No, that's an accurate quote as far as I can tell, but it is taken out of context since he is talking about national exit poll figures, not individual state exit polls. The exit polls conducted by CNN for Florida give the aforementioned figures.
The Exit Polls will have been subject to the problems I describe above in 2000. I am sorry, but your claim that they prove your case is simply misinformed.
The number of persons actually interviewed by the exit pollers will have been very small, and the raw results are systematically distorted to conform to the reported results. Either of those could have introduced an error into the figures you cite.
Perhaps, as much as I have admitted to, but, as I said before, that is certainly not a reason to accept that the election would have turned out differently, since there is even less evidence to support that theory.
Well, I'm sorry, but Bayesian inference says that my theory is more likely.
In order for "No Nader 2000" to have netted Gore the win, there has to have been a net of 537 voters from the pool of 97,488 actual Nader voters. So if just over half of 1% more of Nader's voters would have preferred to vote for Gore than Bush if these were their only options, it doesn't matter how many chose not to vote or picked any of the other real candidates. We have as alternate datapoint in that same Exit Poll at the National Level that 25% of Nader voters preferred Bush and 38% preferred Gore. (I still would not put much stock in this but the sample size as a whole will have been 20,000 so the number of Nader voters ought to have been ~ 600, which is a more reasonable sample.) Had Florida conformed to that nationwide average, Gore would be expected to net about 13,000, as noted above. In order for Nader's absence from the ticket to not cause a Gore victory in Florida, the set of Floridian Nader voters would have to be drastically different in their preferences from the nation as a whole.
We also know that Nader's positions were in general closer to Gore than to Bush.
Now for your (uncited) Exit Poll that indicated Floridian Nader voters actually slightly preferred Bush (and here in good faith is the
2008 CNN exit polling site and a detailed explanation on
why Exit Polls can be highly unreliable).
The problem with it is that it is an extremely unlikely result given the NW Exit Poll cited immediately above and everything we know about the ideological preferences of persons voting for those three candidates.
Given the small size of the sample and the known reweighting, we have to make a choice based on that probability. Which is more likely, that Floridian Nader voters in 2000 were MUCH more conservative than the nation as a whole, or that a tiny sample has made a botched result?
Accepting your data as valid means we have to create a new theory to explain this incongruity in the apparent ideological preferences of Florida's 2000 Nader voters.
I can propose a pretty basic test. There is a Democrat who won the state of Florida in 2000, Senator Bill Nelson. He also massively outperformed Obama in 2012. He seems able to pick up poorer (White) voters in rural in suburban Florida to pad his margin and thus this by being Moderate. We know that a vast number of Bush voters voted for Nelson, it's a mathematical necessity. For the Exit Poll to be credible, we would expect there to be a correlation between Nader's vote share and Nelson's runahead over Gore, since that would show more conservatively or moderately minded voters taking Nader as a protest vote. If there is no correlation or an anti-correlation, then that makes the exit poll being useful much less likely.
