Which states have trended Democrat more than the national average since 2000?

[Lunar]

Three questions I'm honestly curious about, sorry if this is redundant on the board somewhere:

1) Which states have trended Democrat more than the national trend since 2000 (the ones that have shifted the most towards the Democrats)
2) Which states have trended Democrat slightly but less than the national average
3) And which states, despite the Republican brand problems, have bucked the national trend and headed for the Republicans?

I tried to make my own guesswork-based map, but I realized I was in confusions towards way too many states, including the South, New England, and Midwest.

This is 2000-2008.  Cheers

[Lunar]

Central question:  How do you want Nader votes treated?

Good question.  Incorporated to the extent that they are 'Democratic' but dissatisfied with that particular Democratic candidate and/or enamored with Nader's cult of personality.  It's hard to say exactly what I mean by this, I hope that's an adequate explanation.  It's like if Huckabee ran as an independent and did well in the South, it wouldn't necessary show a shift away from Republicanism overall.

I'm really leaning towards identification over registration but of course the latter is more easy to consolidate into factual and convenient numbers.

[Lunar]

True dat. 

Hmm, I guess I mean identification on the national level.  I've been putting myself to sleep lately every night pondering what states have accelerated their Democratization (funny with a big 'D') and what states less so.

At the very least we should have a general idea of what states, like Florida, are becoming more out of reach for the Democrats every year, right?  What states have rapidly growth among demographics that belong to the Republican party?  Gosh even this approach befuddles me because I don't think evangelicals outside of Mormons have huge migration trends.

It's just so obvious that Virginia and Colorado are flying down the Democratic highway while California is sort of stagnating, seems like we should be able to do something.  Maybe without quantifying, can we just guess enough to get a consensus that roughly makes sense?

[Lunar]

Identification causes trouble in that it brings along with it registration in several states, most notably Oklahoma.  You still manage to find a trend, of course.  In fact, it may be sharpened by the fact that there is a "softer" Democratic base to work from.  But I don't think they're "honest" numbers to begin with.

I know what you're trying to quantify, and I'm just not sure it's possible.  Even federal elections aren't totally federalized, if you catch my drift.

What if we looked at the states that have had Bush's approval numbers decline the least and most relative to 2004?