I'm not sure that GDP per capita is the right measure for assessing correlations between average income and education, which is what I presume you're trying to

I'm not sure that GDP per capita is the right measure for assessing correlations between average income and education, which is what I presume you're trying to get at. After all GDP per capita is often skewed by the very rich. Take for example some stats from the US in 2004: GDP per capita: $39,934 (Source: http://www.econstats.com/weo/V015.htm http://www.econstats.com/weo/V015.htm) Median income per household member: $23,535 (Source: http://pubdb3.census.gov/macro/032005/hhinc/new02_001.htm http://pubdb3.census.gov/macro/032005/hhinc/new02_001.htm) Obviously a major discrepancy there. The median income per householder is determined via a sample census survey and would more accurately reflect average income than GDP per capita, which is constructed by dividing GDP by the population. Probably of more interest would be median income per householders over 18 to exclude all the kids. Then you could check the correlations between income and education. Notwithstanding, I'm not that convinced there would be a very strong correlation across countries, in part because of the multitude of other factors that would influence income. There are also other factors to consider regarding the construction of the statistics, such as purchasing price parity, the definitions each country has used for education level, etc. Trying to attain ceteris paribus (all things being equal) for this would be a nightmare.