Monday :: Nov 3, 2008

Daily Polling Report: Election Eve

by CA Pol Junkie

Today is the Grand Finale of polling, with 33 polls of 10 competitive states. The big picture is that Barack Obama is overwhelmingly likely to be elected President of the United States tomorrow. National polls give Obama leads of 5-11 points, with slight movement toward Obama as we close out the race. State polls show some slight movement toward McCain in a couple cases, but not even on the right order of magnitude to introduce doubt about the outcome tomorrow.

I don't think tomorrow night will be suspenseful on the presidential level for the following reasons:

- The Republican brand is deeply tarnished
- The country is in crisis after 8 years of Republican rule
- Barack Obama is likable and charismatic
- Obama has clear themes to his candidacy
- John McCain is uninspiring and has high negative ratings
- McCain has no theme to his candidacy
- Obama's ground game organization is the strongest organization ever
- McCain put minimal effort into his organization
- A bandwagon effect may get people to vote for the winner and a historic candidate: would you want to tell your grandkids you voted for the other guy in 2008?

The only thing McCain could conceivably have in his favor is the Bradley Effect, which may not even exist. Exactly why would millions of voters suddenly switch to McCain? There's a reason 538 is giving McCain a 1.9% chance of being elected, and personally I think that overstates his chances.

On Wednesday morning, I expect there to be two themes which explain Obama's rise: the power of ground organization and African-Americans swinging states to Obama thanks to historic turnout. Barack Obama's superior organization in just about every state propelled him to victory in the primaries. His campaign built upon that infrastructure and learned from that experience to improve for the general election. It was a comprehensive plan from registering voters to identifying them to turning them out, and the campaign is reaping the fruits of many millions of hours worked.

African-Americans were key to his primary and general election strategy. Pollsters have had difficulty polling African-Americans, as they reliably turned out in greater numbers and with more support for Obama than the pollsters anticipated in the primaries. We're seeing history repeat itself in the general election early voting, and the pollsters aren't able to keep up with the new reality. While we could see the Bradley Effect in a few states, I think the bigger effect is very likely to be the Reverse Bradley Effect, where African-Americans give Obama more support than they told the pollsters.

The FINAL prediction according to the magic spreadsheet is Obama 378, McCain 160. My personal take on each state is included with the polls. I think Obama will overperform the polls enough in Montana and Georgia to win those states' electoral votes, for a final result of Obama 396, McCain 142.

Obama Base (264 EV): California, Oregon, Washington, Hawaii, New Mexico, Iowa, Minnesota, Wisconsin, Illinois, Michigan, DC, Maryland, Delaware, Pennsylvania, New Jersey, New York, Connecticut, Rhode Island, Massachusetts, Vermont, New Hampshire, Maine

Competitive states, cumulative electoral votes, and new polls:

Colorado (Obama +7.9) 273 EV

Rasmussen 11/2 (10/26):
Obama 51 (50)
McCain 47 (46)

Early Voting:
PPP (as of 10/31, 65% have voted): Obama 58, McCain 41

CA Pol Junkie's View: This race is 2/3 over, and McCain can't beat Obama 2:1 among the remaining voters. This state will almost certainly go for Obama.

Virginia (Obama +7.0) 286 EV

Public Policy Polling 10/31-11/2 (10/21-23):
Obama 52 (52)
McCain 46 (43)
Reuters/Zogby 10/31-11/2 (10/23-26):
Obama 50.7 (52.0)
McCain 44.6 (44.8)
Rasmussen 11/2 (10/26):
Obama 51 (51)
McCain 47 (47)

CA Pol Junkie's View: Polls have tightened by about 3 points in the last week. Early voting is limited, so it all comes down to election day. Obama has a massive organization here against very little from McCain. Pollsters are probably also underestimating the number of votes Obama will get from the state's large African-American population, so I expect Obama to take Virginia in the upper single digits.

Nevada (Obama +6.8) 291 EV

Public Policy Polling 10/31-11/2:
Obama 51
McCain 47
Reuters/Zogby 10/31-11/2 (10/23-26):
Obama 50.7 (48.2)
McCain 42.9 (44.0)

Early Voting:
PPP (71% have voted): Obama 57, McCain 43

CA Pol Junkie's View: Early voting started off very strong for the Democrats, but tailed off at the end. Democrats still had about 10% better early voting turnout than Republicans. Although the remaining voters will have a strong Republican skew, 14 points is too much ground to cover among only 29% of voters who haven't already voted, so this one is very likely to go for Obama.

Ohio (Obama +5.5) 311 EV

Public Policy Polling 10/31-11/2 (10/22-23):
Obama 50 (51)
McCain 48 (44)
Quinnipiac 10/27-11/2 (10/22-26):
Obama 50 (52)
McCain 43 (41)
Reuters/Zogby 10/31-11/2 (10/23-26):
Obama 50.2 (49.7)
McCain 43.9 (45.1)
Univ. of Cincinnati 10/29-11/2 (10/18-22):
Obama 52 (49)
McCain 46 (46)
Rasmussen 11/2 (10/26):
Obama 49 (49)
McCain 49 (45)
Survey USA 11/1-2 (10/26-27):
Obama 48 (49)
McCain 46 (45)
Strategic Vision 10/31-2 (10/20-22):
McCain 48 (48)
Obama 46 (45)

Early Voting:
Quinnipiac (??% have voted): Obama 64, McCain 26
PPP (30% have voted): Obama 65, McCain 34
Survey USA (30% have voted): Obama 60, McCain 36
Rasmussen (??% have voted): Obama 62, McCain 37

CA Pol Junkie's View: One of the Democrats' twin demons, Ohio thankfully takes a less prominent role in 2008. McCain's ground game is at its best here, but Obama's is still better. His campaign knocked on 1,000,000 doors here TODAY. Polls have closed in the last week and we could see some Bradley Effect in the southeast Appalachian zone of the state, but the very strong early voting and outstanding ground game for Obama should pull it through for him. At least we have plenty more paths to victory if we are disappointed here yet again.

Florida (Obama +3.4) 338 EV

Public Policy Polling 10/31-11/2 (10/16-19):
Obama 50 (48)
McCain 48 (47)
Quinnipiac 10/27-11/2 (10/22-26):
Obama 47 (47)
McCain 45 (45)
Reuters/Zogby 10/31-11/2 (10/23-26):
Obama 47.5 (49.7)
McCain 46.2 (46.4)
Rasmussen 11/2 (10/26):
McCain 50 (47)
Obama 49 (51)
Strategic Vision 10/31-11/2 (10/20-22):
Obama 49 (46)
McCain 47 (48)

Early Voting:
Quinnipiac (??% have voted): Obama 51, McCain 38
PPP (56% have voted): Obama 56, McCain 43
Rasmussen (??% have voted): Obama 55, McCain 44

CA Pol Junkie's View: The more cynical among us might see the the Democrats as Ahab in pursuit of an elusive white whale here, but we really are in very good position to win. This is another state where the GOP has very strong organization, and it showed with absentee ballots. Democrats outdid them with early in-person voting, though. Increased turnout of African-Americans and younger Cubans moving toward the Democrats make this a tough one for Republicans to win again. Even the Republican pollster Strategic Vision thinks Obama has the edge here. I would give Obama better odds here than in Ohio, though he should win both.

North Carolina (Obama +2.6) 353 EV

Public Policy Polling 10/31-11/2 (10/16-19):
Obama 50 (48)
McCain 49 (47)
Reuters/Zogby 10/31-11/2 (10/23-26):
McCain 49.3 (46.4)
Obama 47.7 (49.7)
Rasmussen 11/2 (10/29):
McCain 50 (48)
Obama 49 (50)
Survey USA 10/30-11/2 (10/18-20):
McCain 49 (47)
Obama 48 (47)

Early Voting:
Survey USA (57% have voted): Obama 56, McCain 41
PPP (63% have voted): Obama 55, McCain 45

CA Pol Junkie's View: Forget the pollsters' bottom line, especially Survey USA for once again undercounting African-Americans. They are 21% of registered voters, 26% of early voters, but only 20% of Survey USA's sample. Survey USA also seems to think Obama will only win the demographic 88-9. The key variable is white vote for Obama, and Rasmussen and Survey USA have it at 35-36%. With most of the vote in, there isn't alot of uncertainty as to the result. If the projected white vote for Obama holds up, that's enough for Obama to win by 2-3 points.

North Dakota (Obama +1.9) 356 EV

CA Pol Junkie's View: What, no last minute polls here? Obama is strong in the prairies and his volunteer ground game continued here unabated after Obama pulled his paid staff. McCain has no organization and so is counting on history, but I think Obama is the slight favorite.

Missouri (Obama +1.6) 367 EV

Public Policy Polling 10/31-11/2 (10/11-12):
Obama 49.4 (48)
McCain 48.6 (46)
Reuters/Zogby 10/31-11/2 (10/23-26):
Obama 47.4 (48.2)
McCain 45.7 (45.7)
Rasmussen 11/2 (10/26):
Obama 49 (48)
McCain 49 (47)
Survey USA 10/30-11/2 (10/25-26):
Obama 48 (48)
McCain 48 (48)

Early Voting:
PPP (11.5% have voted): Obama 56, McCain 43

CA Pol Junkie's View: It's tough that the state is culturally half southern, has a relatively low African-American population, and limited early voting. McCain treats this as a second tier state, figuring if he loses here he'll lose enough other places so it won't matter. He hasn't devoted anywhere near the resources that Obama has. Obama's ground game pulled out the upset in the primaries here, and I think it will do the same tomorrow.

Indiana (Obama +0.6) 378 EV

Reuters/Zogby 10/31-11/2 (10/23-26):
McCain 49.1 (50.2)
Obama 43.9 (44.0)
Public Policy Polling 10/31-11/2 (10/18-19):
Obama 49 (48)
McCain 48 (46)

Early Voting:
Indiana (22% have voted): Obama 64, McCain 35

CA Pol Junkie's View: Indiana and Georgia may be the shockers of the night. The state has historically low voting rates, which makes it a prime target for a strong organization. Obama is at his strongest here, has advertised liberally, and has half of Chicago helping out. McCain is doing almost nothing. Don't be shocked if Obama does better here than he does in Ohio.

Montana (McCain +0.5) 160 EV

Public Policy Polling 10/31-11/2:
Obama 48
McCain 47
Paul 4

Early Voting:
PPP (38% have voted): Obama 61, McCain 35

CA Pol Junkie's View: Obama may be relatively strong in the prairies and Rockies, but this is a tough state. Ron Paul may help lower the threshold Obama needs to cross, and the early voting shows the power of Obama's disproportionate organization here. I'm going to disagree with the magic spreadsheet on this one, and give it to Obama by a nose.

Georgia (McCain +2.8) 157 EV

Public Policy Polling 10/31-11/2:
McCain 50
Obama 48
Survey USA 11/1-2 (10/11-12):
McCain 52 (51)
Obama 45 (43)
Strategic Vision 10/31-11/2 (10/20-22):
McCain 50 (51)
Obama 46 (45)
Insider Advantage 11/2 (10/27):
McCain 48 (48)
Obama 47 (47)

Early Voting:
Georgia (57% have voted): Obama 52, McCain 47
Early Voting:
Georgia (56% have voted): Obama 50, McCain 48

CA Pol Junkie's View: PPP and Survey USA released polls today from Georgia which include cross-tabulations. The Survey USA poll badly undercounts African-Americans: they are 29% of the registered voters, 35% of early voters and yet only 26% of their sample? Still, the two important pieces of information from each poll are Obama's percentage of the white vote and the projected total turnout. The white vote for Obama is 25% (Survey USA) or 28% (PPP). They agree that the early voters represent 56-57% of the total vote, which benefits Obama's bottom line because it is less than I had been projecting. Entering this data and PPP's measured "other race" vote for Obama into my Georgia spreadsheet, and I project an Obama win by 1.5% even without elevated African-American turnout on election day. An important assumption I make is that African-American support of Barack Obama has been underestimated in the general election just as the pollsters did in the primaries. If my projection comes true, we might end up hearing the term "Reverse Bradley Effect" for years.

Arizona (McCain +3.0) 142 EV

CA Pol Junkie's View: Now, we're getting greedy here. This is one state where John McCain should have a much stronger organization than Barack Obama. Yeah, it could go blue in a wave but I wouldn't bet on it.

West Virginia (McCain +6.7) 132 EV

CA Pol Junkie's View: This is a second-tier state for Obama, and with good reason. Although the Obama campaign has many offices and lots of dedicated volunteers hitting the streets, let's just say this is not a good cultural fit. An inordinate number of West Virginians believe Obama is a Muslim and that makes him evil. This state is not very predictable, but this would be my first guess for a state where McCain does better than his polling average.

McCain Base (127 EV): Alaska, Utah, Idaho, Wyoming, South Dakota, Nebraska, Kansas, Oklahoma, Texas, Louisiana, Arkansas, Mississippi, Alabama, Tennessee, Kentucky, 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.

CA Pol Junkie :: 4:39 PM :: Comments (22) :: Digg It!