Man Without Qualities

Wednesday, October 20, 2004

Undecided Voters Break Towards The Challenger?

Around this time of the election cycle one often hears that last minute undecided voters "break towards the challenger" in the actual election. That is supposedly because the public knows more about the incumbent and follows a voting decision procedure that first asks whether the incumbent deserves re-election. One often hears (usually from Dick Morris) that an incumbent who has not "broken" the 50% barrier in the polls by this point is supposed to be in trouble. Is that pattern likely to be followed in this election?

It seems unlikely that the "break towards the challenger" pattern will be nearly as strong in this election as in the past. For one thing, the same factors that favor the challenger in the general election are also supposed to be reasons why the challenger gets a bigger "post convention bounce" than the incumbent. In this election President Bush's post-Convention bounce was clearly much larger than Senator Kerry's bounce. The mainstream media has been saturating the public with their view of the race for a very long time. It is hard to imagine that the public has much new, positive material to absorb about Kerry-Edwards between now and the election.

In addition, the "break towards the challenger" pattern was very weak in 2002. An MSNBC/Zogby Poll conducted of 500 likely Florida voters over Oct. 8-10, 2002 (about as far out from election day as we are now) in the course of the Florida statewide race for governor of that state revealed a statistical dead heat: Incumbent Jeb Bush (Republican) had 48% of likely voters behind him, Bill McBride (Democrat) pulled 45% and 7% of likely voters said they were "Not Sure." The actual election wasn't even close. Gov. Jeb Bush trounced lawyer Bill McBride, 56% to 43%. Of course, Florida wasn't the only example of a late swing towards Republicans and incumbents in the 2002 election - a swing that went essentially undetected by almost all pollsters.

Sometimes the last minute 2002 swing was towards a Democratic incumbent. On the very eve of her 2002 re-election polls showed Democratic Louisiana Senator Landrieu in a dead heat with her challenger Terrell - with Senator Landrieu not having broken 50% in the polls in any consistent manner. In fact, Senator Landrieu won re-election with 624,214 votes, or 52%, and Terrell had 587,423, or 48%. The remaining "undecideds" did not break against the better-known incumbent in that case, either - even though she had not broken the supposedly magic 50% barrier in the polls.

MORE: More reasons to question whether the "break towards the challenger" effect has much to it.


Comments: Post a Comment