Our Election Seers
With the election less than a month away, let's swing by the booths of the election fortune tellers. No one, of course, holds a candle to our very own Prof. Rahe and his predictions.
The Daily Beast has a neat little "election oracle" that predicts how the midterms are going to shake out. The Beast has Dems narrowly holding on to the Senate (51 to 49) and, of course, Republicans taking the House--but also by a narrow margin (220 to 215).
The election oracle also makes predictions about individual races (for instance, in Nevada, The Beast says that the likelihood that Reid will beat Angle is 60%. The RealClearPolitics poll average has them dead even at 50%-50%. The Beast says that the probability that Barbara Boxer will edge out Carly Fiorina is 60%, while the RealClearPolitics poll average has Boxer in the lead by only 5.4%).
By contrast, Larry Sabato, at the Center for Politics at the University of Virginia, predicts the GOP will gain 7 or 8 seats in the Senate and 47 seats in the House. Sabato rates both the Nevada and the California Senate races as "toss ups."
On the Senate side, the soothsayers at RealClearPolitics have numbers closer to Sabato's. RCP says that there will be 8 GOP pick ups in the Senate. In the House, RCP is more equivocal, saying there are 188 seats that are safely blue, 210 that are safely red, and 137 that are toss ups.
These are just three predictions of the countless ones out there, but they're interesting to compare. The Daily Beast's numbers are different enough from the other two outlets cited that I wonder how The Beast came to its figures. The Beast says of its methodology: "Our WiseWindow technology scrapes millions of posts from all over the Web for a daily picture of public opinion. That is merged with poll data for the most current and frequently updated election predictions on the Web."
That sounds awfully subjective to me--an unscientific gauge of the election. The "millions of posts" on the Web that are The Beast's primary sources for its "election oracle" must be from blogs, online publications, and mainstream media outlets--not exactly an objective pool of writers and thinkers, and certainly not representative of American voters at large. In the end, the voters, not millions of bloggers, will determine the election results.
Meanwhile, Sabato and RCP are more quantitative and empirical in their approach, using polls conducted with potential voters to shape their predictions. That should mean that their numbers will be closer to what actually happens come November 2.
- Comment (9)
- · Quote
- · UnfollowFollow (1)
Comments :
Jul '10
Re: Our Election Seers
I think I like Prof. Rahe's predictions Here at Ricochet more than these.
Aug '10
Re: Our Election Seers
Get ready for an October Surprise. Statists - probably with guidance from George Soros, master of currency manipulation - will engineer some economic event; attempting to stampede the electorate.
Somebody engineered a multi-hundred-billion-dollar run on Treasury bonds and bills - collapsing the dollar - in mid-September of '08, which precipitated the worldwide meltdown.
Before that run happened, Obama's lead had evaporated, and McCain was dead even with him in the polls.
Or it could be a war somewhere. They're getting scared, and will react kinetically.
Jul '10
Re: Our Election Seers
I think that Dick Morris has a pretty good grasp on the missing element in all of the polls. That is the enthusiasm of the voters.
I seem to recall that the polls had Murkowski and Castle winning thier respective primaries, yet the voters seemed to disagree.
Byron York also noted that historically polls tend to underrate the performance of Republicans in general elections. When they lose, they lose less badly than polls predicted, and when they win, they win bigger than polls predicted.
There is also a concentration issue for the Democrats to deal with, that is they are all bunched in certain areas. Nancy Pelosi, no matter how unpopular, will win re-election. Period. You could get film of her having dinner and making deals with Sadam, Mao, and the Devil and she will get 68 percent of the vote in her district. (Look at Wrangle.) That means that outside of those areas of concentration Democrats are dependant on voters beyond thier base, voters that the leadership, from the areas of concentration, have alienated.
.
So I'll make my crazy prediction again right here;
.
House: +124 seats Republican
Senate: All but 2 seats Republican.
Re: Our Election Seers
In ordinary times, Sabato and RCP would be quite close to being right. The folks at The Daily Beast are whistling as they pass the graveyard. The real question is whether there is something underway that the pollsters do not know how to measure yet. Remember: everything that Sabato and the pollsters do is based on recent history, and in recent history there has never been an occasion in which the likes of Joe Miller and Christine O'Donnell have come out of nowhere to defeat entrenched establishment candidates. They have not ever seen anything like the Tea-Party Movement, and it is upsetting calculations right and left. I do not believe that any conceivable October surprise will alter the current momentum.
Here is further evidence. Sunday's Gallup poll regarding likely voters -- which shows the Republicans winning by 13% if there is a high turnout and 18%, if not. One would have to go back to 1894 to find such a gap. Is it exaggerated in the poll? Arguably, yes. How exaggerated? Maybe, not by a whole lot. There is rumbling, and an earthquake is on the way.
Jun '10
Re: Our Election Seers
Universally bad polls in an election race can severely affect fundraising, so considering how important money is, the polls can become a self-fulfilling prophecy. People don't want to just throw their money away. So, if you're a pollster that wants things to look closer than they are, you can do that by manipulating your model for how many Republicans, Independents, and Democrats will be voting. And then, in the last two weeks, you can worry about accuracy and the reputation of your poll.
Re: Our Election Seers
River, I'm going to pretend you're being tongue in cheek, because you're coming dangerously close to conspiratorial thinking which is not ok!
River: Get ready for an October Surprise. Statists - probably with guidance from George Soros, master of currency manipulation - will engineer some economic event; attempting to stampede the electorate.
Somebody engineered a multi-hundred-billion-dollar run on Treasury bonds and bills - collapsing the dollar - in mid-September of '08, which precipitated the worldwide meltdown.
Before that run happened, Obama's lead had evaporated, and McCain was dead even with him in the polls.
Or it could be a war somewhere. They're getting scared, and will react kinetically. · Oct 6 at 7:29am
Re: Our Election Seers
Prof. Rahe, Sabato and company are comparing this year's election to the one in 1962. Do you think 1962 is analogous to 2010?
Jun '10
Re: Our Election Seers
All better now
Edited on Oct 6, 2010 at 11:41amRe: Our Election Seers
No, I do not think that there is any resemblance at all between the current situation in 1962. Keep in mind that for most Americans most of the time foreign policy was the primary concern in 1962. The Cuban Missile Crisis altered things because everyone saw it as vitally important. Today people are focused on domestic affairs -- unemployment, Obamacare, high deficits, and the prospects of high taxes... What could happen to shift their focus? A terrorist attack, perhaps. But Obama has set himself up for blame on this. We have had one successful attack at a military base in the US, one almost successful attack on an airliner, and a third almost successful attack in Times Square. In any case, the Republicans are doing a whole lot better in the polls today than they were doing in early October, 1962. Sabato and company are political scientists familiar with the recent; they need the help of historians. Let me add that he is underestimating what turnout can do.