Ah primary sources, my good men (edit: and of course women, though I used the phrase in the gender neutral, genitive case) - here's the actual paper[1] (31 page

Ah primary sources, my good men (edit: and of course women, though I used the phrase in the gender neutral, genitive case) - here's the actual paper[1] (31 pages). RAND is a think-tank, not quite Koch's Americans for Prosperity or the Cato Institute but caveat lector, and all. Here are some brief assertions claimed: >> Our analysis leverages this substantial variation in the rate at which states are aging to estimate the effect of aging on the rate of state growth in per capita Gross Domestic Product (GDP) ... " One has to reject the null hypothesis that the aging out demographic are even related in part. (I could just as easily hypothesize that most of these retirees are those who were economically positioned to work until that late age and those who were comfortable enough to retire beforehand did so (if you buy the assumption that those who were highly-salaried throughout their career generated a proportional amount of value to said salary. I don't claim this, but it's an easy refutation.)) Page 6 is just aggregate analysis of population trends amongst certain ages binned by state, using Wyoming as their prime example. I'm not sure how much of the GDP/GNP Wyoming contributes to as a percent but I'd wager the trend is not indicative of the U.S. economy on the whole (in fact, you really can't make generalizations like this paper is trying to do correlating age against value lost -- whether it's in institutional knowledge delivered from master to apprentice in trade or actual quantifiable value. {Started skimming here due to the weak arguments.} They use the US census as their data source in section III (Empirical Strategy), which, again, isn't granular enough to make any sort of trend assertion with a high confidence factor. Section IV (Results) doesn't even try to normalize economic growth against any other market metric, domestic or international. {Stopped reading entirely at that point and went back to reading journal articles based on reproducible results rather than vague social theory and whatever these overpaid analysts remember from their 100 level stats class. Sigma summation, greek symbols and subscript don't make you seem smart, unless of course you're typesetting in TeX ;))} [1] https://www.rand.org/content/dam/rand/pubs/working_papers/WR1000/WR1063/RAND_WR1063.pdf https://www.rand.org/content/dam/rand/pubs/working_papers/WR...