Food For Thought: Why Correlation Better Not Be Causation

Tyler Durden's picture

We can only hope that the chart below, which shows a rather disturbing linkage between consumption of chocolate (in kg/yr/capita) and Nobel prize laureates per 10 million of population, is simply a good confirmation of the old adage which everyone these days seems to forget (especially those who look at "black box" models and predict the outcome of the presidential election with 75.304% accuracy in hopes of signing even better book deals) that correlation most certainly is not causation. Of course, for the vast majority of people who have no idea where statistics ends and heuristics begins, there is nothing greater in life than a high R-squared, which is why we expect that the next generation of Americans will be even fatter (if that is even possible) for one simple reason: their proud parents will fed them intravenous chocolate in hopes of creating the next Economics Nobel prize winner.