The Perils of Blogging

I came upon this old post by star G.M.U. economics professor, blogger, and author of the well received book "The Myth of the Rational Voter" Bryan Caplan. He makes the case that much like how the nations with the biggest populations win the most Olympic medals, those same nations are likely to have the smartest people. (He doesn't say it explicitly, but I'm assuming Caplan means that the population distributions in larger countries allow for more superstar athletes and brainiacs...and also more dunces.)

So, in a big debate between two populations of different sizes, the

"expectation is that the advocates of the most popular viewpoints will prevail. The smartest ten Christians are going to be way smarter and way slicker than the ten smartest Zoroastrians, and will run circles around them in a debate.

Now here's the interesting thing. If virtually everyone just argues for whatever position he was born into, a truth-seeker should hold the gladiators for popular views to higher standards. If the smartest Zoroastrian holds his own against the smartest Christian, the rational inference to make is: "The Zoroastrian position is more likely to be true, because it tied despite the fact that it probably had a weaker defender."

Carrying it further, Caplan adds:

...as an academic, it's hard not to notice that liberals dominate at the university. And the simple truth is that people at top schools are smarter than people at lower-ranked schools. The result is that in any intellectual debate, the best liberals are usually smarter than the best non-liberals.

If my analysis is right, however, this actually counts against the liberal view. Weren't most professors liberals long before they had any arguments for their position? And wouldn't it take overwhelming intellectual firepower to drive them to apostacy? Then it's no wonder that the smartest liberal academics are smarter than the smartest non-liberal adademics. They have the same kind of inherent competitive advantage that China has over Luxembourg.

The upshot is that if an academic debate seems tied, the non-liberal view is more likely to be right. And if the liberal view actually seem to be losing, it's a safe bet that it's wrong.

It's a fascinating argument, but I find it a bit lacking -- and since comments on the post seem to be closed, I'm going to discuss it here.

The big problem is that Caplan's original claim is wrong. India, Indonesia, Brazil, Pakistan and Bangladesh, 5 of the 7 most heavily populated countries, are nowhere near the top of the all-time Olympic medals list.

But even if you take out those not-as-developed nations, the more populous countries still don't come out on top.

This weighted list took the medals each nation earned through 2000 and assigned points to the different levels -- 10 for gold, 3 for silver, and 1 for bronze. Then each nation's medal tally was divided by it's population for a rough measure of that country's Olympic success.

Using this method, the United States is ranked 18th and China is ranked 69th. The top 10 is littered with small countries like Finland, Sweden and Hungary.

The approach is probably too crude to provide good information when comparing countries close in rank (say between #3 and #4), but, as a broad analysis, it does illustrate that population size doesn't guarantee achievement.

UPDATE (12:30 EDT)
Some more reading on medal counts and population size:

• Population- and GDP-adjusted medal counts


• From 2002 paper: "For the Summer Games, nations with larger populations unambiguously send more athletes, but there is no evidence of that relationship in the Winter Games." (SSRN)


• From 2004 paper: "At the margin, population and income per capita are needed to generate high medal counts." (pdf)


• From 2006 paper: "GDP per capita, a relatively cold winter climate, and the age dependency ratio all have statistically significant impacts on national medal counts. Population, health expenditure per capita, the host/neighbour nation effect, and the cricket and rugby effects all appear to hold no explanatory power. " (pdf) FYI, the cricket and rugby effects refer to a hypothesis that the popularity of those sports -- which are not Olympic events -- in certain countries crowds out interest and achievement in other sports, thus explaining why certain countries don't medal.