Selected letters for October 27

Date: Tue, 02 Nov 2004 14:22:31 -0500

From: Manuel Frank

Sam, attached is the chapter I sent a week or two ago (in rich text format).

Below is a short description of the interest factor (in the rich text file) and a separate factor, contentment, which is causing the differential voter turnout for Kerry:

You will note in the attached file, that I have a specific prediction based on survey data from September (that period is significant because as you know, in every election interest tends to rise as a matter of course). In any event, the prediction of 129.25 million voter of eligibility is statistically and theoretically supported.

Some information and general points not included in the file I've attached:

(1) Because Pollsters use historical guides to estimate likely voters and not contemporaneous psychological factors, the error rate due to incorrect estimates of the probable electorate can be higher than expected due to sampling error.

(2) This election will see a significant voter turnout among Americans between 18-29 years old. I expect 50-55% turnout, up from the low to mid 30% range in 2000.

(3) I expect African American turnout to at least equal the demographic percentage representation not the slight discount we saw in 2000. That is, at least 11% of the composition of the electorate will be African American.

(4) Because of the above two points, I suspect that most pollsters have under-represented those two demographics.

(5) Another factor not mentioned in the previously attached file (but based on research I have conducted), is that poll measures of thermometer ratings and likes and dislikes, indicate that Democrats are more discontented than are Republicans. I estimate that because of this variable alone, Kerry has a 3% voter turnout differential. That is, both Democrats and Republicans will vote at significantly higher rates than in recent elections, yet, Democrats are 3% more likely to vote.

(6) I estimate that 66.5% of Kerry's support will vote among the eligible voter population, and Bush's voter turnout support will be about 63.5%. I am predicting 63.5% of the voter eligible population will vote. 129.25 million; 95% confidence interval 123-135 million votes.

Thank you for your contributions to the American political discourse.

Dr. Manuel M. Frank

2 Deibler Dr.

Freeville, NY, 13068