Rasmussen ve un electorado de +3% demócrata

Byron York (The Washington Examiner) ha hablado con Scott Rasmussen:

For all the complexities of polling, says Scott Rasmussen, there are some fairly simple numbers to remember when thinking about this year’s presidential race.  “For the last 20 years, between 37 and 39 percent of voters on Election Day have been Democrats,” says the pollster.  “Republicans have ranged from 32 to 37 percent.  Right now, our sample looks like 36 percent Republican versus 39 percent Democrat.”

The bottom line, Rasmussen continues, is that there is most likely a two, three, or four percentage point advantage out there for Democrats.  That’s what it’s been for nearly a generation; that’s probably what will happen on November 6.

Given that, and factoring in independents, Rasmussen’s national surveys show Barack Obama leading Mitt Romney by a small margin.  The president has a two-point advantage in the latest Rasmussen national tracking poll, and comparably small margins in the super-swing states of Florida, Ohio, and Virginia.  “I think the race is tilting, just barely, in Obama’s favor, with the potential to shift between now and Election Day,” he says.

After some polls, particularly one from Quinnipiac and the New York Times, showed huge Obama leads in the swing states — nine points in Florida and ten in Ohio — there’s been a contentious debate about the relationship between state polls and national polls.

(...) When there are national polls showing a very tight race and big swing state polls showing a blowout, something is likely wrong.  If the national results are close on November 6, it’s very unlikely that Ohio and Florida will be blowouts.  And if Ohio and Florida are blowouts, it’s very unlikely the national race will be close.  “When all is said and done,” says Rasmussen, “it is impossible for me to conceive of a circumstance where there is a huge discrepancy between those key states and the national numbers.”

And whatever the numbers are at this moment, Rasmussen expects them to move by Election Day.  In the last three elections, he notes, the polls moved against the incumbent party in the final weeks of the race.  That’s not an unbreakable pattern, and it might not happen this time, but it suggests Romney will gain on Obama, at least a bit, before November 6.  Of course, some major, unexpected event might move things more.

Meanwhile, Republicans across the country continue to express skepticism, scorn, and in some cases outright contempt for the polls.  Last week in Ohio, voter after voter at Romney-Ryan rallies complained about the polls, with most saying they just don’t believe them.  It’s something every pollster, left, right, and center, is hearing every day.

“When polls appear to be in dispute,” says Rasmussen, “partisans go to the ones they like best and say they are right and everything else is wrong.  Then they rationalize it.  You rationalize things to fit what you want the world to be.”

Rasmussen es probablemente el encuestador más serio del país. Es quien más se acercó al resultado real en las últimas dos presidenciales. Su encuesta final de 2004 pronosticaba Bush 50.2% - Kerry 48.5% y el resultado final fue Bush 50.7% - Kerry 48.3%. En 2008, pronosticó Obama 52% - McCain 46% y el resultado final fue Obama 52.9% - McCain 45.7%. Por estas fechas hace cuatro años, la última semana de septiembre de 2008, tenía a Obama ganando por 6 puntos (50% - 44%).

En 2008 veía un electorado de +7% demócrata en su composición, que es como terminó siendo.