Daily Polling Report 10/5
by CA Pol Junkie
In the last three weeks, Barack Obama has improved in national polling by about 9 points. State polling followed along so that if the election were held today, Barack Obama would win an electoral landslide. The problem is that we are 30 days away from the election, with alot of ugly stuff forthcoming from the McCain campaign. McCain is desperate to win, but does he have a chance?
In an analysis of convention bounces, Mark Blumenthal includes handy charts of the polls from the last 7 elections with poll averages and a final prediction from the polls. Here are the approximate poll averages a month out and on election day, compared with the actual results:
1980 October: Reagan +3.2 November: Reagan +2.3 Result: Reagan +9.6
1984 October: Reagan +19.0 November: Reagan +16.7 Result: Reagan +18.2
1988 October: Bush +5.6 November: Bush +9.0 Result: Bush +7.7
1992 October: Clinton +12.7 November: Clinton +4.7 Result: Clinton +5.6
1996 October: Clinton +15.2 November: Clinton +10.6 Result: Clinton +8.5
2000 October: Gore +0.4 November: Bush +2.6 Result: Gore +0.5
2004 October: Bush +1.7 November: Bush +0.7 Result: Bush +2.5
In each year except for 1988, the leader lost ground as we got closer to the election. The average amount by which the trailing candidate gained was 2.3 points. In the extreme case, Bush gained 8 points on Clinton in 1992. If McCain were to gain that much he would get a very narrow lead. There isn't a strong independent candidate in this year's race, so McCain would need to do some combination of A) get the votes of those who are currently undecided, B) bring over some Obama supporters to his side, C) excite more of his potential supporters so more of them become likely voters, or D) turn off potential Obama supporters so they don't vote.
Rasmussen says that in hard support, Obama leads McCain 45-38. McCain needs over 70% of the soft supporters and truly undecideds to tie Obama in the popular vote, but Brian Schaffner notes that 80-90% of soft supporters stick with their candidate. The math just doesn't work for McCain, especially when you consider that Obama likely can still win or tie in the electoral college while tied or slightly behind in the popular vote. McCain is still trying the political equivalent of an on-side kick with an all-out character assault on Barack Obama. Even if he were able to change the dynamics of the campaign, however, McCain's ground campaign still stinks so he would have a hard time taking advantage. He also has to do this without keeping his own unfocused campaign from unraveling.
We're seeing poll physics in Minnesota, where each outlier has an equal and opposite outlier. Where Survey USA had McCain ahead 1 point, The Star-Tribune poll has Obama ahead 18. The truth is doubtless closer to the middle. Obama continues his double-digit lead in the Pennsylvania tracking poll while Mason-Dixon indicates movement toward McCain in Colorado.
The current prediction is now Obama 380, McCain 158 with Colorado as the state that puts Obama over the top and West Virginia as the closest state. The latter is now theoretically for Obama according to the spreadsheet thanks to Obama gaining about a point in the national tracking polls over the weekend. If the current trend continues, Minnesota and Pennsylvania will be lumped in with Obama's base.
Obama Base (229 EV): California, Oregon, Washington, Hawaii, New Mexico, Iowa, Illinois, Wisconsin, Michigan, DC, Maryland, Delaware, New Jersey, New York, Connecticut, Rhode Island, Massachusetts, Vermont, Maine
Competitive states, cumulative electoral votes, and new polls:
Minnesota (Obama +10.2) 239 EV
Star Tribune 9/29-10/3 (9/10-12):
Obama 55 (45)
McCain 37 (45)
Pennsylvania (Obama +10.1) 260 EV
New Hampshire (Obama +8.7) 264 EV
Colorado (Obama +6.8) 273 EV
Mason-Dixon 9/29-10/1 (8/13-15):
Obama 44 (45)
McCain 44 (42)
Ohio (Obama +5.2) 293 EV
Virginia (Obama +4.6) 306 EV
Florida (Obama +4.5) 333 EV
Nevada (Obama +3.7) 338 EV
Missouri (Obama +2.9) 349 EV
North Carolina (Obama +2.4) 364 EV
Indiana (Obama +1.5) 375 EV
West Virginia (Obama +0.1) 380 EV
Montana (McCain +5.6) 158 EV
McCain Base (155 EV): Alaska, Arizona, Utah, Idaho, Wyoming, North Dakota, South Dakota, Nebraska, Kansas, Oklahoma, Texas, Louisiana, Arkansas, Mississippi, Alabama, Tennessee, Kentucky, Georgia, South Carolina
The poll averages are created by a magic spreadsheet. Self-selected (Internet and mail) polls are ignored; no favoritism is done among the remaining pollsters. Polls are adjusted to today's conditions by shifting them by the amount of change in the average of Gallup and Rasmussen tracking polls. The weight of polls in the averages decreases geometrically each day such that 7 day old polls have 1/2 weight and 14 day old polls have 1/4 weight. The weight of state tracking polls is divided by the number of days in the sample. This method is very responsive to recent changes in both state and national polling.