1. If John Kerry had won 13,300 more votes in Iowa, 11,000 additional votes in New Mexico and 21,600 more votes in Nevada (a total of fewer than 46,000 votes), then the Electoral College would have turned out 269-269.
I can readily imagine ongoing demographic changes turning New Mexico and Nevada into Democratic states very soon, but what is going to stop wannabe urbanites from fleeing rural Iowa for Chicago?
In percentage terms, Ohio was about as close as these three states, but the vote margin was 137,000 (to date, those provisional ballots still aren't tabulated)
2. Obviously, Kerry also narrowly won a couple of states. New Hampshire and Wisconsin were won by a total of just over 20,000 votes. Add those to Bush's column and he gets to 300 electoral votes. Pennsylvania and Michigan were about as close as Ohio.
Demographically, I'd guess that these states are almost sure to remain "swing states" into the foreseeable future.
3. It looks like Bush got fewer votes in 2004 than he did in 2000 in only 3 states: California, Washington and Maine. Democrats won them all in both elections. I wonder if the two border states on the northern corners of the mainland vote more like Canadians than any other states?
The AP numbers I'm studying show Kerry also received fewer votes in California 2004 than Gore got in 2000, so I wonder if the numbers are incomplete or if turnout was actually down in the most populous state?
4. Bush's percent of the vote went down from 2000 to 2004 in only Vermont and Virginia. The latter must be a good sign for Democrats for 2008 and beyond. Again, there are underlying demographic shifts at work that help explain this outcome.
While Bush got a greater share of the vote in most other states in 2004 than he did in 2000, his percent of the vote didn't change in DC, Nevada, North Carolina, Oregon, South Dakota and Wyoming. I'm not sure there's much to learn from this information, though this may be another sign about Nevada's future.
5. Bush currently has about 3.6 million more votes than Kerry (some are still being tabulated, but not that many). In most states, largely because of higher turnout, Bush 2004 beat Bush 2000 and Kerry 2004 beat Gore 2000. It is not always true that Kerry 2004 beat Gore 2000 plus Nader 2000.
In any event, if one subtracts out Kerry's vote increase over Gore 2000 from Bush's growth, one can identify the greatest sources of Bush's gains in 2004. Put simply, one can find the states that delivered the popular vote margin of victory in 2004.
Here are a few interesting states:
New York, Bush improved his vote margin by 500,000So far, I haven't switched a single Electoral vote and I've found 1.75 million of Bush's 3.6 million margin...or about half. In some states, NY, CT & NJ, this meant narrowing the loss. In others, it meant winning bigger (TX, AL, GA, IN).
New Jersey, Bush improved by 300,000
Texas, Bush improved by 250,000
Georgia, Bush improved by 230,000 votes
Alabama, Bush improved by 200,000
Indiana, Bush improved by 170,000 votes
Connecticut, Bush improved by 100,000
In these 7 states, the winner got at least 7% more of the vote than the loser in both 2000 and 2004. Typically, the margin of victory was much higher. These are mostly 60-40 states.
Put differently, Bush won a hell of a lot more votes in 2004 in states that were not at all contested. I could have added Tennessee, where Bush added another 280,000 votes to his margin and Oklahoma, for another 200,000.
Most of the highly contested swing states didn't move much in Bush's direction, with one exception: Florida. There, Bush improved his margin by 400,000 votes over 2000.
In the states Kerry seriously contested, the country had a very close election. That is why I began above with the points about the really close swing states. The closest elections occurred in Iowa, Nevada, New Hampshire, New Mexico, and Wisconsin, as well as the big three of Michigan, Ohio and Pennsylvania. These were 8 of the 9 states Kerry spent most of his time and money and he very nearly won the election. Florida was the 9th state and the only swing state featuring a big move toward Bush.
Where Kerry did not campaign, the President racked up huge vote margins. I've now listed 10 states with significant Bush movement and found 2.6 million of his 3.6 million victory margin in the popular vote.
Remove the 5 states that were virtually tied and that means the other 35 states explain only 1 million of the 3.6 million margin of victory for the President.
Thus, it looks like incumbency was worth a point or two in virtually every state in 2004 and Kerry's strategic decision not to campaign across the country likely depressed his potential vote total all over the map. Votes in large areas of the country were simply not in play, meaning that residents of those states had four years of Bush to weigh against the relatively unknown Kerry.
Move fewer than 200,000 votes and Kerry would have won the Electoral College 289-249, despite losing the popular vote by perhaps 3.4 million votes.
I think this is a good reason to abandon the Electoral College in the near future. Voters in solid red or blue states just didn't matter in this election. Urban issues were rarely discussed because big cities are mostly in solid Democratic states. Agriculture issues were rarely discussed because farming areas are mostly in solid Republican states.
The swing staters were asked to vote on gay marriage and national security. These issues were important to many voters, but it wasn't a particularly rich policy debate. Remember the Vice Presidential debate that featured two questions about gay marriage and none on national health insurance?
Finally, Kerry's strategy may have hurt the Democrats in various congressional contests and Senate races. For example, Democrat Daniel Mongiardo lost in Kentucky by a very small margin (less than 23,000 votes, 51-49%). Kerry really didn't campaign in the state and incumbent Senator Jim Bunning ran TV ads linking Mongiardo to Kerry.
There are lots of interesting lessons to be culled from the 2004 voting data.