Wednesday, November 19, 2008

Outliers

Author: Macolm Gladwell

Review: Tyler Cowen

The book is getting snarky reviews but if it were by an unknown, rather than by the famous Malcolm Gladwell, many people would be saying how interesting it is. The main point, in economic language, is that human talent is heterogeneous and that the talent of a particular person must mesh with the capital structure of his or her time if major success is to result. The book is best read as a supplement to Ludwig Lachmann's Capital and its Structure. The main enduring insight of both Lachmann and Gladwell is simply how much we live in a world of complementarity rather than substitutability.

Nowhere in the book does the name Dean Keith Simonton (check out the headings to these links) appear nor does the phrase "multiplicative model of human success." A lot of the content here has already been done with more rigor and empirical support and also in readable form I might add. Everyone should read Simonton, noting that his hypotheses fare better in the arts than in politics.

If you ask too much from Outliers it will fall apart. It is too easy to find contingency in the world and Gladwell doesn't begin to look for a theory of which contingencies are interesting or not. For instance arguably Ludwig van Beethoven would not have been a great composer if:

1. An extra butterfly had died two million years ago.

2. The outcome of the Thirty Years' War had been different.

3. The Germany of his time had not had fortepianos.

4. His parents had conceived their child one second earlier.

5. Haydn had not paved the way.

#3 and #5 seem more interesting than #1 and #4 but that's because some contingencies just don't help us understand the world very much. Gladwell never gives us enough theoretical structure to see why his contingencies are the relevant ones. Simply showing the contingencies in personal histories is not, taken alone, very enlightening.

Gladwell's contingency stories skid out of control. At one point it seems the main claim is that the steady accumulation of advantages is what matters, but once you ask which advantages end up "counting," the claim collapses into tautology.

There is also a "PC" undercurrent in the book of "don't write anyone off" but if everything is so contingent on so many factors, maybe writing people off isn't such a big deal. It could go either way. It depends.

Gladwell deliberately steers us away from the contingency of genetic endowment (even for a given set of parents, which sperm got through?), but if you hold everything else fixed you can assign a very high marginal product to the genetic factor if you wish, usually up to 100 percent of a person's outcome. That mental exercise is verboten but somehow it is OK to hold the genetic endowment constant and vary some other historical factor and regard that as a meaningful contingency. See the discussion of Beethoven above, especially #4 on the list.

Gladwell descends into the swamp of contingency but he is unwilling to really live in it and take it seriously or, alternatively, to find a way out.

In reality the complementarity concept is easier to work with and also more fruitful for thinking about policy implications or for that matter the implications for management or talent training. Success is fragile but foster competing cultures based on clusters of talent motivated by rivalry and emulation. Don't filter out the eccentrics or the risk takers. That's about where David Hume ended up but Gladwell never gets anywhere close.