Subject: RE: Query: Rules Governing TV Surveys
From: "Douglas Johnson" <djohnson@NDCresearch.com>
Date: 6/21/2006, 5:04 PM
To: richard@shepardlawoffice.com, election-law@majordomo.lls.edu
Reply-to:
djohnson@NDCresearch.com

Many voters in exit polls and phone surveys say they would prefer "None
of the Above" or, lacking that, a generic "someone else", if they are
given that option. This is a very different response than saying they
support one of the other candidates in the race.

I suspect that another potential factor may be that Rasmussen may not be
offering an "undecided" option, or they may encourage undecided voters
to declare which candidate they are "leaning" toward. In that case,
undecideds may resist that effort by choosing the "Someone Else" option.

High-propensity voters choosing stating a "None of the Above" or
"Someone Else" preference usually will move to one of the
perceived-as-leading candidates on election day. Low-propensity voters
stating "None of the Above" or "Someone Else" will often simply not vote
on election day.

I have not checked, but I suspect the PPIC and Field surveys leading up
to California's 2003 Recall election of Arnold Schwarzenegger may show
this trend. And that election also points out the flaw in the "include
all candidates" idea: there were something lik 135 candidates for
Governor on the 2003 ballot. Any poll attempting to list them all -- or
even 20 of them -- would have lost the attention of so many respondents
as to become factually incorrect about the leading candidate numbers.

As a side note, I have heard that in countries where "none of the above"
is an option or where the winner is required to get a majority of vote
slips turned in, "None of the Above" or blank ballots are said to have
led to no one winning in some rare elections. Unfortunately I can't cite
specifics, only reports I've heard from others in the political
consulting world, but others on this listserv may know of specific
examples of this.

Hope that helps,

- Doug

Douglas Johnson