The 2006 election is less than a week away. Based on the polls, Democrats seem to be popping Champagne, believing they will be taking control of the House of Representatives and possibly the Senate. Conversely, Republicans claim the Democrats are sipping their bubbly prematurely and will have to drown their sorrows after Nov. 7. And so the spin goes, as pundits from both sides of the political aisle analyze their crystal balls and the polls.
Over the past months, pollsters have been working hard to gauge the direction of this year’s election. Besides the polling done by political campaigns, news organizations contract for polling and use the data to manufacture news — all in hopes of being the first to provide answers about each race or issue. The problem: These answers may apply the day the poll is taken, but can change a multitude of times before the election actually takes place.
This campaign season must be a banner year for the polling business. For months we’ve been hearing the projections about whether the Democrats will take control of the House of Representatives and the Senate, each poll breaking down specific political races so commentators can reconfigure the “what if” scenarios of wins and losses.
Obviously, we won’t know the real results until Election Day, and maybe after, but that hasn’t stopped news organizations from creating buzz about this topic for the last five or so months. By hiring pollsters to research various races, the news media are able to create a story and “break” the news about the results. Then this story can be used and analyzed for at least the next 24-hour news cycle. Finally, that information can be recycled in the proceeding months when additional polls are done on the same topics and the trends are analyzed. All this is done in the name of providing the public with information.
But how reliable is polling? Polling is based on probability and statistics. In most cases, phone numbers are chosen at random for a sampling of the polling area. According to John Zogby of Zogby International, which has been tracking public opinion since 1984, the sampling size determines the margin of error (MOE). A sample of 400 will have MOE of plus or minus 5 percent, 600 is plus or minus 4 percent, and 1,000 reduces the MOE to plus or minus 3 percent. The margin of error is the percentage range, plus or minus, that the results will vary if the same poll is repeated.
In an election scenario, however, it’s difficult for pollsters to predict voter turnout, which is critical to the accuracy of the results. And as Michael Barone reports on RealClearPolitics.com, “Americans have fewer landline phones than they used to, and the random digit dialing most pollsters use does not include cell-phone numbers.” Also, a growing percentage of call recipients are refusing to be interviewed.
The exit polling for the 2004 presidential election proved these results can be far from accurate. According to the National Election Archive Project, President Bush officially won the 2004 presidential election by 2.5 percent, but the exit polls showed John Kerry winning the race by 3 percent. This is a dramatic discrepancy.
But after all the money is spent on polls, and all the analysis and pontification has evaporated like the handful of air that it is, only one thing remains sure in this and all elections. It’s not the polling numbers or the voices of the pundits that matter, it’s your voice that counts. And only one poll matters, the one that’s tallied at the end of Election Day. So, go to the polls this Nov. 7, vote and have your voice be heard.
Kathleen Antrim is a columnist for The San Francisco and Washington, D.C. Examiner newspapers, is the author of “Capital Offense,” and writes for NewsMax Magazine. She can be heard regularly on Hot Talk 560 KSFO in San Francisco on “The Lee Rodgers and Melanie Morgan Show.” For more information go to: www.KathleenAntrim.com.