Annie Duke on the Power of Quitting

Episode Summary

Executive Summary: Russ Roberts and Annie Duke discuss why quitting is a skill, not a moral failing. Duke argues that grit and quit are complementary virtues: perseverance helps when goals remain worthwhile, but quitting is essential when new information shows a path is no longer valuable or safe. Using Everest, poker, startups, cab drivers, and mega-projects, she shows how sunk costs, identity, and goal-setting can trap people in bad decisions.

Main Topics: Quit vs. grit as complementary virtues (Priority: 5/5): Duke argues that quitting and persevering are not opposites but two sides of the same decision. The right choice depends on whether continuing still serves the underlying goal. Moral and practical cases for quitting (Priority: 5/5): She distinguishes ordinary quitting from situations where quitting becomes a moral imperative, such as dangerous Everest climbs or failing startups that should release workers to better opportunities. Mental accounting, sunk costs, and finish lines (Priority: 5/5): The conversation explains how people mentally measure progress from the finish line rather than the start, making them reluctant to stop when they are 'in the losses.' Poker as a model for loss-cutting (Priority: 4/5): Duke uses poker to show how experts fold more selectively, while amateurs overplay hands because of sunk cost fallacy, optimism bias, and difficulty accepting losses. Goals can motivate and distort (Priority: 4/5): Goals help people persist through hardship, but they can also create harmful short-termism, causing people to keep going even when conditions have changed or risks have become unacceptable. Monkeys and pedestals: minimizing sunk-cost traps (Priority: 4/5): The 'monkeys and pedestals' framework advises testing the hardest unknown first before building supporting infrastructure, so people can quit bad ideas before committing too much capital. How quitting can expand optionality and serendipity (Priority: 3/5): Duke emphasizes that saying yes to some uncertain opportunities and quitting others creates room for backup plans, better alternatives, and unexpected human or professional gains.

Key Arguments: Quitting and sticking are the same decision seen from opposite directions; the question is calibration, not ideology. Grit is valuable only when the goal remains worthwhile; when new information makes continuing unsafe or unproductive, quitting is virtuous. In some cases quitting is a moral duty, especially when others depend on you or when continuing exposes people to avoidable harm. People resist quitting because they mentally account from the finish line and dislike closing accounts in the losses. Sunk costs should not determine future action; only the expected value of the next unit of effort or money should matter. Poker illustrates good quitting because professionals fold marginal hands and cut losses, while amateurs overplay weak situations. Goals can improve performance, but they can also trap people into pursuing the wrong endpoint at any cost. The best way to avoid bad quitting decisions is to identify kill criteria early and test the hardest unknown first. Optionality matters: keeping room to explore new paths can reveal better opportunities and protect against change. Quitting is often reversible; many paths can be revisited later, so stopping is not always a permanent door slam.

Data Points: Podcast date: October 27, 2022 - EconTalk episode introduction Everest turnaround time: 1 p.m. - Expedition rule described as the safe cutoff for descent Everest departure time: midnight - Climbers leave Camp 4 to summit Everest predicted time to summit: 3 hours - Expedition leader’s estimate that triggered the climbers’ decision to turn around Cab-driver efficiency gain from rational strategy: 15% more money - Camerer/Thaler study comparing actual behavior to rational driving patterns Cab-driver gain from fixed hours strategy: 8% more money - Alternative benchmark mentioned in the cab-driver study Amateur poker players’ starting hands played: over 50% - Amateurs play too many Texas Hold’em starting combinations in a nine-handed game Professional poker players’ starting hands played: 15% to 25% - Observed range for stronger players Poker table size: 9 players - Used to explain why over 50% of hands is far above a fair share Virgin Hyperloop progress: one-sixth of full speed - Project had not solved the safety/stopping problem California bullet train budget: $33 billion projected budget - Initial 2010 estimate for the high-speed rail project California bullet train bond: $9 billion - 2010 bond floated for the rail project California bullet train revised budget: $80 billion - Budget after problems with mountain tunneling became apparent California bullet train latest projection: over $100 billion / about $115 billion - Later projected cost as described in the discussion Siobhan O'Keefe marathon injury: fibula snapped at mile 8 - Example of goal-driven persistence despite serious injury Marathon distance: 26.2 miles - Implied by discussion of half marathon vs full marathon Half marathon distance: 13.1 miles - Used to illustrate how identical effort can be framed differently

Pivotal Quotes: "The opposite of a great virtue is also a great virtue." — Annie Duke: Core thesis about balancing grit and quitting "Quitting is a skill and it's one that you should get good at." — Annie Duke: Argument against treating quitting as cowardice "We think about waste as a backward-looking problem when it's really a forward-looking problem." — Annie Duke: Explanation of why sunk costs distort decisions

Implications: Listeners should treat quitting as rational portfolio management of time, money, and identity. For careers, startups, and projects, define kill criteria early, test the hardest unknown first, and preserve optionality instead of honoring sunk costs.