An updated graph that Claudia Goldin had me make two and a half decades ago. The nonfarm unemployment rate since 1890:
Then it was 1890-1990, now it is 1869-2015.
- J.R. Vernon (1994): http://delong.typepad.com/1-s2.0-0164070494900086-main.pdf
- C.D. Romer (1986): http://delong.typepad.com/spurious-volatility.pdf
- BLS (2015): http://data.bls.gov/cgi-bin/surveymost?ln
- with spreadsheet at: http://tinyurl.com/dl20160405
The assumption--debateable--is that "unemployment" is not a farm thing--that in the rural south or in the midwest or on the prairie you can always find a place of some sort as a hired hand, and that "unemployment" is a town- and city-based nonfarm phenomenon.
I confess I do not understand how anyone can look at this series and think that calculating stable and unchanging autocorrelations and innovation variances is a reasonable first-cut thing to do...