bup quoted some of the data, but I was actually going on a statement from Nate Silver yesterday:
So since 1968 at least, the challenging party has never finished with a bigger margin than what they had right after their convention other than Obama in 2008.
And I would argue that the collapse of Lehman and general financial meltdown counts as a “unknown unknown” - the sort of external event that drives elections outside their typical path. And I think that, barring one of those, Obama is a solid favorite right now.
Statistically speaking, it’s wildly improbable that a dozen randomly selected polls are all going to be sitting at the edge of their confidence interval, all in the same direction. The only way they’re all wrong in that fashion is if the sampling isn’t representative (a possibility I’d never rule out, especially with the increase in people with no land line - I know polling firms are trying to compensate for that, but are they doing so in an accurate fashion?)
A.) There is a serious problem on this board with reading comprehension skills. I do believe I said “even though it’s old”. Hell, you even quoted it. Way to point out something I pointed out myself. I guess that “gotcha!” didn’t work out, huh? Anyway, at least I produced something disputing the quoted claim. As it is, I’m still waiting for the cite which says that undecideds tend to break along the same ratio as the decides. I’ll wait. Or do certain people of certain persuasions get to make claims without cites? (Actually I find it a bit odd how I have to disprove the claim made without proof with something more palatable to you even though the onus is typically on the person who makes the positive claim to prove that as true*). Now perhaps you’ll even take the chance to find said source, since you decided to-- apparently-- rush to someone else’s defense or something.
B.) What you quoted from me has nothing to do with what I said or responded to. In fact, I’m not even sure what point you were attempting to make there.
I don’t think “convention bounce” is going to be much of a thing anymore. I full;y expect Obama to get a similarly small convention bounce.
The thing is that conventions used to represent a significant degree of added exposure for a candidate. Back in the day, before everyone had the Web and 24-hour news channels, the degree to which you were exposed to the Presidential candidates was limited. In the 1980 election, if you wanted to know what was up with Carter and Reagan during the campaign, you’d get 10-20 minutes at most during the six o’clock news, plus your newspaper. It was the same in 1984; in 1988 more people had cable news but it wasn’t yet as popular as it would become, and as late as 1996 most people didn’t use the Web. That was more or less the extent of the news media at the time. When the convention rolled around, that represented a big chunk of Candidate A getting his mug in front of the public. Their awareness of him and his VP pick could actually increase.
Today - I mean., Christ, you see Obama and Romney on 24/7. Turn on your computer and there they are. The cable news networks are running crap about the campaign morning, noon and night. People didn’t need the convention to hear all about Romney and Ryan; they’ve been hearing all about it for what feels like a million years. The convention barely represents an increase in in-your-face electioneering at this point.
Perhaps, Rick. But McCain got a very nice bounce in 2008. As did Bush in 2004. The quality of the conventions (and the candidates) still matter.
As to undecideds, IIRC the latest research indicates that it is very different for races with high name recognition like POTUS. I’ll try to find the cite tomorrow.
While I think that’s truer than it has been, I still think there are a number of Americans who do not “tune in” until about now. Not everyone uses the internet regularly. Not everyone watches cable news shows.
Speaking as a Pubbie.
If Romney loses, the Republicans will look at the numbers. 97+% of religious fundamentalists, social conservatives and tea-partiers voting for Romney vs. 20-30% of independent voters voting for him. Their reaction to those numbers will be
“How do we get that 3% of our base to vote Republican?”
You didn’t get the point. The point is that you seemed to be asserting by implication that we could take the results of a 23 year old study and fully apply its conclusions to the present day. Given how much the electorate, the process, and society have changed since 1989, I think you bear some responsibility to explain why you think the conclusions should be exactly the same.
This discusses the so called “incumbent rule” - the tendency for undecided voters to break more for the challenger in races featuring an incumbent.
It expounds on the previous cite from Omg a Black Conservative and notes that while the so called “Incumbent rule” has shown signs of weakening, a study updating prior work shows the incumbent rule strengthening in Presidential races.
The"Incumbent rule" refers only to how undecideds tend to break between the last poll and election day. There are relatively few undecideds left at that point so even a large break for the challenger may not translate to more than a percentage point or two in terms of overall vote tally.
The Mark Blumenthal article refers to a Guy Molyneux piece, The Big Five-Oh written for the American Prospect which argues, “An incumbent who fails to poll above 50 percent is in grave jeopardy of losing his job.”
At the moment Obama is polling 47% in the Gallup poll and only touched 50% briefly back in April.
Given this work I’ve referred to was from 2004, I checked the results for the 2004 election against the last pre-election poll of likely voters (as was the protocol of the study referenced). I didn’t include the 2008 race because there was not an incumbent,
In 2004, Bush v. Kerry. Last poll was Oct 31, 2004.
Last poll Bush 49% Kerry 47% Nader/Others 1%
Actual vote Bush 50.74% Kerry 48.27% Nader 0.38%
That looks like about 3% were undecided going into election day. I can’t find poll numbers more precise than rounded to the nearest percentage. Hard to tell if this is a near even split of undecideds or if Kerry got more of them.
Where, exactly, is the cite stating that undecideds breaks much the same way as decideds?
Seriously now. What kind of intellectually dishonest crap is this? How about you-- and everyone else who wants to take offense to my cite-- go out and find a cite substantiating the original claim? Because, as it stands, you don’t have a leg to stand on here except to scream “Well, that’s old!”.
For those keeping an eye on this sort of thing, Nate Silver’s model now puts Obama at a 76.3% chance of winning and a projected total of 310.9 electoral votes. It puts Obama at a 69.7% chance of winning if the election were held today.
To expound: even for “True Incumbents”, like Obama, the best Romney should realistically hope for is a 2-to-1 split of undecideds in polls taken in the last month of the campaign. For polls in September there is a great breakdown plot that shows that most of the “incumbent rule” split is due to regression to the mean - they are actually pretty solidly on the trend line.
… not to mention that a substantial fraction of undecideds are actually apathetic and won’t vote at all.
I’ve seen abstracts of psych studies that show people who admit leaning one direction in a decision (of any kind) really have made their decisions subsconsciously, and only rarely will actually go the other way.
A bad or misleading cite is worse than no cite at all. An unsophisticated reader may actually believe it. Until then a claim is merely unsubstantiated and retains all its potential for confirmation or denial.
