Bush 48.39% - 281| Kerry 46.92% - 257
June 22, 2004
In explaining this analysis, I've said a number of times that the numbers aren't intended to say specifically that if the election were held today, Bush would win Ohio by 0.07% of the vote. Sure, I hope the numbers approximate the actual results when I do my last analysis on the Monday before the election. But I think the real value of this work is to show trends. That's also a benefit of the other sites analyzing the electorate and trying to gauge the direction of poll movement.
At the bottom of the page is a comparison of the electoral vote margins predicted by a number of different sites. It seems apparent that during May, the trend was in Kerry's favor. And based on the news in May, that makes some sense. But in June, the focus shifted off of Iraq momentarily to Ronald Reagan and then back to Iraq which suddenly had a new President and Prime Minister, and the primary face of Iraqi opposition, Moqtada al Sadr, was shutting down much of the violence of his movement.
Falujah calmed down and the enemies of America in Iraq continued to show their barbarism and that they deserved to be treated as harshly as possible. As a result, the trend seems to be turning back in Bush's favor.
During the past week, Harris released a poll showing Bush ahead by 10 among likely voters, and that poll naturally effects this week's composite average. But it looked like the rare poll that must have fallen outside the margin of error. So, we'll see in next week's poll whether this apparent pro-Bush trend holds up. Early indications are that it might, as a FoxNews poll has been released (after the June 22nd date of this analysis), showing Bush leading by 7 among likely voters.
This week's results include the results of polls in 10 states, most notably Florida (Bush +7), New Jersey (Kerry +6) and North Carolina (Bush +5).
One more thing. I've received complaints that I shouldn't be showing two digits past the decimal, because this suggests that I'm representing these digits as significant figures. I don't think that at all. But I am calculating an electoral vote each week, allocating all 538 electoral votes in order to identify trends developing - no matter how small. Thus, if I only reported whole integers, I'd report some states as being tied, yet I'd be allocating their electoral votes per my analysis. So, I'll show that, in Ohio this week, it really is tied, but because my formulae give me a 0.07 margin in Bush's favor, I'm calling the state for Bush.
But you are reading it right when you say, "Not so fast, Ohio is tied." Just remember, I'm trying to show trends, that's why you'll find the graphs below.
Harris (Bush 51 | Kerry 41)
Investor's Business Daily / CSM / TIPP (Bush 44 | Kerry 41)
ABC / Washington Post (Bush 44 | Kerry 48)
Pew Research (Bush 46 | Kerry 42)
Rasmussen Week Average (Bush 45.9 | Kerry 45.7)
Iowa Electronic Market (Bush 51.0 | Kerry 46.5)
Fox News/Opinion Dynamics (Bush 42 | Kerry 42)
Investor's Business Daily / CSM / TIPP (Bush 43 | Kerry 40)
AP/Ipsos (Bush 46 | Kerry 45)
LA Times (Bush 42 | Kerry 48)
Rasmussen Week Average (Bush 45.9 | Kerry 45.1)
Iowa Electronic Market (Bush 50.1 | Kerry 47.3)
Bush State Gains
New Mexico
Ohio
Kerry State Gains
None
Toss Ups (slight lead of less than 2%)
New Hampshire (Kerry)
New Mexico (Bush)
Ohio (Bush)
Oregon (Bush)
Pennsylvania (Kerry)
West Virginia (Kerry)
Wisconsin (Kerry) *NEW*
For the rest of the data, maps and graphs, see the Composite Poll page.
