This paper is an example of spurious correlations that appear because of selection-distortion effect and makes the authors come up with "believable" causes ti justfiy them.