Last poll: 47D, 32R
This poll: 45D, 35R
No problem there IMO, especially given the effects of the RNC. I am NOT a fan of reweighting by party.
Last poll: 68 White, 24 Black
This poll: 65 White, 26 Black
No problem there either, and this would tend to work against Bush. We might notice that Maryland Lt. Gov. Michael Steele was the chair or whatever for the RNC. Bush's performance among blacks went from 11% to 21%.
Last poll: 50 Military/Vet, 45 non
This poll: 52 Military/Vet, 44 non
Lost of old Midshipmen and other vets in Maryland. The slight shift works in Bush's favor, but more importantly Bush's performance among military/vet increased from 45-51 (in favor of Kerry) to 55-42 Bush!
The internals don't mirror the actual registration numbers in Maryland, which means SUSA is not weighing by party.
http://www.elections.state.md.us/pdf/March2004.pdf
According to the March 2000 figures the breakdown is as follows (note numbers = 101.1% because of rounding):
55.8% D
29.9% R
14.4% I/O
If you weigh the results, Kerry would have a decent-sized lead. (I don't have time to do the weighing).
However, even if you don't weigh, Kerry's lead has been eroded in Maryland since SUSA's methodology remained the same.
Hold on. What does it mean by "military/vet"? No way is half of any state's population made up of veterans.
Very interesting analysis, as usual.
52% military or vets among MD voters? That's almost as high as West Virginia!
J,
If Bush pulls 15% of the African-American vote nation-wide this race is approaching Reagan-Carter territory.