October 7, 2004


Bush wins, according to econ models: Plug in income and inflation for vote projection output (Emily Church, Oct. 6, 2004, CBS.MW)

An Economy.com model gives Bush 373 electoral college votes, well ahead of the 270 electoral votes he needs to win. Kerry lands only 10 states, according to the model -- Illinois and California are the only two outside a handful of East Coast states.

The model seeks to guess voter behavior by looking at per capita income growth and inflation trends in the states. It also plugs in some non-economic factors such as a political party's share in the last presidential election.

The model does suggest "President Bush is going to win a big victory," said Gus Faucher, a senior economist from Economy.com, a consultant. [...]

A second econometrics model -- the President Vote Equation designed by Yale University professor Ray Fair -- has consistently shown a Bush win. In July, the equation predicted a 57.48 percent vote for Bush.

The model uses U.S. growth and inflation figures among its key inputs. Fair expects to update the PVE one more time before the election at the end of October once the third quarter U.S. GDP estimate is released. But he said he doesn't expect the PVE projections to change much.

"I expect we're pretty close on the final (third quarter) numbers," he said.

The Brothers Judd 2004 Presidential Prognostathon has gone live, you can make your picks and they'll be saved as your contest entry.

Posted by Orrin Judd at October 7, 2004 12:13 AM

Let's hope these models (and not the polls which show the race deadlocked) are correct. A Bush win by 3% or less will lead to Florida 2000 all over the country.

Posted by: AWW at October 7, 2004 6:55 AM


What polls? They all have Bush at or over 50% approval ratings, even when they skew Democratic.

Posted by: oj at October 7, 2004 8:34 AM

If you look at theindividaul state polls the Bush percentage in the deep south, farm state, and the rocky mountain states is in the massive landslide range. If thses polls are correct then Bush will get over 60 percent of the vote in twelve to fifteen states.

Since a Presidential election is really 51 individual elections this popular vote distribution may work to "undervalue" the Bush electoral vote total.

Posted by: Earl Sutherland at October 7, 2004 9:39 AM

OJ - a whole bunch of polls are showing Bush with a slight lead 2-4 pts and there are some showing Kerry ahead but these are being questioned. As I state above if Bush wins with 53% or less you know the Dems are going to challenge the results. As Hugh Hewitt says if it aint close they can't cheat.

Posted by: AWW at October 7, 2004 11:27 PM

If he has 53% he'll have won 400 electoral votes. How many states can you challenge 5% margins in to get it back down to a Kerry win?

Posted by: oj at October 7, 2004 11:34 PM