I checked out Peter Leeson’s paper “Two cheers for capitalism?”. It’s potentially interesting, and at least based on some real-world data (good). Unfortunately it reads more like a polemic than a piece of serious science (bad).
There is not a null hypothesis,control group, p-value, or confidence interval to be found anywhere in the paper. No Bayesian statistics either. Nor are there any formal references, just a list “for further reading” (a lot of them being self-references). Initially I found it hard to believe that this got through peer review, but it appears that Society accepted it http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1438364
Now it’s not a bad idea to use comparative method, and there are some interesting possible conclusions from the data. But I would expect a more robust analysis than Leeson gives. In particular to explore a causal relationship you need to look at confounding factors, and the 5 countries which became “less” capitalistic by his measure (Myanmar, Rwanda, Ukraine, Venezuela, and Zimbabwe) had a heck of a lot of confounding factors. You certainly can’t pin their problems purely on the loss of a point or two of “economic freedom”. And in any case a sample size of 5 is likely too small to conclude anything.
Leeson should really have compared the countries which became more capitalist against the countries which stayed roughly the same on his scale (an obvious control group). My prediction is that both groups grew in average welfare (by his indicators) over the period in question. So we have to check whether the difference in welfare growth was significant between the groups against the null hypothesis that it wasn’t. Leeson doesn’t do that comparison though… I wonder why not?
Also, eyeballing Leeson’s static cross-section graphs on page 17 suggests some fairly good welfare correlations up to 7 points of capitalism on his scale (at least in 1995), but beyond that it’s pretty much a random scatter plot. The rich countries all have (it appears) at least 6 points (mixed economy?), so I can’t see a great case that they benefit from becoming “more” capitalistic. Further, the correlation between capitalist score and democratic score looks very weak (again almost a random scatter plot). This is probably because purer forms of capitalism are not actually very popular. So, I don’t see a strong reason here for more than the customary two cheers after all.
On the lower end of the scale, poor countries probably can benefit from an injection of capitalism (a reason for at least one of the two cheers). But then Marx would have found that unsurprising, as would Lenin (see http://en.wikipedia.org/wiki/New_Economic_Policy) and the modern Chinese communist party. Remember that socialism was always supposed to happen after capitalism had built up a huge stock of wealth, and hit a point of diminishing returns in terms of human welfare (then collapsed under its own contradictions). That’s all out of fashion these days of course. (And, disclaimer, I’m neither a Marxist nor a socialist.)
I checked out Peter Leeson’s paper “Two cheers for capitalism?”. It’s potentially interesting, and at least based on some real-world data (good). Unfortunately it reads more like a polemic than a piece of serious science (bad).
There is not a null hypothesis,control group, p-value, or confidence interval to be found anywhere in the paper. No Bayesian statistics either. Nor are there any formal references, just a list “for further reading” (a lot of them being self-references). Initially I found it hard to believe that this got through peer review, but it appears that Society accepted it http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1438364
Now it’s not a bad idea to use comparative method, and there are some interesting possible conclusions from the data. But I would expect a more robust analysis than Leeson gives. In particular to explore a causal relationship you need to look at confounding factors, and the 5 countries which became “less” capitalistic by his measure (Myanmar, Rwanda, Ukraine, Venezuela, and Zimbabwe) had a heck of a lot of confounding factors. You certainly can’t pin their problems purely on the loss of a point or two of “economic freedom”. And in any case a sample size of 5 is likely too small to conclude anything.
Leeson should really have compared the countries which became more capitalist against the countries which stayed roughly the same on his scale (an obvious control group). My prediction is that both groups grew in average welfare (by his indicators) over the period in question. So we have to check whether the difference in welfare growth was significant between the groups against the null hypothesis that it wasn’t. Leeson doesn’t do that comparison though… I wonder why not?
Also, eyeballing Leeson’s static cross-section graphs on page 17 suggests some fairly good welfare correlations up to 7 points of capitalism on his scale (at least in 1995), but beyond that it’s pretty much a random scatter plot. The rich countries all have (it appears) at least 6 points (mixed economy?), so I can’t see a great case that they benefit from becoming “more” capitalistic. Further, the correlation between capitalist score and democratic score looks very weak (again almost a random scatter plot). This is probably because purer forms of capitalism are not actually very popular. So, I don’t see a strong reason here for more than the customary two cheers after all.
On the lower end of the scale, poor countries probably can benefit from an injection of capitalism (a reason for at least one of the two cheers). But then Marx would have found that unsurprising, as would Lenin (see http://en.wikipedia.org/wiki/New_Economic_Policy) and the modern Chinese communist party. Remember that socialism was always supposed to happen after capitalism had built up a huge stock of wealth, and hit a point of diminishing returns in terms of human welfare (then collapsed under its own contradictions). That’s all out of fashion these days of course. (And, disclaimer, I’m neither a Marxist nor a socialist.)