumed consent and prompted choice raise supply.
- Saving the planet: energy use responds to feedback, incentives, and better default settings.
- Freedom
- School choice: simplified information and forms let disadvantaged families actually choose well.
- Medical liability: legal and insurance rules can be redesigned to reward safer, more honest care.
- Privatizing marriage: letting couples write their own terms expands freedom rather than threatening it.
- Extensions and Objections
- A dozen nudges: the same tools apply across everyday settings, from finance to health to environment.
- Objections answered: worries about paternalism, manipulation, and slippery slopes get direct replies.
- Libertarian paternalism: steer people toward better outcomes while banning no option and coercing no one.
- The real third way: genuine reform preserves freedom of choice instead of choosing between mandates and neglect.
- Humans and Econs
- Acknowledgments
- Institutional Support
- University of Chicago: research funded by the Graduate School of Business and Law School.
- John Templeton Foundation: grant to the Center for Decision Research supported the book's research.
- Center for Decision Research: served as the grant's institutional home.
- Editorial and Production
- Sydelle Kramer: the agent whose advice guided the book throughout.
- Michael O'Malley: the editor whose suggestions shaped the manuscript.
- Dan Heaton: copy editor who cleaned up the writing with style and good humor.
- Kim Bartko: invaluable on the book's artwork and cover design.
- Research Team
- Two summers of research assistants: Balz, Dizard, Fronk, Johnson, Liu, Reynolds, Tokson, Wells.
- John Balz: singled out for double thanks after tolerating the authors for two summers.
- Vicki Drozd: kept everything running and made sure the research assistants got paid.
- Scholarly Colleagues
- Colleagues: supplied insights, hints, and nudges beyond the call of friendship and duty.
- Named thinkers: Kahneman, Gilbert, Gilovich, Karlan, Madrian, Mullainathan, Sunstein's peers and many more.
- Toughest readers: France Leclerc and Martha Nussbaum gave the authors' wisest advice.
- Everyday Debts
- Ellyn Ruddick-Sunstein: patience, good cheer, and both sense and amusement about behavioral economics.
- Noodles on 57th Street: fed the authors and hosted years of planning; they'll be back next week.
- Institutional Support
- Introduction
- Introduction · I
- The Cafeteria Experiment
- Carolyn's experiment: rearranging food displays shifted consumption of many items by up to 25 percent
- Context over preference: children, like adults, are greatly influenced by small changes in presentation
- No neutral option: any cafeteria must organize food somehow, so some influence is unavoidable
- Choice Architecture
- Choice architect: anyone responsible for organizing the context in which people make decisions
- Everything matters: small, apparently insignificant design details can have major behavioral impacts
- Schiphol fly: a urinal etching captured attention and cut spillage by 80 percent
- No neutral design: like buildings, every choice arrangement subtly shapes how people act
- Libertarian Paternalism
- Libertarian: people should be free to choose, and to opt out of arrangements they dislike
- Paternalistic: institutions may legitimately steer choices toward longer, healthier, better lives
- As judged by themselves: influence counts as helpful only if choosers would endorse the outcome
- Weak and nonintrusive: choices are not blocked, fenced off, or significantly burdened
- Nudges, Econs, and Humans
- Nudge defined: any choice-architecture feature altering behavior without forbidding options or changing incentives
- Easy to avoid: to count as a nudge, the intervention must be cheap and simple to resist
- Fruit at eye level: a nudge; banning junk food is a mandate, not a nudge
- Econs vs Humans: textbook agents reason flawlessly; real people predictably err
- Planning fallacy: systematic unrealistic optimism about how long projects will take
- Status quo bias: people stick with defaults—an inertia that choice architects can harness
- The Cafeteria Experiment
- Introduction · II
- Nudges and the Limits of Econ Thinking
- Nudge defined: any factor that alters Human behavior while being ignored by Econs
- Econs vs. Humans: Econs respond only to incentives; Humans also respond to framing and defaults
- Power of defaults: seemingly trivial menu changes raise savings, improve health care, supply organs
- Combined toolkit: incentives plus nudges improve lives while preserving freedom to choose
- The False Assumption
- Just Maximize Choices: a policy mantra whose only imagined alternative is a single government mandate
- False assumption: that almost all people, almost always, choose in their own best interest
- Chess novice: inferior choices reflect inexperience and can be improved with helpful hints
- Context determines quality: choices are good with experience, information, fast feedback; poor without
- Two Misconceptions
- First misconception: influence over choices cannot be avoided — some agent must set starting points
- Unintentional nudges: biweekly pay raises saving rates without any designer intending it
- Second misconception: paternalism always coerces — food order forces no diet on anyone
- Government distrust: real risks of mistake and overreach argue for nudges over mandates
- Choice Architecture in Action
- User-friendly design: iPod and iPhone let users do what they want with minimal effort
- Open enrollment: even administrators forget deadlines, so defaults must choose for the absent-minded
- Status quo vs. back to zero: insurance favors status quo; flexible spending accounts favor zero
- Design principles: send reminders and minimize costs imposed on those who space out
- Libertarian paternalism: neither left nor right, costs little, and crosses partisan divides
- Nudges and the Limits of Econ Thinking
- Introduction · I
- Part I humans and Econs
- Two Minds and Their Biases (1 Biases and Blunders · I)
- Two Systems of Thinking
- Automatic System: rapid, instinctive, gut-level; rooted in the brain's oldest parts
- Reflective System: deliberate and self-conscious; handles arithmetic and major life decisions
- They talk at once: Automatic cries "I'm going to die," Reflective answers "planes are safe"
- Practice relocates skill: athletes and true bilinguals run on trained Automatic intuition
- Teen drivers and voters: untrained Automatic plus slow Reflective makes both dangerous
- Biased Yet Confident Minds
- Shepard's tables: identical tabletops look different; seeing is not believing
- Errors are systematic: predictable mistakes, delivered with confidence, not random failure
- Bat and ball: the intuitive 10 cents is wrong; 5 cents is right
- Widgets and lily pads: 100 minutes and 24 days feel right; 5 and 47 are correct
- No check: even bright students accept first answers without consulting reflection
- Spock and Homer: Econs always reflect, Homers never do; design should protect the Homers
- Rules of Thumb and Three Heuristics
- Necessity of shortcuts: busy, complex lives demand quick judgments that usually work
- Tversky and Kahneman: identified anchoring, availability, and representativeness with their biases
- Two-system origin: heuristics emerge from the interplay of Automatic and Reflective thought
- Anchoring: Starting Points That Stick
- Insufficient adjustment: we start from a known number and adjust too little toward truth
- Milwaukee: Chicago residents guess high, Green Bay residents guess low; true answer 580,000
- Irrelevant anchors work: phone digits shift guesses about Attila the Hun by centuries
- Order shapes judgment: asking about dating first raises its correlation with happiness to .62
- Anchors as nudges: charity tiers, jury awards, and opening offers all exploit them
- Availability: What Comes to Mind Feels True
- Heuristic defined: risk is judged by how readily examples come to mind
- Vividness inflates: tornadoes overestimated, asthma underestimated twentyfold
- Recency drives fear: earthquake insurance spikes then decays as memories fade
- Policy distortion: governments fund feared risks rather than the likeliest ones
- Corrective nudge: reminding people of bad outcomes raises fear; good outcomes raise confidence
- Representativeness: Similarity Mistaken for Probability
- Similarity heuristic: probability judged by resemblance to a stereotype of the category
- Stereotypes sometimes right: tall men really are likelier to play professional basketball
- Linda problem: feminist bank teller seems likelier than bank teller — a logical impossibility
- Gould's homunculus: knowing the right answer does not silence the Automatic System
- Randomness misread: people detect meaningful patterns in coin tosses and chance
- Two Systems of Thinking
- Seeing Patterns, Clinging to Defaults (1 Biases and Blunders · II)
- Misperceiving Randomness
- Pattern-seeking: We build informal tests only after seeing the evidence, so randomness masquerades as design
- London bomb maps: Clustered V-1 strikes looked precisely aimed; statistical analysis proved the distribution random
- Manufactured tests: A horizontal grid shows a pattern, a diagonal one shows randomness — nothing dictates the right test
- Hot hand myth: Recent makes predict nothing; announcers' "he's on fire" calls had zero predictive power
- Cancer clusters: In a population of 300 million, some neighborhoods will show elevated rates by chance alone, yet each report triggers fear and investigation
- Optimism and Overconfidence
- Above-average effect: 90% of drivers and 94% of professors rate themselves better than average
- Unrealistic optimism: MBA students crowd the top deciles; less than 5% expect to finish below median
- High stakes ignored: Nearly all couples put divorce odds near zero; entrepreneurs rate their own success 90% versus 50% for peers
- Risk denial: Smokers, students, and the elderly underestimate personal risk of harms they know statistically
- Nudge potential: Reminding people of a bad event can puncture optimism and prompt preventive steps
- Loss Aversion
- Loss aversion: Losing something hurts roughly twice as much as gaining the same thing pleases
- Mug experiment: Owners demand about twice what buyers will pay — value depends on ownership, not the object
- Betting test: Most people need about a $200 potential win to accept a $100 loss
- Inertia: Reluctance to incur losses makes people refuse trades and stick with current holdings
- Mug versus candy bar: With equally valued goods, only one in ten agreed to switch
- Status Quo Bias
- Status quo bias: A general tendency to stick with the current situation, deeper than loss aversion alone
- Retirement paralysis: The median TIAA-CREF professor made zero allocation changes; some still listed their mothers as beneficiaries
- Free-subscription trap: Automatic renewal exploits inattention — Sunstein paid for unread magazines for a decade
- "Yeah, whatever" heuristic: Viewers stay on a channel though switching costs one thumb press
- Default power: Loss aversion plus mindless choosing gives designated defaults a large market share
- Implicit endorsement: Consumers read defaults as advice from the employer, government, or scheduler who set them
- Framing
- Framing effect: "90 of 100 alive" and "10 of 100 dead" are identical facts with opposite emotional pull
- Experts too: Doctors recommend the operation more when told survival rates than mortality rates
- Credit card lobby: Companies fought to make the cash price a "discount" rather than the card price a "surcharge"
- Loss-framed campaigns: "You lose $350 by not conserving energy" beats the equivalent savings message
- Mindless acceptance: Because people answer questions as posed, frames are powerful and must be chosen with care
- Nudge-Ability
- Bounded attention: Busy people rely on rules of thumb and cannot deliberate deeply about every choice
- Nudge-able: Even life's most important decisions shift with presentation in ways standard economics cannot predict
- Lake Shore Drive: Stripes spaced ever closer create the sensation of accelerating speed, prompting drivers to brake
- Gentle steering: A purely visual signal changed behavior without coercion
- Misperceiving Randomness
- Temptation, Mindless Choosing, Self-Control (2 Resisting Temptation · I)
- Dynamic Inconsistency and Temptation
- Dynamic inconsistency: people prefer A cold but later choose B when tempted.
- Cashew bowl: removing an option can make people better off, contradicting Econ dogma.
- Options paradox: more choices are not always welfare-improving when self-control is at stake.
- Preference reversal: guests wanted a few nuts at 7:15, whole bowl by 7:30.
- Economists' blind spot: rational-choice theory cannot explain gratitude for removed temptation.
- Hot-Cold Empathy Gap
- Temptation defined: consume more when hot than when cold.
- Hot-cold empathy gap: in cold states we underestimate arousal's effect on behavior.
- Arousal naïveté: dieters, shoppers, lovers misjudge future selves under influence.
- Ulysses contract: precommitment—wax ears, tie to mast—lets cold self bind hot self.
- Hot states not always bad: fear can be overcome, dessert and love can be wise.
- Planner versus Doer
- Two selves: far-sighted Planner (Reflective/Spock) vs myopic Doer (Automatic/Homer).
- Internal conflict: Planner promotes long-term welfare; Doer chases immediate temptation.
- Neuroeconomics: brain regions for temptation and restraint can battle in severe conflict.
- Self-control strategies: lists, alarm clocks, friends help Planner rein in Doer.
- Clocky: alarm runs away, forcing sleepy Doer out of bed.
- Mindless Choosing
- Automatic pilot: people eat, drive, and shop without active attention.
- Mindless eating: we eat whatever is in front of us, even stale popcorn.
- Wansink popcorn: big bucket led 53% more eaten despite disliked stale popcorn.
- Bottomless soup: self-refilling bowls kept people eating without noticing quantity.
- Choice architecture: plate and package sizes are nudges that shape consumption.
- Practical nudge: smaller plates, small packages, keep tempting food out of fridge.
- Self-Control Strategies
- Self-control by outsiders: Planners use incentives, bets, and precommitments to bind Doers.
- Alarm clock battle: Planner sets 6:15; Doer snoozes until 9:00 unless restrained.
- Thaler's thesis bet: $100 checks cashed if chapters late; party without David punishes delay.
- Salience wins: losing $100 and missing party hurts more than abstract retirement contribution.
- Weight-loss pact: Romalis and Karlan agreed to lose thirty pounds over nine months.
- Dynamic Inconsistency and Temptation
- Commitment Devices and Mental Accounting (2 Resisting Temptation · II)
- Commitment Contracts and Self-Control
- Planner vs. Doer: Dean and John bet money to force weight loss, knowing willpower alone would fail.
- Weigh-in contract: One day’s notice, over target pays a fine; only one lapse in four years.
- StickK model: Formal commitment contracts let Planners constrain Doers through self-imposed penalties.
- External enforcement: People seek devices that make bad choices costly and good choices automatic.
- Ulysses strategy: Binding oneself in advance is rational when future temptations are predictable.
- Government as Commitment Device
- Pure vs. libertarian paternalism: Bans on heroin, prostitution, and drunken driving are intrusive; less coercive tools preserve choice.
- Cigarette taxes: Discourage smoking without forbidding it, helping self-control while raising revenue.
- Casino self-exclusion: Voluntary ban lists are libertarian because no one must sign up and refusal is nearly costless.
- Daylight saving time: Relabeling hours nudges everyone to wake earlier, adding evening light and saving energy.
- Franklin’s candle savings: Proposed in Paris but not adopted until World War I.
- Market Self-Control and Its Limits
- Christmas savings club: Weekly deposits locked for a year, near-zero interest, redeemed before Christmas.
- Economically irrational: No liquidity, high transaction costs, low return—impossible for Econs, sensible for Humans.
- Illiquidity as feature: Like Ulysses, the inability to withdraw prevents spending before the goal.
- Credit cards displaced clubs: Households can finance Christmas instead of saving in advance.
- Competition can undermine goals: Cinnabon aromas outsell fruit and yogurt, showing markets cater to demand, not wisdom.
- Mental Accounting and Fungibility
- Mental accounting: Implicit household system to evaluate, regulate, and process the budget.
- Violates fungibility: Money is labeled and treated as non-transferable despite economic equivalence.
- Hoffman’s mason jars: Rent, utilities, and food jars; food jar empty while other money sits nearby.
- Organizational budgets: Departments cannot spend another account’s surplus, mirroring household jars.
- United Way insurance: Professor deducts parking fines from intended charity gift to self-insure minor mishaps.
- House Money and Costly Self-Control
- House money effect: Gamblers treat winnings as separate money and take greater risks.
- Windfall and investment gains: Unexpected money or stock winnings are spent on luxuries and risky bets.
- Simultaneous debt and savings: Households hold low-interest savings and high-interest credit card debt.
- Costly self-control: Separate accounts prevent paying off cards only to run them up again.
- Credit card limits: Lenders extend more credit to those near limits, worsening the trap.
- Saving, Fun, and Policy Accounts
- Dennis’s fun account: Social Security checks fund enjoyable purchases, not roof repairs.
- Rainy-day and fun accounts: Near-sacrosanct savings plus freely available entertainment accounts benefit many.
- Public policy implication: Direct increased savings into accounts where spending is not too tempting.
- Tax refund as Easter account: Refunds are treated as forced savings and a windfall, not interest-free loans.
- Charity adjustment: Professor’s intended gift must cover all mishaps, so United Way is not deprived.
- Commitment Contracts and Self-Control
- Conformity, Social Norms, and Nudges (3 Following the Herd · I)
- Social Influence and Human Vulnerability
- Humans vs. Econs: Econs ignore fashion; Humans are easily nudged by others’ statements and deeds.
- Jonestown: Social pressure can make entire groups commit unthinkable acts, including mass suicide.
- Two mechanisms: Social influence works through information from others and peer pressure to avoid disapproval.
- Social contagions: Teen pregnancy, obesity, broadcast fads, roommate effort, and judicial votes spread through peers.
- Learning and error: Most people learn from others, but major misconceptions also come from others.
- Asch's Conformity Experiments
- Line-judgment task: Alone, people almost never err; amid unanimous wrong answers, they err over one-third of the time.
- Unanimity and prevalence: Conformity appears in 20–40% across 17 countries; unanimous groups are strongest.
- Public vs. anonymous: Conformity falls with anonymous answers and rises when others will see one’s response.
- Perception: Brain imaging suggests conformists may actually see the situation as the group does.
- Odd conclusions: Private support for “subversive activities” was 12%; with unanimous group consensus, 48%.
- Sherif, Group Norms, and Tradition
- Autokinetic effect: A stationary light appears to move; individual estimates are uncertain and vary.
- Group convergence: In small groups, estimates converge into a stable group norm.
- Arbitrary starting points: Different groups commit to different judgments from modest or arbitrary initial variations.
- Confident confederate: A firm ally’s estimate can inflate or lower group judgment; a small nudge has large effects.
- Internalized tradition: Norms persist alone, a year later, in new groups, and even after original members are gone.
- Collective Conservatism and Pluralistic Ignorance
- Collective conservatism: Groups cling to established patterns even as new needs arise.
- Perpetuation without basis: Once established, practices like ties persist absent practical justification.
- Pluralistic ignorance: People follow a practice because they wrongly think most others approve of it.
- Hidden dissent: Many Soviet bloc citizens despised communism but did not know how many shared that view.
- Dislodging norms: A small shock or nudge can break a practice sustained only by misperceived consensus.
- Nudging with Social Norms
- Choice architects: Can use social influence to encourage beneficial behavior.
- Exercise example: A city can nudge exercise by publicizing that many people already exercise.
- Texas anti-litter: “Don’t Mess with Texas” used Cowboys players, Willie Nelson, and Texas pride instead of moralizing.
- Results: Litter fell 29% in the first year and 72% over six years without mandates, threats, or coercion.
- Slogan success: About 95% of Texans knew it; it became America’s favorite slogan.
- Spotlight Effect and Cultural Contagion
- Spotlight effect: People overestimate how much others notice them, which fuels conformity.
- Barry Manilow test: Subjects guessed 46% would identify their T-shirt; only 21% did.
- Music market: 14,341 participants rated and downloaded unknown songs; half saw others’ download counts.
- Eight worlds: Social-influence group was split into independent worlds, each seeing only its own downloads.
- Key question: Did others’ choices nudge downloads, and could different songs become popular in different worlds?
- Social Influence and Human Vulnerability
- Herding, Priming, and Social Nudges (3 Following the Herd · II)
- Success Is Largely Unpredictable
- Music download experiment: In eight parallel worlds, the same songs became hits or failures almost at random
- Social influence drives demand: People download what others have already downloaded, far more than quality dictates
- First movers decide fates: A song's success hinged on the early choices of initial downloaders
- Beware retrospective certainty: Stars and governors are often indistinguishable from the equally talented who never made it
- Unintentional Social Nudging
- Seattle Windshield Pitting Epidemic: In 1954, reports of pitted windshields spread contagiously across Washington
- Pits were always there: The damage turned out to be ordinary, but nobody noticed until reports made it salient
- Obesity is contagious: We mirror companions' eating habits whether or not they intend to influence us
- Group size inflates intake: People eat 35% more with one companion and 96% more in groups of seven or more
- Consumption norms rule: Light eaters overeat among heavy eaters; heavy eaters restrain themselves among light ones
- Strategic Use of Social Influence
- Marketers exploit conformity: Advertisers tout "most people prefer" and "growing numbers are switching"
- Politicians manufacture bandwagons: Claiming that voters are flocking to a candidate can help make it true
- Kerry's 2004 surge: The Democratic nomination shifted on a perception of momentum, not independent voter judgment
- Cascade of imitation: Voters chasing the likely winner can start a stampede that heads in any direction
- Social Nudges as Choice Architecture
- Tax compliance (Minnesota): Only telling taxpayers that 90% of Minnesotans comply reduced cheating
- Preserving petrified wood: "Please don't remove" beat warnings about how many past visitors had taken samples
- College drinking norms: Correcting exaggerated beliefs about peer drinking statistically reduced abuse
- Energy use (San Marcos): Neighborhood-average feedback cut high users' consumption but boomeranged low users upward
- Emoticons defeat the boomerang: A happy face kept below-average users from increasing; an unhappy face pushed high users down further
- Priming and Channel Factors
- Priming: Subtle cues activate the Automatic System and ease associated thoughts and actions
- Mere-measurement effect: Asking about intentions changes behavior—25% more voting, 35% more car purchases
- Channel factors (Lewin): Removing a small obstacle often facilitates behavior better than persuasion
- Tetanus shot study: Mapping the route and planning a time raised inoculation from 3% to 28%
- Incidental cues shift conduct: Briefcases breed competitiveness; cleaner scent tidies; iced coffee makes strangers seem colder
- Success Is Largely Unpredictable
- 4 When Do We Need a Nudge?
- The Golden Rule of Libertarian Paternalism
- Choice architecture is unavoidable: someone always designs the environment, so only the quality is in question
- Golden rule: offer nudges most likely to help and least likely to inflict harm
- When nudges are needed: decisions that are difficult, rare, lack prompt feedback, and resist translation into understandable terms
- Fraught Choices: Now Versus Later
- Self-control problems arise when choices and consequences are separated in time
- Investment goods: exercise, flossing, dieting — costs now, benefits later, so people do too little
- Sinful goods: smoking, alcohol, doughnuts — pleasure now, costs later, so people do too much
- New Year's resolution test: few vow to floss less or smoke more; both types are prime nudge candidates
- Difficulty and Frequency
- Degree of difficulty: picking a mortgage demands far more help than picking a loaf of bread
- Frequency: practice improves hard tasks, but life's most important choices offer few trials
- Higher stakes, less practice: college, spouse, career, and retirement come once or twice at most
- Boundary: this concerns nudges, not government telling people whom to marry or what to study
- Weak Feedback
- Learning requires immediate, clear feedback after each attempt — putting differently when you can see the balls
- Feedback comes only on options chosen, so people never learn about the paths they reject
- Long-term processes hide harm: a high-fat diet gives no warning until the heart attack
- When feedback fails, a nudge may substitute for the learning experience people cannot get
- Knowing What You Like
- Familiar tastes are learned by sampling, but unfamiliar preferences must be forecast
- Translation problem: choices are hard when they cannot be mapped onto imaginable experiences
- Mutual funds and health plans: even experts struggle to predict retirement spending or specialist access
- Fewer options can help: Alinea and Charlie Trotter's let the chef decide, authorizing what you'd never order
- When prediction fails, more options and self-choice deliver less rather than more
- Markets: A Mixed Verdict
- Competition usually protects: quarter lovers still pay twenty-five cents each when banks compete
- Price signals quality as long as irrational consumers do not predominate
- Competition sometimes fails: twenty-dollar extended warranties for two dollars of expected value persist
- Firms cater to frailties rather than eradicate them — who profits from talking you out of a purchase?
- Competing firms oppose each other: cigarettes versus quit-smoking, fast food versus diet advice
- Government is populated by Humans too, which is why choice architecture, not prohibition, is the tool
- The Golden Rule of Libertarian Paternalism
- Designing Choice Architecture for Human Fallibility (5 Choice Architecture · I)
- Stimulus-Response Compatibility
- Bad architecture: doors with big wooden handles shout "pull" even when they must be pushed
- Stimulus-response compatibility: signals should match the desired action, or performance suffers
- Automatic System wins: even warned, students kept pulling — the cue cannot be screened out
- Stroop test: naming the ink color of a mismatched color word slows responses and raises errors
- Everyday failures: remotes where volume and accidental "input" buttons are the same size
- Norman's stoves: The Design of Everyday Things shows knob layouts that map clearly to burners
- Defaults: Padding the Path of Least Resistance
- Defaults win: inertia, status quo bias, and the "yeah, whatever" heuristic drive people to the pre-set option
- Unavoidable: every choice node requires a rule for what happens if the chooser does nothing
- Dead man switches: chainsaws and mowers stop when released — the safe outcome is the default
- Automatic renewal: magazines keep subscribers who never read them
- Opt-in versus opt-out: the military recruiter rule produced radically different outcomes depending on the default
- Mixed motives: software defaults may be helpful (regular install) or self-serving (marketing emails)
- Required (Mandated) Choice
- Mandated choice: forcing an active decision defeats inertia on emotionally charged matters
- Nuisance cost: most people prefer a sensible default to reading an arcane manual
- Complexity limit: required choosing suits simple yes-or-no decisions, not complicated ones
- Restaurant analogy: no diner wants to hand the chef a recipe for every dish ordered
- Expect Error
- Forgiving design: good systems assume users will err and make recovery easy
- Le Métro versus Chicago: symmetric fare cards work either way; credit cards have one correct orientation
- Postcompletion errors: after the main task, people forget the ATM card or the copier original
- Forcing functions: requiring card removal before cash makes forgetting the card impossible
- Physical constraints: wrong-size diesel nozzles and unique anesthesia connectors eliminate mistakes
- Look right: London pavement signs counter tourists' lifelong instinct to look left
- Habit, Routine, and Forgetting
- Drug compliance: once-daily, same-time dosing turns pill-taking into an Automatic System habit
- Regularity over frequency: every-other-day schedules defeat most people's memory
- Placebo pills: birth control packs add dummy pills so the daily routine never breaks
- Missing attachment: Gmail prompts "Did you forget your attachment?" — an error-catcher in software
- Give Feedback
- Feedback improves performance: well-designed systems say when people are doing well or erring
- Digital cameras: instant review eliminates film-era errors like an unloaded roll or forgotten lens cap
- Fake shutter click: an artificial sound restores the missing confirmation that an image was captured
- Anticipatory warnings: laptops urge plug-in or shutdown when the battery is dangerously low
- Stimulus-Response Compatibility
- NUDGES — Feedback, Mappings, Incentives, Structure (5 Choice Architecture · II)
- Feedback That Actually Works
- Warning fatigue: too many warnings get ignored, rendering them useless
- Terror alerts: DHS color codes demand vigilance but tell citizens nothing actionable
- Pink-then-white paint: sensing paint reveals where you've painted; feedback solves the task
- Understanding Mappings
- Mapping: the relation between a choice and the welfare it actually delivers
- Easy task: ice cream flavors map directly onto predictable, known experience
- Hard task: cancer treatment trades a one-third impotence risk against 3.2 extra years
- Treatment lottery: choice hinges on which specialist is seen; watchful waiting goes underused
- Comprehensible units: translate megapixels into print sizes, apples into glasses of cider
- RECAP: Making Prices Transparent
- Hidden pricing: credit cards, mortgages, and cell plans bury fees consumers cannot compute
- RECAP: Record, Evaluate, Compare Alternative Prices — disclosure without price regulation
- Price disclosure: publish fee schedules and formulas in spreadsheet-like format
- Usage disclosure: annual electronic statement of real usage; comparison sites follow
- Structuring Complex Choices
- Compensatory strategy: with few options, weigh every attribute and trade off strengths
- Elimination by aspects: with many, set cutoffs and eliminate attribute by attribute
- Paint wheel: ordering two thousand colors by similarity beats alphabetical listing
- Collaborative filtering: Netflix uses tastes of people like you to filter vast catalogs
- Caution: serendipity matters; sometimes sample what people unlike you enjoy
- Incentives and Salience
- Four questions: who uses, who chooses, who pays, who profits?
- Incentive conflicts: U.S. health care splits the stakes among payer, doctor, and suppliers
- Salience: the key modification — do choosers actually notice the incentives they face?
- Opportunity cost: car owners forget the $10,000; taxi meters stay in their face
- Directing attention: cost-displaying phones, thermostats, and calorie meters shift behavior
- The NUDGES Mnemonic
- iNcentives: put the right incentives on the right people
- Understand mappings: translate options into welfare-relevant terms
- Defaults: supply sensible starting points
- Give feedback: make consequences visible
- Expect error: design for human mistakes
- Structure complex choices: organize options to aid choosers
- Feedback That Actually Works
- Two Minds and Their Biases (1 Biases and Blunders · I)
- Part Ii money
- Retirement Saving and Defaults (6 Save More Tomorrow · I)
- The Retirement Saving Problem
- Negative savings rate: In 2005 Americans saved nothing on average, spending and borrowing more than they earned.
- Retirement shortfall: Many workers, even high earners like Tony Snow, fail to save or enroll in 401(k)s.
- Social Security strain: Low private saving worsens the coming solvency problem requiring tax hikes or benefit cuts.
- Tax incentives insufficient: IRAs and 401(k)s exist, but eligible people often fail to use them fully.
- Why Standard Saving Theory Fails
- Econ assumption: Standard theory assumes people calculate lifetime earnings and smooth retirement consumption.
- Impossible math: Even economists rarely solve the complex saving optimization problem.
- Willpower gap: Theory assumes no temptation distracts from saving, but Humans face sports cars and vacations.
- Historical shift: Longer lives and dispersed families made self-funded retirement necessary, starting with Bismarck in 1889.
- Defined-Benefit vs. Defined-Contribution
- Defined-benefit ease: Traditional pensions and Social Security require almost no decisions and are forgiving.
- Defined-benefit limits: Payouts depend on formulas, vesting, and staying with one employer; costly to administer.
- Defined-contribution rise: 401(k)s are portable and flexible but demand active saving and investing choices.
- Unforgiving design: Employees must join, choose rates, manage portfolios, and handle proceeds at retirement.
- Are People Saving Enough?
- Expert disagreement: Economists dispute ideal post-retirement income; some urge full replacement, others economical living.
- Asymmetry of error: Saving too little is harder to fix than saving too much, so nudges should favor saving.
- Clear undersavers: Nonparticipants and older workers saving low percentages are definitely saving too little.
- Stated intentions: 68% of 401(k) participants say they save too little, suggesting openness to help.
- Enrollment Failures
- Free money missed: Employer matches are virtually free money, yet about 30% of eligible employees fail to enroll.
- Extreme failures: In UK plans fully paid by employers, only 51% joined; some older US workers forgo matches.
- Procrastination: Many workers delay joining for months or years, spacing out rather than making reasoned decisions.
- Clear blunder: Failure to join is often a mistake, not a sensible choice to use money elsewhere.
- Defaults and Active Choice
- Two proposed nudges: Automatic enrollment and Save More Tomorrow are the chapter's central remedies.
- Automatic enrollment: Changing default to opt-out raised participation from 20% to 90% immediately, over 98% in three years.
- Dropout evidence: Very few auto-enrolled workers quit, suggesting defaults align with their real preferences.
- Required choosing: Forcing a yes/no decision raised participation by about 25 percentage points.
- Simplified forms: Preselecting rate and allocation and asking only to check yes raised early participation from 9% to 34%.
- The Retirement Saving Problem
- Retirement Saving Via Smarter Choice Architecture (6 Save More Tomorrow · II)
- Removing Friction, Limiting Overload
- Channel factors: dig a smooth path and the seemingly tiny barriers that block people disappear
- Choice overload: the more funds offered in a plan, the lower the participation rate
- Refusal to choose: excess complexity makes some people decline to decide at all
- The Perils of Low Defaults
- Sticky defaults: automatic and forced-choice plans set 2–3 percent rates, and many workers stay there
- Wrong investment default: remaining in a conservative default fund costs employees dearly over time
- Cursory decisions: 58 percent spend under an hour choosing contribution rate and investments
- Round-number heuristic: people pick 5, 10, or 15 percent with no rational basis
- Match threshold: many contribute only enough for the full employer match, so raising the match ceiling raises saving
- Why Education Falls Short
- Financial literacy test: a free employer education program moved average scores only from 54 to 55
- Intention gap: everyone at a seminar wanted to save more, but only 14 percent joined, versus 7 percent of non-attendees
- Benefit fairs: attendance similarly produced only small gains in tax-deferred account participation
- Designing Save More Tomorrow
- Future self-control: restrictions are easier to accept when they begin later, never today
- Loss aversion: people hate watching their paychecks shrink
- Money illusion: losses are felt in nominal dollars, ignoring inflation
- Inertia: automatic increases harness the same force that normally prevents saving
- Synchronized timing: hikes are tied to pay raises, so take-home pay never actually falls
- The First Implementation
- Consultant access: about 90 percent of employees accepted a one-on-one meeting with an adviser
- Advice rejected: suggested 15 percent rates, then 5-point increases, which only 25 percent accepted immediately
- Reluctant savers: of those who couldn't afford the cut, 78 percent joined escalation at 3 points per raise
- Three trajectories: non-attendees stayed near 6 percent, advice-takers rose to 9 percent, escalators quadrupled from 3.5 to 13.6 percent
- Ratchet effect: leavers stopped escalating rather than asking to reduce their contributions
- Scaling Up, and the Government's Role
- Private adoption: Vanguard, Fidelity, TIAA-CREF and others installed escalation; 39 percent of large employers by 2007
- Ease of enrollment: signing up on an obscure web page suppresses participation; having forms on hand boosts it
- Automatic enrollment: Safelite's combined plan enrolled 93 percent by default with only 6 percent opting out
- Vanguard evidence: participating savers jumped from 23 percent to 78 percent once enrollment became automatic
- Government as enabler: IRS rulings and the 2006 Pension Protection Act rewarded auto-enrollment with regulatory waivers
- Blunder counterexample: Social Security's Break-Even Age calculator ignores interest, taxes, and spouses; a choice architect could ask the right questions
- Removing Friction, Limiting Overload
- Naïve Investing and Portfolio Mistakes (7 Naïve Investing · I)
- The Portfolio Allocation Problem
- Defined-contribution shift: employees gain control, options, and responsibility for investment choices.
- Portfolio selection: choosing how to invest is even harder than deciding how much to save.
- Asset allocation: mix stocks, bonds, and other assets to balance risk and return.
- Risk-return trade-off: riskier stocks usually earn more, but may lose value.
- Linked decisions: willingness to bear risk affects how much one must save.
- Stocks, Bonds, and the Equity Premium
- Historical returns, 1925–2005: Treasury bills returned 3.7%; bonds 5.5%; large-company stocks 10.4%.
- Inflation caveat: Treasury bills' 3.7% return barely exceeded 3.0% annual inflation.
- Equity premium: stocks' higher return compensates investors for greater risk than government bonds.
- Stock volatility: stocks can fall over 30% in a year; global indexes dropped 20% in one day in 1987.
- Econ benchmark: trade risk and return based on preferences about retirement income.
- Human limitation: investors cannot calculate the trade-off, so short-term swings and rules of thumb dominate.
- Loss Aversion and Monitoring Frequency
- Loss aversion: people hate losses about twice as much as they enjoy equivalent gains.
- Daily monitor: Vince checks his portfolio daily, feels losses acutely, and grows nervous about stocks.
- Long-horizon sleeper: Rip invests all in stocks before a twenty-year sleep, calm because stocks almost surely rise.
- Monitoring lesson: risk attitudes depend on how often investors check their portfolios.
- Practical mistake: frequent monitoring makes investors hold too little stock; long-horizon data makes them choose nearly all stocks.
- Market Timing: Buy High, Sell Low
- Backward timing: new Vanguard participants allocated 58% to stocks in 1992, 74% in 2000, 54% in 2002.
- Chasing peaks: investors bought heavily when prices were high and sold after prices fell.
- Tech-fund example: participation rose from 12% in 1998 to 37% at the 2000 peak, then fell to 18%.
- Belief driving it: investors came to think stocks only go up or dips are buying opportunities.
- Market timing error: new participants adjusted allocations in the wrong direction.
- Rules of Thumb and Naïve Diversification
- Markowitz's confession: even a portfolio-theory founder split contributions fifty-fifty between bonds and equities.
- TIAA-CREF default: over half of participants chose an even split; Sunstein never rebalanced, now over 60% in CREF.
- Drift risk: without rebalancing, the asset mix changes with relative returns.
- Diversification heuristic: "When in doubt, diversify"; the 1/n rule divides assets evenly across options.
- Halloween experiment: children offered two candies at one house each took one of each; sequential choosing did so only 48%.
- Menu effects: fund choices shape allocation; plans with more stock funds led participants to invest more in stocks.
- The Portfolio Allocation Problem
- Retirement Portfolios, Company Stock, Defaults (7 Naïve Investing · II)
- Lifestyle Funds and Misuse
- Lifestyle funds: Conservative, moderate, and aggressive blends let participants choose one diversified portfolio.
- Designed as defaults: If fees are reasonable, they make excellent default options for different risk tolerances.
- Common misuse: Few participants put all retirement money in one lifestyle fund, as intended.
- Unwitting aggression: A conservative-fund investor spread money across stock-heavy funds, ending 77% in stocks.
- Company Stock Concentration
- Enron lesson: Charlie Prestwood’s $1.3 million Enron nest egg vanished; he now lives on pension and Social Security.
- Widespread risk: Five million Americans hold more than 60% of retirement savings in company stock.
- Double exposure: Employees can lose job and retirement savings together if their employer fails.
- False safety: Half of surveyed 401(k) participants thought company stock was as safe as money market funds.
- Performance chase: Employees extrapolate strong past company-stock performance into the future.
- Match as advice: Required company-stock matches lead workers to put 29% of their own contributions in company stock.
- Law, Risk, and Fixes
- Single-stock discount: A dollar in company stock is worth less than half a dollar in a diversified mutual fund.
- ERISA exemption: Company stock escapes the diversification rule in defined-contribution plans.
- Perverse incentive: The exemption encourages concentrated employer stock despite correlated job and savings risk.
- Clean fix: Treat company stock like any other investment, letting fiduciary risk deter heavy allocations.
- 10% cap: A nonlibertarian alternative would limit company stock to 10% of a retirement portfolio.
- Personal rule: If over 10% of retirement money is in your employer, diversify quickly.
- Sell More Tomorrow
- Sell More Tomorrow: Employees may gradually sell company stock into a diversified portfolio over time.
- Dual purpose: Gradual sales avoid flooding the market and avoid signaling management doubts.
- Opt-in or opt-out: Firms can offer the program either way.
- Defaults and Automatic Enrollment
- Default necessity: Automatic enrollment requires a specific default asset allocation.
- Too conservative: Money-market defaults plus low savings rates can produce poverty in old age.
- Legal fear: Firms once chose conservative defaults to avoid lawsuits and lacked safe-harbor cover.
- New guidelines: Department of Labor safe harbor now permits sensible default funds.
- Good defaults: Lifestyle, target-maturity, and managed-account funds can match participants by risk or retirement date.
- Structuring Choices and Feedback
- Tiered choice: Default for passive investors, small balanced-fund set for involved, full menu for sophisticated.
- Expect error: Combine automatic enrollment with Save More Tomorrow and automatic rebalancing.
- Mappings: Translate savings rates and returns into housing images—hovel, condo, pool—to make outcomes concrete.
- Feedback: Annual reports show current trajectory and how raising savings changes retirement lifestyle.
- Incentives: Conflicts between employer and employee, like company stock, require enforcing fiduciary law.
- Need for help: Most participants cannot manage portfolios alone; choice architecture and nudges help.
- Lifestyle Funds and Misuse
- Borrower Confusion and Mortgage Choice Architecture (8 Credit Markets · I)
- The Complexity Trap in Mortgage Shopping
- Homer Simpson's lease: borrowers accept payment terms they never really parse — naïveté is common, not exceptional.
- Once simple: fixed 30-year rates and 20% down made comparing loans a matter of picking the lowest rate.
- Truth in Lending Act: requiring uniform APR disclosure was excellent choice architecture for comparison.
- Options explosion: variable rates, interest-only loans, teaser rates, points, and prepayment penalties made mortgages harder to shop than retirement portfolios.
- Shu's study: even top MBA students struggled to identify the best loan in simplified tasks.
- Evidence on Who Loses in Lending
- Woodward's FHA findings: African-American borrowers pay $425 more; Latino borrowers pay $400 more.
- Education gap: borrowers in high-school-educated neighborhoods pay $1,160 more than those in college-educated ones.
- Broker premium: brokered loans cost about $600 more than direct-lender loans.
- Complexity penalty: points and seller contributions raise costs, and do so more on brokered loans.
- General lesson: complicated markets disproportionately disadvantage unsophisticated shoppers and expose them to self-interested advice.
- Subprime Lending: Neither Predator Nor Saint
- Two extremes: calling all subprime "predatory" ignores risk pricing; dismissing all criticism ignores real exploitation.
- Genuine value: subprime credit opens homeownership and second chances to borrowers otherwise shut out.
- Reality check: microfinance rates above 200% still leave borrowers better off, so high rates alone aren't predation.
- The predatory pattern: brokers solicit borrowers, offer points few can evaluate, and profit as borrowers defer.
- Broker incentive gap: those serving wealthy clients build reputations; those serving the poor often seek quick profits.
- Disclosures That Fail and Bans That Overreach
- Good-faith estimate: required within three days but often withheld until signing, buried in a pile of papers.
- HOEPA disclosure: meant to warn high-risk borrowers, yet never says "high-risk" and is simply signed.
- Well-intentioned failure: Truth in Lending summarized terms clearly, but fine print defeats the truth.
- Wilkins's proposal: ban exotic mortgages and require waiting periods to simplify comparison.
- Why it doesn't qualify: banning contracts blocks mutually beneficial loans, like variable-rate loans for those selling soon.
- RECAP: Better Architecture for Borrowers
- Two salient categories: lenders report fees and interest separately, with all fees summed into one number.
- Zero-fee advantage: borrowers paying no up-front fee consistently get the best deals — only interest remains to compare.
- Payment schedule: disclosure includes rates and future payments, revealing what happens when teaser rates end.
- Worst-case scenario: showing how high payments could rise prepares borrowers for rate changes.
- Machine-readable data: detailed RECAP reports let third parties build comparison services, improving online shopping, especially for women and minorities.
- The Complexity Trap in Mortgage Shopping
- Borrowing Traps and Choice Architecture (8 Credit Markets · II)
- Student Loan Market Failures
- Rising costs: Tuition, room, and board exceed $50,000 yearly at many private universities.
- Debt default: About two-thirds of four-year college students graduate with debt.
- Two loan types: Private loans from financial institutions; federal Stafford loans are need-based and subsidized.
- Complex application: FAFSA and College Board profile each exceed 100 questions, taking hours.
- Private loan risk: Direct-to-consumer loans can be used for any expense, encouraging overborrowing.
- Misaligned Advice and Rent-Seeking
- Predatory solicitations: Lenders pitch loans as easy as ordering pizza, hiding cheaper federal options.
- Preferred lender lists: Colleges accept gifts, stock, and donations from lenders to label them preferred.
- Conflicted advice: Some aid offices push preferred lenders or let lenders staff call centers under college name.
- Rent-seeking: Guarantees and subsidies make loans so profitable that lenders compete via influence, not price.
- Confused customers: High stakes plus confusion make borrowers vulnerable to self-serving experts.
- Nudges for Student Loans
- Simplify FAFSA: Cut and standardize questions; combine the application with annual tax returns.
- RECAP for loans: A comparison spreadsheet would let borrowers easily compare federal and private loans.
- Early savings: 529 plans and automatic deposits help families reduce future borrowing.
- Ohio tax study: H&R Block tax professionals complete FAFSA from tax returns for likely aid-eligible families.
- Credit Cards: Convenience and Self-Control
- Dual function: Credit cards provide payment and short-term liquidity; debit cards only payment.
- Debt growth: Average family credit card debt rose from $2,697 in 1989 to about $8,000 in 2007.
- Interest burden: Typical 18% rates mean over $2,000 yearly interest on $12,000 debt.
- Self-control erosion: Cards replace pay-as-you-go budgeting and enable spending beyond cash.
- Spending boost: People bid twice as much for Celtics tickets when paying by credit card rather than cash.
- Anchoring risks: Minimum payments and credit limits serve as anchors that encourage more spending.
- Nudges for Credit Cards
- Annual fee statement: Require card issuers to send yearly list totaling all fees.
- Salience: A report showing $2,153 interest, $247 late fees, $57 currency fees creates a wakeup call.
- Hidden fees: Foreign transaction fees and shortened grace periods are buried and ignored.
- Automatic full payment: Companies should allow automatic payment of the full bill.
- Minimum payment default: The printed minimum anchors users toward paying too little.
- Conclusion: Choice Architecture in Credit Markets
- Needless complexity: Mortgages, student loans, and credit cards are more complicated than necessary.
- Exploitation: Complexity and confusion let lenders and advisers exploit borrowers.
- Freedom with nudges: Government should respect choice but improve architecture to prevent bad choices.
- Discrimination: Women and African-Americans pay more at car dealerships, but not online.
- Student Loan Market Failures
- Sweden's Choice Architecture Lessons (9 Privatizing Social Security: Smorgasbord Style · I)
- The Swedish Experiment
- Bush's 2005 proposal: partial Social Security privatization failed, but the idea keeps resurfacing
- Sweden's parallel launch: the same debate produced a real system worth studying for lessons
- Core lesson: freedom of choice has limits; choice architecture determines whether choosers succeed
- Verdict: Swedish officials did well on some design points, but one error degraded citizens' portfolios
- Pro-Choice Design
- Just Maximize Choices: give people every option, then let them do whatever they want
- Portfolio rule: participants could select up to five funds from an approved list
- Open entry: any fund meeting fiduciary standards could join, leaving 456 funds initially, over 1,000 since
- Information overload: fees, performance, and risk delivered as a book; funds free to advertise
- Econ assumption: free entry plus unfettered competition and abundant choice would be uncontroversial — if Swedes were Econs
- Choosing the Default
- The range: from no choice (A) through discouraged, encouraged, or neutral defaults (B–D) to required choosing (E)
- No-choice rejected: eliminating all choice conflicts with libertarian paternalism
- Required choosing rejected: forcing choice is harsh on the ill, absent, or clueless, and hard among 400+ funds
- Neutrality illusion: merely describing the default — expert-designed, low-fee — already influences its market share
- Sweden's pick: option B, with a massive ad campaign pushing active choice
- Who Actually Chose
- Active choosers: two-thirds selected their own portfolios; one-third landed in the default, the single largest fund
- Predictors: more money at stake, plus women and younger participants, raised active choosing
- Campaign decay: after advertising stopped, participation collapsed — only 8 percent of 2006 entrants chose
- Comparison: young workers chose 56.7 percent actively in 2000 versus 8 percent later
- How Good Were the Choices
- Default fund quality: 65 percent foreign stocks, 60 percent indexed, fees of just 0.17 percent — experts approve
- Excess risk: active choosers held 96.2 percent stocks, chasing the recent boom
- Home bias: choosers put 48.2 percent in Swedish stocks, though Sweden is ~1 percent of the world economy
- Buying what you know: familiarity is not edge; workers buying employer stock showed no profitable trading ability
- The Swedish Experiment
- Sweden's Privatization and Choice Architecture (9 Privatizing Social Security: Smorgasbord Style · II)
- How Active Choices Compared With the Default
- Home bias: Swedish investors held roughly 48 times the rational weight in Swedish stocks
- Active management: only 4.1 percent of actively chosen portfolios were indexed
- Higher fees: active choosers paid 0.77 percent versus 0.17 percent for the default fund
- Compounding cost: $60 more per year for every $10,000 invested
- Worse returns: the default lost 29.9 percent in 2000–2003; active portfolios lost 39.6
- Enduring gap: by July 2007 the default was up 21.5 percent, active choices only 5.1
- The Accident of Timing
- Trend following: actively chosen portfolios held over 96 percent stocks as the bull market ended
- Hot-fund magnet: Robur Aktiefond Contura's 534 percent past return drew 4.2 percent of the pool
- Past is not prophecy: that same fund then lost 69.5 percent in three years
- Confused signals: investors struggle to distinguish past returns from forecasts of future ones
- Political timing: privatization wins approval after bull markets and stalls after bear markets
- Inertia Preserves Early Mistakes
- Status quo bias: only 1.7 to 3.1 percent of participants changed portfolios each year
- Familiar pattern: this inertia mirrors behavior in U.S. 401(k) plans
- Lasting consequence: launch-date accidents harden into decades-long allocations
- Advertising: Dream or Nightmare
- The dream: ads educate consumers about fees, diversification, and long-run investing
- The nightmare: ads push big bets, tout past performance, and almost never mention fees
- Reality: Cronqvist found the ads resembled the nightmare far more than the dream
- Celebrity pitch: Harrison Ford urged Swedes to "pick a better pension"
- Effect: advertising steered people toward higher fees, more risk, and more home bias
- Doing Badly Without Nudges
- Default done well: Sweden's carefully designed default fund earned a five-star rating
- Core failure: officials obeyed the Just Maximize Choices mantra instead of shaping choice
- Predictable harm: availability bias and inertia produced exactly the results a nudge could prevent
- Legitimate nudge: when decisions are hard, unfamiliar, and feedback-free, guidance is warranted
- Better script: "If unsure, consult an expert or take the default designed for people like you"
- Sweden's correction: the government no longer actively encourages self-directed choosing
- Lessons for a U.S. Plan
- Greater scale: a U.S. system over thirty times larger would generate thousands of funds
- Bewilderment: long lists please Econos but paralyze Humans
- Start with a default: mostly index funds with managers chosen by competitive bidding
- Stage the choice: first a yes-or-no on the default, then a small set of age-based blended funds
- Full menu last: only those rejecting every simpler option reach the comprehensive list
- Governing principle: the more choices you give people, the more help you must provide
- How Active Choices Compared With the Default
- Retirement Saving and Defaults (6 Save More Tomorrow · I)
- Part Iii health
- Part D's Daunting Choice Architecture (10 Prescription Drugs: Part D for Daunting · I)
- Health as a Nudge Domain
- Social contagion: obesity and weight loss spread through social networks; shifting norms shifts behavior.
- Framing works: warning of increased risk from skipping cancer screenings motivates more than stressing reduced risk.
- Doctors as choice architects: physicians who understand how Humans think could lengthen lives considerably.
- Part D: Consumer Choice as Policy
- Compassionate conservatism: Bush paired an expensive entitlement with private plans and free consumer choice.
- Half-trillion-dollar subsidy: the largest overhaul in Medicare's history, passed narrowly by Congress.
- Not laissez-faire: government set minimum coverage and approved plans — constrained free choice.
- Six design features: voluntary enrollment, late penalties, menus of 45–66 plans varying by state.
- Doughnut hole: coverage stops mid-spending then resumes; no economist would recommend such a policy.
- Four Defects of the Architecture
- No guidance: a crowded menu with nothing to help participants select wisely.
- Wrong defaults: nonenrollment for most seniors, random assignment for six million others.
- No matching: officials resisted pairing people with plans using their drug histories.
- Vulnerable failed: the poor and poorly educated gained least from the program.
- Confusion at Enrollment
- McAllen, Texas: 47 plans in a poor town; doctors described total chaos among patients.
- Overwhelmed helpers: pharmacists, physicians, and Medicare hotlines were as confused as seniors.
- Conflict of interest: AARP sold insurance while also advising seniors which plan to pick.
- Bush's counsel: telling seniors to seek help among 46 choices is nearly the same as no help.
- Enrollment Success, Learning Doubtful
- Headline numbers: by January 2007 fewer than 10 percent of beneficiaries lacked drug coverage.
- Illusory learning: roughly two-thirds enrolled automatically or easily, not through informed choice.
- Approval rose: unfavorable views fell, and three of four liked their personal experience.
- Persistent gap: four million remained uncovered, mostly poorly educated people above poverty.
- Unclaimed subsidy: a quarter of 13.2 million eligible low-income seniors skipped nearly free coverage.
- Still baffling: after a year, 73 percent of seniors and 90 percent of doctors called it too complicated.
- Random Defaults for the Most Vulnerable
- Dual eligibles: six million of the poorest, sickest seniors were assigned to randomly chosen plans.
- Who they are: disproportionately African-American, Latino, and female, averaging ten or more prescriptions.
- Random harm: some were defaulted into plans covering as little as 76 percent of their key drugs.
- Maine's alternative: "intelligent assignment" matched 45,000 people to plans — the shrewd exception.
- Health as a Nudge Domain
- Choice Architecture for Medicare Part D (10 Prescription Drugs: Part D for Daunting · II)
- Random Defaults and Their Costs
- Random assignment: dual eligibles are defaulted into drug plans by lottery, and Humans mostly stay
- Health toll: 22 percent of dual eligibles stopped medications over plan-management problems
- Official defense: needs change over time — yet last year's drug use predicts next year's use well
- Market contradiction: markets reward better products; no one picks cars by lottery or wants equal shares
- Costly misallocation: lowest-cost fitting plans save $5–$50 per drug monthly; random plans cost about $700 more yearly
- Two Readings of Low Switching
- 2007 enrollment: 2.4 million changed plans, but 1.1 million were moved unilaterally by government
- Active choosers: excluding those moved, only about 6 percent switched plans themselves
- Econ reading: low switching means diverse plans are fitting diverse needs — all is well
- Human reading: inertia and status quo bias keep people in whatever plan they got
- Deciding evidence: even randomly assigned enrollees rarely switch, though their plan cannot be best
- A Maze No Senior Can Navigate
- Fifteen dimensions must be compared, with prior authorization and quantity limits unknowable before enrolling
- Medicare Web site: no spell checker, dosage and generic-drug traps; four testers produced four different answers
- Constant churn: nearly 40 percent of 225 plans changed prices by more than 5 percent
- Passive seniors: most don't use the Web, relying on impersonal insurer, government, and AARP mailings
- Good Choices Are Rare
- McFadden's test: four equal-value plans, no network or authorization worries, yet two-thirds still failed to minimize costs
- Core failure: seniors didn't connect plan features to their own health, drug use, and risk attitudes
- Unchecked claim: government asserts seniors choose well while withholding the data to verify it
- Intelligent Assignment: Maine's Model
- Matching, not lottery: ten benchmark plans scored against three months of each person's prescription history
- Auto-switch rule: participants under 80 percent drug coverage were moved, retaining the right to cancel
- Scale and result: 22 percent of dual eligibles switched; all now in plans covering 90–100 percent of drugs
- Isolated: insurers' legal challenges and Washington skepticism limited intelligent assignment to Maine alone
- RECAP and the Lessons of Choice Architecture
- RECAP report: annual itemized drug and fee list, with electronic pricing schedules importable by third-party tools
- Salience as nudge: vivid cost information breaks status quo bias; delay warnings target non-enrollees
- Wisconsin experiment: personal savings letters lifted switching to 27 percent, ten points above brochures
- Private response: Experion's plan assistant and CVS partnership show firms would build RECAP-ready services
- Central lesson: Just Maximize Choices fails in complex markets; more options demand enlightened choice architecture
- Random Defaults and Their Costs
- 11 How to Increase Organ Donations
- The Shortage and Its Bottleneck
- Scale of scarcity: over 90,000 Americans on waiting lists by 2006, growing 12 percent a year
- Supply ceiling: 12,000–15,000 potential deceased donors yearly, but fewer than half become donors
- Leverage: each donor can supply up to three organs, so a thousand more donors saves 3,000 lives
- Real obstacle: consent from surviving family members, not medical or technical capacity
- Explicit Consent: Willing but Not Registered
- The regime: citizens must take concrete steps—license box, donor card—to declare willingness
- Kurtz and Saks (Iowa): 97 percent supported transplantation, yet only 43 percent had checked the box
- Action gap: of those personally willing, 64 percent marked a license and only 36 percent signed a card
- Lesson: registration costs deter otherwise willing donors; inertia favors whatever the default is
- Routine Removal: Effective but Intrusive
- The policy: the state owns body parts of the dead and certain hopeless patients, taking organs without permission
- Evidence: Georgia's routine corneal removal raised transplants from 25 in 1978 to over 1,000 in 1984
- Objection: it violates the widely held principle that people decide what happens to their own bodies
- Presumed Consent: Shifting the Default
- Design: everyone is presumed a donor but may opt out easily—"easily" is what keeps it libertarian
- Shared structure: under both explicit and presumed consent, non-default choosers must actively register
- One-click thought experiment: with trivial costs, Econs would behave identically—Humans still would not
- Johnson and Goldstein: one-click opt-in drew 42 percent donors, opt-out 82 percent, neutral 79 percent
- Austria versus Germany: 99 percent consent under presumed consent against just 12 percent under opt-in
- Limits of the Default Rule
- Not a panacea: donation requires a complete procurement, matching, delivery, and transplant infrastructure
- Spain's lead: nearly 35 donors per million versus just over 20 in the U.S., despite America's opt-in system
- Abadie and Gay: holding everything else constant, presumed consent raises a country's donation rate about 16 percent
- France's blur: technically presumed consent, yet physicians routinely ask families and follow their wishes
- Family refusals: roughly 50 percent in the U.S. without a donor card, about 20 percent in Spain, 30 percent in France
- Mandated Choice and the Power of Norms
- Mandated choice: license renewal requires ticking a donation preference; the application is rejected otherwise
- Procrastination trap: an "unsure" option makes it "mandated choice if you feel like it" and invites delay
- Family deference: an explicit, active "yes" is likelier to be honored than a silent failure to say "no"
- Illinois First Person Consent: 2.3 million registered donors, online sign-up, and family cannot override the wish
- Norms as nudges: the site cites 97,000 waiting nationally, 4,700 in Illinois, 87 percent approval, 60 percent registered
- Visible commitment: links letting people signal civic concern publicly extend the same influence that drives environmental action
- The Shortage and Its Bottleneck
- Incentives, Feedback, Environmental Nudges (12 Saving the Planet · I)
- Environmental Regulation and Choice Architecture
- Command-and-control: mandates preferred technologies or emissions limits, allowing little flexibility.
- Mixed results: regulation has cleaned air but is costly, wasteful, and sometimes counterproductive.
- Global choice architecture: environmental outcomes emerge from consumers, firms, and governments interacting.
- Nudge potential: improved choice architecture might reduce greenhouse gases, though gentle nudges can seem inadequate.
- International focus: attention has shifted to ozone depletion and climate change, with climate lacking effective controls.
- Two Market Failures: Incentives and Feedback
- Externalities: polluters impose harms on others who cannot feasibly negotiate with them.
- Transaction costs: high costs of voluntary agreements justify some government intervention.
- Tragedy of the commons: each dairy farmer gains from more cows while sharing pasture damage.
- Misaligned incentives: polluters pay little for environmental harms their consumption causes.
- Missing feedback: people rarely know or feel environmental consequences; hidden costs lack salience.
- Better Incentives: Taxes and Cap-and-Trade
- Pollution taxes: price harmful behavior so markets determine how to reduce it.
- Carbon tax: higher gasoline taxes would induce fuel-efficient cars, less driving, and lower emissions.
- Cap-and-trade: set a pollution cap and let emitters trade rights to pollute.
- Efficiency and freedom: incentive-based systems are more effective, efficient, and choice-preserving.
- Libertarian paternalism: polluters may continue if they pay the harm; firms favor cap-and-trade for flexibility.
- Political Hurdles and the Acid Rain Model
- Transparency problem: incentive systems expose cleanup costs, making them politically harder to pass.
- Cost illusion: efficiency standards sound free while carbon taxes sound expensive.
- Mental accounting: pair carbon-tax or auction revenues with tax cuts, Social Security, Medicare, or health insurance.
- Acid rain program: 1990 Clean Air Act amendments created emissions trading for acid deposition.
- Proven success: near-perfect compliance, active permit market, and over $20 billion projected savings.
- Feedback, Disclosure, and Information
- Getting prices right: most important but politically difficult when gasoline prices are salient.
- Hidden costs: pollution harms are unseen while pump prices are immediate and visible.
- Nudgelike steps: while waiting for price fixes, use politically palatable feedback improvements.
- Better disclosure: improve information to consumers, making environmental consequences visible.
- Strong motivator: disclosure may seem weak, but information can surprisingly motivate change.
- Environmental Regulation and Choice Architecture
- Environmental Disclosure and Energy Nudges (12 Saving the Planet · II)
- Disclosure as a Regulatory Tool
- Risk labels: FDA, EPA, OSHA, and Congress require warnings for cigarettes, drugs, pesticides, asbestos, saccharin.
- OSHA Hazard Communication Standard: mandates employer hazard programs and training, making workplaces safer without altering behavior.
- Political disclosure: NEPA requires environmental information before major government projects, activating political safeguards.
- Public reaction: if citizens get riled up, government faces pressure; if they yawn, inaction is justified.
- Toxic Release Inventory Success
- Emergency Planning and Community Right to Know Act: 1986 post-Chernobyl law began as modest EPA bookkeeping.
- Toxic Release Inventory: firms report stored or released hazardous chemicals; data is public on EPA’s Web site.
- Unmandated effects: without requiring behavioral change, the law spurred large reductions in toxic releases.
- Environmental blacklist: media and green groups target worst offenders, a social nudge companies fear.
- Competitive avoidance: firms reduce emissions at low cost to avoid bad publicity, lower stock prices, and listing.
- A Greenhouse Gas Inventory
- GGI proposal: government should require major emitters to disclose greenhouse gases.
- Visibility and tracking: the inventory would show U.S. sources and changes over time.
- Likely responses: states, localities, media, and interest groups could act on the disclosed list.
- Cautious optimism: not massive alone, but a low-cost nudge that would almost certainly help.
- Fuel Economy and Green Signaling
- Fuel economy disclosure: since 1975, new cars must meet standards and post large-print expected MPG to promote competition.
- Mapping problem: mileage numbers are opaque; translating them into dollars is more effective.
- EPA label revision: new stickers highlight estimated annual fuel cost and class MPG range.
- Visibility signals: five-year costs, rear-car display, and Prius-only hybrid identity show green signaling matters.
- Split incentives: builders pay for efficiency while owners pay utilities, so homes lack hotel-style key-card switches.
- Making Energy Use Visible
- Invisible energy: people do not know when they are using a lot, so feedback is crucial.
- Ambient Orb: glowing red or green cut peak-period energy use by 40 percent in weeks.
- Wattson device: displays energy use and can transmit data for comparisons with other users.
- Cascade of conservation: Thompson suggests daily consumption visible, perhaps public, could spread.
- Optional competition: public sharing should not be required, but willing conservation contests are unobjectionable.
- Core lesson: making energy use visible nudges reductions without mandating them.
- Voluntary Programs and Profitable Nudges
- Voluntary agreements: EPA asks companies to follow standards, requiring no one to act.
- Green Lights and Energy Star: EPA voluntary programs promoted efficient lighting, office products, appliances via pledges, logos, publicity, awards.
- Theory puzzle: competitive markets should already cut costs, so EPA programs should not have worked.
- Managerial blind spot: busy managers avoid championing small, boring energy savings with little career reward.
- Practical success: both programs diffused low-cost efficient technologies, saving money and reducing emissions.
- Broad lesson: government information can help firms do good and do well, even without mandates.
- Disclosure as a Regulatory Tool
- Part D's Daunting Choice Architecture (10 Prescription Drugs: Part D for Daunting · I)
- Part Iv freedom
- 13 Improving School Choices
- The Case for School Choice
- Friedman's logic: competition among schools improves education; if schools compete, kids win.
- Equity argument: wealthy families already buy school choice, so vouchers put poor children nearer parity with privileged ones.
- Critics' fear: vouchers drain students and money from public schools and may subsidize parents who already pay private tuition.
- Libertarian caution: a general preference for choice does not justify any particular plan; the proof is in the pudding.
- Evidence That Choice Helps
- Competition works: Hoxby finds public schools facing competition produce higher student achievement per dollar spent.
- Measured gains: test scores improve 1 to 7 percent a year, most for younger, low-income, and minority students.
- Lottery winners: students who win seats at better schools score higher and are suspended less than lottery losers.
- Softer verdict: choice is no panacea, but existing evidence shows it holds considerable promise.
- Complex Choices and Mental Shortcuts
- Worcester's failure: of 4,700 students eligible to transfer under No Child Left Behind, exactly one switched.
- "Friendly discouragement": districts qualified their language, stressed program limits, and staged tedious multistage application processes.
- Status quo bias: the known neighborhood school, failing or not, beats an unfamiliar school half an hour away.
- Byzantine data: Charlotte's 100-page booklet, written by schools themselves, omitted locations, test scores, attendance, and demographics.
- Waiter's answer: telling parents "a good school depends on each child" helps as much as a waiter saying "everything."
- Better Information, Better Choices
- Charlotte fact sheet: a simple sheet ranked available schools by test scores and acceptance rates, in the spirit of RECAP.
- Effect on parents: recipients doubled the weight they gave school quality when choosing.
- Resulting choices: selected schools averaged 70 percent higher test scores, matching choices of families earning $65,000 more.
- Incentive Conflicts and Matching
- Controlled choice: Boston's algorithm maximized first choices while prioritizing neighborhood and sibling applicants.
- Strategic misrepresentation: ranking to game the system paid off, since first-choice applicants beat everyone ranking a school lower.
- Who learns the game: affluent, well-networked parents cracked the strategy first; others made the fatal error of listing a popular school second.
- Strategy-proof fix: Roth's mechanism, borrowed from hospital-resident matching, stopped penalizing parents naive about competition.
- Nudging High Schoolers Toward College
- The San Marcos nudge: completing a community-college application became a graduation requirement, tantamount to admission.
- Mapping, not preaching: counselors sold college as Mercedes-versus-KIA earnings rather than high-mindedness.
- Removing friction: free on-site admissions testing, financial aid information, and weekend tax help for parents.
- Results: college-going rose 11 points, to 45 percent; over forty-five Texas schools plus programs in Florida and California followed.
- The Case for School Choice
- 14 Should Patients Be Forced to Buy Lottery Tickets?
- The Hidden Mandate: Buying the Right to Sue
- Bundled liability: every health care customer is forced to buy the right to sue, whether wanted or not
- Public policy bars waiver: courts void malpractice waivers as against public policy — pure nonlibertarian paternalism
- Priced like a deductible: the right to sue is insurance; larger rights cost more, as with FedEx liability limits
- Stigler's fable: if students could sue professors for wrong teaching, education would cost more and teach worse
- What the Liability System Costs
- Direct costs: about 85,000 malpractice lawsuits yearly, totaling $11–29 billion in estimates
- Hospital burden: liability exposure accounts for an estimated 5–9 percent of hospital expenditures
- Defensive medicine: doctors order unnecessary treatments or refuse risky beneficial ones to avoid suits
- Discouraged error reporting: liability fears suppress disclosure, undermining patient safety improvement
- The Deterrence Illusion
- Not experience-rated: premiums don't rise with suits, so financial risk barely pushes doctors to perform better
- Poor fit: fewer than 2 percent of negligently injured patients sue; payouts often go to the undeserving
- Apologies avert suits: because admitting fault prevents lawsuits, the deterrent effect shrinks further
- A Lottery Ticket Nobody Chose
- Erratic awards: pain-and-suffering and punitive damages are unpredictable from case facts
- Negative expected value: roughly 40 percent goes to lawyers, leaving about 60 cents per dollar
- Affordability barrier: bundling suit rights with treatment prices out the sick who cannot pay for both
- Competition compensates: patients who relinquish rights will demand real discounts in return
- Designing Waivers: Libertarian Paternalist Proposals
- Offer both plans: insurers should sell coverage with and without the right to sue
- Procedural safeguards: waivers must be written, explicit, informed — never buried in fine print
- Defaults decide: if waiver is the default, most patients will stick with it
- Ambitious alternative: allow suit only for intentional or reckless wrongdoing, with an option to buy stronger rights
- Other routes: contractual caps on noneconomic damages, or no-fault systems like New Zealand's, Denmark's, Sweden's, and Finland's
- The Hidden Mandate: Buying the Right to Sue
- Getting the State Out of Marriage (15 Privatizing Marriage · I)
- The Privatization Proposal
- Core proposal: the word marriage vanishes from law; no government issues or recognizes marriage licenses
- Civil union: the only legal status states confer — a domestic partnership between any two people
- Private ceremonies: marriages performed by religious and other private organizations, strictly private matters
- Self-governing bodies: within broad limits, marrying organizations set their own rules of eligibility
- No one-size-fits-all: couples choose the organization that best suits their needs and beliefs
- No official endorsement: government stops blessing particular relationships by calling them marriages
- What Marriage Actually Confers
- Legal status: as a matter of law, marriage is no more than an official status with entitlements and mandates
- Tax: large rewards when one spouse earns far more; a stiff penalty when both earn substantially
- Entitlements: laws like the Family and Medical Leave Act cover spouses but not partners
- Death and ownership: estate-tax-free bequests, automatic community property rights mere partners lack
- Surrogate and evidentiary: spouses may decide for the incapacitated and keep marital communications privileged
- The Symbolic Stamp
- Expressive benefits: the state binds material rights to official recognition, which many value most
- Private-only marriage: lacks validation no matter how deep the couple's commitment
- Interracial analogy: offering full benefits under "civil union" instead of "marriage" would itself be unconstitutional
- Official legitimacy: marrying delivers a stamp of approval no private ceremony can supply
- Why License Marriage at All?
- Business partnership model: coauthors contract and cooperate without state-sanctioned oaths or monogamy
- Informal commitments: bind us seriously even when no legal sanction backs them
- The open question: why not treat domestic partnerships like any other partnership and privatize?
- Licensing No Longer Fits
- Discriminatory past: state marriage was long entangled with sexual and racial inequality
- Original function vanished: licenses once legalized sex and childrearing, both now permitted without them
- Policing exit: the institution's real historical purpose was making abandonment difficult
- Precommitment: marriage works like Ulysses at the Sirens — self-binding against impulsive error
- Covenant marriage: only 1–3% choose it, mostly couples already committed; the default stays sticky
- Dissoluble contract: exit is now easy, so official marriage no longer guarantees stability
- The Married/Single Dichotomy
- Two boxes only: the official split imposes real economic disadvantages on the unmarried
- Indefensible inequalities: same-sex partners denied medical decisions and untaxed bequests
- Reality is varied: committed monogamy without marriage; marriages neither intimate nor monogamous
- Let choice rule: leave relationships to individuals and to private organizations' judgments
- The Privatization Proposal
- Privatize Marriage, Nudge Divorce Fairly (15 Privatizing Marriage · II)
- Protecting Children Without Official Marriage
- Children's interests: a legitimate concern, but marriage is a crude protective tool
- Direct mandates: when children's interests are at stake, required support is perfectly appropriate
- Automatic enforcement: absent parents could be auto-enrolled in monthly payment plans
- Equal protection: civil unions and private arrangements can protect children just as well
- Default Rules and the Vulnerable
- Defaults first: good default rules are the best protection for dependents after dissolution
- Steering without coercion: libertarian-paternalist defaults preserve choice while protecting the weakest
- Neither necessary nor sufficient: official marriage does not guarantee good default rules
- Post-divorce asymmetry: women's economic prospects typically fall while men's rise
- Sticky presumptions: rules favoring primary caretakers tend to stick, shielding the vulnerable
- The Licensing Monopoly
- Not mere facilitation: the state does more than permit marriage; it monopolizes the legal form
- Sharp limits: licensing restricts who may enter and how, unlike ordinary contract law
- Unique privileges: only the state confers marriage's material and symbolic benefits
- Liberty cost: for those who prize liberty, this monopoly is hardly an unambiguous good
- Commitments Without a State License
- Commitment without license: people stay bound to friends, churches, and employers without state sanction
- Private enforcement: religious bodies, clubs, and associations bind members through their own rules
- Nothing lost: civil unions and private institutions can still formalize commitment
- Open question: is government licensing under the name "marriage" necessary at all?
- The Balance Sheet and Privatization
- Low benefits: official marriage's contribution to commitment is modest; in many ways an anachronism
- Real costs: unnecessary polarization and bitter fights over fundamental definitions
- Both factions satisfied: religious groups define marriage freely; same-sex couples gain equal legal status
- Separation: marriage should belong to private institutions, not the state
- Room for experiments: privatization increases freedom for individuals and religious organizations alike
- Nudging Couples Toward Settlement
- Unrealistic optimism: nearly everyone believes they won't divorce, so prenups are avoided
- Wealth gap: sophisticated couples get prenups and lawyers; most simply take their chances
- Current mess: divorce law is often unintelligible even to specialists
- Self-serving bias: both spouses and their lawyers see "fair" as skewed their own way
- Impasse: unclear law produces long, intense, life-ruining disputes
- Anchoring formulas: sentencing-like guidelines narrow expectations and nudge couples toward settlement
- Protecting Children Without Official Marriage
- 13 Improving School Choices
- Part V extensions and Objections
- 16 A Dozen Nudges
- Automating Generosity and Paperwork
- Give More Tomorrow: donors commit to modest gifts that automatically rise each year, defeating inertia-driven undergiving
- Pilot evidence: asking donors to increase in two months rather than immediately raised giving 32 percent
- Easy exit: a quick call or email cancels the pledge, so inertia works for charity
- Charity Debit Card: accepted only by charities, it itemizes donations and can file the deduction with the IRS
- Automatic Tax Return: pre-filled returns for non-itemizers could save 225 million hours and $2 billion yearly
- Commitment Devices for Self-Control
- Stickk.com: stake money on a goal; succeed and reclaim it, fail and it goes to charity
- Weaponized stakes: losers' money can be sent to a hated rival's fan club or opposing political party
- Peer pressure: nonfinancial commitments broadcast your progress and failures to family and friends
- CARES: Philippine smokers bank their cigarette money and reclaim it only after passing a urine test
- Gambling self-bans: addicts enroll on state lists barring them from casinos and from collecting winnings
- Reflective control: commitments let the far-sighted self bind the impulsive one in advance
- Small Salient Incentives
- Dollar a day: teenage mothers earn a dollar for each day not pregnant — trivial cost, striking results
- Destiny Health Plan: Vitality Bucks reward workouts, children's sports leagues, and normal blood-pressure checks
- Design lesson: tiny, recurring, immediate rewards can outweigh large but distant ones
- Beyond Bans
- Motorcycle helmets: unhelmeted riders would need a special license, an extra course, and proof of insurance
- Cost without coercion: such hurdles deter some riders while intruding far less than an outright ban
- Air-conditioner filter light: a conspicuous warning prevents costly breakdowns, as oil lights already do in cars
- Rigging Vices to Be Unpleasant
- Bitter polish: products like Mavala make nail-biting immediately unpleasant
- Disulfiram: drinking triggers sickness, disciplining the Automatic System through the body itself
- Precommitment: the Reflective System can buy products that make the bad habit punish itself
- The Civility Check
- Angry-email rule: draft it and wait a day; often you calm down or forget to send it
- Civility Check: software detects an uncivil message and asks whether you truly want to send it
- Stronger default: holds the email twenty-four hours unless you deliberately resend — friction as cooling-off
- Better angels: the Reflective System can be nicer as well as smarter than the Automatic System
- Automating Generosity and Paperwork
- Defending Libertarian Paternalism Against Objections (17 Objections · I)
- The Slippery Slope
- Slippery-slope fear: modest nudges may lead to coercion, bans, even prison terms.
- Cigarette example: warning labels can escalate to taxes and public-smoking bans.
- Direct merits: judge proposals by whether they help people, not hypothetical slide.
- Opt-out safeguard: low-cost exit reduces slope steepness and preserves freedom.
- Inevitable choice architecture: government cannot stand aside; defaults and rules must exist.
- Evil Nudgers and Bad Nudges
- Architect self-interest: choice architects may nudge for their own benefit, not users'.
- Private-sector risk: profit can reward exploiting consumer confusion, e.g., teaser mortgage rates.
- Public-sector risk: officials may be captured by lobbyists and partisan framing.
- Both sectors suspect: worry about all choice architects, public and private alike.
- Rules of engagement: reduce fraud, promote competition, limit interest-group power, align or monitor incentives.
- Transparency as Safeguard
- RECAP disclosure: make fees and usage clear so consumers can compare.
- Environmental disclosure: low-cost monitoring of pollution and emissions.
- Government sunlight: require officials to disclose votes, earmarks, and lobbyist money.
- Brandeis principle: "sunlight is the best of disinfectants"; democracies need more.
- The Right to Be Wrong
- Freedom to err: people may choose badly, and opt-out rights protect that right.
- Learning limits: mistakes can teach, but not by falling into pools or buses.
- Warning signs: unsophisticated choosers benefit from salient cautions without coercion.
- Protect the duped: worry more about borrowers misled into unaffordable mortgages.
- Sophisticated can adapt: investors should evaluate risk, aided by disclosure.
- Punishment, Redistribution, and Choice
- Forced-exchange objection: extreme critics reject any transfer, even to help the poor.
- Redistribution defended: good society trades off protecting unfortunate and encouraging initiative.
- Minimal cost to Econs: nudges help Humans while costing capable choosers little.
- Shared health costs: if the needy impose costs, Econs sharing modestly is fair.
- Required choosing: libertarians prefer active choice, but forced choosing is not always best.
- Drawing Lines and the Publicity Principle
- Lollapalooza sign: "DRINK MORE WATER" nudges hydration without compulsion.
- Effective framing: "more water" and "lose water" use loss aversion better than bland advice.
- Subliminal contrast: invisible advertising steers choices without consent or transparency.
- Publicity principle: visible, defensible nudges differ from covert subliminal manipulation.
- The Slippery Slope
- Transparency, Neutrality, and Paternalist Limits (17 Objections · II)
- The Publicity Principle
- Insidious objection: nudges may let government maneuver people toward its preferred ends.
- Transparency antidote: Rawls’s publicity principle bans policies officials could not publicly defend.
- Respect rationale: secret manipulation treats citizens as tools, not ends.
- Lying analogy: manipulation resembles lying by treating people as means.
- Applied test: Save More Tomorrow, automatic enrollment, defaults, and campaigns should disclose methods and motives.
- Subliminal limits: disclosed subliminal ads may still be objectionable because invisible influence cannot be monitored.
- Neutrality and Ballot Design
- Unavoidable architecture: some choice architecture, and thus influence, cannot be avoided.
- Ballot order effect: first-listed candidates gain about 3.5 percentage points; letting incumbents choose order is unfair.
- Neutrality rule: ballot design should be neutral, often achieved by randomizing candidate order.
- Constitutional rights: voting, religion, and speech require government not favor particular outcomes.
- Selective neutrality: random ballots are good, but defaulting Medicare participants to sensible plans can be good.
- Effects test: evaluate nudges by whether they help or hurt people, not by neutrality alone.
- When Nudging Helps
- Hard choices: nudges most useful for difficult, complex, infrequent decisions with poor feedback.
- Nudger expertise: helpful nudging requires outsiders know better than Nudgees what serves them.
- Preference differences: nudging works when tastes are similar or easily detected.
- Mortgages vs. soft drinks: complex mortgages suit expert nudges; soda preferences do not.
- Self-dealing risk: incompetent or corrupt Nudgers make nudging dangerous; watch special risks.
- Government discretion: ballot design by politicians is bad; expert Medicare defaults may be good with disclosure.
- Asymmetric Paternalism
- Definition: design policies helping least sophisticated people while imposing smallest costs on most sophisticated.
- Libertarian variant: libertarian paternalism keeps costs on the sophisticated near zero.
- Sunlamp timer: require automatic shutoff if switch cost is low and burn risk high.
- Cooling-off periods: mandatory waiting periods protect against self-control problems and high-pressure sales.
- FTC rule: door-to-door buyers get three days to rescind; cost-benefit decides when such laws make sense.
- Divorce and marriage waits: for infrequent, emotional decisions, sober reflection periods make sense.
- Limits of Paternalism
- Nonlibertarian mandates: OSHA bans, Social Security, and anti-discrimination laws override choice and hurt some people.
- Slippery slope: where opt-outs vanish, heavy-handed regulators make slippery-slope concerns real.
- Preference for liberty: favor interventions more libertarian and less intrusive; accept some cooling-off periods.
- Burkean tradition: long-standing practices may embody collective wisdom, but often just reflect inertia, procrastination, imitation.
- Tradition tested: evaluate traditions case by case, not as blanket objection to nudging.
- The Publicity Principle
- 18 The Real Third Way
- Two Core Claims
- Small situational features: tiny social cues can massively shape behavior, whether noticed or not.
- Choice architecture is unavoidable: every arrangement of options guides decisions, for better or worse.
- Libertarian paternalism: nudge people toward better choices while preserving their freedom to choose.
- Broad relevance: the book’s applications in savings, health, marriage and more only hint at possible uses.
- Beyond Government: Private Nudges
- Private choice architects: workplaces, universities, clubs, religious groups, and families can nudge beneficially.
- Libertarian paternalism as practice: small private exercises can improve lives without coercion.
- Creative expansion: understanding nudges should inspire applications in domains the book did not cover.
- A Real Third Way
- Artificial Third Way: twentieth-century talk often remained abstract and failed to resolve polarized debates.
- Real Third Way: libertarian paternalism can break through intractable democratic arguments.
- Middle ground: it avoids both rigid mandates and dogmatic laissez-faire.
- Polarized sloganeering: ordinary people find tired debates increasingly abstract and unhelpful.
- Rethinking Left and Right
- Democratic tendency: after the New Deal, enthusiasm for rigid national requirements and command-and-control regulation.
- Republican response: mandates are often uninformed or counterproductive because one size cannot fit all.
- Freedom as principle: conservatives often define themselves by opposition to intervention and support for choice.
- Sensible Democrats: recognize mandates can fail in a diverse, creative, rapidly changing society.
- Sensible Republicans: accept that markets need government to protect property and enforce contracts.
- Markets as tools: environmental protection, retirement, and aid can enlist markets rather than reject them.
- Freedom With Gentle Guidance
- Distinction: opposing all government intervention differs from insisting intervention promote freedom.
- Good choice architecture: uses incentives and market mechanisms while preserving choice.
- Public officials can nudge: like private actors, they can guide without making the final choice for people.
- Complexity argument: modern life and rapid change undermine rigid mandates and dogmatic laissez-faire.
- Emerging developments: should strengthen both commitment to freedom and support for gentle nudges.
- Two Core Claims
- 16 A Dozen Nudges
- Nudge
- Core Conclusion and Practical Takeaways
- The Core Diagnosis
- Humans, not Econs: real people rely on rules of thumb and predictably err; the rational agent exists only in theory
- Biases are systematic: anchoring, availability, and representativeness produce confident, repeating mistakes
- Inertia rules: loss aversion and status quo bias make people stick with whatever is already pre-set
- Present bias: temptation defeats willpower, so commitment devices outperform good intentions alone
- Social proof steers: conformity and others' choices quietly drive what we buy, eat, and believe
- The Core Design Principle
- Choice architecture is unavoidable: someone always arranges the options, so the only question is how well
- Libertarian paternalism: steer people toward better outcomes while banning no option and coercing no one
- Nudge defined: any feature that alters behavior without forbidding choices or changing economic incentives
- Easy to avoid: a true nudge is cheap and simple to resist
- Golden rule: offer nudges most likely to help and least likely to inflict harm
- Where Nudges Matter Most
- Hard, rare, slow-feedback choices: retirement, mortgages, and health plans offer no learning and need help
- Defaults decide: the pre-set option captures enormous market share through pure inertia
- Small details, large effects: presentation and framing can shift behavior by 25 percent or more
- Asymmetric paternalism: help the least sophisticated choosers at near-zero cost to everyone else
- Practical Tools
- Set good defaults: automatic enrollment, opt-out organ donation, and sensible lifestyle funds exploit inertia for good
- Save More Tomorrow: tie automatic savings increases to pay raises so take-home pay never actually falls
- Make costs salient: RECAP disclosure exposes hidden fees; annual statements deliver a wakeup call
- Give feedback: warn before errors, prevent postcompletion failures, and make invisible energy use visible
- Use commitment devices: stake money, burn bridges, or precommit to bind your future impulsive self
- NUDGES mnemonic: incentives, mappings, defaults, feedback, expect error, structure complex choices
- Mindset Shifts
- The real third way: libertarian paternalism breaks the deadlock between rigid mandates and dogmatic laissez-faire
- Markets as tools: competition can be enlisted for public goals like retirement saving and environmental protection
- Design for Humans: officials and experts are Humans too, so build systems that expect and forgive error
- More choices demand more help: complexity paralyzes; good defaults and clear guidance beat endless menus
- Combine both levers: incentives and nudges together improve lives while preserving freedom of choice
- The Core Diagnosis
opening map…