Decision ‘08

Okay, THERE’s The Convention Bounce

I said yesterday I thought Democrats bemoaning the lack of a convention bounce were speaking too soon, and sure enough, today’s Gallup Daily Tracking shows Obama by six. I also see that Obama is planning to lay out a strategy for governing, as I have insisted he must (no link at the moment, but I saw a story on it earlier)…but I’m not the only one who has noticed that it is missing in action. Here’s the LA Times saying that Obama’s agenda so far is a no-show:

This week’s Democratic convention sought to relaunch Barack Obama’s presidential campaign by doing three things: Healing the party’s internal rift, showing voters who Obama is, and spelling out more clearly what he would do as president, especially on the economy. But at the halfway point, the convention still seemed, at best, to have accomplished Step 1.

I’ll update later in the evening, of course…

5 Responses to “Okay, THERE’s The Convention Bounce”

1 Scott Says:
August 28th, 2008 at 9:01 pm
First, my apologies if this has been mentioned previously; I think that many of you, including the MSM, are missing what underlies the polls. Given the disparity between self-identified Republicans and Democrats, Obama should win easily. Clinton Democrats have been saying that they will vote for McCain at a shocking rate. So not only is Obama losing votes, he is having his votes cancelled out by Democrats! Obama wins in a landslide if this changes even a little. Each Clinton vote he flips back to Dem will gain him two votes. The Convention has been orchastrated to “unite” the party and do just that. They are not talking to you in Republican Land. They’re not even talking to Independents -they don’t need any more of them.

The negative VP bounce was a result of more Clinton votes moving away from Obama when they felt offended that she wasn’t the pick, but they will come back. You are now seeing Clinton voters starting to move back - McCain can try to woo them, but he won’t have much luck. Clinton doesn’t have to say the perfect thing to get them back, hurt feelings will eventually heal, and as the election nears Democrats will come home. Over next few days there will be a big “bounce,” but it’s really more like a correction. McCain won’t ever be able to catch up.
2 Mark Says:
August 28th, 2008 at 11:42 pm
Scott, you make a good point…I’ve said all along that the jilted Hillary fans will almost all vote for Obama eventually…
3 Chris J. Breisch Says:
August 29th, 2008 at 7:07 am
Following up on my comment in the previous thread:

Ok, I’ve done the math. And no, I didn’t leave my math degree at home.

The previous day’s track was 45-44, which means a three day total of {134-136}-{131-133}. If we’re trying to figure what the “minimum bounce” is, then we’d take 136-131, for a spread of 5 points.

The current track is 48-42, which gives a three day total of (143-145}-{125-127}. Again, if we’re trying to figure what the “minimum bounce” is, we’d take 143-127, which has a spread of 16 points.

So, the difference is 11 points. Using the same logic to find the “maximum bounce” gives us 19 points. Which means that Wednesday’s poll was between 11-19 points better for Obama than the Sunday one. Now the Wednesday poll doesn’t really go late enough Wednesday to get much reaction from Wednesday night’s speeches, so the real difference is Hillary’s speech on Tuesday.

Now pardon me while I make some assumptions, but I’ll tell you what’s wrong with them at the end.

So, Hillary’s speech caused a minimum of an 11 point bounce. Either it was a much better speech than everyone has thought, or there really were a lot of Hillary supporters either on the fence or supporting McCain and she brought them back into the fold.

That’s what you would think if you analyze the results and look at Wednesday vs. Sunday.

And that may very well be correct.

But, as Jacques likes to point out, this is the problem with looking at just two points on a graph, and not taking all factors into consideration.

Gallup doesn’t post MOE for a single day in their track, that I’m aware of, but we can guess that it’s about 4 points.

So, that 11-19 spread is really 7-23, right? No, you’re forgetting that it’s 4 points on each day. Ok, so it’s 3-27 then, right? Umm..no, not quite. It’s 4 points for each candidate, which brings the resulting “bounce” to something between (-5)-35 points. Now, yes, there’s a nice bell curve there and I’m not going to even try to convince you that a -5 point bounce is anywhere nearly as likely as my original number of 15. There’s a couple of order of magnitudes of difference in the likelihoods. I don’t have my calculator handy, and it’s been far too long since I did confidence intervals and probability curves, or I’d give you some better numbers.

But, what it does tell us is that you should be suspicious if a 3-day tracking poll # is suddenly quite a bit different than the day before. If today’s 3-day track shows the same (or actually bigger), then we can infer that Hillary’s speech really did sway some people. My suspicion though is that Wednesday is a bit of an outlier, and that we may even see the gap shrink a bit. That seems a safe assumption given that so far Rasmussen’s track doesn’t show that kind of a bounce.

We’ll know for sure in a couple of hours though.
4 Chris J. Breisch Says:
August 29th, 2008 at 7:11 am
Oh, I’m fudging the math some. The three day totals that I put above aren’t really how those percentages are calculated.
5 Chris J. Breisch Says:
August 29th, 2008 at 12:32 pm
Ok, today’s Gallup track is finally out and the “bounce” is up to 8. Oddly, that probably means that Thursday wasn’t quite as good as Wednesday, but still much better than Sunday. If Thursday had been as good as Wednesday, the bounce should’ve gone up more than that, to at least 9, by my calculations.

You’ll expect another “bounce” on Saturday when Friday is included (since today’s won’t have Obama’s speech in it yet), which will be perfect timing to solidify the bounce, since Wednesday will be rolling off then.