Media Ignore Poll Showing Schwarzenegger in Lead
NewsMax.com Wires
Friday, Aug. 29, 2003
WASHINGTON – If you watched the 24-hour news channels Wednesday, one of the things you did not see was hour-by-hour coverage of the California electorate's tremendous swing toward millionaire movie star Arnold Schwarzenegger.
According to a new KPIX/KCBS poll of 500 registered voters statewide, support for the recall of Democrat incumbent Gov. Gray Davis has also soared dramatically.
The poll, conducted for the CBS affiliate in San Francisco by Survey USA, shows that 45 percent of the voters want Schwarzenegger to be the state's next governor. And 64 percent told pollsters they would vote to recall Davis on Oct. 7.
In second place among the more than 130 candidates trying to become the state's next governor is Democrat Lt. Gov. Cruz Bustamante, who has the support of 29 percent of the electorate.
The KPIX/KCBS poll is the first public statewide survey taken since 2002 GOP gubernatorial candidate Bill Simon dropped out of the race.
The results are dramatically different from what a seemingly more in-depth Los Angeles Times survey of 1,351 "likely" voters in California showed when it was released last Sunday.
In that poll, support for the recall had dropped from what it had been in previous polls, with 50 percent of respondents saying they would vote for the recall and 45 percent saying they would vote against it. The Times had Bustamante leading the pack with 35 percent, Schwarzenegger in second place at 22 percent.
Making up the Outcome They Desire
The Times' survey was deemed significant enough that the 24-hour news channels and most of the major newspapers covering the race devoted a great deal of attention to it Monday and Tuesday. As they spun it, the data, when compared with earlier polls showing support for the recall growing and Schwarzenegger dominating the field, now showed Schwarzenegger and the Republicans were fading, Bustamante was surging and voters were beginning to question the wisdom of removing Davis from office.
Talk about a self-fulfilling prophecy.
For some time the elite opinion journalists have been writing pieces questioning the wisdom of California's recall. Removing the governor, potentially replacing him with a political novice when the state's deficit is estimated to be greater than the combined deficits of the other 49, has been portrayed as an irrational act, something the voters would grow to regret.
The Times' poll confirmed these assumptions and was therefore embraced by many who think the recall makes a mockery of the state's political process. The pundit class that, like me, specializes in instant analysis of complex developments used the data to reinforce their belief that the recall, as a matter of policy rather than a matter of journalism, should fizzle.
The KPIX/KCBS poll, which should be considered as reliable as any other public media poll taken in this most volatile of environments, shows a 14-point swing in favor of the recall in just a few days. Support for Schwarzenegger has increased by 23 percent, and Bustamante has slipped 6 points. If the coverage of the L.A. Times' poll set the standard, then the KPIX/KCBS poll at least rates the printing of an "Extra."
In fact, the KPIX/KCBS poll may be as flawed as some have argued the Los Angeles Times' poll may be. For all their impact on the U.S. electoral system, polling is not an exact science. A survey can be manipulated to produce a desired outcome or be badly constructed to produce a faulty or inaccurate measure.
Media polls are susceptible to error. The way the sample is constructed, the language of the questions, the fact that they are done quickly, the weighting given to certain voter blocs, even the manner used to determine what constitutes a "likely voter" are but a few ways errors can creep in.
"The haste with which many media sponsored polls are drafted, put in the field and publicized creates its own 'margin of error.' If the conclusions, meaning tomorrow's headlines like 'Voter Support for Recall Drops,' are already written, then the polling process is just a fishing expedition used to produce evidence to support the conclusions already reached," The Polling Company's President Kellyanne Conway said.
In a situation such as the California recall, one without precedent in the modern information-gathering era, the ability to accurately assess voter attitudes and intentions is difficult. No one can be sure to any degree of certainty just who will turn out to vote on Oct. 7 or how those voters will behave if they do.
The recall, if it teaches us anything, demonstrates just how polling is an inexact science. The data, rather than driving a story or being a story unto itself, should be used only to further the voters' knowledge. Election contests are not, contrary to the way they are generally portrayed, horse races. They should not be covered like they are. Analysis by Peter Roff, UPI political analyst and 20-year veteran of the Washington scene. |
