Swing States in 2008

[ATFFL]

Shira, while you're method might have had a little use this year, it's useless for 2008:

George W Bush is not allowed to seek re-election in 2008

It is not Bush specific.
The "State minus National" number is the best indicator showing how Republican or how Democratic a state is.  It completely eliminates the impact of the national outcome.
If the election were absolutely close (49.5% to each candidate) then the number the GOP and the Dem got in a state would have reflected their real power in that state.
If this is not exactly the case, then you should make the correction.
In last election the real power of the GOP in a state is the number they got there minus 1.5%, because nationally Bush got 1.5% above 49.5%.   
If Bush got 55% in a state then it is a 5.5% victory. These 5.5% should be separated to 1.5% + 4%, where 1.5 is the national component and 4 is the state-local component, which reflects the GOP power in that state.

Except for the fact you have causality all backwards.  The national numbers do not cause the states to move, the states cause the national numbers to move.

Lets say that the Kerry GOTV machine cranked things up in NY and CA for some reason.  If he had gotten an additional 4 million votes in those 2 states and won the PV -- but no other state had a single vote difference -- you would be on here talking about how dramatically some states swung to Bush by comparing the state result to the national average. 

There is no national popular vote.  It is a fiction created by the media.  If one party had 5 million votes, and the other had 30, the party with 50 wins if they win 1-0 in states with enough EVs.

There are 51 local elections.

[ATFFL]

Lets say that the Kerry GOTV machine cranked things up in NY and CA for some reason.  If he had gotten an additional 4 million votes in those 2 states and won the PV -- but no other state had a single vote difference -- you would be on here talking about how dramatically some states swung to Bush by comparing the state result to the national average. 

Impossible situation. Had Kerry gained 4 million in these two states he would have proportionally gained in all other states.
Nothing in presidential race is CA specific.  If Kerry had gained, for example 7% in CA then he would have gained 5% - 9% in each and every state including UT and WY.

That makes less than zero sense.  You are saying when my friend in Alabama decided to vote people in each and every other state decided to vote as well?  Why?  How?

There is no national race and the majority of election events are local and not national.  The first swift boat ads were national events because of the media coverage and Kerry making them a national issue.  The later ones I only saw thanks to the Internet.  Moveon.org ads were never national events, they only aired in a few key battleground states.

Despite all the media coverage stump speeches remain a local event.  This is why every candidate uses the same stump speech for weeks or more at a time.

GOTV drives are inherently local.  If I drive a dozen friends to the polls in NC to vote for Bush I do not drive a dozen people in each and every state, just my own.  Even a nationwide GOTV drive, such os the ACT one and Moveon's drive, are a collection of local drives.  Just because they drive 100,000 voters to the polls in Florida does not mean the Ohio drive will succeed in getting a single voter to the polls.

If what you were saying is true, if all states voter turnout was driven by the national numbers, turnout in each state would be the same.  An increase in California would be perfectly mirrored in each and every state.  It simply does not happen.

[ATFFL]

Could you recall one single case, where between two consecutive elections, a party gained 6% or more in state A and lost 6% or more in state B?

Not quite, but I can give you an example of an 8% gain with a 5% loss, which is actually more extreme than what you asked for.  THough to do so I have to reach back all the way in time to a few weeks ago.

Vermont:
2000: Gore 50.63%
2004: Kerry 58.94%

Alabama:
2000: Gore 41.59%
2004: Kerry 36.86%

Here is one from a little further back:

Lousiana:
1960: Nixon 28.59%
1946: Goldwater 56.81%

Massachusets:
1960: Nixon 39.55%
1964: Goldwater 23.44%

[ATFFL]

Could you recall one single case, where between two consecutive elections, a party gained 6% or more in state A and lost 6% or more in state B?

Not quite, but I can give you an example of an 8% gain with a 5% loss, which is actually more extreme than what you asked for.  THough to do so I have to reach back all the way in time to a few weeks ago.

Vermont:
2000: Gore 50.63%
2004: Kerry 58.94%

Alabama:
2000: Gore 41.59%
2004: Kerry 36.86%

In Vermont the Dem practically gained only 2.63%. The 50.63% is not the real number of Gore. In 2000 Nader got over 6% in VT.
You can see it with Bush who droped 2.56% during the same period. (Buchanan got few fractions of a percent in 2000).
In Alabama Bush gained 5.54% from 2000 to 2004. 
I will later look into the 60-64 example  you have brought, but as to Vermont-Alabama
we have -2.56 vs. +5.54 and not -6 vs. +6.

I'm sorry, but those people who voted for Nader voted for Ralph Nader, not Al Gore and not John Kerry.  Nader was even on the ballot again this year so the people could have voted for him again if they chose.

Actually, if you add all other votes to George Bush I can say he took 100% of the vote, which puts him up at the level of Saddam Hussein and the old Communist Party in the Soviet Union.  So if we were a one party state Bush would have 100% of the vote.  Yay for ignoring those troubling other candidates!

[ATFFL]

Nader went from 6.92% in 2000 to 1.44% in 2004 in Vermont. Where did that 5.48% go? The most logical answer is they voted for Kerry.

Which would mean the democrat gained votes.  They did not vote democrat in 2000, no matter how you slice it.  In 2004 they did.  Thus the democratic party gained votes.

[ATFFL]

Shira is basically talking about votes shifting between the two main parties though.

Why?  We are in a two party dominated system, but we also have third party spoilers. 

She was talking about one party gaining 6% in one state while losing 6% in another.  No mention of taking it all from the other major party. 

[ATFFL]

Shira is basically talking about votes shifting between the two main parties though.

Why?  We are in a two party dominated system, but we also have third party spoilers. 

She was talking about one party gaining 6% in one state while losing 6% in another.  No mention of taking it all from the other major party. 

When you do comparison analysis, you have to eliminate aberrations like Perot, Nader and Buchanan (to a lesser extend). You have to decide what would had been the results without these ‘aberrations’. In the case of Nader and Buchanan it is very simple.

They are not abberations, they are candidates.  Their level of success indicates a change in the mood and preferences of he electorate, not statistical noise.

If the US army wants to compare casualties from war to war should they decide there are only 2 categories, killed and alive?  Since all the people who were wounded are not dead we can call them alive and pretend our casualty rate is much, much lower.  Sure the data will be useless for what we were supposed to be comparing, but it makes us look good.

When I grade a multiple choice test I should only accept two of the four possible answers.   I'll assume the other two incorrect ones are abberations and give the kids credit for the right answer.  They may love me for it, but their education won't benefit from it.

I bet scientists would love it if they could decalre the search for the Grand Unification Theory over and done with and ignore all the remaining ununified forces by calling them abberations.

Life would be great if we could just ignore all those little abberations tht come up and pretend they don't exist.  Sadly, we can't.  To ignore the third party candidates and say they are nothing more than stasticial abberations is to discount the people who voted for them whether out of protest or out of genuine belief in their views.  Every vote for Ralph Nader was a vote that did not go to John Kerry, Al Gore, George W. Bush, Michael Badnarik or any other candidate who appeared on a ballot anywhere in this nation this year or four years ago.