Some Random Economic Fun – Alcohol Consumption and GDP Per Capita

by robekulick

I recently noticed that Luxembourg has close to the highest GDP per capita and the highest alcohol consumption per capita in the world. This made me wonder what the overall connection between alcohol consumption per capita and GDP per capita is in the world. So in a particularly geeky moment, I plugged the data into my statistical software and voila!

First the sources:

I’m using IMF GDP per Capita data from:

http://www.imf.org/external/pubs/ft/weo/2011/01/weodata/weorept.aspx?pr.x=47&pr.y=6&sy=2010&ey=2010&scsm=1&ssd=1&sort=country&ds=.&br=1&c=512%2C941%2C914%2C446%2C612%2C666%2C614%2C668%2C311%2C672%2C213%2C946%2C911%2C137%2C193%2C962%2C122%2C674%2C912%2C676%2C313%2C548%2C419%2C556%2C513%2C678%2C316%2C181%2C913%2C682%2C124%2C684%2C339%2C273%2C638%2C921%2C514%2C948%2C218%2C943%2C963%2C686%2C616%2C688%2C223%2C518%2C516%2C728%2C918%2C558%2C748%2C138%2C618%2C196%2C522%2C278%2C622%2C692%2C156%2C694%2C624%2C142%2C626%2C449%2C628%2C564%2C228%2C283%2C924%2C853%2C233%2C288%2C632%2C293%2C636%2C566%2C634%2C964%2C238%2C182%2C662%2C453%2C960%2C968%2C423%2C922%2C935%2C714%2C128%2C862%2C611%2C716%2C321%2C456%2C243%2C722%2C248%2C942%2C469%2C718%2C253%2C724%2C642%2C576%2C643%2C936%2C939%2C961%2C644%2C813%2C819%2C199%2C172%2C184%2C132%2C524%2C646%2C361%2C648%2C362%2C915%2C364%2C134%2C732%2C652%2C366%2C174%2C734%2C328%2C144%2C258%2C146%2C656%2C463%2C654%2C528%2C336%2C923%2C263%2C738%2C268%2C578%2C532%2C537%2C944%2C742%2C176%2C866%2C534%2C369%2C536%2C744%2C429%2C186%2C433%2C925%2C178%2C869%2C436%2C746%2C136%2C926%2C343%2C466%2C158%2C112%2C439%2C111%2C916%2C298%2C664%2C927%2C826%2C846%2C542%2C299%2C967%2C582%2C443%2C474%2C917%2C754%2C544%2C698&s=PPPPC&grp=0&a=#download

And Alcohol per Capita Data from the World Health Organization

http://apps.who.int/whosis/database/core/core_select_process.cfm?strISO3_select=ALL&strIndicator_select=AlcoholConsumption&intYear_select=latest&language=english

I ended up with clean data for 170 countries overall. Taking the correlation between GDP per Capita and alcohol consumption per capita I found a correlation coefficient of 0.4856. Correlation coefficients run between -1 and 1 so that is a pretty impressively strong correlation, especially considering how many factors effect alcohol consumption.

I also performed the correlation calculation omitted Muslim countries that report almost no alcohol consumption. Doing this I find a correlation coefficient of 0.5660. The correlation becomes a little stronger, but it’s interesting that it doesn’t really change that much.

I also performed a regression of the natural log of gdp per capita on natural log of alcohol consumption per capita. I restricted this to the non-muslim countries to avoid zero values (which cannot be handled by the natural log function). For those of you who are asking why on earth I did this, the reason is that using the natural log allows for a very easy interpretation of regression coefficients. Specifically, using this method, I found that a 1% increase is GDP per capita is associated with an almost 0.5% increase in alcohol consumption per capita. This result is highly statistically significant with t = 8.79 (If you don’t know what this means, don’t worry about it, it just means this isn’t a fluke).

So I think this is all pretty impressive. I would assume the direction of causality goes from GDP to alcohol consumption (unless someone wants to propose an economic theory as to why high alcohol consumption boosts GDP!) It appears to be the case that people in richer countries drink more and substantially more at that. So I guess we should all plan on moving to Luxembourg!

Advertisements