Monday, May 24, 2010

Outliers, by Malcolm Gladwell



This is a Very Popular Book with a Big Message. Gladwell purports to study outrageously successful individuals--those who are "outliers" from the main sequence of humanity--to try to discover what makes them successful. Surprise! It's not just talent!

As you might expect, I am underwhelmed.

When you get right down to it, Gladwell comes to the astonishing conclusion that success requires talent, plus hard work, plus opportunity and luck. Sometimes it requires picking the right ethnic/cultural background. Sometimes it requires being born at the right time. Sometimes it's being good at something that turns into a growth industry after you are already good at it.

Basically--it's out of any one person's control.

But Gladwell isn't interested in drawing that message. Instead, he tries to draw lessons from the stories he tells in order to create a moving call to social change. Many people who are successful are the lucky recipient of opportunities. So let's make sure everybody gets opportunities! Or, as Gladwell puts it:

We cling to the idea that success is a simple function of individual merit and that the world in which we all grow up and the rules we choose to write as a society don’t matter at all.
Really? Does anybody really think that Bill Gates might still be The Richest Man In The World if he had been born into an illiterate nomadic tribe of Mongolian sheep herders? Does anybody think that the incredible run of Victorian era robber barons might have had something to do with being in the right place at the right time to exploit vast natural resources in a period before environmental and business regulation? That your brilliant kid will still win a Nobel Prize in Chemistry if he goes to the local vo/tech community college instead of Harvard?

I didn't think so.

Gladwell misses the real point. It's not that we don't believe that the rules don't matter, so much as we can't predict which rules will generate which successes. Sometimes rules have to be arbitrary, because we need some rules. And if those rules happen to advantage Kid A over Kid B, well, that's maybe unfortunate for Kid B, unless Kid B turns out to be perfectly placed as a result of that rule to take some advantage Kid A can't take--maybe even decades later.

We can't really predict a priori what the "right rules" are, after all. Sure, in retrospect, a Gladwell can pick out a year--say 1955--when it might be advantageous to be born, if one wants to be a computer billionaire like Bill Gates and Steve Jobs. But what about all the millions of babies who were born in 1955 who didn't turn out to be either Gates or Jobs? And can society really write rules that will guarantee another Bill Gates? After all, it's not like we can require that all babies in the country be born in 1955.

Which is the essential problem with Gladwell's "method"; there isn't one. He has a rough sort of thesis, and then he cherry picks his way through pop sociology and history to locate anecdotes that support it. There is no critical examination of whether his ideas are actually true; rather, because they are good stories and generally well told, they feel like they must be meaningful. In the end, however, this book hardly creates a blueprint for a better society.

Start at the beginning: the first chapter describes a small town in the mining hills of Pennsylvania--Roseto. The town is insular and almost entirely populated by emigrants from a single village in Italy. Despite adopting American diets and work habits, there is almost no heart disease in this town. Nearby towns have disease rates equal to or exceeding American averages, but sure enough, in this town of Italian emigrants, there is zero incidence of heart attacks below the age of 65. A pair of medical researchers studied this town, and ran through myriad theories as to why this might be, but by the end of the chapter, the conclusion is that the only factor must be that the nature of this particular community is the protection against heart disease.

To which one can only mutter a skeptical "really?" Do we really think that there is absolutely no other possible explanation for this anomaly? That it's a "magical Italian" way of life that involves strolling around in the evenings talking to your neighbors that is the silver bullet? How can we know that every single other explanation has been tested--do we really think that medical science is so complete that we understand what every last cause of heart disease might be, and that each one was tested? Personally, I am more likely to consider that there is some factor we don't understand and so can't control for than I am to believe that this small town is just magically immune from heart disease.

Plus--even if this is true--how is it even reproducible?

We then move to Canadian hockey players. Gladwell sets up a straw man assumption: that no matter where you are born in Canada, if you are a great hockey player, the system will find you. Does anyone really, actually believe that? I have some great ocean-front property in Haiti to sell you if you do.

Anyway, lo and behold, Gladwell discovers that Canada's youth hockey leagues have a January 1 age cut-off, and that kids born in January are generally bigger by that date than kids born in December. So the bigger, more mature kids get funneled into more elite teams, where they get better coaching and more ice time. The end result is that a staggering 40% of kids on national championship teams are born between January and March, while less than 10% of kids are born in the last three months of the year.

Gladwell asserts that if Canada had two leagues, with the second cut-off date being June 1, they would have twice as many hockey stars to chose from! But wait--that assumes that Canada could simply and easily double the resources it has to support elite teams: rinks, coaches, audiences, ice time, etc. Nor does Gladwell consider the social costs--whether having twice as many hockey players chasing the same limited number of professional spots (because the NHL is also not likely to be economically feasible at twice the size) isn't rather a waste of lives that could be better spent pursuing some other goal.

It's not that an arbitrary cut-off date has no effect, it's that there has to be some limit of some sort, and January 1 is no more or less arbitrary than any other date. Why have one cut-off date rather than two? That's an idea, but why stop at two? Why not twelve different leagues, so kids never compete against anyone more than a month older than they are? I can think of at least two reasons: 1) there's just not enough resources to support that fine a distinction (as I suggested above), and 2) there are some benefits to kids testing themselves against others who are bigger, stronger, etc.

Or, maybe Gladwell's "insight" has already been adopted, just at a different level. After all, the fact that Canada has leagues for each year means that Canada has three times as many elite hockey players than there would be if leagues grouped by age, such as "under 5," "6-8," "9-12," etc. Why does it need to double that number again?

The point is, in order to run a viable hockey league, the line has to be drawn somewhere. Gladwell doesn't really quibble with the fact of a cut-off date, merely about where that cut-off is made. Even if there was suddenly a second league, with a cut-off date of June 1, what could we expect would happen? The next "Malcolm Gladwell" could come along and note that the new system now discriminates against kids born in May and December, so there should be additional leagues with cut-off dates of April 1 and September 1.

There is also the risk that the quality of "elite" coaching might drop, if there is suddenly twice as much demand for it. Rather than adding opportunities for more kids, we are taking those opportunities away from the ones who currently have it. Do we have a better, more vibrant league if more kids are mediocre and fewer are exceptional? How many hockey geniuses are there, really? Maybe the December born kids are the real phenoms, while the kids who are born January-March are just winnowed out later? Who are the actual losers in a system where kids play hockey--or don't play hockey anyway. Gladwell doesn't ask any of these questions; he's moved on.

Gladwell uses Canadian hockey to segue into another theme--the 10,000 hour expert. It is the contention of at least one scientist that it takes 10,000 hours of practice to become an expert at anything: music, hockey, computer programming. For example, the Beatles were invited to Hamburg when they were young, where they played seven days a week, 8-10 hours a night. Gladwell contends that this is what made them into The Beatles as a phenomenon, and it wouldn't have happened without all those hours of playing.

Okay, but what about all those other bands who played in Hamburg who didn't become The Beatles? Or what about the Rolling Stones, or the Beach Boys, or Led Zeppelin, who didn't play Hamburg? Or what about the fact that The Beatles actually didn't become The Beatles until after they brought on Ringo Starr, who hadn't been in Hamburg? Gladwell doesn't gloss over those criticisms so much as he just doesn't even acknowledge that there is any such criticism possible. The Beatles were famous. The Beatles played in Hamburg. Therefore, the Beatles became famous because they played in Hamburg. Post hoc ergo propter hoc. And, I guess it means that all rock bands should go to Hamburg in order to get famous.

It ain't necessarily so.

Gladwell's thesis starts getting even squishier after this point, as he considers the genesis of Bill Gates. Prior to 1968, computer programming was an arduous and boring process of creating punch cards that had to be turned into a mainframe administrator and run in a batch. If there were any errors in the punch cards, the entire batch would be rejected, and the programmer would have to go through the entire batch to find the error, correct it, and resubmit the cards. Even without errors, a single batch could take hours to run, and there were always errors. (For no good reason, my college roommate, who was majoring in accounting, was still running these damn punch cards at the University of Minnesota in 1984!!)

However, around 1968, a couple of American universities discovered a way to bypass the damn punch cards, and developed a method of using terminals to program. Bill Gates happened to live near one of these universities, and also happened to go to a school with a well connected parents' organization that got one of these terminals for the school's computer club. Gates was immediately obsessed, and through a couple of other lucky breaks, he managed to spend thousands of hours in middle and high school with this new technology, programming computers.

Gladwell seems to think that he is breaking new intellectual ground: Bill Gates wasn't just a gifted computer programmer--he also had unusual opportunities to program, he had his 10,000 hours young, and he got them at the dawn of the personal computing era. More obviously, he was lucky to be at the right place at the right time, with both interest and talent for computers. How innovative is this? All the rest of kids in that computer club aren't the Richest Man in the World after all--there was something unique about that kid in that place with those opportunities. Giving every kid in America unlimited access to computer programming in 1968 wouldn't mean every kid in America would become the Richest Man in the World.

What Gladwell doesn't acknowledge is that there really aren't universally applicable lessons to be gleaned from his anecdotes. Every kid simply cannot have every opportunity, nor does every kid who has any particular opportunity become an Outlier. Would Bill Gates have founded Microsoft if he'd been forced to play hockey in Canada? Would the Beatles have become famous rock stars if they'd been keying in computer code all through eighth grade?

In fact, we can't predict which opportunities will be seized by which kids, nor can we predict which things are actually opportunities. Gladwell takes us to the mean streets of early 20th century New York, where recent Jewish immigrants developed their own garment shops. The grandchildren of these garment workers often grew up to be doctors and lawyers. The anecdote for this section is the story of Joseph Flom, the first associate hired by the firm that became Skadden Arps. As a fat Jewish boy, Flom was not going to be hired by the white shoe law firms of 1950s New York. So he worked for the start-up firm of Skadden Arps, doing whatever work came in the door. Some of that work was hostile corporate take-overs, which the white shoe firms refused to touch. So he spent years perfecting this recondite area of law, which suddenly became very lucrative in the 1980s. By the time the white shoe firms decided they could touch this area of law, they simply couldn't catch up.

Gladwell goes on to point to a number of factors that positioned Flom to become the mergers and acquisitions shark he became: factors that were outside anybody's control. He was born in the 1930s, when birth rates had dropped due to the Depression. As a result, the schools he attended were not overcrowded, and were comparatively new. The quality of teaching was especially good, as there were highly educated teachers who simply couldn't get college jobs and so taught high school. He was able to get into law school and support himself because there were more jobs than workers. The white shoe firms didn't realize how lucrative M&A could be. M&A got lucrative before everybody else caught on. None of these factors is reproducible generally, and not every fat Jewish kid born in 1930 became Joseph Flom.

But look how much depended on luck. M&A became the money bubble of the 1980s, and Flom happened to do that. But he could just as easily been practicing divorce law, which white shoe firms also didn't touch, or DUI, or criminal defense, or any of a number of other things that wouldn't have become a license to print money. Joseph Flom didn't start doing M&A law because he saw there was a lucrative future market in it--he just happened to be the guy who was in the right place at the right time, with the right experience. He could just as easily not been--the 1980s could have been about derivatives rather than hostile take overs. In which case, Skadden Arps would not be the firm it is today, and we wouldn't be reading about Flom at all.

Gladwell has an appealing writing style, and he teases out interesting stories. He also seems to only ever see the rosy side of whatever story he is writing, and so one gets the sense that he honestly believes his insights into Canadian hockey will easily double the number of Wayne Gretzkys produced every year. Yet his steadfast refusal to examine his "findings" with any sort of critical eye means that they remain so many rainbows and unicorns--lovely to imagine, but without any role in the real world.

No comments: