I think it's more likely that 34 years were chosen because 1980 - 34 = 1946 makes sense for a post-WW2 analysis. The graph doesn't vary wildly for most of the t

I think it's more likely that 34 years were chosen because 1980 - 34 = 1946 makes sense for a post-WW2 analysis. The graph doesn't vary wildly for most of the time (most of the change is a fairly smooth motion), so it seems unlikely that 30 or 40 years would change the conclusion. Note that a sharp rise in the curve can also be explained by bad years dropping out of the 34-year window. This would probably explain the increases in the 1%'s weath around 1992 and 2003 (1958 and 1969-70 recessions, respectively). See https://en.wikipedia.org/wiki/List_of_recessions_in_the_United_States https://en.wikipedia.org/wiki/List_of_recessions_in_the_Unit...