Projected Electoral College Vote: Swing States, 10/06/04


Here’s a surprise: The WSJ/Zogby Swing State Polls has Kerry significantly ahead in electoral votes. A new projection of the Electoral College map, based on online polling done in 16 battleground states by Zogby Interactive, shows that under the survey’s current trends, and assuming Mr. Kerry wins states where he has razor-thin leads now, he would have 322 electoral votes and the president would have 216:

Click for larger graphic

Here’s an excerpt from the Journal:

A strong performance in the first of three presidential debates appears to have helped Sen. John Kerry among likely voters in 16 battleground states, according to the latest Zogby Interactive poll. Mr. Kerry now holds leads, albeit some of them razor-thin, in 13 states, an improvement from the 11 he held three weeks ago.

Notably, Mr. Kerry overtook President Bush in Ohio, a critical swing state where he had been trailing since mid-July. However, the senator’s position there is tenuous — the lead, by 0.3 percentage point, falls well within the margin of error and is one of the closest in the latest poll. Mr. Bush, meanwhile, also dropped Nevada but continued to lead in Missouri, Tennessee and West Virginia — though the spread narrowed in all three.

In total, Mr. Kerry holds leads outside the margin of error in six states, including Michigan and Pennsylvania; none of Mr. Bush’s leads are outside the margin. The margin of error varies between +/- 2.2 and +/- 4.3 percentage points for each candidate.

The WSJ does note, however, the dynamic and unsettled this race is: “The race remains volatile, and reading conditions across battleground states can be more art than science, given wide variations in polls from state to state, and the sometimes-conflicting opinions of party strategists. The exercise is made all the more difficult this year by higher levels of voter registration and intense partisan feelings in both political camps.”

Makes for an interesting compare and contrast with the recent Gallup numbers . . . Its worth recalling that Zogby was the most accurate pollster in the 2000 presidential election, while Gallup had then Governor Bush with a double digit popular vote lead right before the actual election (Bush lost the popular vote by at least 540,000 votes).

Final thoughts: The table at the site (here) shows five of the states credited to the Dems — Arkanasas, Tennessee, Ohio, Nevada, Florida — with a Kerry lead of less than the margin of error (MOE) suggests the race is tighter than this study makes it appear . . .

Back out the states within the MOE, and we pull out 68 electoral votes for Kerry (AK, FL, NV, OH, WI) and 22 electoral votes for Bush (MO, TN), for a total 254 to 194, in Kerry’s favor.

Battlegrounds States Poll – August 23, 2004

Electoral College Analysis

Battlegrounds Contract as Race Tightens
Bush and Kerry Give Up Hope On Handful of States and Slug It Out More Intensely in Others
By Greg Hitt
The Wall Street Journal, October 7, 2004; Page A4,,SB109710396352238596,00.html

Print Friendly, PDF & Email

What's been said:

Discussions found on the web:
  1. BOPnews commented on Oct 7

    Projected Electoral College Vote: Swing States, 8/06/04

    Here’s a surprise: The WSJ/Zogby Swing State Polls has Kerry significantly ahead in electoral votes. A new projection of the Electoral College map, based on online polling done in 16 battleground states by Zogby Interactive, shows that under th…

  2. Mememomi commented on Oct 7

    Political winds gathering in a new direction

    Strong evidence the debates are indeed giving positive traction to Kerry/Edwards.

  3. Outside The Beltway commented on Oct 7

    Zogby: Kerry 322, Bush 216

    Barry Ritholtz passes on a word that a WSJ/Zogby poll has Kerry way up in the electoral college race, contradicting virtually every other poll known to mankind. For example, Slate’s projection two days ago had it Bush 348, Kerry 190.

    A new WSJ.c…

  4. Jim Durbin commented on Oct 7

    I take Zogby Interactive seriously about the same way I believe that 99% of the citizens of Philadelphia felt John Edwards won the VP Debate.

    It’s embarassing for the WSj to use a self-selecting internet poll.

  5. bryan commented on Oct 7

    Actually, it appears that at least Zogby “tries” to get around the obvious non-randomness of his sample by adding weightings and rotating which of the respondents are used in any given survey. But the fact of the matter remains that these are overwhelmingly self-selected survey participants. It’s not like the typical “online poll,” but it suffers more traditional methodological problems.

  6. tal commented on Oct 11

    All telephonic polls, and I mean all, have the bias of not taking into account cell phone users. To be more precise, people who exclusively use cell phones. These are, I assume, mainly young adults. Probably Zogby catches those youngsters online.

Read this next.

Posted Under