By playing, you agree to uListen's Terms and Privacy Policy and YouTube's Terms of Service.
Tim FerrissMaster Economist on Strategic Quitting and Valuable Decisions on the Margin — John List
Summaries, chapters, transcripts, and FAQs on this page are made by uListen, not YouTube.
Great conversations run long. Their best ideas don't have to.
uListen is free on iPhone, iPad, Android, Chrome, and Claude.
See howCHAPTERS
Chapter summaries
Clawback incentives: paying upfront to leverage loss aversion
Tim Ferriss asks John List to define clawback incentives and explain why they can outperform traditional end-of-period bonuses. List describes how giving the reward first and taking it back for nonperformance leverages loss aversion and can raise effort across settings.
- •Clawback incentives reverse traditional bonus timing: reward first, reclaim if goals aren’t met
- •Loss aversion is the behavioral mechanism: giving something up feels worse than not receiving it
- •List contrasts clawback vs standard incentives in employer settings like Lyft
- •He describes using clawbacks in Chinese manufacturing plants and a pre-K setting
- •He also uses clawbacks at home, including with potty training and other behaviors
Incentives with kids: focus on inputs rather than outputs
List explains how he designs incentives for his children, emphasizing controllable inputs over outcomes that are partly luck or context-dependent. He contrasts this with workplaces where outputs are easier to measure and ultimately matter for the organization.
- •He prefers incentivizing inputs (hours studied, books read) over outputs (test scores) for kids
- •Outputs can be unfair when outcomes depend on curves or other people’s performance
- •He uses incentives across schoolwork and sports behaviors
- •He notes the workplace often forces output-based incentives because inputs are hard to observe
- •He uses experiments at home and mentions twins as a natural treatment/control opportunity
Do clawbacks backfire? Durability and side effects over time
Tim raises concerns that clawbacks might breed resentment or stop working with repeated use. List addresses two key risks—depreciation and negative side effects—and explains what the data show so far.
- •Important questions for any nudge: does it keep working, and does it have side effects?
- •List says clawback effects depreciate somewhat but not dramatically
- •Some workers like clawbacks as a commitment device that brings benefits forward in time
- •He connects this to self-control problems in health, education, and other long-term tradeoffs
- •Some participants would even pay to use clawbacks, though others dislike them
Inside a Chinese factory experiment: weekly bonuses and quality output metrics
Tim asks for specifics on how clawbacks were tested in a Chinese manufacturing plant. List details the weekly reward cycle, experimental design, and the output-based metric used to compare clawbacks to standard incentives.
- •The reward period was weekly: bonus given Monday and cleared Friday
- •The study ran about six months and moved people between incentive systems
- •Standard incentive paid after performance; clawback paid upfront then reclaimed if targets missed
- •Metrics were outputs: number of non-defective items produced
- •List reiterates why output metrics are common in worker experiments versus home settings
From fundraising field experiments to tipping: what generosity data can and can’t predict
List tells the origin story of his charitable-giving research, which began as a practical fundraising assignment. He describes how fundraising “rules of thumb” lacked data, motivating field experiments, and then connects these insights to tipping behavior where patterns differ.
- •He began charitable giving research after being tasked to raise money for a new center
- •He insisted on running the fundraiser as a field experiment and requested upfront money
- •Fundraising practitioners offered strong claims (like match ratios) without supporting data
- •He frames his broader agenda as using the world as a lab to study generosity
- •He compares charitable giving patterns to tipping and notes results don’t always transfer
Uber tipping data: who tips, who doesn’t, and how anonymity changes behavior
List explains how tipping was rolled out in the Uber app during 2017 and what the large datasets revealed. He shares key facts about tipping frequency, gender patterns among riders and drivers, and the importance of being removed from face-to-face interaction.
- •He says tipping was introduced partly to help retain drivers during the #DeleteUber period
- •Only a small fraction tip every trip, and many riders never tip
- •Men tip more than women in Uber tipping data, opposite of charitable giving patterns
- •Men tip younger female drivers more than male drivers; the difference shrinks for older drivers
- •Tipping rates are lower when tipping is anonymous and separated in space/time from the driver
Designing tipping prompts: presets, choice architecture, and avoiding “too high” defaults
Tim asks about user-interface experiments like preset tip buttons. List describes how Uber tested different preset configurations and what generally works to increase tipping without triggering zero tips due to unrealistic options.
- •List discusses experimenting with preset options (dollar or percent choices)
- •He says you want users to choose a preset rather than type an amount
- •If preset options are too high, riders may choose zero; too low can leave money on the table
- •He notes the “optimal” scheme may depend on consumer traits and past tipping behavior
- •He recounts turning down Amazon’s chief economist role due to inability to publish findings
The economics of apologies: measuring damage from bad trips and what repairs it
List recounts a personal bad-trip incident that led to research on how poor experiences reduce future spending. He explains how Uber estimated lost revenue from bad trips and tested apology interventions to recover business.
- •List describes an Uber trip error that nearly derailed his keynote timing and prompted a complaint
- •He says bad trips lead to substantial lost future revenue when compared to statistical “twins”
- •He explains why the team used observational matching rather than creating bad trips experimentally
- •Apology treatments included an email apology versus an apology plus a cash coupon
- •He reports that cash with the apology matters, restoring about a third of the negative effect
Nice in the lab, ruthless in the market: reputation effects and real-world behavior
Tim asks about List’s remark that few people won’t “screw each other,” prompting a discussion of lab-versus-market behavior. List describes experiments inspired by his experience as a baseball card dealer, showing how observation and reputational checks change conduct.
- •List’s early exposure to trading came from being a baseball card dealer in the late 1980s
- •He tested whether lab findings about social preferences hold in real market settings
- •When people know they’re in an experiment, they behave unusually nicely
- •In market settings without observation, he says people act more self-interested and exploitative
- •When quality can be certified and reputations checked, nicer behavior returns due to long-run incentives
Critical thinking and theory of mind: anticipating others using “primitives”
List defines critical thinking in two parts—data reasoning and abstract strategic reasoning—and ties the latter to theory of mind. He explains a practical method: identify another person’s preferences, beliefs, and constraints to predict behavior and choose an optimal response.
- •Critical thinking includes avoiding errors in how data are generated and interpreted (e.g., confirmation bias)
- •Theory of mind is the ability to put yourself in others’ shoes to anticipate responses
- •He connects this to game theory and competitive strategy across firms and settings
- •He recommends focusing on decision “primitives”: preferences, beliefs, constraints
- •He gives an example where a strong norm (never tipping) functions as a constraint that limits what incentives can accomplish
Poker as a training ground: randomization and an “exogenous” bluffing rule
The conversation turns to poker as a way to think about prediction and strategy. List describes how humans struggle to randomize and shares a tactic using the second hand of a clock, plus a story about winning a poker bracelet in Australia.
- •List says poker requires controlled randomization, especially for bluff frequency
- •Humans are bad at generating random sequences and tend to force 50/50 patterns
- •He uses an external mechanism (the clock’s second hand) to decide when to bluff
- •He recounts entering a tournament in Alice Springs and winning a bracelet
- •He argues poker has more skill than many people assume, though luck plays a role
Why ‘The Voltage Effect’ book: scaling failures, the ‘voltage effect,’ and replication advantages
List explains why he wrote a book on scaling, tying together audience questions about stalled progress on social problems, his preschool experience, and lessons from working with firms. He also discusses why field experiments can improve credibility and replicability compared to lab-only work.
- •He frames time allocation as opportunity cost and says scaling became his next focus
- •He cites frustration that many interventions show promise but don’t make a dent at scale
- •He describes policymakers dismissing his preschool results as “not scalable”
- •He argues big-data partnerships enable faster replication and learning across contexts
- •He contrasts lab experiments’ limits with field experiments’ realism (people, circumstances, selection into markets)
Scaling as an ‘Anna Karenina’ weakest-link problem: five vital signs and real examples
List explains why scaling is not about finding a single “silver bullet,” but instead avoiding multiple failure modes. He introduces five vital signs for scalability and illustrates them with examples including Jonas Salk’s polio vaccination, COVID vaccination delivery challenges, and teacher hiring constraints.
- •The “voltage effect” is when strong small-scale results shrink to near zero at scale
- •Scaling is a weakest-link problem: scalable ideas succeed in similar ways; failures happen for varied reasons
- •He introduces five vital signs and walks through them using Jonas Salk as an example
- •He contrasts successful delivery systems (routine pediatric visits) with harder adherence problems
- •He notes scaling can fail due to supply constraints, like needing 30,000 high-quality teachers instead of 30
Unintended consequences in scaling: tipping rollout, network effects, and driver-trip incentives
List provides concrete examples of spillovers and unintended effects that appear only after scaling. He discusses how tipping’s apparent wage gains disappeared when everyone received tipping, contrasts that with positive network effects, and then addresses a rider question about strategic driver behavior and how incentives can reduce it.
- •In a beta test, tipping increased participating drivers’ earnings and labor supply
- •After full rollout, increased driver supply offset gains, leaving hourly earnings unchanged
- •He contrasts this with positive network externalities where value rises with adoption (Facebook example)
- •He explains strategic trip rejection behavior and notes a gender difference in acceptance patterns among drivers
- •He describes tools to mitigate bad behaviors: paying for pickup time, cancellation fees, and bonuses based on number of trips
Four ‘after launch’ tools: scalable incentives, marginal thinking, optimal quitting, and culture building
List outlines the book’s “back nine” for sustaining high performance after launch. He explains scalable incentives (including non-cash nudges), marginal thinking with an ad-spend example, reframes quitting as pivoting or calling an audible, and discusses building culture with early design choices that shape equity and cooperation.
- •Scalable incentives include behavioral framing (like clawbacks) and non-financial levers (norms, peers)
- •He cites tax-payment nudges, including a Dominican Republic letter experiment tied to large revenue gains
- •Marginal thinking example: ad channels should be judged by the next/last driver acquired, not historical averages
- •Optimal quitting: his shift away from a golf-pro dream and a coin-flip experiment showing many people are happier after change
- •Culture building: workplace cooperation can spill into community behavior; job ads stating wages are negotiable can change negotiation behavior and wage outcomes
