For my econometrics class, I am comparing the relationship between economic freedom and prosperity, and I just got my first regression results for 2001 for 155 nations. The results are very preliminary, but the evidence is clear: there is an extremely high correlation between economic freedom and prosperity, explaining over 73% of the variation in wealth. This means that 73% of the difference between the wealth of nations is explained by the economic policies of their government, with only 27% accounting for differences in natural resources, location, climate, culture, other nations, etc.
This fact alone is not very surprising (unless you’re a socialist, in which case you’re probably ignoring these results), but it is interesting to see which specific factors affect per capita GDP the most. Not surprisingly, property rights and the fiscal burden (taxes) of government have the greatest effect, and significantly monetary policy (inflation) -which shows that (as Lenin said) the best way to destroy capitalism is to go after the currency. Factors which (to my surprise) do not affect prosperity individually are foreign investment and regulation – which may not be true if these variables are significant jointly -I’m not sure yet.
After my analysis of economic factors is complete, I am going to see what effect non-economic factors such as political freedom, government welfare, and population control have on prosperity.
(Note: while the black market correlation is highest, this is more a result of government regulation than a cause, which is why I don’t consider it a factor. Data comes from the 2001 CIA Factbook and 2001 Heritage Inst. Economic Freedom Index. The 2001 data was used because 2002 GDP’s are not available for all nations yet.)
Here is the regression output:
Model 3: OLS estimates using the 155 observations 1-155
Dependent variable: indGDP
Mean of dependent variable = 8854.32
Standard deviation of dep. var. = 9169.89
Sum of squared residuals = 3.43936e+009
Standard error of residuals = 4853.58
Unadjusted R-squared = 0.7344
Adjusted R-squared = 0.719847
F-statistic (8, 146) = 50.4624 (p-value < 0.00001)
Durbin-Watson statistic = 2.13614
First-order autocorrelation coeff. = -0.0694149
Excluding the constant, p-value was highest for variable 11 (BK)
(Variables are explained here: htp://www.heritage.org/research/features/index/2002/chapters/chap5.html)