Tag Archives: gdp ratio

Does Heritage’s EFW index correlate with GDP/capita?

The Heritage Foundation has published an “Index of Economic Freedom” in some form or another since 1992. (Or, at least that’s how far back their dataset goes.)

The EFW index grades the following 10 characteristics:

  • Business Freedom (low regulatory burden on business)
  • Trade Freedom (free trade)
  • Fiscal Freedom (low taxes)
  • Government Spending (low gov spending / GDP ratio)
  • Monetary Freedom (price stability, low price controls)
  • Investment Freedom (low restrictions on capital flow)
  • Financial Freedom (low financial services regulation)
  • Property Rights
  • Freedom from Corruption
  • Labor Freedom  (essentially, a lack of strong unionization)

Each of the above is scored 0-100, and then those scores are averaged for each country to yield an overall score. (You can read more about the methodology on the Heritage page if you’re interested.)

The first thing I did was to pull the data into Mathematica and append a column of GDP/capita using the CountryData[] function. Then, I filtered the data so that all the entries remaining in the array contained valid scores/numbers for each column, so that I wouldn’t end up with random zeros in the data. The only four years with totally complete data were 2005-2008, so those were the only years I studied. GDP was normalized in adjusted USD via the exchange rate as of last Friday (1/20/11, when I tabulated the data).

And, yes, I’m aware that GDP/capita does not necessarily equate to the “wealth” of a nation, per se, but it is most widely available/published metric that suits this purpose.

So, first, a log-plot of EFW score and GDP/capita (in 2008):

GDP/capita against overall EFW score. Correlation of 0.68 and a slope of .095 (on a log axis).

The plots for 2007-2005 look very similar. But, in case you’re not convinced, here are the ANOVA tables for 2005 and 2006:

ANOVA for 2005

ANOVA for 2006

ANOVA for 2007

ANOVA for 2008

I would like to point out the outrageously small P-values for the linear fits of these datasets. If we can conclude anything, it is that economic freedom MATTERS. We can go back and forth all day about whether or not there are too many extrinsic correlations between wealth and freedom for us to interpret the data appropriately, but these numbers suggest that Heritage’s index is a good predictor of the wealth of a country, causal relationships notwithstanding.

I’m still working on interpreting the relationships between the EFW sub-categories and wealth; it would appear that most of the subcategories do not correlate as well with wealth than the overall index does. (Property Rights and Freedom from Corruption correlate better than anything else, with values of .724 and .799, respectively, in 2008.) Additionally, I would like to look at whether or not adjustments in EFW over time affect GDP/capita, because it would help establish (or disprove) a causal relationship between the two. I’m also open to suggestions about picking a new measure of “wealth” via publicly available econometric data, if anyone has suggestions. GDP/capita looks sort of like productivity, except it does not take into account how much of the population is working. Perhaps GDP/working population or just raw productivity would be more accurate. (Since government spending is part of GDP and is a negative factor in EFW, it creates noise as a consequence of it being a factor in the EFW score. It doesn’t help, either, that deficit spending will have the effect of making those countries look wealthier.)

The raw data (without econometric data) is available for download on the Heritage site. Contact me personally if you want my filtered sets with econometric data.

Tagged , , , , ,