Thursday, May 22, 2008

Time Series Econometrics and Happiness Research

An article by Johns and Ormerod in the current edition of the Real World Economics Review (formerly the Post-Autistic Economics Review), examines well the importance of time-series econometrics for happinness research. The conclusion is that well-being evidence is currently not robust enough to guide policy-making.

"From a statistical perspective, any calculation of a correlation between a variable which exhibits a trend and one which does not is fraught with inherent problems. (In technical terms, by definition time series happiness data is integrated of order zero, and GDP is integrated of order one)."

1 comment:

Anonymous said...

that very intersting!