>http://dx.doi.org/10.1016/j.intell.2005.04.002

>correlational study

Without trying to adjust for confounders, this study makes as much sense as "heat is bad for GDP".