Today, on Steven Landsburg's blog, there was an interesting (albeit highly simplified) read about which sorts of freedoms have stronger correlations with wealth. I wish there was a link to the paper itself, because it seems like there may be a few dubious assumptions being made. Some are cosmetic, but some are deeper.
For example, the assumption that happiness can be measured in GDP per capita terms seems somewhat trivial. As discussed in the comments section, utility (mathematical measure of happiness) is generally measured in ability to purchase things (leisure time included), so to use the terms "wealth" and "happiness" interchangeably doesn't present any real issue. Essentially, that issue is merely semantic. As anyone who knows me, I'm not what Dr. Gregory House would call an "anti-semantic bastard," so I'm not bothered by the diction.
A much larger assumption is that of causation between economic freedom (capitalism) and per capita GDP. It's easy to tell a heuristic tale to explain something with causation, but it's not as easy to find empirically. For example, take Thomas L. Friedman's Golden Arches Theory of Conflict Prevention. Would you be willing to bet that if a McDonalds was reopened in Iran, the possibility of an armed conflict involving Iran and any countries on this list would be up in smoke? Clearly, McDonalds doesn't cause peace. Is there a confounding variable lurking somewhere in the third chart in the aforementioned blog post? I'm tempted to believe that there is not because I cannot think of any (not to say that none exist), and I'm not a strong believer in coincidence (although I do believe that there are unknown variables that cause phenomena).
-FSG
You can google the Economic Freedom stuff and see another problem. They measure Economic Freedom with five indicators, and they aren't necessarily related. It's perfectly legitimate (or at least commonly promoted) to believe that Free Trade and Sound Money Supply contribute to societal wealth where Deregulation and Small Government might not. I would like to break it down into the five indicators and run a regression analysis (possibly include the other liberties as well), maybe I'll do this after I move.
ReplyDeleteI also reject the notion that political and civil liberties are not economic liberties. That's asinine. The ability of a woman to wear a dress in public, the ability of a ethnic minority to own a business, and the ability of a writer to criticize the government allow for markets for goods and services that would not exist otherwise.
Kenny
I agree...It would be ideal to see some sort of multivariate regression. Also, I'm not completely convinced that political and civil liberties are independent from economic liberties (addressing your concerns from the second paragraph). Perhaps along with an OLS sort of regression, GMM may be justifiable (if you want to think about GMM again...).
ReplyDeleteIf you do that sort of analysis, I will need to see it! You'll also be implored to do a bit of guest posting on the topic.