From prhode@email.unc.edu Sun Aug  6 17:26:19 2000
Date: Sun, 6 Aug 2000 10:33:02 -0400 (EDT)
From: Paul Rhode <prhode@email.unc.edu>
To: Koleman Strumpf <cigar@email.unc.edu>
Subject: Re: your mail

As a check on your elderly 65+ results I calculated the dissimilarity
index on state-level data from HS (CD_ROM) over the 1870-1970
period.  There were slight increases in 1920s and the 1960s but overall the
movement was down, down, down.

1870	0.1458
1880	0.1400
1890	0.1265
1900	0.1028
1910	0.0964
1920	0.0792
1930	0.0860
1940	0.0772
1950	0.0668
1960	0.0653
1970	0.0662


The share of the elderly population was initially small (<4 percent) but
eventually grew near 10 percent.

