Editorial: Politics, polling and predictions

The end of filing for candidates for the 1998 elections was Monday, and it generated a flurry of activity, including the entry of Las Vegas Mayor Jan Jones in the race for governor.

Although there were some surprises during the filing period about what candidates decided to run -- and those that opted not to -- one certainty you can count on between now and Election Day is the proliferation of political polling. Most of the polling done by candidates and parties is done for the internal benefit of the campaigns themselves, to help them identify public opinion and then map their strategy.

But campaigns will often leak positive polling results to the media. Too often, news organizations will publish the results without placing the poll in perspective -- disclosing who paid for it and how the questions were specifically worded. When this happens, it's a good bet the campaign has slanted the questions to skew the results in favor of its candidate.

With this in mind, the public and the media should be suspicious of poll results paid for by specific campaigns, political parties or interest groups with an axe to grind. Even independent polling doesn't always get it right. After all, polling is anything but an exact science, as the New York Times reminded us on Monday in a story about the 50th anniversary of one of polling's biggest blunders: the 1948 presidential race.

Respected pollsters had Thomas Dewey easily beating President Harry Truman in 1948 by as much as 15 percentage points; the voters fooled the pollsters as Truman ended up winning by 4.4 percentage points.

Based on polling alone, some candidates might seem invincible today, but remember that in the coming weeks and months some of these candidates might be surprised on Election Day and end up as Nevada's version of "President Dewey."

archive