Megan McArdle on Failure, Success, and the Up Side of Down
What this covers
Megan McArdle and Russ Roberts discuss why failure matters economically and personally, treating it not as an unfortunate obstacle to success but as the mechanism through which people and institutions actually learn. McArdle defends the counterintuitive claim that the real determinant of a society's strength is not how well it avoids failure, but how quickly and thoroughly it recovers from it. The conversation moves across psychology, economics, law, and history to show how survivorship bias distorts our understanding of failure stories; how bankruptcy law, parole systems, and feedback mechanisms either accelerate or impede learning; and why the willingness to try, recover, and try again has become a competitive advantage in uncertain economies.
The discussion draws on specific institutional examples and research findings. Vernon Smith and Bart Wilson's asset-market experiments demonstrate how experience inoculates traders against bubble-thinking. Judge Alm's probation program illustrates how Swift Certain Consistent Punishment—swift feedback that does not cripple—reshapes behavior better than rare severity. McArdle examines labor scarring, Self-Handicapping, and the aversiveness of job search, arguing that systematic process goals work better than willpower alone. A recurring theme is the distinction between Minimize the Downside and recklessness: the conversation advocates planning for failure's costs while maintaining the courage to take intelligent risks. The pair also challenge received wisdom on unemployment benefits, the moral status of bankruptcy, and whether helicopter-parenting creates a generation too risk-averse for a slowing-productivity economy.
McArdle argues that failure is an indispensable source of information and growth at both the individual and societal level, and that societies (and people) succeed not by avoiding failure but by recovering from it well—through resilience, consistent feedback, and institutions like generous bankruptcy that minimize the downside of trying.
- Failure is how we acquire information, mirroring how science learns by disproving hypotheses
- America's generous bankruptcy system and entrepreneurial tolerance of failure are hidden strengths that encourage risk-taking
- Consistent, immediate, non-crippling punishment (and feedback) works better than rare, severe punishment for changing behavior
Failure provides irreplaceable information; we learn what works by discovering what doesn't through testing and trying.
- Failure is valuable because it is how we acquire information; we cannot plan our way around it, and just as science progresses by testing theories that turn out to be false, individuals and entrepreneurs learn what works by discovering what doesn't.
“It's valuable because that's how we get information. We like to think that there is some way that we can just plan our way around failure”
- The fastest way to learn whether a thesis is correct is to try to disprove it rather than to seek confirming evidence, because most theories worth testing are false and confirmation is uninformative.
“You go out and look for new stories that show that there aren't 300 ducks in Chicago. Right? Because that's actually the fastest way to find out whether you are correct or not, is to see if your thesis can be disproven.”
- Duncan Watts argues that when we explain why something like the Mona Lisa is famous, we are not actually identifying causes of its fame but merely describing the object itself, ignoring contingent factors such as the painting only becoming famous after it was stolen in the early 20th century.
“what Duncan Watts, the sociologist, says is we are telling a story why we think the Mona Lisa is famous, but what we are actually doing is just describing the Mona Lisa.”
- In Vernon Smith and Bart Wilson's asset-market experiments, bubbles reliably form among new traders but disappear when the same people trade together two or three times as they develop an instinct that 'I've seen this before'; replacing the experienced traders with new ones brings the bubbles back, showing the information value of having lived through failure.
“every time in these markets they see asset bubbles. And the only way to get rid of the bubble is to have the same people trade together two or three times.”
People protect themselves from failure psychologically by avoiding risk, but this avoidance is more dangerous than controlled risk-taking.
- People engage in self-handicapping—deliberately undermining their own performance (e.g., not trying, getting drunk before a test)—to protect their self-image, because as long as you haven't genuinely tried, failure isn't a referendum on you; experiments show people will even choose a performance-inhibiting drug after ambiguous feedback.
“It's this thing called self-handicapping, which can really go to astonishing lengths.”
- The opposite of failure is not safety, it is nothing; success and failure are the same process, so refusing to try in a risky economy—like GM workers who hunkered down in seemingly safe jobs—can be more dangerous in the long run than taking risks, because the apparent island of safety can be wiped out with no backup plan.
“the opposite of failure is not safety, it's nothing. Success and failure are the same process.”
- Being willing to fail does not guarantee success, but not being willing to fail guarantees you won't succeed; the price of admission to success is the willingness to fail, while smartly minimizing the downside through planning rather than taking stupid risks.
“the price of admission to success is the willingness to fail. That, being willing to fail does not guarantee that you will you succeed; but not being willing to fail guarantees that you won't.”
- The desire for a sure thing is itself the warning sign of danger; con men work by convincing victims they've found a sure thing, and the financial crisis showed that the collective belief that we had found a sure thing was the most dangerous economic position in 70 years.
“What you know from con men is that the easiest way to cheat someone is to convince them that they've found a sure thing. And what we know from the financial crisis is that people who thought sure thing, that was the most dangerous place we've been in as a society”
Experiencing failure makes people more capable and fearless; those who succeed without failure tend to be brittle and risk-averse.
- People who have succeeded too much without experiencing failure tend to be brittle, as illustrated by General McClellan who refused to attack, versus Ulysses S. Grant who had failed in business and the army but was fearless and willing to try despite risk.
“what you see with people who have succeeded too much is that they are often quite brittle. The first time--the famous example of this is General McClellan”
- Once you have failed publicly and spectacularly, you become fearless because the worst has already happened and is behind you, freeing you to take further risks.
“Once you've failed publicly and spectacularly it's like the best thing that can ever happen to you because after that you are fearless. The worst has already happened.”
- Steve Jobs's failure at NeXT—where his obsessive perfectionism made the beautiful computer too expensive and slow to market, so it kept getting superseded—taught him to dial back that tendency somewhat when he returned to Apple, showing that even bull-headed optimists genuinely learn from their failures.
“I think he did learn that he had an overwhelming tendency to do this, to be obsessive and try to make everything perfect. And he did dial that back a little bit... when he went to Apple because he saw how it had come out in Next.”
- Resilience in the face of failure can be taught and self-taught, largely through reframing failure to oneself as a normal part of life rather than a referendum on one's worth, and through reading about others' failures to normalize it.
“you can teach yourself to do it. You can teach yourself to be better at it... Literally, it's not like 10 easy exercises. It's just reframing it to yourself.”
- Carol Dweck's distinction holds that fixed-mindset people treat every challenge as a referendum measuring their innate talent, while growth-mindset people treat each challenge as a chance to learn; the latter aligns with how brains actually learn by strengthening or weakening connections based on whether predictions pan out.
“fixed-minded people think of any challenge as a referendum on them... And growth people, on the other hand, are people who think of each challenge as a way to learn.”
Minimizing failure's downside through planning—not avoiding risk itself—frees people to take career and economic risks rationally.
- When taking a risk, you should minimize the downside by planning for failure—for example, buying only as much house as you could afford if both earners lost their jobs and had to take jobs paying half as much—so that the cushion frees you to take further career risks.
“when we bought our house, we said, Well, what would happen if we both lost our jobs and had to take jobs that paid half as much? That's how much house we can afford.”
- Because U.S. bankruptcy exemptions are set by states, they form a natural experiment showing that states with more generous exemptions also have higher rates of entrepreneurship—prospectively because would-be founders aren't terrified of losing everything, and afterward because failed entrepreneurs aren't shackled to crippling debt that prevents them from starting again.
“the states that have more generous exemptions also have higher rates of entrepreneurship. And this is actually not shocking”
- America has the most generous bankruptcy system in the world—contrary to the assumption that Scandinavian countries are friendlier to debtors—allowing consumers under Chapter 7 to walk in and have their debts discharged; this developed historically when small Western farmers who owed Eastern bankers obtained easy bankruptcy after the 1898 code.
“America has the most generous bankruptcy system in the world. Which is something a lot of people don't know... in Scandinavia it's much tougher.”
Job search avoidance and unemployment trap people; removing urgency worsens outcomes while systematizing effort and lowering wage expectations speeds reemployment.
- Extended unemployment benefits increase unemployment not mainly because people game the system, but because job search produces such anxiety and depression that people avoid it; benefits remove the urgency that would otherwise force them to confront hard decisions like taking a lower-paying job.
“this to me is an explanation for why unemployment, extended unemployment benefits, seem to increase unemployment: is not that people... abuse the system... I think what's happening is that that anxiety and depression is so terrible that they want to avoid it. So they put it off.”
- There is a phenomenon called labor scarring whereby after roughly six months of unemployment, an applicant's ability to be called in for an interview drops radically, meaning that the rational short-term decision to delay difficult job search becomes a terrible long-term decision.
“what we see is something called 'labor scarring,' where people who are in the job market for, it looks like about 6 months, there's a radical drop-off in their ability to get people to call them in for an interview when they send in their resume.”
- Because job search is so aversive that willpower fails, the unemployed should take the decision out of their own hands by systematizing it—setting and ticking off daily process goals (calls, want ads, informational interviews) and focusing on the controllable process rather than the uncontrollable outcome, plus finding peers who share the misery for support.
“make a system and focus on the process, not on the outcome. You can't control when you get a job, but you can control how much time you spend doing things you really don't like doing.”
- Economic research shows three things speed reemployment: lowering your reservation wage (willingness to work for less), spending more time on job search, and labor mobility (being willing to move, which widens the pool of openings)—yet people frequently fail to do these because job search feels miserable.
“who finds jobs faster, people who lower their reservation wage... And people who spend more time on their job search. And then people who move. Labor mobility enhances the ability to find a job”
- In sales, the single best predictor of how much you sell is how many people you talk to; even salespeople who hate cold-calling overcome it by setting numerical contact targets and systematizing the act, because the volume (the denominator) is the foundational driver of results.
“the way they learn to do it is usually by setting numerical targets... Because they know it really is the single best predictor of how many you sell. That's what they'll tell you, is just how many people do you talk to.”
Institutions normalize failure by making it visible and required to confront, transforming it from shameful to expected and manageable.
- Institutions like the military's after-action reports and doctors' Morbidity and Mortality conferences are valuable both because they extract information from what went wrong and because, since everyone has been through them, they normalize the idea that making even serious mistakes is normal and must be confronted rather than shunned.
“because everyone has gone through this--there is no one who is a successful doctor who has not been through a Morbidity and Mortality Conference... Because everyone has been through it, it normalizes the idea that error is obviously something you try very, very hard to avoid... but that having made a mistake, even a very bad mistake, is normal”
- Healthy response to failure requires holding two ideas at once: that the bad outcome does not mean you are a bad or stupid person, while simultaneously recognizing that the specific thing you are doing is not working and must be stopped or changed.
“just because this went wrong doesn't mean that I am a bad, stupid person who is never going to succeed... but at the same time figuring out how to move on, not trying the same thing over and over again”
- Honestly telling others about one's failures, despite the fear of being judged a loser, typically prompts people to share their own stories in return, creating a powerful bonding moment and revealing how many people attach 'the best thing that ever happened to me' even to events like cancer, prison, breakups, or job loss.
“when you do tell the stories and people tell their stories back, it's really this incredible powerful bonding moment. And I wish more people would do it.”
- Judge Alm's program opens by telling probationers 'Everyone in this room wants you to succeed,' which transforms the dynamic from adversarial cops-and-robbers into a collaborative one where offenders control a certain outcome (show up, test clean), giving them a framework of agency that middle-class people grew up with but the criminal-justice population often lacked.
“the first thing he says is: Everyone in this room wants you to succeed. So, what it does is it puts it, instead of in this adversarial cops-and-robbers thing... it gives them control.”
Punishment and policy should enable recovery and reintegration rather than crippling people; consistency matters more than severity.
- Effective punishment should satisfy three conditions: the pain should not be crippling, it should always follow from the behavior (consistency), and it should enable the person to move on by keeping them connected to family, labor market, and community rather than locked away learning to be a better criminal.
“First thing, the pain should not be crippling... Second, it should always happen... And then the third thing is that it should enable you to move on.”
- Judge Alm's HOPE-style probation program, which imposes a brief jail stay for every single violation rather than nothing-nothing-then-a-huge-sentence, has cut the rate of people ending up with prison terms in half, demonstrating that the key to effective punishment is consistency and immediacy, not severity.
“every single time you violate, you are going to jail. But only for a few days. And it has cut the rate of people who end up with prison terms in half.”
- Forgiveness is more important and much cheaper than we usually think; society spends too much time worrying about abuse of second chances and too little about the people whose lives are damaged by punishment, so policy should aim to punish in ways that maximize rehabilitation and minimize damage to both society and the person.
“I end by talking about forgiveness and how important forgiveness is, and how much cheaper it is than we usually think. We usually spend too much time worrying about abuse and too little time worrying about the people whose lives are affected when we punish them.”
- Beyond creative destruction, what matters for a successful society is what happens after people and companies fail; a good society helps them get up, learn, and move on rather than staying stuck in what went wrong.
“a successful society is going to be one that helps them get up, learn what they can, and move on rather than sort of staying stuck in what went wrong”
America's legal and cultural openness to failure—bankruptcy law, startup culture—creates better outcomes than societies treating failure as personal shame.
- Unlike many other countries where failure is seen as personal inadequacy, America—especially Silicon Valley—treats entrepreneurial failure as a resume booster, with some employers preferring to hire people who have failed because they learned on someone else's dime and developed risk-sensing instincts.
“in America it's viewed as a resume booster. And that really is--especially in Silicon Valley--it really is unique and it really is a secret strength of that system”
- High bankruptcy rates, as seen in Memphis (around 1% of the population per year), are largely the price paid for easy availability of credit; while there are downsides like predatory 'Tote the Note' car lots, most people benefit from the system, and it is generally a decent tradeoff.
“mostly what it ends up looking like is that this is the price we pay for free availability--for easy availability of credit. And that's a pretty good tradeoff.”
- Raising children to be so self-protective that they cannot take any academic risk by age 15 is both psychologically damaging and socially harmful, because demographic forces are slowing productivity growth and society now needs more risk-takers and innovators, not fewer.
“we are looking at a lot of demographic forces and other forces that are causing productivity to slow down. So, we need more innovation and people out on the frontier than we did 10 years ago. And we are grooming kids to have less of them.”