A Question on Polling
Given that we select the president based on the Electoral College rather than the popular vote, it seems to me that it would make more sense for Gallup or Rasmussen to be doing continual state polling just like they do national polling now. That way, electoral college predictions would be based on current data and a consistent methodology. It would likely be more difficult, but would it be astronomically more difficult? Is there any compelling reason why they don’t do this?
6 Comments
Sorry, the comment form is closed at this time.
Our Humble Logo
Adam’s Shared Items- Red Line rehab plans announced
- Occupy and General Strike Links, Other Links, 11/2/11
- SHOW OF FORCE
- Inequality, Economic Stagnation and Product Innovation
- Halloween Costumes, Foreclosures, Creditors’ Bargains and how the 1% View Our Laws on Debt
- New Facebook Feature Allows User To Cancel Account
- Why Quan’s Got to Go
- A Softer World: 733
- “There is room where the fences used to be”
- The virtue of Philip Blond
- Fear, American Style: What the Anarchist and Libertarian Don’t Understand about the US
- Petition to Google: don't kill Google Reader
For opinion polls, they need to sample at least 1000 voters to have a valid sample, to do this for each electoral college would thus require upwards of 50,000 in the sample, but they can do a national poll with 1000.
Comment by Gabe | September 11, 2008
Paul/Kucinich (or Kucinich/Paul) in 2008!
Comment by Toby | September 11, 2008
If I was Obama’s campaign manager …
Act as if the election is being held in Ohio in 10 days. Camp out there, a few side trips to Michigan and PA to shore up support in neighboring states. One side trip to Indiana, perhaps, near the end of the ten days. But basically, get on the ground with voters in Ohio. At least 7 out of 10 full days. Figure out what makes them tick. Figure out why he lost to Hillary there and what can be done to appeal to her base. Get a grip on race dynamics there. Make sure everybody as in everybody in ‘our demographics’ is registered. Make an appearance in every single Ohio precinct. Literally. Pull all the ads off the air in Ohio for those 10 days and put the money elsewhere. Make sure to make the local evening news every single one of those 10 days. Ohio in September is Iowa in late December, early January. Quitely, but forcefully encourage early voting there. Win Ohio, win the election (unless Washington state is lost – so a two day trip to Washington, Oregon and a fundraiser in California immediately after the 10 day Ohio campaign).
So, come late September, if an Obama victory in Ohio isn’t nearly guaranteed, it’s plan ‘B’: How to win the election w/out winning Ohio. Perhaps another single state strategy (Virginia or Indiana would dom so long as Oregon nor Washington wavers)? No. Take it to the Republican ticket strength. Let’s win the west. Two out of three of Nevada, Colorado, New Mexico. Let’s really put Montana and North Dakota in play with serious ad money and a couple of trips. A major speech in the heart of Arizona on McCain’s 5 greatest failures. Do not as in do not lose New Hampshire. A two day fundraising trip to the Northeast with serious time in New Hampshire in early October. …
Comment by old | September 11, 2008
Those New Hampshire bastards are the ones who really cost Gore the 2000 election. If they’d acted like a normal northeastern state, the Florida situation wouldn’t have even mattered.
Comment by Adam Kotsko | September 11, 2008
Old will be my campaign manager if I ever run for President…
Comment by Dave Belcher | September 11, 2008
thanks dave!
Comment by old | September 11, 2008