Megan McArdle’s “The Up Side of Down”: Book Review, Notes + Analysis

Poor Ash’s Almanack > Book Reviews > Effective Thinking

Overall Rating: ★★★★★★ (6/7) (standout for its category)

Per-Hour Learning Potential / Utility: ★★★★★★ (6/7)

Readability: ★★★★★★★ (7/7)

Challenge Level: 1/5 (None) | ~270 pages ex-notes (320 official)

Blurb/Description: Journalist Megan McArdle delivers an eminently readable yet surprisingly thorough, empirical, multidisciplinary, research-backed exploration of the important topic of failure: preventing it and picking ourselves up from it.

Summary: Over time, I’ve gravitated away from “pop” business books, because a disproportionate number of them seem to suffer from the issues discussed in Rosenzweig’s “The Halo Effect” (Halo review + notes) and Tetlock’s Superforecasting (review + notes) – a hedgehog-like focus on one big idea that fails to account for counterfactuals and survivorship bias.

Megan McArdle’s “The Up Side of Down” is a notable exception: it was the book I was reading at the moment I decided to quit my job and launch ACM, and it holds up to repeated scrutiny.

With the topic of failure as a linking (though not overly constrictive) organizing principle, McArdle journeys through psychology research, case studies ranging broadly from parole systems in Hawaii to bankruptcy in Tennessee, and her personal experiences in an accessible, engaging, humorous way.

By her own admission, Megan McArdle has failed a lot in her life… but she took her own advice and got unstuck, writing the masterful The Up Side of Down along the way. Photo credit: The Washington Post

Highlights: This is above all a practical, useful, mental-models-rich book that everyone should have on their shelf.  McArdle has done a ton of research (see the notes) and synthesizes a lot of important concepts, applying them across a lot of different domains, coming to a lot of very smart conclusions.  These conclusions are communicated effectively, too: McArdle is a very capable, very engaging writer.

While McArdle unfortunately isn’t likely to be invited to the cool-kid table and spoken about with the same hushed reverence as Taleb gets from all his groupies, the difference between something like Antifragile (AF review) and The Up Side of Down is that Antifragile is only useful as kindling, whereas The Up Side of Downactually has intellectual merit.

The discussion of the parole system in Hawaii – the culmination of earlier threads including medical failures and the swiss cheese model of causality – provides extremely insightful and usable takeaways vis-a-vis process vs. outcome, clear and consistent feedback, etc.

Lowlights: There’s not much to dislike… there are a few sections that are of less interest than others, and some places where McArdle could synthesize other concepts but doesn’t bring them up, which are only reasons this is 6 and not 7 stars (hopefully you can tell I’m a really big fan of this book).  

In places I can (and do) quibble with some of her conclusions and presentation – for example, in the introduction, she seems to allege something she doesn’t actually seem to believe – but that’s standard fare for any book and not a demerit.  The one modestly annoying bit is that McArdle sort of equivocates on the definition of “failure” throughout the book. It still feels totally cohesive, though.

Mental Model / ART Thinking Points: confirmation biasschemasunk costscommitment bias,growth mindsetagency / learned helplessnessarms racemargin of safetylocal vs. global optimizationincentivesfeedbackhindsight biasmulticausalityloss aversionprobabilistic thinkingloss aversionendowment effectutilitysocial proofoveroptimismtrait adaptivity

You should buy a copy of The Up Side Of Down if: you have a pulse.  

Reading Tips: None in particular relating to the book; however, this book is one of those you’ll get a lot more out of if you go down the rabbit hole and actually some of the books noted below.

Pairs Well With:

Misbehaving by Richard Thaler ( M review + notes).  This is the definitive tour de force through cognitive biases and behavioral economics: it’s funny, insightful, and provides multiple levels of learning.

Deep Survival by Laurence Gonzales (DpSv review + notes).  McArdle references this book vanishingly briefly; it deserved a more extended comparison, because there are a lot of parallels between her discussion of certain phenomena and Gonzales’s.  Gonzales seeks to answer, via research, the question: in life-or-death situations, who survives and why?

To Engineer is Human” by Henry Petroski (TEiH review + notes).  Petroski addresses failure from an engineer’s perspective, providing a concise and thoughtful exploration of margin of safety.

The Design of Everyday Things by Don Norman (DOET review + notes).  Norman takes the swiss cheese model of causality, among topics, and heads off in a different direction than McArdle: how can we design systems to minimize failures?

 Superforecasting by Philip Tetlock ( SF review + notes) and “ The Success Equation by Michael Mauboussin ( TSE review + notes).  McArdle touches on Tetlock’s research, as well as the idea ofluck vs. skill and process vs. outcome, and these books serve as great introductions thereto.

Reread Value: 4/5 (High)

More Detailed Notes + Analysis (SPOILERS BELOW):

IMPORTANT: the below commentary DOES NOT SUBSTITUTE for READING THE BOOK.  Full stop. This commentary is NOT a comprehensive summary of the lessons of the book, or intended to be comprehensive.  It was primarily created for my own personal reference.

Much of the below will be utterly incomprehensible if you have not read the book, or if you do not have the book on hand to reference.  Even if it was comprehensive, you would be depriving yourself of the vast majority of the learning opportunity by only reading the “Cliff Notes.”  Do so at your own peril.

I provide these notes and analysis for five use cases.  First, they may help you decide which books you should put on your shelf, based on a quick review of some of the ideas discussed.  

Second, as I discuss in the memory mental model, time-delayed re-encoding strengthens memory, and notes can also serve as a “cue” to enhance recall.  However, taking notes is a time consuming process that many busy students and professionals opt out of, so hopefully these notes can serve as a starting point to which you can append your own thoughts, marginalia, insights, etc.

Third, perhaps most importantly of all, I contextualize authors’ points with points from other books that either serve to strengthen, or weaken, the arguments made.  I also point out how specific examples tie in to specific mental models, which you are encouraged to read, thereby enriching your understanding and accelerating your learning.  Combining two and three, I recommend that you read these notes while the book’s still fresh in your mind – after a few days, perhaps.

Fourth, they will hopefully serve as a “discovery mechanism” for further related reading.

Fifth and finally, they will hopefully serve as an index for you to return to at a future point in time, to identify sections of the book worth rereading to help you better address current challenges and opportunities in your life – or to reinterpret and reimagine elements of the book in a light you didn’t see previously because you weren’t familiar with all the other models or books discussed in the third use case.

Page xii: McArdle opens the book with the classic saw: what’s the secret to success?  Good judgment. How do you get good judgment? Bad judgment!

McArdle notes that one of our biggest societal fears is failure.  (She gets to the growth mindset momentarily…)

Pages xiv – xv: McArdle here notes a mixture of things that I agree and disagree with, although in some senses I view it as a matter of semantics.

With the context that kids are sent out into the world all but wearing bubble-wrap, McArdle alleges that:

“since we cannot succeed simply by not failing, we should stop spending so much energy trying to avoid failure or engineer it away.”  

I don’t think this is literally true; Munger et al note that the best way to manage risk is, you know, to avoid it: margin of safety and all.  McArdle, in fact, even demonstrates this herself later, by buying a much smaller house than she could theoretically afford (we’ll get to this).

So I think what McArdle is getting at here is more one of the points Henry Petroski gets at in To Engineer is Human (TEiH review + notes) or Gonzales cites in “ Deep Survival (DpSv review + notes) – if you only do things that are perfectly safe, you’ll never do anything super interesting.  (Opportunity costs.)  And, as Petroski notes, it’s from pushing the boundaries that you learn new things.

So I think it’s in that context that McArdle advocates, Silicon Valley-like, that we should:

 “encourage people to fail early and often – by making sure that their failures are learning opportunities, not catastrophes.”  

McArdle notes that it’s important to learn to “identify mistakes early, so they can be corrected.”  (The First Rule of Holes is Stop Digging.)  She cites business examples of sunk costs and commitment bias; I would add to this, topically, Thaler’s discussion of “dumb principals” in Misbehaving (M review + notes).

McArdle also goes directionally Don Norman, noting that we should “overcom[e] our natural instincts to blame someone” – although she argues this less in the Norman sense of “the design/system is the problem,” but rather in the sense of “mistakes are unavoidable, but we should learn from them and move on.”

Remember Norman’s perspective from The Design of Everyday Things (DOET review + notes):

“How is it that so many people are so incompetent? Answer: they aren’t.  It’s a design problem […]

physical limitations are well understood by designers; mental limitations are greatly misunderstood.  

We should treat all failures in the same way: find the fundamental causes and redesign the system so that these can no longer lead to problems.”

And, of course:

It's not possible to eliminate human error if it is thought of as a personal failure: if the system lets you make the error, it's badly designed. If the system induces you to make the error, it's really badly designed. - Don Norman Click To Tweet

Page 2: An interesting example of culture as well as constraints: Kindergarteners beat engineers at building a marshmallow-supporting structure… of course, their designs were terribly unaesthetic and probably not very durable and wouldn’t exist in the real world.

Page 5: McArdle here discusses the role of dopamine in mediating a lot of things, from learning to addiction.  She goes into it more later; the intriguing bit here – related to defensive pessimism on a neurological level – is that dopamine levels actually rise in anticipation of something happening and plummet if they don’t.  (Combined with loss aversion, this is the basis for defensive pessimism.)

See the  margin of safety mental model, as well as Shawn Achor’s discussion of this phenomenon in Before Happiness (BH review + notes).  As Achor states somewhere in pages 96f – 101:

“expectations create brain patterns that can be just as real as those created by events in the real world.”

Pages 7 – 11: McArdle here cites a paradoxical phenomenon that is totally true thanks to feedback x  incentives: highly intelligent kids tend to be bad students because they can coast on natural ability and never have to work.  (This shows up in the NFL too; sometimes it can backfire on athletes who never have to learn, for example, to run precise routes because they’re faster and bigger than everyone else.)

McArdle cites Carol Dweck and the growth mindset, which is phenomenally important, although I would note Dweck’s “Mindset” (Mndst review) is a mediocre book at best because unlike McArdle, Dweck tries rather unconvincingly to pretend that natural talent is mostly nonexistent, similar to how Mukherjee tries to pretend in “The Gene” (TG review + notes) that IQ / “g” doesn’t exist.

McArdle reviews the highlights of the growth mindset and also brings up the idea of “impostor syndrome,” which is the opposite of fundamental attribution error (and, interestingly, people can display both at the same time… it’s a bit context-dependent.)  McArdle also here reviews agency / learned helplessness and cites Dweck’s research on the priming effects of putting people in a fixed vs. growth mindset.

These are real phenomena, by the way, that are critically important to understand.

Many experiments show that being exposed to fatalistic stimuli renders us ineffective; one I interpret this way was overviewed in Olds/Schwartz’s The Lonely American (TLA review + notes).

College students were put together in groups for 15 minutes then randomly assigned to be accepted or rejected by the group… then, one group is told that they’ll be alone for life and won’t have successful/lasting relationships.

The experimenters (Twenge and Baumeister) found that this led to self-defeating behavior such as being “more likely to […] procrastinate […] when given the opportunity to prepare for a test.”  There was a decline in effort on cognitive tests and a state of mind that “avoids meaningful thought […] and is characterized by lethargy.”  And quitting sooner on challenging tasks.

Although Olds/Schwartz cite focus on the social exclusion angle, I view it as a clear example of learned helplessness (with potentially other factors interacting).  If you were told – with a dose of authority bias tinting your perception – that you’ll be alone for the rest of your life (with the implication being that you can’t do anything about it), that’s a pretty strong fixed-mindset, anti- agency, pro- learned helplessness framing.

And Martin Seligman’s 1972 paper on the topic observes:

In dramatic contrast to a naive dog, a typical dog which has experienced uncontrollable shocks before avoidance training soon stops running and howling and sits or lies, quietly whining, until shock terminates.

The dog does not cross the barrier and escape from shock. Rather, it seems to give up and passively accepts the shock. On succeeding trials, the dog continues to fail to make escape movements and takes as much shock as the experimenter chooses to give.

Page 13: McArdle notes that our education system subtly reinforces the idea that smart people don’t make mistakes; she cites the example of science class, where current theories are presented as obvious facts and old theories are presented in a way that makes the proponents seem hilariously out of touch with reality.

Reality is, of course, a lot more complicated; plenty of really smart scientists went down rabbit holes.  There are lots of great examples of this. Two of my favorite books touching on the topic are Richard Rhodes’ The Making of the Atomic Bomb (TMAB review + notes) and David Oshinsky’s Polio: An American Story (PaaS review + notes).  In both cases, the progress of science was actually halted by very, very smart scientists making mistakes.

McArdle isn’t the first to make this observation.  Thomas Kuhn’s The Structure of Scientific Revolutions (Kuhn review + notes) raises a similar idea.  Kuhn believes that textbooks hide the existence of revolutions by only presenting ideas from previous paradigms insofar as they are still valid or useful.

Pages 17 – 18: McArdle hilariously riffing on the arms race of early childhood education: if you thought that high school resume padding for Ivy League applications was bad… two-year-olds are now being queried about their aspirations.  McArdle’s friend:

“Right now we’re working on not eating used gum off the street.”  

Page 20: McArdle here notes (as she’ll go into more later) that one of the differences between being rich and poor is whether mistakes are fatal; in other words, growing up with some means provides a margin of safety – specifically a redundancy; the sort of “alternate load path” that Petroski discusses in TEiH.

This is easy to visualize and worth thinking about in a little more detail: for most readers of this website, a flat tire or other minor car trouble would be an annoyance (perhaps even a major one), but probably wouldn’t have any effect on what you eat for dinner any day this month.  On the other hand, people living paycheck to paycheck have no alternate load path.

This is one of the reasons Thaler notes in Misbehaving (M review + notes) that those living in poverty do less mental accounting and are more rational (in a narrow sense) in the way they view fungibility.  They don’t have the luxury of having a mental bucket for “car expenses” and “food expenses” – if they have unexpected car expenses, it has to come out of one of the other proverbial mason jars.

It’s probably easy for most readers of this website to view this as an “other people problem,” but research demonstrates that a surprising number of Americans – even highly-educated, highly-earning white-collar professionals – are underprepared for financial adversity.  

Page 23: McArdle points to save-games in video games as a great example of being able to practice parts that give you trouble… thereby enhancing learning thanks to clear and consistent feedback, a theme throughout this book.  I’ve definitely had this experience.

Page 28: Interesting discussion here of a lab simulation of the California blackout.

Page 30: This experiment by Vernon Smith on “equilibrium prices” contradicts newer research on the Coase theorem due to the endowment effect – see Thaler’s “ Misbehaving ( M review + notes).  I’d be curious to square these two up… but it’s not a burning desire.

Pages 37 – 38: Really fascinating discussion here of attempted experimental creations of markets; McArdle concludes that a prerequisite for markets is trust, which experimentally is a function of relationships / social connection.

The take-home is that there’s a big local vs. global optimization problem here: the conditions that are universally global can’t necessarily be reached by stepwise hill-climbing.  In other words, capitalism may be the system that works best, but it’s not necessarily the easiest to achieve because of the valley you have to go through to get there.

A somewhat similar trend, by the way, plays out with regard to traffic in the third world.  According to most of my friends who’ve been there, traffic rules are rarely followed.

It is obvious to most Westerners that traffic functions more efficiently and safely if everyone follows the rules… I’m not talking about occasional speeding; I’m talking about stopping at red lights and such: basic things that seven-year-olds know you’re supposed to do.

But obviously if nobody else is following the rules, it’s not locally adaptive to follow the rules… and there’s not always a clean/easy path from that local quandary to the global rules that allow capitalism to prosper.

This is basically the point.

Page 39: Among certain circles (there’s probably a decent degree of overlap with egalitarianism), it’s popular to point out that we’re not that far removed from monkeys.  (The Moral Animal, etc.)

Well, the flip side is nonlinearity: that supposedly small difference is actually pretty damn large.  McArdle here notes that chimps can’t collaborate very well… humans collaborate very well.

Page 41: There’s a really intriguing angle here on the difference between agrarian and hunter-gatherer morality: McArdle notes that for hunter-gatherers, sharing is utility-maximizing and provides a margin of safety: it’s more important to always have food than to have a huge surfeit at certain times and none at other times (which is the nature of hunting, an inherently probabilisticendeavor where luck can trump skill sometimes.)

Page 44: McArdle summarizing some experimental research: people tend to share when the payoffs are very uncertain, and tend to keep their own when the payoffs are very linear.  This squares up well with loss aversion and marginal utility.

Page 46: McArdle suggests that modern morality (and thus, in some senses, our view of failure) is driven by the fact that relative to hunting, farming provides more predictability but also more constraints.

Page 49: McArdle cites the different culture in the U.S. and Europe regarding entrepreneurship… societal perception of failure is part of it; the other part (as we’ll see later) is liability law.

Intriguingly, sort of like how both conservatives and liberals are inherently intellectually inconsistent relative to libertarians and populists, the U.S./Europe have dynamically inconsistent views: in the U.S., entrepreneurial success is viewed as having a lot of luck involved but poverty not so much; vice versa in Europe.

Page 50: McArdle does make the Don Norman point here of the design problem in California’s energy rules; she also notes (as we’ll see soon) that:

“markets work best when formal rules are supported by a strong set of informal moral rules — [a] cultural operating system.”

Pages 54 – 55: McArdle here notes that predicting big movie hits and flops is very difficult; it’s easy with hindsight bias but not ex ante.  This is a similar discussion to Mauboussin’s wonderful discussion of the Music Lab experiment on pages 128 – 131 of “ The Success Equation ( TSE review + notes).

Page 58: McArdle here references Tetlock’s research on predictions (see “Superforecasting” (SF review + notes) – it’s worth it.)  She notes that one of the bigger points isn’t “experts suck,” but rather “wow, the world is complex and probabilistic.”

She further notes that contrary to popular opinion, engineering problems are “easy” – it’s unintuitive, but as one of my professors once said, you don’t see a lot of planes falling out of the sky.  On the other hand, a lot of problems that seem “easy” – like getting people to take their medication – turn out to be quite difficult.

McArdle doesn’t reference Sunstein/Thaler’s “Nudge” here (NDGE review + notes), but she should: it works toward a solution of that problem.

I’d also note she’s not the only one to point out this phenomenon – it’s a true/valid point in many disciplines.  For example, in Peter Thiel’s Zero to One (Z21 review + notes), he states:

“people overestimate the relative difficulty of science and engineering, because the challenges of those fields are obvious.  What nerds miss is that it takes hard work to make sales look easy.”  

Pages 59 – 62: McArdle here references storytelling in reference to Gladwell’s “The Tipping Point” – she cites a sociologist named Duncan Watts.  It turns out Watts did Music Lab. Watts, by the way, wrote a book called “Everything Is Obvious” (which is now on my shelf… haven’t read it yet.)  This is a review of the MusicLab experiment including social proof and path-dependency and hindsight bias, though I prefer Mauboussin’s explanation.

Page 63: McArdle brings up the idea of a/b testing here… noting the ethical challenges in running these sorts of experiments in many situations.  It’s a real problem; see Oshinsky’s Polio: An American Story (PaaS review + notes) for an example thereof.

Page 69: McArdle cites Jim Manzi, who wrote “Uncontrolled (now also on my shelf, unread as of yet.)  The problem is that just because an experiment demonstrates something but that doesn’t mean it’s true.  He notes similar replication problems to the famous Ioannidis paper “Most Published Research Findings are False.”  See my notes to pages 147 – 155 and 161 of Jordan Ellenberg’s “How Not To Be Wrong ( HNW review + notes) for more on this topic.

McArdle provides the example of “liar loans,” which were, at the time, reasonable to make based on historical data… and then all of a sudden the trend wasn’t your friend.  This is related to both sample size and context-dependency.

See also the backtest overfitting problem.

Page 71: McArdle references the New Coke fiasco as an example of when experiments didn’t pay off.  See “For God, Country, and Coca-Cola” if you like.

Pages 73 – 74: Why?  McArdle answers: the question that’s tested isn’t always the one that matters: a more institutionalized version of the substitution issue where we have a cognitive tendency to oversimplify and answer easier questions.  The famous Pepsi “taste tests” were wrong because they only measured which drink people preferred in small doses, not which one tasted better in large quantities and drunk repeatedly (which history has documented thoroughly is Coca-Cola, for most people, but not all – one of my best friends is a Pepsi guy!)

Great example of  dose-dependency.

Similarly, McArdle notes that a ‘healthy food’ program in a school district performed well in small tests, but flopped in the real world because kids had more choices than they did in the lab, and also supply chain / logistics problems.

Pages 76 – 77: In a section subtitled “Learning the Right Lesson,” McArdle points to a “third way” – it’s not about blindly accepting experiments or proclaiming them useless – but rather, realizing they’re an iterative process.

She also cites the famous IBM (maybe apocryphal?) story about a young man who Tom Watson didn’t fire because the man’s mistakes were a few million bucks worth of education.  

Again, this isn’t how most businesses are run… see pages 188 – 190 of Thaler’s “ Misbehaving ( M review + notes) for how hindsight bias, among other factors, causes executives to disproportionately punish those who make mistakes.

Pages 79 – 81: So this is rare, but I think McArdle does something better than Don Norman.  In Norman’s god-level “ The Design of Everyday Things ( DOET review + notes), he presents a very technical, detailed categorization of “mistakes” and “slips” that quite honestly makes my eyes glaze over.  I’m sure it’s useful… it’s just not easy for me to remember or apply in everyday life.

McArdle provides a much more helpful, memorable three-part categorization.

An “accident” is something unpredictable – “wrong place, wrong time” – you’re walking down the street on a clear day and lightning strikes you out of the blue.  Basically, there are no useful lessons to learn.

A “mistake” is when there’s something you should have done differently… you did something you shouldn’t have.  For example, forgetting to check left before crossing the street.

A “failure” is when a mistake leads to a tangible bad outcome.  For example, getting run over by a bus, like Alex Rogo’s dad. (That’s an inside joke that literally nobody besides me will laugh at.)

McArdle does note the idea of checklists here; see Gawande’s “ The Checklist Manifesto ( TCM review + notes) for more on this.

Pages 84 – 85: McArdle discusses the swiss cheese model of causalityhere, similarly to Norman in DOET.  She notes that our systems – highways, for example – are typically designed with lots of margin of safety.

This leads to the counterintuitively bad n-order impact of not providing clear and consistent feedback when we make mistakes, which means we can keep making them (without consequences) until one day they suddenly matter…

She comes back to this topic several times.  I touch on a similar point in my  salience  mental model, examining how the  n-order impacts of good decisions can lead to problems becoming less visible… until they’re problems again.

Pages 88 – 89: McArdle cites Perrow’s “Normal Accidents – which is on my shelf with a high priority.  As with Gonzales in “ Deep Survival” ( DpSv review + notes), she references the “tightly coupled” nature of complex systems.

Pages 94 – 95: McArdle here provides a nice visual bullet-point list of the ten “causes” of failure in her mother’s case; she concludes by noting an example of salience / vividness: she’d written about medical errors before, but it didn’t prepare her for seeing her own mother in the hospital.

Page 97: She brings it home to a discussion of process vs. outcomesalience, etc:

“As I discovered when I myself had to spend ten days administering IV antibiotics at home, the reason that handwashing is so hard to do consistently is that it’s not actually that risky to forgo it.  The odds that any one slip will cause an infection are extremely low, well under 1 percent.

And since it’s tedious and often must be done multiple times while touching a single patient, it’s very tempting to skip it sometimes.  Over thousands of repetitions, this kills people.

But most of us don’t judge our actions over thousands of repetitions.”

She actually comes back to this idea of making payoffs more certain later (Hawaii prison).

Pages 100 – 101: McArdle highlights the Van Halen brown M&Ms bit… it wasn’t arrogance: it was ensuring attention to detail for complex, potentially dangerous shows.  This actually reminds me a lot of Petroski’s discussion of designed failure – “leak before break” – on page 119 of “ To Engineer is Human ( TEiH review + notes).

Pages 102 – 103: McArdle highlights an n-order impact of antibiotics: they allow us to be less vigilant.  She also circles back to the point discussed earlier about the pitfalls of not receiving clear and consistent feedback.

Page 106: Interesting bit on hedonic adaptationcontrast bias, and a few other things: most people, when asked, will say that the best thing that ever happened to them involved a catastrophe.  See also the Stephen Colbert GQ profile – “it would be ungrateful to not take everything with gratitude.”

Page 108: McArdle talks about her own breakup… interestingly, her boyfriend’s attitude was similar to the detrimental one discussed by Tavris/Aronson in “ Mistakes Were Made (but not by me) ( MwM review + notes) or, to a lesser extent, Olds/Schwartz in the less-good “ The Lonely American ( TLA review + notes): the tendency of some people to form relationships of “convenience” rather than “commitment,” and the fixed-mindset approach of not working to improve.

McArdle here provides a good example of loss aversion that all of us have felt at one point or another…

Pages 112 – 114: McArdle cites something called “normalcy bias” – the act of pretending everything’s fine when the plane isn’t crashing.  It’s not entirely clear whether or not this is any different than status quo bias in a special circumstance… anyway, the point is that people have an astonishing tendency to mill about and not react when their world is literally on fire.

McArdle cites Amanda Ripley’s “The Unthinkable: Who Survives when Disaster Strikes – And Why”  – which is now on my shelf, not yet read. 

On the topic, Gonzales’s “ Deep Survival ( DpSv review + notes) is fantastic.

Page 115: McArdle quips on lake wobegon syndrome, i.e. overoptimism:

“There’s a scientific name for people with an especially accurate perception of how talented, attractive, and popular they are – we call them ‘clinically depressed.”

McArdle notes that of course, this optimism is usually an adaptive trait, but in some circumstances it can become non- adaptive.

As I discuss in the notes to “ Deep Survival ( DpSv review + notes), this overoptimism is a trait with dose-dependency.

Pages 121 – 124: McArdle here takes a brief detour into economics, touching on sunk costs and utility as well as how loss aversion frames schema.  She also notes the fact that we’re risk-seeking when it comes to losses and risk-averse when it comes to gains.  See Thaler’s phenomenal, otherworldly “ Misbehaving ( M review + notes) for the best overview of cognitive biases and basic economics around.

Page 128: McArdle notes social proof, which, again, is usually an adaptive trait, but can lead to bad outcomes, such as participating in a housing bubble.  She doesn’t go super deep into it here; see Tetlock’s “ Superforecasting ( SF review + notes) for more on how groups can function badly… or well.

Page 134: McArdle here notes all journalists dream of “covering a story that might actually get you shot.” Reminds me a bit of Dan Harris’s commentary in several places in “ 10% Happier” ( 10H review + notes).

Page 136: McArdle here cites Hallinan’s “Why We Make Mistakes” (WWMM review), which I was excited to read, but it proved very disappointing… not recommended.

Pages 138 – 139: IS IT THE GORILLA EXPERIMENT???  BY GOLLY YES IT’S THE GORILLA EXPERIMENT.  Another confirmation of Samir’s First Law of Psychology Books: they all must reference at least one of the gorilla, marshmallow, or Milgram experiment.  (There are plenty of counterexamples but I’m using confirmation bias to exclude them.)

McArdle does a good job of discussing inattentional blindness here; she also discusses commitment bias in the context of disasters.  Her twist here is that one of the people counting didn’t even notice his own brother was one of the people passing the basketball.  ahahahahaha.

I think Gonzales’s “ Deep Survival (DS review _ notes) is a great read-across here… particularly around pages 79 – 81, where he brings up the gorilla experiment in a survival context:

“the implicit assumption is that you know what you’re doing and know what sort of perceptual input you want […] such a closed attitude can prevent new perceptions from being incorporated into the model.”  

Gonzales goes on to note:

“Gorillas are not helpful in completing the task [of counting the number of passes.]  … Gorillas are irrelevant and would displace the task in working memory. So the brain, efficient system that it is, filters out the gorilla so that you can keep counting.  Seeing the gorilla would be a mistake. You’d lose count.”

Gonzales further explores that plans are important for survivors, but you also need to keep your eyes open for opportunities that might not be in the plan, that, like the gorilla, you’re likely to miss if you’re exclusively focused on the plan.

Page 143: McArdle here quotes “ Deep Survival” ( DpSv review + notes) on bending the map.  She got me to read it again, actually; I’d read it years ago and shamefully forgotten about it.  She should’ve quoted it more extensively!

Pages 144 – 145: McArdle cites Feynman’s famous O-ring experiment here; she notes the halo effect problem of killing the messenger, as well as the sunk costs and commitment bias that kept the plan going forward.  She analogizes between this and “bending the map” in Deep Survival.

Page 147: McArdle here notes, briefly, confirmation bias.

Page 154: My dad was laid off a lot when I was a kid; somehow I manage to do the mental gymnastics and still get excited about M&A “synergies.”  Nice example of gaining perspective, i.e. another schema here: a tenured Nabisco salesman called the day he was laid off (post a Kraft buyout) the “worst moment of my life.”

Pages 155 – 156: McArdle notes that some people get “unstuck” after failures… and wants to figure out why.  (Note that she does sort of equivocate about failures in the sense that it’s hard to call what happened to the Nabisco dude a “failure” – he didn’t make any mistakes the way McArdle defines them.  This is why I state that the book is sort of loosely organized around failure.)

Page 157: McArdle spills the beans on investment banking… ahem.  

“I spent my summer sitting in a well-upholstered office until 11 p.m. every night, not exactly to pretend that I was working […] but to demonstrate that I would have been willing to work late if there was, you know, some work.”

  product vs. packaging, anyone?  This is a nice time to quote Jason Fried – the author of the highly sensible “Remote: Office Not Required” –  via Cal Newport’s “ Deep Work” ( DpWk review + notes):

“very few people work even 8 hours a day.  You’re lucky if you get a few good hours in between all the meetings, interruptions, web surfing, office politics, and personal business that permeate the typical workday.  Fewer official working hours helps squeeze the fat out of the typical workweek.”

Page 158: McArdle’s post-MBA job ended up not being available thanks to a hiring freeze, so she started working at Ground Zero…

Page 159: … and found out that employers don’t like people with gaps on their resumes.  It’s inversion and product vs. packaging – they’re looking for any way to cut down on the deluge.

Pages 161 – 163: McArdle notes that GDP is quantifiable, but perhaps not the best utility measure of lived experience of the economy.  For that, she points to employment, particularly given the U.S.culture of people asking what you do and equating identity with job.

McArdle notes the hedonic adaptation bit – a function of  contrast bias — see Achor’s “ The Happiness Advantage ( THA review + notes) – and points out that unemployment is the exception to the rule: people recover from divorces, deaths in the family, traumatic injuries, etc… but they don’t come back for unemployment.

Unemployment, for various reasons, leads to lower social connection.

Page 166: Nice bit on trait adaptivity here: McArdle’s blog, which would’ve been persona non grata for many journalism jobs, were viewed as a plus by The Economist, which hired her.

Also, amusing, McArdle’s dad asks her to calculate the expected value of becoming a journalist and uh, McArdle says that it “was roughly similar to the expected value of becoming addicted to crack.”  Thankfully for all of us, she ignored that little expected value calculation and graced the world with her writing 🙂

Page 169: A la Covey 4DX, here is a nice discussion of lead measures and lag measures.  If you want a job (lag measure), then the appropriate lead measure for that is looking for a job.

Pages 170 – 173: McArdle fleshes this out in a bit more depth: other correlated factors include being willing to work for less, and move if there’s no jobs where you are.  

She addresses the question of whether or not employment should be a big enough incentive to get people to look for a job.  The answer is no because looking for a job is depressing and loss aversion makes us not want to do it.

What we should be doing is what salespeople do: “smile and dial” is the lead measure for sales, it turns out.  McArdle notes that this is hard and even experienced salespeople hate doing it.

Two thoughts here: first would be Kip Tindell’s “man in the desert” selling as discussed in Uncontainable (UCT review + notes), which basically re-frames selling in a more positive way.

Second would be Howard Schultz’s rather depressing experience trying to pitch Starbucks to investors when he wanted to expand it beyond Seattle… as he notes in “ Pour Your Heart Into It” ( PYH review + notes), 217 of 242 prospective investors rejected him.  217 rejections is, you know, kind of a lot.

Pages 174 – 175: McArdle discusses the process vs. outcome angle here; this is another nice example of reframing: success is now measured not in sales, but in following the process, which will necessarily lead to sales.  (Also, of course, social proof.)

Pages 178 – 180: Nice examples of n-order impacts here as it results to the American push for homeownership (reduced mobility) and laws that make it hard to fire people in Europe (which leads to less hiring).

Pages 188 – 189: McArdle notes our brains’ inherent pattern-seeking storytelling behavior.  She notes (my analogy) that our brains function like one of Leroy Jethro Gibbs’ rules: there are no coincidences.  When we see correlation, we also see causation.

Page 191: ahahahahahaha.  McArdle knows it’s a bad idea to argue with conspiracy theorists, but y’know, Planner-Doer.

Page 193: McArdle notes – as others have – that if you can’t falsify a belief with any evidence, it’s usually not a great belief.

Page 200: McArdle here talks about feedback loops as well as our overconfidence tendency: everyone knew it was a possibility that housing prices could fall; they just didn’t weight it high enough.

There is a fun quote on this from Nate Silver’s “ The Signal and The Noise” ( SigN review + notes):

“[Moody’s 50% adjustment] might have been fine had the potential for error in their forecasts been linear and arithmetic.  

But leverage, or investments financed by debt, can make the error in a forecast compound many times over, and introduces the potential of highly geometric and nonlinear mistakes.  

Moody’s 50 percent adjustment was like applying sunscreen and claiming it protected you from a nuclear meltdown.

Pages 208 – 209!: McArdle reviews some of the research on learned helplessness and agency: it’s adaptive for us to believe we have control even when we don’t, because:

“Both humans and animals who are subjected to an unpleasant stimulus, such as a loud noise or electric shock, exhibit less stress if they believe they can control the negative events […] they performed much better on subsequent cognitive tests, like proofreading, when the switch [that allowed them to turn off airplane noise] was there.”

See also the classic Seligman paper.  She also cites the famous Whitehall studies, which showed that even controlling for weight, diet, and cardiovascular fitness, having a sense of control extends lifespan:

“people in the [Whitehall] study who reported low control over their work lives had three times the mortality rate of those who reported that they enjoyed high levels of control.”

Stress  agency.

Pages 210 – 213: McArdle compares confirmation bias exhibited by prosecutors and conspiracy theorists.  If this brief discussion intrigues you, Tavris/Aronson go into more depth – a lot more depth – on confirmation bias and self-justification in “ Mistakes were Made (but not by me)  MwM review + notes).


Pages 222 – 223: This part of the book (about the Hawaii parole system) may be one of the most important parts of the book.

Remember the hand-washing bit, where McArdle noted the challenge of “getting away” with bad behavior N times in a row, only to be punished the N + 1th time?  Well, Judge Alm in Hawaii described probationers’ experiences – both with their parents and the legal system – as :

“nothing… nothing… nothing… nothing… nothing… bam!  Five year prison term.”

 It is obviously hard to “learn” from that kind of a system.

McArdle also provides a (mildly heartbreaking) example of shooting an adult dog for peeing on the floor; she references hyperbolic discounting here too and – waitforitwaitforit wait for it – the marshmallow experiment, aw yeah.

Pages 224 – 228: Marshmallow experiment videos are fun.  But anyway, McArdle goes into the idea of self-control / willpower: McArdle argues that most people focus more on “punishment” and less on “certain,” which appears to be the more important half.  It’s not the severity of the punishment; it’s the guarantee of the punishment.

McArdle notes that the parole system was overstretched because of paperwork requirements that were overwhelming; Judge Alm used structural problem solving to reduce the activation energyrequired for parole officers to bring in suspects.  Alm also realized that drugs were causing most of the crime (McArdle provides a brief, insightful explanation here), so addressing that was critical.

Alm basically guaranteed (via random drug testing) that if they used drugs, they would get caught… McArdle applies the same idea to parenting.

Pages 230 – 232: Bringing the point home, McArdle notes that inconsistent feedback leads to people believing in luck as the driving factor of results.  In fact, kids raised in “reliable environments” did better on the marshmallow test.  Why does this matter?

It’s, unsurprisingly, agency again:

“Successful people have what psychologists call ‘self-efficacy’ or an ‘internal locus of control’: they feel that outcomes mostly depend on what they do.  People who believe that they can control their fate are more likely to have happy futures even if they’re wrong about the extent of their control.”

Alm returned a sense of agency to probationers: McArdle quotes him as asking, “Who is in charge of whether I succeed on probation?  I am.”  He uses words like “decisions,” “choices,” and “control,” and is encouraging/upbeat.

This part of the book is worth rereading every year or so.  I don’t say that about a lot of parts of a lot of books.

Page 234: McArdle references marginal utility here; specifically, she notes that the marginal utilityof prison as a deterrent is pretty low at this point.  In fact, an unintended n-order impact is that if a lot of people you know are in prison, you’ll think being in prison is pretty normal, and won’t be worried about impressing your neighbors.

Pages 237 – 238: McArdle again punches home that “occasional mercy is not merciful.”  This is a bit like Munger in several places (navy captains, someone caught stealing from the register.)

McArdle also notes that punishment should be focused on teaching, not revenge,” though she notes that one’s particularly hard.  Again, see Mistakes were Made.

Pages 241 – 242: The 80/20 from Dave Ramsey: debt is bad, mmkay?  (A modest mortgage is okay, and also save a lot.)

Pages 243 – 244: McArdle tries using cash for a bit as a structural problem solving approach to spending less money.  It is well-documented that people spend more using credit cards than cash, both due to lower activation energy (using cash is annoying and hard), and due to the salience /vividness of having to count out the cash and no longer having it, vs. some electrons from some server moving to some other server.

This isn’t just for big purchases, either; research I did on vending-machine cashless payment provider USA Technologies (USAT) suggested a material uplift in vending machine sales that had credit card readers.

Back to McArdle – she notes, as I have elsewhere from personal experience, that it’s not just the poor who live paycheck to paycheck – McArdle saw this kind of behavior among “nice Washington professionals.”

Anyway, see also Thaler in Misbehaving on mental accounting.

Pages 248 – 253: Contrasting America’s bankruptcy system to that in Europe (and providing some amusing historical context on how it got to be that way), McArdle comes to the counterintuitive conclusion that the concept of limited liability and easy personal bankruptcy is a big “ margin of safety” for entrepreneurs – i.e. you can lose it all, but you can’t lose more than it all and be burdened by a bad investment for the rest of your life, which is the case in Europe.  As McArdle vividly demonstrates, you can’t really just “hand over the keys” in Europe – the onus of responsibility is put on the borrower, not the lender – and so entrepreneurship opportunities are foregone.

Pages 260 – 262: interesting utility and n-order impact here: McArdle notes that in many cases it’s cheaper to give homeless people apartments than it is to treat them at the revolving door of the emergency room.

The point?  A pithy rhetorical question: “Do we want to be right, or do we want to be rich?”  Which could just as easily be referencing confirmation bias and ideology.

Page 263: McArdle notes here the culture angle she noted earlier; it would be bad for everyone if everyone made the local optimization decision of not trying to pay off their debts and just declaring bankruptcy.

Pages 267 – 268: McArdle concludes with a nice bit about margin of safety.


First Read: 2016

Last Read: spring 2018

Number of Times Read: 2

Planning to Read Again?: definitely


Review Date: spring 2018

Notes Date: spring 2018