Index of Personal Freedom – when good statistics go bad.
June 20, 2011
A recent article by Andrew Leonard titled “Why do liberals hate freedom so much” caught my attention. (http://www.salon.com/technology/how_the_world_works/2011/06/15/why_do_liberals_hate_freedom)
It’s an interesting discussion focusing on a document recently published by the Mercatus Center at George Mason University: “Freedom in the 50 States: An Index of Personal and Economic Freedom.” As one might expect from the Koch funded libertarian think tank, the premise is that states with fewer regulations are “freeer” than those with more regulations. The paper (http://mercatus.org/sites/default/files/50States_2011_Embargoed_Copy.pdf) is resplendent with impressive “data”, statistical analysis, charts and, of course, conclusions.
What really struck me is how credible that type of pseudo-analysis can appear and be totally wrong. The issue is that all of the analysis is based on false premises. It doesn’t matter that the technical computations are accurate if the source data is opinion based or overtly biased.
One way to confirm the relevance of the paper is to compare its results with other data sets. It doesn’t take long to see that the freedom rank doesn’t seem to correlate with any other data. So maybe that an extreme statement. It doesn’t correlate with the US census data that ranks states for various parameters.
The following table was created from the “Freedom Rank” and data from the US Census website, http://www.census.gov/compendia/statab/rankings.html.
The following series of graphs do comparisons between the “Freedom Ranking” and various other state rankings. The most striking observation is that there really is no correlation between the “Freedom Ranking” and any other ranked parameter. The only ranking that showed a hint of correlation was that of Doctors per 100,000 population. That particular correlation was negative, meaning that there were more doctors in states that were, according to the Mercatus Center, less free. So apparently doctors choose to practice in states that are more heavily regulated. I’m not sure how that supports the Koch’s libertarian perspective of freedom.
In his article, Leonard raises the question of population distribution. Building on his comments, if there were any significance to the freedom scale then one might expect that there would be a disproportionate percent of the population seeking to live in the free states. In fact, the population is almost evenly divided between the top 25 “freeest” states (46%) and the other 25 repressed states (54%).
As Benjamin Disreli is quoted as saying, “There are lies, damned lies and statisticians”.