Trends - they're a favorite talking points.  Well, I've gone through my share and attacked a few, but I was doing research for a separate diary when I found some very interesting data in the process that somehow has gone unnoticed.  By and large, independents trend toward the conservative candidate.

I was looking at data about party affiliation over time, when I found an interesting chart from Wikipedia derived from registered voters in 2004.  That is, among the 169 million registered voters at the time, 72 million (44%) were Democrats, 55 million were Republicans and lastly, 42 were independents (decline a party or third party).  That seems right, but I wanted more and found a nice seven point scale that the Center for National Election Studies has used since 1952 which can be found here: http://www.electionstudies.org/nesguide/ text/tab2a_1.txt Self-identification might be a better means of analyzing how people vote than their party affiliation.

I took the liberty of running some analysis on the data from 1980 on and here's some statistics jargon if you would like it (numbers given are percentages of the total population surveyed; apolitical responses are excluded due to their small weight and lack of likelihood to vote):

Strong Democrat Analysis:
Mean: 17.92
SD: 1.382

Weak Democrats:
Mean: 19.077
SD: 2.62

Independent Democrats:
Mean: 13
SD: 2.04

Independents:
Mean: 10.85
SD: 1.344

Independent Republicans:
Mean: 11.615
SD: 1.386

Weak Republicans:
Mean: 14.385
SD: 1.2609

Strong Republicans:
Mean: 11.923
SD: 2.2159

I lump the groups into three separate categories: Democrats (Strong and Weak), Independents and Republicans (Strong and Weak).  Ultimately, these are our relative averages.

Democrats:
Mean: 37
SD: 3.2404

Independents:
Mean: 35.462
SD: 2.6336

Republicans:
Mean: 26.308
SD: 2.2504

From this data, we can see that Democrats tend to be the largest group but also have the most spread and fluctuate more than Independents or Republicans.  Furthermore, while they are a large group, they require the Independent vote to successfully win an election.

Now, let's look at some elections data using the above means as percentages of the total voting population:

1980 Analysis:

Voter Turnout: 86,450,000

In a Theoretic World the Affiliation Breakdowns Would (Roughly) Be (Based on 1980 self-identification survey percentages):
Democrats: 26,800,000 (41% - About 2 deviations higher than the mean)
Independents: 29,400,000 (34% - Less than 1 deviation lower than the mean)
Republicans: 19,800,000 (23% - About 1 deviation lower than the mean)

Actual Results (Rounded to the Nearest 100,000):
Reagan: 43,900,000 (50.75%)
Carter: 35,500,000 (41.01%)
Others: 7,000,000 (8.23%)

Variables: John Anderson (I) ran to the left of Reagan winning 5,700,000 votes.

Analysis: Reagan and Carter both exceed their bases, but Reagan much more starkly.  Assuming that both bases voted 100% for their respective candidate (they didn't but it makes the math easier), Independents would have needed to break 8.2 million for Carter verses 24.1 million for Reagan at the very least (1:3).  Democrats have the advantage of higher self-identification with the party.

1992 Election Analysis:

Voter Turnout: 104,400,000

In a Theoretic World the Affiliation Breakdowns Would (Roughly) Be (Based on 1992 self-identification survey percentages):

Democrats: 37,600,000 (36% - Less than 1 deviation lower than the mean)
Independents: 39,700,000 (38% - More than 1 deviation higher than the mean)
Republicans: 26,100,000 (25% - Less than 1 deviation lower than the mean)

Actual Results:

Bill Clinton: 44,900,000 (43.01%)
George Bush: 39,100,000 (37.45%)
All Others: 20,500,000 (19.56%)

Variables: Ross Perot (I) ran a rightist campaign that is frequently sited as a spoiler for Bush.

Analysis: Both Bush and Clinton exceed their camps, Bush more starkly than Clinton.  Perot's campaign likely pulled Republicans and Independents.  Assuming 100% party loyalty (once again, not possible) Independents broke 7.3 million to Clinton verses 13 million to Bush (about 1:2).

2004 Election Analysis:

Voter Turnout: 122,300,000

In a Theoretic World the Affiliation Breakdowns Would (Roughly) Be (Based on 2004 self-identification survey percentages)

Democrats: 40,400,000 (33% - More than 1 deviation lower than mean)
Independents: 47,700,000 (39% - More than 2 deviations higher than mean)
Republicans: 34,200,000(28% - About 1 deviation higher than mean)

Actual Results:

George W. Bush: 62,000,000 (50.73%)
John Kerry: 59,000,000 (48.27%)
All Others: 1,300,000 (1%)

Variables: No significant spoiler.

Analysis: Bush and Kerry both exceed their camps, Bush exceeding by almost 30,000,000.  Assuming 100% party loyalty, independents broke for Bush 27.8 million verses Kerry's 18.6 million (1.5:1).

Conclusion: Independents tend to lean right in almost every situation but periodically a spoiler may steal enough away to cause a serious dent in the Republican camp.

I haven't gone so far as to run tests on this data, but I suspect there's a significance in there somewhere.  I always assumed independents could be swayed both ways but, by and large, they seem to favor the right.  What implications does this have for us long term?

There's more...

Hello all,

Back to the issues and how Edwards is leading and promoting ideas that go to solving several problems in this country. He announced a new plan to promote the development of green collar jobs, investing in TRAINING high schoolers and college students in the renewable energy field so we can slowly start pulling up to Europe on renewable energy. Have a look

There's more...

A poll released today by Surveyusa shows Hillary Clinton and Rudolph Giuliani tied for the preferences of Ohio registered voters.

There's more...