‘Pssst! Bluebird in the fifth.”
“Pssst! Martinez on the 31st.”
Is there really any difference between a tip at the track and a poll predicting the likely winner of an upcoming election?
Not much — except the latter might actually affect the outcome. That’s legitimate turf for editorial boards, which issue endorsements, but something that journalists who report the news should strictly avoid.
News organizations pride themselves on informing voters about where candidates stand on relevant issues. Issues, though, can be pretty mind-numbing stuff.
So, to liven things up, they periodically throw in polls, showing where support for the various candidates stands at that point in the contest. They call it the “horse-race factor.”
At the track, of course, bettors can see, for themselves, which horse is in the lead. During a political campaign, though, determining who is ahead depends on what the voter reads.
Take, for instance, Tuesday’s primary elections for Florida’s soon-to-be-vacated United States Senate seat.
Less than a week before the vote, this newspaper, using information that Mason-Dixon Polling & Research Inc. collects for the Sentinel and WESH-Channel 2, reported that Democrat Betty Castor was ahead of Peter Deutsch among likely voters 45 to 31 percent. Similarly, it stated that Republican Mel Martinez was ahead of Bill McCollum 33 to 27 percent.
A poll done for the South Florida Sun-Sentinel and Jacksonville’s Florida Times-Union had Castor leading Deutsch 40 to 32 percent but McCollum leading Martinez 35 to 31 percent. Another poll, done for the Miami Herald and the St. Petersburg Times and published just two days before the election, had Castor with twice as much support as her rival and McCollum with a two-percentage-point edge.
The predictions on the Democratic side, in which Castor won by five furlongs and fetlock, came pretty close to the mark. Those in the Republican race, which Martinez won in similarly lopsided fashion, not only missed the numbers but in all cases but one missed the winner.
Only the Sentinel’s Mason-Dixon survey called both races correctly. So what good are differing polls? Aren’t they just educated guesses?
Mason-Dixon’s managing director, Brad Coker, acknowledged that polling is “as much of an art as a science.”
Because of diverse population groups and the need to interview some voters in Spanish, he said, “Polling Florida is somewhat more difficult than the average state.”
So why do it?
The only poll that counts is the one taken on Election Day, and calling attention beforehand to what may be a phantom lead could have a jump-on-the-bandwagon effect — or the converse for someone who does poorly.
“People like to see what their neighbors think,” Coker explained.
Publishing candidates’ statements and activities likely affects support for their campaigns, too, but journalists do that from the legitimate perspective of observers. By originating an element with similar potential and injecting it into the race, though, journalists become participants.
Coker defended newspapers’ conducting polls, noting that candidates frequently commission — and release — their own surveys, some of which Coker characterized as “garbage,” designed to sway voters’ opinion.
“Our objective is to find out what the real numbers are,” he said, and “help filter through the garbage.”
I’m all for straightening out misimpressions, but if voters listen to the candidates, consider the issues and make up their own minds, I’m not sure how they benefit from knowing what other voters say they plan to do.
Readers — and voters — would be better served if news organizations stuck to the issues and left the horse race to the bookies.
