Thursday, October 28, 2004

Analysis of the Polls a Fortnight Before the Election, and My Exact Predictions

I shall look at the "battleground" states, and compare the current polling with two previous listings. Bush leads will be in blue; Kerry leads in red; ties in green. New Hampshire, Hawaii, and Florida are now also considered toss-up states, and so have been added to my analysis.

From Rasmussen state-by-state polls of electoral votes:

Pennsylvania (21)

47%-47 (Oct. 2 poll)
Kerry 47 Bush 46 (Oct. 13)
Kerry 49 Bush 46 (Oct. 24)

Ohio (20)

Bush 48 Kerry 47 (Oct. 3)
Bush 49 Kerry 47 (Oct. 14)
Bush 50 Kerry 46 (Oct. 26)

Michigan (17)

46-46 (Sep. 30)
Kerry 49 Bush 46 (Oct. 13)
Kerry 51 Bush 46 (Oct. 24)

Wisconsin (10)

Bush 49 Kerry 46 (Oct. 1)
No new poll for my second post
Kerry 48 Bush 47 (Oct. 18)

Minnesota (10)

46-46 (Sep. 26)
No new poll . . .
Bush 49-46 (Oct. 26)

Colorado (9)

Bush 48 Kerry 44 (Oct. 2)
No new poll . . .
Bush 50 Kerry 45 (Oct. 20)

Iowa (7)

Bush 48 Kerry 45 (Sep. 26)
Kerry 50 Bush 46 (Oct. 12)
Bush 48 Kerry 46 (Oct. 24)

Nevada (5)

Bush 47 Kerry 45 (Sep. 24)
No new poll . . .
Bush 49 Kerry 47 (Oct. 27)

New Mexico (5)

46-46 (Aug. 18)
No new poll
Bush 48 Kerry 44 (Oct. 27)

Florida (27)

Bush 48 Kerry 48 (Oct. 25)

New Hampshire (4)

Kerry 49 Bush 47 (Oct. 20)

Hawaii (4)

No information.

Rasmussen currently has Bush leading the electoral college 222-203 (538 total; 270 needed to win), and among likely voters 48.9% to 46.9%.

I'm predicting that Bush will take Ohio, Minnesota, Colorado, Iowa, Nevada, New Mexico, and Florida. Kerry will take Pennsylvania, Michigan, New Hampshire, Wisconsin, and Hawaii. Assuming the other projections in Rasmussen stay the same, that is a Bush victory: 296-242 (54 points and 32 states to 18; I predicted a 60-point margin of victory and 32-18 on September 10th). If Bush loses Florida (with all else the same), it's a tie: 269-269. If he loses Minnesota, Iowa, and New Mexico, he wins 274-264. If he loses Florida and Minnesota, he loses: 259-279. If he loses both Florida and Ohio he loses 249-289. But that seems unlikely. If he loses Ohio and Minnesota, he loses: 266-272. But he can reverse that by taking either New Hampshire or Hawaii (winning, 270-268).

If Kerry can manage to take Minnesota, Iowa, Nevada, and New Mexico, with all else the same in my prediction, that would be a tie too. Bush grabbing New Hampshire or Hawaii could break either tie (273-265). The House determines who wins in a tie, and it is controlled by Republicans. But I am sticking to Bush winning both Florida and Ohio. The most likely battleground states that he would lose are Minnesota and Iowa. This would still be a victory for him: 279-259. Kerry needs some major victories and upsets to pull this off, if Rasmussen polling is a good indication of actual voting. Florida is clearly the most important state still up for grabs. If Kerry can "steal" Florida from Bush, with all else the same, it is a tie (but then Bush would win with a House vote).

My actual predictions for each state, then, are as follows:

President George W. Bush (32 states; 296-242 in the electoral college, or +54 points; September 10th prediction: 32-18 [states not specified], and +60 points; I also predicted on that date a 53-45% margin of the popular vote, with 2% for Nader, and I am sticking to it. Vice-President Cheney predicted 52-47 a few days ago)

Alabama
Alaska
Arizona
Arkansas
Colorado
Florida
Georgia
Idaho
Indiana
Iowa
Kansas
Kentucky
Louisiana
Minnesota
Mississippi
Missouri
Montana
Nebraska
Nevada
New Mexico
North Carolina
North Dakota
Ohio
Oklahoma
South Carolina
South Dakota
Tennessee
Texas
Utah
Virginia
West Virginia
Wyoming

Senator John Kerry (18 states and Washington D.C.)

California
Connecticut
Delaware
Washington D.C.
Hawaii
Illinois
Maine
Maryland
Massachusetts
Michigan
New Hampshire
New Jersey
New York
Oregon
Pennsylvania
Rhode Island
Vermont
Washington
Wisconsin

I have long predicted that Bush would carry the South. Now I will predict that he will take every single Southern state. The only reason Florida is even close is because of the large transplanted liberal Northern retiree population. All Kerry can win are the liberal east and west coast states and industrial midwestern states with large cities and lots of factories (labor unions) and working class people and minorities (which usually vote Democratic).

The The Real Clear Politics website polls for the battleground states are now as follows (after my two previous listings):

Penn.

Kerry 49 Bush 45
(same)
Kerry 49 Bush 47 (average: 10/22-10/28)

Ohio

Bush 49 Kerry 47
Kerry 48 Bush 47
Kerry 48 Bush 46 (average: 10/20-10/28)

Mich.

Kerry 49 Bush 46
Kerry 50 Bush 44
Kerry 48 Bush 45 (average: 10/20-10/28)

Wisc.

Bush 49 Kerry 44
Bush 47 Kerry 44
Bush 47.2 Kerry 46.6 (average: 10/14-10/28)

Minn.

46 46
Kerry 47 Bush 43
46 46 (average: 10/19-10/28)

Colo.

Bush 50 Kerry 45
(same)
Bush 49 Kerry 45 (average: 10/14-10/27)

Iowa

Kerry 48 Bush 47
(same)
Bush 47 Kerry 46 (average: 10/23-10/28)

Nev.

Bush 49 Kerry 45
(same)
Bush 50 Kerry 46 (average: 10/19-10/27)

N.M.

47 47
(same)
Bush 48 Kerry 45 (average: 10/15-10/28)

Florida

Bush 49 Kerry 46
Bush 49 Kerry 47 (average: 10/20-10/28)

New Hampshire (4)

Kerry 47 Bush 45 (average: 10/14-10/21)

Hawaii (4)

Bush 45 Kerry 44 (average: 10/13-10/20)

RCP currently has Bush leading the electoral college 232-207 and among likely voters 48.4% - 46.1%.

RCP's last five major polls listed all favor Bush:

Reuters/Zogby (1206 LV)
10/25 - 10/27
48% - 46% - 1% (Nader) Bush +2

ABC/Wash Post (1747 LV)
10/24 - 10/27
49% - 48% - 1% Bush +1

TIPP (792 LV)
10/24 - 10/27
47% - 44% - 2% Bush +3

ICR (741 LV)
10/22 - 10/26
48% - 45% - 2% Bush +3

CNN/USAT/Gallup (1195 LV)
10/22 - 10/24
51% - 46% - 1% Bush +5

Notable Bush Trends in Some Recent Polls for the Battleground States

Fla.

Zogby (10-28) 48-47
Quinnipiac (10-26) 49-46
LA Times (10-26) 51-43

Pa.

Quinnipiac (10-26) 49-47

Iowa

ARG (10-27) 48-47
CNN/USAT/Gallup10/23-26 50-46

Minn.

Zogby (10-28) 46-45
Humphrey Inst (10-27) 47-44
Rasmussen (10-26) 49-46

Mich.

Zogby (10-28) 47-45

N.M.

Zogby (10-28) 49-43

Nevada

Zogby (10-27) 51-44

Hawaii

SMS Research (10-20) 46-45
Honolulu Advertiser (10-18) 43.3-42.6

Possible Bush "upsets" in Michigan or Pennsylvania could change the outcome considerably, as those states have 17 and 21 electoral votes, respectively.

Real Clear Politics states in its October 25th commentary on the electoral college:

Iowa, Wisconsin, Minnesota and New Mexico are fully in play: eight days before the election Bush holds leads in the RCP State Averages in all four of these states. This is seriously complicating Kerry's strategy in getting to 270 Electoral Votes. Conventional wisdom for months, including RCP's, had been that whoever won two of the "big three" Ohio, Pennsylvania and Florida would almost certainly become President.

While it may still be likely that whoever carries two out of those three will win the election, it is not the cut and dry proposition it was earlier. President Bush can offset a loss in Ohio (and New Hampshire) by carrying Wisconsin and either Iowa, New Mexico or Minnesota. He can offset a loss in Florida (and New Hampshire), by winning three of those four states. Winning Wisconsin, Iowa, Maine's 1 Electoral Vote and holding New Hampshire would also allow President Bush to gain reelection while losing Florida.

. . . The problem for Senator Kerry is he has no backup plan to not winning in either Florida or Ohio. The problem for President Bush is that Kerry is still very much alive in both those states. All of Bush's backup Electoral scenarios will be irrelevant if he loses FL and OH and Kerry hangs on to PA and MI.


No comments: