The media storyline of record turnout was so pervasive on Tuesday that even
the Weather Channel was reporting, "clear skies for the record California
turnout."  What really happened?

As of 6am California time, with 96.8% precincts reporting, total number of
ballots cast is 7.9 million.  When all precincts have reported, about 8
million will probably have cast a ballot.  With so few precincts left to
report, and with most of the large counties completely accounted for, it is
unlikely that the raw number of votes will exceed the 8.6 million cast in
the 1998 election, though the number has already exceed the 7.7 million cast
in the 2002 election.  In terms of raw turnout, this is a remarkable special
election, but it is not, as the media reported it, a record turnout for a
governor's election.  These reports were based on pre-election polling.
Moral of the story: people lie about intention to vote on pre-election
polls, a fact that is well-known in survey research.

Of course, these are raw numbers.  There are more toilets in California than
in 2002 simply because the population is growing.  As a percentage of the
eligible population, the turnout rate in the special election was about
37.9%.  This will be slightly higher than the 37.2% turnout rate of 2002 and
will certainly be less than the 45.4% in 1998.  Note that the Secretary of
State's office reports the turnout rate as a percentage of the registered
voters.  This is not a good statistic to use for between-state and
historical comparisons because registration laws and the accuracy of the
registration rolls vary across states and over time.  A preferred statistic
is turnout as a percentage of eligible voters.

According to the Secretary of State's office 7,572,867 people voted for the
first question on recall, while 7,947,717 cast a ballot.  Approximately
370,000 people did not vote on the recall question!  Los Angeles, one of the
counties that used the punch cards, accounts for almost half of the
discrepancy: 1,789,766 people voted for recall while 1,964,853 people cast a
ballot, about 91.1% of the people in Los Angeles voted on the first
question, or a difference of 170,000 persons.  Of course, Los Angeles is
also one of the most populous counties in the state.  My guess is that
Alameda has touch screens, and the election returns show that the people in
Alameda supported Davis, like in Los Angeles.  In Alameda, of the 391,727
people who voted, 388,819 cast a ballot in the recall, for a rate of 99.3%.
The approximately 700,000 difference on the recall cannot be overcome even
if all undervotes went to Davis, though it is possible that more people
voted against the recall: 3,465,666, than voted for Schwarzeneggar:
3,528,093, when undervotes are factored in.

==================================
Dr. Michael P. McDonald
Assistant Professor
Dept of Public and International Affairs
George Mason University
4400 University Drive - 3F4
Fairfax, VA 22030-4444

Office: 703-993-4191
Fax: 703-993-1399
Efax: 561-431-3190

mmcdon@gmu.edu
http://elections.gmu.edu/