Presidential election maps relative to the national margin

[Gustaf]

If I understood this correctly, here's 1996, using his system...

[Gustaf]

Oh, come on...
Percent margin compared to national percent margin for the GOP:

Colorado:

2004: +2%

2000: +9%

1996: +10%

1992: +1.5%

1988: 0%

1984: +10%

1980: +14%

On average Colorado was roughly 7.5% more Republican than the national average during the period 1980-2000. And it was 2% more Republican this election. No trend here, none at all...

Nevada:

2004: +0%

2000: +4%

1996: +8%

1992: +2%

1988: +13%

1984: +16%

1980: +26%

Once again, ladies and gentlemen, no trend is evident...

Ohio:

2004: -0%

2000: +4%

1996: +2.5%

1992: +3.5%

1988: +3%

1984: +0%

1980: +1%

Here, I'm tempted to agree. There is no clear trend, Ohio remains slightly more GOP than the nation, just like Pennsylvania is a few points more Democratic than the average.

[Gustaf]

Percent margins are not relevant? So...since both the Democrats and the Republicans improved they're trending in all directions? Or are you saying that you know how the Nader voters and Perot voters would have voted.

Besides, I was one of Shira's critics, just so you know.

I'm not saying that the comparison I made is a perfect tool that provides the ultimate truth. But it's at least an indication. Please explain why your "analysis" is superior. (If I understand it, you're looking at the change in percentage for one side and disregarding national trends?)

[Gustaf]

People vote for who they want to win, not how much more Republican or Democrat they want to be than the national margin.

A trend is a consistent movement within a given area. If a Democratic state suddenly starts electing more and more Republicans, and voting more and more GOP in presidential elections, it's trending Republican. Doesn't matter if other states are becoming even more Republican. According to your comparison, that state would be "trending Democrat," which I think you would concede is inaccurate.

You are effectively disregarding the actual opinions of people in a given area, and saying that because the nation moved Republican faster, they moved Democrat, when in reality, the opposite is true.

You have a point. Would you say that every state in the country was trending Republican in 1972? Or that Maine, for instance, was trending republican in 1968 since Nixon improved on GOldwater's result? You're ignoring the fact that candidates change and the results change with them.

In this case you're ignoring the fact that more and more people in these states vote Democrat. At a higher rate than more people are beginning to vote Republican.

[Gustaf]

So you account for the increased Republican margin how exactly?

[Gustaf]

The Republican candidate got a larger percentage of the vote. I don't have to 'account' for that; it's a fact.

You said that the increase in the Democratic vote share didn't count because it was the result of "reduced vote splitting". So...how do you account for the increase in Republican votes, since that apparently qualifies as a trend?

[Gustaf]

There you go with assumptions that may or may not be correct. The point is that the Democrats gained more votes in Nevada and Colorado than Bush did. How is that a positive trend for Republicans? And if you're assuming that there was a left-winged majority in 2000 that wasn't there before and then disappeared in this election in those states, please deevlop this theory further...

[Gustaf]

Your theory that there was a swing to the left in 2000 breaking a conservative trend and then reversing back to conservative. Why did this happen?

[Gustaf]

You said that the Nader votes should be seen as Democrats. I assume that we can view Buchanan voters as republicans? If you lump together libertarian reform and Republican candidates v Democratic and green for Colorado you get:

2004: Left 47.62%, Right 52.05%
2000: Left 47.65%, Right 52.04%
1996: Left 46.09%, Right 53.21%
1992: Left 40.13%, Right 56.74%
1988: Left 45.28%, Right 54.53%

For Nevada:
2004: Left 48.46%, Right 50.85%
2000: Left 48.44%, Right 50.84%
1996: Left 44.95%, Right 53.34%
1992: Left 37.36%, Right 61.28%
1988: Left 37.92%, Right 59.87%

This is definitely not a Republican trend, IMHO.

[Gustaf]

The results of 1992 and 1996 are worthless because of Perot, a tax-hiking wing-nut. Nice try counting him as a right winger, though.

Remove them, and you have one landslide election, and then 2000 followed by 2004.

I said movement, not trend. You're the one trying to show a Demcoratic 'trend.'

I'm using another model. But what does the landslide mean? I thought you WANTED absolute numbers, and not relative ones?

I'll make one more attempt at explaining the point of using the national margin as a reference point. Just for you.

There are certain inherent gographic factors. Such as the people of Utah being Mormons or the coal mines in West Virginia, the Catholics in RHode Island and so on. This creates leans to one side or the other. So Utah will pick a GENERIC Republican over a GENERIC Democrat and so on. But candidates aren't generic. So in an actual campaign other things decide. Such as Reagan being so charming or Bush sr. raising taxes, etc, etc. This creates a national result. Those, however are not trends.