A Conversation about False Recall

 

TruthIsAll

 

Analyst

I see that you are still using False Recall to explain the 2004 Final National Exit Poll 43/37% returning Bush/Gore voter mix as well as the 2008 Final 46/37% returning Bush/Kerry mix. You have agreed that they are mathematically impossible since they imply that there were millions more returning Bush voters than were alive.

 

Naysayer

Correct. The mixes are not possible mathematically, but False Recall on the part of returning Gore and Kerry voters was why they misreported their past vote in favor of the previous winner Ė Bush in both cases.

 

Analyst

Even though Bush had a 48% approval rating in 2004 and 22% in 2008?

 

Naysayer

Yes. That is what the NES surveys have consistently shown.

 

Analyst

So you are saying that Final National Exit Poll weights and vote shares are the result of actual survey samples, and therefore include returning Gore and Kerry voters who misreported their vote at more than double the rate of Bush voters?

 

Naysayer

That is correct. The Final NEP returning voter mix and corresponding vote shares are based on actual sampled results.

 

Analyst

OK, then letís compare the preliminary 12:22am NEP (13047 respondents) to the Final (13660 respondents), assuming that they were based on actual sample results. Kerry won the preliminary by 51-48%. Bush won the Final by 51-48%.

 

Naysayer

Fair enough.

 

Analyst

I assume that you are aware that only 3168 respondents were asked how they voted in 2000 but all were asked whom they just voted for - and Kerry won the total sample by 51-48%. How do you explain that?

 

Naysayer

The exit poll was not completed at 12:22am.

 

Analyst

But there were only 613 additional respondents after 12:22am. It is mathematically impossible for the mix to change from 41/39 to 43/37% with just a 4.6% increase in respondents. It is also impossible that Kerryís share of female voters could decline from 54% to 51% with just 613 additional respondents.

 

Naysayer

That is not the way to look at it.

 

Analyst

Why not?

 

Naysayer

The exit pollsters had to re-adjust the mix and vote shares based on voter turnout, among other things.

 

Analyst

So you now claim that the NEP was adjusted to force a match to the incoming vote count.

 

Naysayer

That is correct. It's Standard Operating Procedure.

 

Analyst

But you previously stated that the Final 43/37% Bush/Gore returning voter mix was a sampled result and that Gore voters misstated their vote and told the pollsters they voted for Bush in 2000. And now you claim that the pollsters forced the Final National Exit Poll to match the recorded vote. That is a clear contradiction.

 

Naysayer

You just donít get it, do you? You have never contradicted the NES study which indicates that voters misreport their past vote.

 

Analyst

You are changing the subject.You have just conceded that the 43/37 mix was not a sample; that it was an artifice to match the Final NEP to the recorded vote - along with implausible increases in Bush shares of returning and new voters. When are you going to give up the false recall canard?

 

Naysayer

And when are you going to stop trying to convince everyone that Kerry won the election and that the early exit polls were correct? You are setting back the cause of election reform. Your work is an embarrassment. Thatís why I have been debunking your posts since 2005. It's my job to expose your crappy analysis.

 

Analyst

Itís your job? Oh. But it's not just my "crappy" analysis that you find fault with. You have also tried to debunk analyses of scores of researchers who have determined that the election was stolen. You laud Farhad Manjoo of Salon, who tried to debunk the RFK Jr.. Rolling Stone article "Was the 2004 Election Stolen"?

 

RFK presented powerful evidence that Ohio was stolen - along with a number of other key battleground states. Farhad claimed authoritatively that the election was NOT stolen. His "analysis" was full of factual errors which only betrayed his ignorance. And you have defended it. It's almost as if he relied on your input in writing that hit piece.

 

But I digress. Does it even matter what the voters said about their past vote?

 

We know that Bush had 50.45 million recorded votes in 2000. We know that very close to 5% of Election 2000 voters passed on before the 2004 election and therefore 2.5 million Bush 2000 voters died. Therefore, the maximum number of returning Bush voters was 48 million Ė assuming 100% turnout of living Bush 2000 voters in 2004. The maximum returning Bush voter share of the 122 million votes recorded in 2004 was 39.2% (48/122).

 

So why do we even care what returning voters said about their past vote? We only want to know how they voted in 2004. We can test different turnout scenarios. Even if 98% of Bush voters turned out and just 90% of Gore voters turned out, Kerry still wins by 7 million votes - assuming 12:22am vote shares. Assuming equal 98% turnout, Kerry wins by 10 million.

 

Assuming a feasible turnout of returning voters, Bush vote shares have to beincreased to implausible levels above those in the Final NEP to match the recorded vote.

 

New voters: from 41% at 1222am to 46% (2% MoE).

Bush voters: from 90% to 92% (1% MoE).

Gore voters from 8% to 15% (1% MoE).

 

So which argument are you making?

Is it 1) massive Gore voter defection or 2) Gore voter false recall?

 

1) Defection: You require implausible Bush vote shares (far beyond the MoE) applied to a feasible return voter mix. This means that you must also believe that at 12:22am (13047) and in the Final (13660), respondents must have misreported their 2004 vote when they said they voted for Kerry.

 

2) False Recall: Returning Gore voters misreported their past 2000 vote when they said that they voted for Bush in 2000. Yet they told the truth about their 2004 vote.

 

You have also agreed that the 43/37% mix is mathematically impossible:

 

a) For returning Bush voters to comprise 43% of the 122 million who voted in 2004there had to be at least 4.6 million phantom Bush voters.

 

b) The change in the returning Bush/Gore voter mix from 41/39% at 12:22am to 43/37% in the Final was also mathematically impossible since 13047 (95.5%) of the 13660 respondents were already counted at 12:22am. It was also mathematically impossible that Kerry's share of females could decline from 54% at 12:22am to 51% in the Final.

 

The premise is that the returning voter mix was based on an actual sample of the 3168 respondents who were asked how they voted in 2000. It's also the basis of your False Recall argument. But at the same time you acknowledge that the exit pollsters always force the exit poll to match the recorded vote Ė a contradiction.

 

You cannot have it both ways.

 

It appears that False Recall is your primary canard, since that is the one that you constantly refer to. Perhaps it's easier to sell that myth than the ridiculous Gore defection rates that would be required to match the vote assuming fairly equal 2000 voter turnout. But you continue to promote both myths. You switch from one to the other depending on which is contextually most appropriate for misleading your diminishing audience.

 

Naysayer

So what if itís a contradiction? Youíre just as wrong now as you have always been. See you in 2010 when I will once again kick your ass.