Time to Panic? - Today's "Incumbent Rule" Results



Today's "incumbent rule" results are dramatically different. Using the latest polls listed at 2.004k.com this evening, assigning undecided voters 3:1 to Kerry as the challenger, the Electoral College result would be a 269-269 tie sending the presidential election into the House of Representatives.

The only difference from yesterday's 296-242 result is two states: Florida (27 EV) and Minnesota (10 EV).

This is a problem with 2.004k.com -- it lists only the "latest" poll from each state, and of course the latest poll is not often the best. In both states, there are contemporaneous polls which contradict the "current" poll.

For example, here are the latest four polls from Florida (results in this order: Bush-Kerry-Nader-undecided):

50-45-n-5 Rasmussen Reports 10/27
46-48-1-5 The Florida Poll 10/23-27
44-44-1-11 Quinnipiac U. 10/22-26
46-49-1-4 ARG 10/23-25

Applying the incumbent rule, 3 out of the last 4 Florida polls would give the state to Kerry. But 2.004k.com lists only the last one as the most "recent."

Here are the two Minnesota polls done in the last week:

47-44-5-4 U. of Minnesota 10/21-26
42-46-3-9 St. Cloud State 10/17-26

I have read the internals of both polls. The U of M poll oversamples Republicans (44-42 over Democrats). But the St. Cloud St. poll oversamples Democrats 40-30.

Here's the latest list of closest states and how the incumbent rule applies:




I have been asked by a reader how the results would be different if the undecideds split for Kerry 2-1 or even 3-2. A 2-1 split does not change the results (although Kerry's margin in Iowa drops from 1.0 to 0.4). A 3-2 split shifts Iowa from Kerry to Bush, and results in a 276-262 win for Bush.

Remember, none of this accounts for any turnout bonus Kerry might receive. It is my subjective opinion that Democratic turnout will set records, particularly with young voters and new voters, none of whom are accounted for in the turnout models used by the polls. In particular, I think turnout will most affect the result in Florida.

Subjectively in my judgment Bush has a better chance in Iowa than he does in Wisconsin or Minnesota. If Kerry loses Florida, wins Minnesota, but loses Iowa, he wins the election with 272 electoral votes. That today's predicted outcome in Slate.

Massive Democratic turnout in Florida wins the election for Kerry - that's my prediction.

As an aside, Bush was campaigning in New Hampshire today -- a silly decision in my judgment. Why would they take valuable campaign time in a state where Bush is down by 8 points and where many voters (who live in the Boston media market) are familiar with Kerry?

Because the Bush campaign realizes they're going to lose Ohio, and are looking to cobble together a group of small states to replace its 20 electoral votes.

But this strategy is premised on holding Florida. If Bush loses Florida he loses the election. I would not want to be an incumbent running for election needing to hold Florida given these numbers.

Posted: Fri - October 29, 2004 at 10:17 PM        

©