A theoretical prediction

[Shira]

The states in the following list underwent significant changes between 1988 and 2000.  If this trend continues at the same pace, these are going to be Bush’s numbers (at least theoretically) in these states:

IA - 51.30%
MN – 47.85%
OR – 48.37%
WI – 49.02%
MI – 44.93%
TN – 50.17%
VA – 51.32%
AZ – 50.30%
NV – 48.72%
FL – 46.50%
NH – 45.15%

This is more or less my prediction IF the elections are EVEN. However, if Bush for example wins nationally by 4%, then add 2% to each of the above numbers. (half the nationwide margin of victory)

The above numbers were very simply derived. Take AZ for example. In 1988 Bush the first got 59.95% in AZ, while in 2000 G.W.Bush got there 50.95%. From the 1988 number subtract 3.8% which is half of Bush’s nationwide margin of victory over Dukakis. To the 2000 number add 0.81% which is Buchanan’s number in AZ . The result is that the net decline of the REPs in AZ during these 12 years was 4.39%. Assuming that the pace of decline has not been changed, Bush is going to lose 1.46% in AZ (4.39 in 12 years or 4.39/3 in 4 years) so the likely number of Bush in AZ is going to be 50.30% (Bush2000 + Buchanan – Expected Decline).

Despite the fact that the above figures are theoretical, I strongly believe that they are the best prediction (subjected to the correction that should be done in the case of non-even elections).

[AuH2O]

I mostly agree, but you would probably add 3% based on a 4% national victory. And of course it works better for some places than others.

This election will be very telling in terms of party trend in the Midwest. If the gains continue for the GOP, expect a 'Northern strategy' in 2008 to complete the realignment.