Statistics indicate that more people are born in the fall (in the USA anyway), with the quick explanation being that we have more sex during the winter holidays (9 months earlier).  GE takes a cut at this notion by comparing average temperatures and deviations from the national average birthrate. Statistically, however, using annual data adds enough noise in my mind to make drawing conclusions kind of tough. Anyone want to dig up the monthly data (even for one state) and do a lagged scatter plot? Hmmmm… how would you seasonally adjust this data?

(one minor complaint: the 3d scale is interesting for comparing states, but you can’t tell what the values are for any of them because of the angle)