Continuing Conversation... McArdle on Failure, Success, and the Up Side of Down

EconTalk Extra
by Amy Willis
McArdle on Failure, Success, a... Diane Coyle on GDP...

Megan McArdle spoke to Roberts this week about her new book, The Up Side of Down. Let's keep the conversation flowing; we love to hear from you, your students, your kids, your friends...

Questions below the fold.

Check Your Knowledge:

1. What does McArdle mean by "self-handicapping," and why does she suggest people use it so frequently? Do you self-handicap? Why?

2. Roberts says, "Sunk costs are sunk, and it's best to move on" in the context of personal relationship failures. In what other circumstances would this be good advice, and why?

3. Roberts pushes back on the value of failure by suggesting "survivorship bias." What does he mean by this, and how might it suggest that the value of failure may not be as strong as McArdle asserts?

Going Deeper:

4. McArdle and Roberts agree that the culture of entrepreneurship differs from the larger society in that failure is honored. What evidence do they suggest to illustrate? To what extent do you agree that this subset of our culture is indeed different? What do you make of their example of Steve Jobs?

5. McArdle points admiringly to the U.S. bankruptcy system as the most generous in the world. Why does she regard the system as "a neat little natural experiment," and what value does she argue it provides?

6. Roberts notes his concern with the notion of expected punishment, as used in Law and Economics. What constitutes expected punishment, and what do Roberts and McArdle think is wrong with this formula?

Extra Credit:

Perhaps individuals are not alone in being susceptible to failure. In 2011, Roberts interviewed another "fan" of failure, Tim Harford. Harford's interview focused more on failure at the macro level, though some of the examples he discussed are also present in McArdle's interview. To what extent is failure recognized in the world of regulatory policy? Which of McArdle's "lessons" on failure apply equally to institutions and individuals? Should failure be tolerated less in such sectors as nuclear power and warfare? Why?

Comments and Sharing

TWITTER: Follow Russ Roberts @EconTalker

COMMENTS (10 to date)
Greg G writes:

I'm not willing to tackle all of these but for:

2. Allow me to nominate foreign military adventures as the arena where sunk costs lead to the most catastrophic decisions.

Greg G writes:

3.Survivorship bias:

When you fail a bunch of times and then hit it big, the story makes your persistence look heroic and the story gets told more.

When you fail a bunch of times and continue to fail, the story makes you look dumb and you tell it a lot less.

As a result the number of failures leading to future failures is under reported.

That said, I do agree with Megan's basic message here. I know myself that I learned a whole hell of a lot more from the bad investments I made than from the good ones.

Michael writes:

#2....professional sports teams do this all the time. Sign someone to a big long term contract and after 2-3 years realize their mistake. Buy out the remaining years and eat the loss rather than let said player continue to be a drag on team performance. Ilya Bryzgalov and the Philadelphia Flyers come immediately to mind.

Chris T writes:

I'll take a crack at a couple of these:

1. Self-handicapping is the mental process you go through to protect your own ego. You throw up another barrier so that you can convince yourself you didn't really try, or that your failure wasn't your fault. It was an externality responsible for the outcome ("blaming the equipment", for example).

I did this in my previous academic professional life a fair bit. Now that I've switched fields, it very easy for me to say things like "I have never done that before, but I think I can learn it with a bit of practice".

3. Another example of sunk costs that occured to me was gamblers trying to win back their losses. They'll be down some amount, but insist that they just have to keep going a little longer, to win it back. But the probability of winning typically doesn't change (the odds favour the house), so the likelihood of the desired outcome is still low, and they end up digging themselves in deeper.

6. The expected punishment in the law-and-order context is simply the expected value of the outcome: probability of the outcome multiplied by the effect. So the chance that you'll be caught, time the punishment. It's bad because most people think only of the probability of getting caught, and then because that's low, consider the payoff. The chance of getting caught in a theft might be low, and also you get to keep a bunch of money that isn't yours.

The way to flip this reasoning is to make sure the probability of the outcome is very high, but the outcome not as bad. This is the goal of the parole scheme that was discussed. The probability is certain; if you violate parole, you go to jail! The fact that the outcome is only a few days jail is irrelevant if you're the kind of person that only considers the probability.

Amy Willis writes:

@Chris T, great answers...But I also found your personal story very interesting. I hear a lot of people say that the academy suffers a lack of risk tolerance and/or taking. Interesting that your private sector (I assume) experience seems so different!

Mark writes:

One might re-cast'survivorship bias' as "history being written by the winners". We'll never hear from the 'losers' of the game.

Pavlov came to mind when listening to this. The importance of closely tying a response to a stimulus remains critical. Perhaps we should frame more regulations with this focus.

Chris T writes:

@Amy -- I think you will find that in academia, it is a very common attitude. People are often discussed in terms of how smart they are, or how "brilliant". Very little consideration is given to how hard someone works to achieve success.

When I left academia, and switched to the business world, it's just taken as given by my colleagues that I have never encountered most situations or techniques. They do, however, know that I'm quite technically gifted, and can learn quickly on my own, so I've had some small success making the transition. Not surprisingly, as with most thing, business just takes a bit of practice. I found just getting in the door was the hardest part!

Thomas A. Coss writes:

I would rather take a beating. A beating would be short and I would be the only one hurt, few would know and healing likely brisk and private, yet now I'm in my third start-up.

Where I live failure is less honored than ignored, a single tailed distribution where those upon whom success has smiled, share their individual surprise in local success bias seminars to those eager to join them.

I'm not certain I'd have it any other way. Despite the abundance of often insipid, "we're looking to invest in the next Google" (I actually had someone tell me that)and MBA trained advice equally impractical, it remains a good beating.

I'm regularly struck by the large gap between the academic ideal and resource constrained reality of where the entrepreneur must live. In this B-schools have much to learn. Still arrogance is no anecdote to self-doubt, I choose humble appreciation for the business environment in which I seek to succeed.

Mike F writes:

Extra Credit: Regulatory failure is recognized quickly and loudly by political forces not aligned with group responsible for regulating an industry. In general, there are so many dimensions associated with an industry that can be rated high to low (eg. safety, speed, cost, competitiveness, comfort, effectiveness, privacy, amenities, ease of use) that some dimension(s) can always be rated poorly or as failing. Everyone agrees that catastrophic failures should be avoided, even at very high cost. So catastrophic failures in the nuclear power industry should have extremely high costs. Ferry boats sinking should have very high costs. But not all errors are catastrophic. Some risks need to be encouraged in the interests of progress. Improvements in the nuclear power industry to improve safety, security, disposability, yield should all be encouraged, even though changes will all have some risk of failure associated with them.

Kurt writes:

Hi Russ,

Thanks for your excellent library of econtalks.

1 recommendation - you should feature more prominently the top/most voted econtalks for the year. They are kind of hard to find and that content is amazing.

That should be a featured content to attract more website hits.

Kind regards,


[Nice idea, Kurt. I've added a link to the Favorites category in the left-hand side bar on every page. That may help. --Econlib Ed.]

Comments for this podcast episode have been closed
Return to top