What I am finding amazing post-election is how much idiotic analysis is being foisted on us, primarily by the mainstream media. Much discussion seems to assume the margin in the election was larger than in fact it was.
This was a close election by any measure. Bush won by 3% in the beauty contest, the national popular vote, but in the electoral college, by just 2.5% in Ohio and 1% in Iowa and New Mexico.
I thought Kevin Drum of Political Animal said it best in his own swipe at the media. And Josh Marshall is running a contest to identify the most ludicrous overstatement of the Republican victory.
Today's New York times had a typically weak analysis of the Republican win in Florida focusing on votes in the I-4 corridor. The Republicans did do well there but in seeking explanations, instead of trite observations about the Republicans being well organized and having many volunteers, I like the results of the Florida exit poll question on government response to the hurricanes. This seems to me a key factor given that 87% expressed approval of government action among whom 57% voted for Bush and 43% for Kerry. The I-4 area was the region hardest hit by the hurricanes this year. See here and here, for example.
Overall, looking at Bush's gains among, for example, women and big city dwellers, and given that he campaigned non-stop on the war on terror, it strikes me that as an overall explanation for his slight edge this was the most important influence. He always had a perceived advantage (however unwarranted as Richard Clarke's book makes clear) and exploited it. Many have attributed Republican gains to moral issues but it appears that there was little difference in this respect between 2004 and 2000. This Slate article argues these points well. Also see this post on the Emerging Democratic Majority site.