Poll Vault

[MattusMaximus]

I just got done reading a great article over at MSNBC about how post-convention bounces are now likely much ado about nothing in regards to prediction markets. Check it out...

Bye-bye to the bounce?

Dated 9-5-08
Republican presidential candidate John McCain didn't benefit from that big of a bounce today in the post-convention prediction markets. Then again, Democratic rival Barack Obama didn't get much of a bounce from his party's convention last week, either.

In fact, the idea that political conventions produce an upward "bounce" of popularity doesn't really apply to the prediction markets, and economists think they know why.

By the way, McCain's surprise selection of Alaska Gov. Sarah Palin to be his running mate didn't move the markets either, for different reasons. ...

It also explains what things do and don't move prediction markets. This helps to explain what we've been seeing (or not seeing) in the markets over the last two weeks. Pretty cool.

 

[David Wong]

For all of you who say polling is meaningless, we have:

Gallup USA Today poll, Jul 25-27: McCain by 4

Gallup daily tracking poll, Jul 25-27: Obama by 8.


Uh, Gallup? Hello?

So they're at it again, today Gallup has McCain up by 3, but the USA Today/Gallup poll that applies their "likely voter" criteria, gives us McCain up by 10. Same range of dates.

Hmmmmm.

 

[Pookster]

The USA Today/Gallup poll of likely voters has McCain up by 10 (54%-44%) ...

http://www.usatoday.com/news/politics/election2008/2008-09-07-poll_N.htm

The convention bounce has helped not only McCain but also attitudes toward Republican congressional candidates and the GOP in general.

"The Republicans had a very successful convention and, at least initially, the selection of Sarah Palin has made a big difference," says political scientist Larry Sabato of the University of Virginia. "He's in a far better position than his people imagined he would be in at this point."

One note about polls just after conventions ...

However, in an analysis of the impact of political conventions since 1960, Sabato concluded that post-convention polls signal the election's outcome only about half the time. "You could flip a coin and be about as predictive," he says. "It is really surprising how quickly convention memories fade."

Time will tell, but for right now, it's McCain's lead to lose.

 

[Pookster]

Has anyone taken into account the fact that Democrats outnumber Republicans now by about a 4-to-3 ratio? And that's not even counting the huge increase in Democratic voter registration I mentioned earlier.

Sure. It's why the poll showed a tie even though Obama was getting less of the Democratic Party than McCain was from the Republicans. The polls generally weight for political party affiliation. I know Rasmussen adjust the affiliation percentages they use to weight the polls very regularly to reflect the changes in party affiliation.

 

[Pookster]

Perhaps Pollster.com is taking the under 30 crowd into account by making cellphone calls? Does anyone know if this is what they do?

I don't know if they do, but Gallup does call cellphones. They just can't use an auto-dialer to call them with.

 

[Pookster]

So they're at it again, today Gallup has McCain up by 3, but the USA Today/Gallup poll that applies their "likely voter" criteria, gives us McCain up by 10. Same range of dates.

Hmmmmm.

I apologize for not noticing you had already posted this. I should learn to read through the new posts before posting.


ETA: I'm as confused with Gallup's conflicting polls as anyone. I wish I knew the differences. Seems like there should be an explanation for the differences somewhere. I haven't seen it yet though.

 

[Pookster]

Rasmussen has McCain up by one today (48%-47%) ...

http://www.rasmussenreports.com/pub...ial_election/daily_presidential_tracking_poll

Forty-one percent (41%) of voters say that they are certain they will cast their ballot for McCain and will not change their mind before November. Thirty-eight percent (38%) say the same about Obama.

Should be a fun two months.

 

[Pookster]

Rasmussen has McCain up by one today (48%-47%) ...

http://www.rasmussenreports.com/pub...ial_election/daily_presidential_tracking_poll

To add:

For a variety of reasons, the Rasmussen Reports daily Presidential Tracking Poll is less volatile than some other polls and always shows a somewhat smaller convention bounce than reported by others. This is primarily because we weight our results by party identification (see methodology). Looking at the data before adjusting for partisan identification, the Republican convention appears to have created a larger surge in party identification than the Democratic convention the week before. If this lasts, it could have a significant impact on Election 2008.

 

[David Wong]

Here's a study somebody did on why Gallup's "likely" voter criteria give such volatile results:

http://poq.oxfordjournals.org/cgi/content/full/68/4/588?ijkey=053EosjdTO0oc&keytype=ref

Part of the problem is that you can't judge who is a "likely" voter this far out.

Please note my beef with them isn't that they show my guy down, it's that their numbers are extremely volatile from one day's polling to the next when they use their "likely" voter methodology. The study quotes a famous 19-point swing in one day during the 2000 race.

 

[Pookster]

Here's a study somebody did on why Gallup's "likely" voter criteria give such volatile results:

http://poq.oxfordjournals.org/cgi/content/full/68/4/588?ijkey=053EosjdTO0oc&keytype=ref

Part of the problem is that you can't judge who is a "likely" voter this far out.

Please note my beef with them isn't that they show my guy down, it's that their numbers are extremely volatile from one day's polling to the next when they use their "likely" voter methodology. The study quotes a famous 19-point swing in one day during the 2000 race.

Thanks for posting that link. Interesting study.

 

[Pookster]

Here's another poll for registered voters done Friday through Sunday - Diageo Hotline Tracker - that I've seen posted by RCP in the past.

They have the race tied (44%-44%).

http://diageohotlinepoll.com/Tracker/Diageo Hotline Tracker release - 09 08 08 data.pdf

 

[Brainster]

Here's a study somebody did on why Gallup's "likely" voter criteria give such volatile results:

http://poq.oxfordjournals.org/cgi/content/full/68/4/588?ijkey=053EosjdTO0oc&keytype=ref

Part of the problem is that you can't judge who is a "likely" voter this far out.

Please note my beef with them isn't that they show my guy down, it's that their numbers are extremely volatile from one day's polling to the next when they use their "likely" voter methodology. The study quotes a famous 19-point swing in one day during the 2000 race.

Interesting article, although I do have to wonder if it's that hard to judge a likely voter in early October when the 19-point swing occurred.

If it's a function of enthusiasm, then at least some of the bump is real. I live in a Republican area, and many people have expressed to me their delight over the Palin pick.

That said, I'm going to go against my confirmation bias and say that 10 points does indeed seem like too much of a lead to be credible at this point. My rule of thumb is always that if you want to believe it and you're still not sure, it's probably too good to be true.

 

[David Wong]

Gallup's own poll has McCain up by 5 today (49 to 44) among registered. That... should be the high water mark of his bounce if previous polls are an indication.

Now Obama's convention bounce faded after like three days, but of course the Palin announcement and the GOP convention helped cut it short. So we'll see if his lasts longer.

Meanwhile, CNN's poll of registered voters has it tied.

 

[Pookster]

If it's a function of enthusiasm, then at least some of the bump is real.

I think you're correct with this. From what I read, the article indicated that the methodology had them bouncing between likely and unlikely to vote based on enthusiasm, when it all likelihood, some were going to end up voting no matter what.

 

[Pookster]

Meanwhile, CNN's poll of registered voters has it tied.

Interesting quotes from the CNN article ...

"Other polls are showing 'convention bounces' for McCain, but ours does not. The reason is probably due to the fact that the CNN poll has had a very low number of respondents who say they are undecided for several months. Bounces tend to occur when undecided voters switch to a candidate, but with so few truly undecided voters in our polls, there was little room for a bounce for Obama or McCain," said CNN polling director Keating Holland.

Another factor may be gender. "Although Sarah Palin is the first woman on a GOP ticket, women nationwide are sticking with Obama -- 52 percent of women are planning to vote Democratic, while 51 percent of men are in the GOP column," Holland added.

ETA link: http://www.cnn.com/2008/POLITICS/09/08/campaign.wrap/index.html

 

[gdnp]

Even though it is only 2 months until the election, it is still not clear to me what the defining issue will be. Fanny and Freddy being taken over and a 0.5% jump in the unemployment rate may push economic issues, whereas the easing of gas prices may make the "drill, baby, drill" chants anachronistic. It remains to be seen how well Obama can turn this back into a referendum on Bush, and obviously Iraq, Iran, Israel, or Russia could move to the forefront.

I am amazed at how many people seem to think this is going to, somehow, be fun. About as fun as riding out a hurricane.

 

[Pookster]

I am amazed at how many people seem to think this is going to, somehow, be fun. About as fun as riding out a hurricane.

From a political science perspective, it's just like Christmas.

 

[BenBurch]

...

I am amazed at how many people seem to think this is going to, somehow, be fun. About as fun as riding out a hurricane.

All you need is enough ice, tequilla, triple-sec, lemon juice, salt, & limes...

 

[Wangler]

Intrade Prediction Market has the following:

Obama for Prez: 54.0, down 2.6
McCain for Prez: 45.9, up 2.4

And the really interesting one:

Recession in 2008: 10.0, down 7.5

Is that just because we're getting close to the end of the year?

 

[BenBurch]

And the really interesting one:

Recession in 2008: 10.0, down 7.5

Is that just because we're getting close to the end of the year?

Yeah, not many more official reports left in the year and that is how it is determined. And lowering oil prices make that result slightly less likely. What we have on the horizon is closer to "stagflation" than classical recession.