|
||
Are the Polls Tightening?
The question on everyone's mind today is - are the polls tightening? As I suggested the other day in my post, Expect McCain to Gain Ground These Final Weeks, the real question isn't whether the polls are "tightening," but whether the fundamental dynamic in the race - a clear and decisive win by Barack Obama - has begun to change. I guess you can say that John McCain turning an eight-point race into a six-point race - which of course is within margin of error - shows the race is tightening. But it still means a landslide win for Obama. So is this concept of tightening at this stage important, salient? I'm not so sure.
Surveying all the main sites - Real Clear Politics, fivethirtyeight, Pollster.com - and DemFromCT's always excellent early morning analysis - there is no evidence of sustained movement to McCain, or any major change in the fundamental dynamic in the race. The averages have it six points today 49-43, and if anything, the news from the states just got a whole lot worse for McCain, as his campaign now admits with CO, IA, NM and VA slipping away, that their map is essentially impossible. Obama's number - the important one to watch - is holding steady in most polls at 49-51. As I wrote yesterday, unless that number starts to drop, there is no way McCain can win at this point.
So the question isn't "tightening" now; it is whether the map and the polling have changed enough to alter the dynamic in the race. And the answer to that today is clearly "no."
All this became clear to me when I watched CNN this morning. They claimed the race was tightening by showing one-point movement in polls in Ohio and Missouri, while showing Obama still winning Florida by a wide margin. The anchor's conclusion by looking at these 3 states was "the race was tightening." But of course by showing Obama winning Florida and with OH and MO within margin of error, what he was really saying - but could not say - was that this data showed Barack winning the election. It will be important to hold analysts and commentators accountable on this point in the final two weeks.
With the national numbers apparently stabilizing, it will be interesting to see what happens in the states these next two weeks. My sense is that this is where the Obama fundraising advantage will really kick in, and if anything, we could see improvement in the states while the national numbers either stay the same or we see McCain bring his base home and get a slight national uptick at the end.
Finally, as the CNN piece above suggests, I think the states to watch now are the small and medium sized states. If Obama holds the Kerry/Gore states, he then just needs to win 3 of 5 currently Obama-leaning states - CO, IA, NM, NV, VA - to win. These are now the true battlegrounds and if McCain cannot chip away there, the race will be over long before Election Day.
Cross-posted at the NDN Blog.
Follow Simon Rosenberg on Twitter: www.twitter.com/ndn729
Loading comments…
Want to reply to a comment? Hint: Click "Reply" at the bottom of the comment; after being approved your comment will appear directly underneath the comment you replied to
To clarify my last post (it's really annoying that you can't edit after posting):
1. Of the red states where Obama has a CLEAR lead Colorado has only 9 EVs, New Mexico has 5, and Iowa has 7. This means Obama has to win a few of these states to guarantee victory, increasing risk.
2. See above. Since NC etc are quite close it would be useful to win Ohio/Florida but this is not certain.
3. I was referring to swings in previous elections at this point eg
http://news.yahoo.com/s/ynews/20081021/pl_ynews/ynews_pl101_11
4. Historically undecideds tend to break against the incumbent party but like I said there's a good chance it wil be different this time.
As Charlie Cook has pointed out, Obama's overwhelming advantage in early voting means that even if McCain surges in the last week, he likely won't have enough voters left to draw on to put him over the top.
and the clock tics.............tic tick tick
The way I see it, McCain would have to win FL (27), PA (21), OH (20), NC (15), VA (13), ND (3), and OH (20) just to barely squeeze out a victory (that is assuming that all the currently blue states stay blue and the currently red states stay red). From what I have read, the flip-flopping of solid blue and solid red states at this point is unlikely.
I'm very nervous about overconfidence on our part. We're creating the sensation that he's already won. VERY risky. McRove is a dangerous, wounded animal and 13 days is a LONG time in politics.
Save the celebrations for No. 5, act like McCain is winning, and pound the opposition into the ground with a landslide.
Obama 364, McCain 171, Ties 3 -- http://www.electoral-vote.com/
Obama 52%, McCain 41% -- http://www.gallup.com/Home.aspx
13 days to go.
Thanks for expanding on your previous argument, but I still don't get it.
i wouldnt get too excited yet. polls for carter showed him leading reagen..and we know who won that race...also, clintons polls said he should win by 19 points and he won by 6...look up the history of polling and you will see they poll overboard pro democrats...then dems LOSE
with 13 days to go...in politics that is a lifetime...so hang on, it is going to be a bumpy ride....
This election, as past history shows, is far from over. Obama has a lead now not because voters are loyal to him-it is based on the economic meltdown.
Ohio is leaning to McCain; Florida will go to McCain. None of these Republican "swing" states are a lock for Obama.
The article didn't mention Ohio or Florida. Obama doesn't need either to win. As to voter loyalty, almost exactly the opposite is true. If you really look at the trend line, with the exception of a couple of weeks in September, Obama has been pretty steady for months and is now starting to trend consistently upwards. In contrast, McCain support has swung wildly up and down. His support seems to be at the mercy of circumstance now.
http://www.realclearpolitics.com/epolls/2008/president/us/general_election_mccain_vs_obama-225.html
To be at the mercy of events unfolding is like being on an out of control rollercoaster----
Obama's chances and the myth of the Undecided Voter:
Everyone should look carefully at the polling data for the Democratic primary in the following states:
NC, MO, CO, IN, FL, PA, WI, SC, VA and GA. (OK, I cherrypicked states that support my argument, but..)
In each of these states, Clinton received the percentage of the vote predicted by the polls, within +/- 3 points. But in each of these states Obama received WAY more votes than the polls predicted.
The standard narrative built by the MSM was that "undecideds" "broke" for Obama at the last minute.
Well, the way I see it, the polls underpolled key Obama voting blocks: people under 40 and minorities. Younger people tend to be more active and spend less of their time at home. Young people and minorities tend to be more suspicious of strangers calling and may be more comfortable with technologies like Caller ID. Young people and minorities tend to work more in sectors of the economy that involve odd hours, so Saturdays may be a work day, or 8 PM they may be on the clock, not at home watching game shows. There are a lot of reasons why these groups tend to be underpolled. I think it is a plausible alternative explanation to the Obama bump in each of these states during the primary.
And if polls were underpolling Obama supporters in those states in the primaries, why wouldn't they do so in the general election polls?
My concern in the recent wave of news coverage given to the tightening race and the problems with exit polls is that.... though there is certainly truth in these factors... I fear the majority of Americans are being prepped to accept another fraudulent election. The message will be:
"Surprise! McCain won! Gee, too bad. Gosh darn, the strangest things happen in elections, don't they? Well, let's get on with it."
I mean, if you've stolen two national elections, what's a third? (And I think the evidence of election fraud in the last two Presidential elections at least merits serious concern.)
This time though, there may be some very unhappy people who won't swallow it again. This might explain why there is a 3,000 man Brigade Combat Team deployed in the U.S. for "crowd control." At the time I had to ponder, what crowd control? But now I think I get it.
Want to sell out your country for a few extra (hundred) dollars in a tax rebate?
Want big government and loss of the freedoms you currently enjoy?
Vote Obama/ Biden '08
want more war, more anger and more hate?
Want big government in your private life and small government in oversite?
Vote McCain/Palin
want coal in your sock this Christmas?
---Vote Nader
Patriot act, wiretapping, dissent suppression, nationalization of banks....
whose watch did this happen under again? Refresh my memory...
Want to sread the wealth among the rich people only?
Vote Palin-McCain Administration.
Go back to sleep. We'll wake you in eight years.
"Want to sell out your country for a few extra (hundred) dollars in a tax rebate?"
Uh huh. Yours is the party in favor of loopholes that allow corporations to locate outside the U.S. to avoid having to pay any taxes (but maintain 100% of their influence in Washington lobbying). Talk about selling out your country.
The scariest scenario for me is if the networks call the race for Obama very early in the day based on exit polls in the East, and as a result Obama voters west of the Mississippi stay home.
I encourage everyone to vote, no matter what the media is saying on election day! Let's make this a landslide!
Even as an avowed pessimist (and not for good reason in view of 2000/04) it is clear that Obama is a firm favorite for Nov4.
HOWEVER, there are a few points I would make.
1. The states where Obama has a clear lead are small states with few electoral votes. He should win enough to pass 270 but things may be touch and go if the polls tighten.
2. With regards to the big states I'm not sure Florida and Ohio are as promising as some people think. Based on the RCP average they are both very close although showing an Obama lead. However at the state level many polls are from unreliable small pollsters The Ohio average, for instance, is massively skewed by one poll with a small sample size showing an unrealistic +9 lead.
Although Rasmussen are Republican leaning, their previous polls haven't shown any obvious bias this cycle and they are at least an experienced polling organization which, for instance, use a reliable sample size. The last 2 polls have Obama behind in Florida and Ohio.
3. Although, as far as I know, no one has won from this far behind at this stage there have been large enough swings for McCain to catch-up.
4. Further to 3, I have a nagging concern that this election will go against history and undecideds will break for McCain. One reason is, of course, race but I think many will also think McCain is a "safety first" option.
Your points made -
Now here are responses:
1. I don't know what states you are referring to that have "few electoral votes" but of the current "battlegrounds" NC is not small, CO is not small, VA is not small, and MO is not small.
2. You apparently are not thinking through the electoral math. Obama doesn't need OH or FL to win. To win either of those would be a coup de grace, but really, he just needs VA or NC or MO or CO to cross 270. Something tells me he is likely to win at least one of those four states. check out fivethirtyeight d o t c o m for a clear analysis of the pros and cons of each poll, if you still feel insecure.
3. No, you are confusing swings by McCain in individual state polling... You really need to train yourself to think in terms of electoral votes... this isn't a popularity contest. I don't think there have been any significant swings in projected electoral vote count ... not nearly as volatile as in 2004. See electoral-vote d o t c o m to compare data.
4. Undecideds don't "Break". That is a nagging myth and misunderstanding. polls are also potentially under-polling Obama supporters.
1. Of the red states where Obama has a CLEAR lead Colorado has only 9 EVs, New Mexico has 5, and Iowa has 7. This means Obama has to win a few of these states to guarantee victory, increasing risk.
2. See above. Since NC etc are quite close it would be useful to win Ohio/Florida but this is not certain.
3. It's really irritating that you can't edit after posting. I was referring to swings in previous elections at this point eg
http://news.yahoo.com/s/ynews/20081021/pl_ynews/ynews_pl101_11
4. Undecideds tend to break against the incumbent but like I said there's a good chance it wil be different this time.
Princeton Professor of Neuroscience Sam Wang's highly lauded blog site offers precise meta-analysis of the electoral map and votes using proprietary means of aggregating state, not national polls (I'm simplifying his very selective and precise method).
Wang's analysis is not a predictor of the race's outcome, but rather, an accurate (less than a 1% margin of error) daily snapshot of the electoral landscape.
Check it out for yourselves:
http://election.princeton.edu
Love that EV map. Gives a much better interpretation of the relative political weights of different parts of the country.
Absolutely. The other day, for a half hour, HuffPost seemed to use Sam's findings as the front page headline, map and all. Was used sort of like an alert, or quick update.
Of all the aggregated polling sites, Sam Wang's strikes me as the most comprehensive and precise. The fact that Prof. Wang does not burden his analysis with the need to predict contributes greatly, I believe, to his accuracy.
You must be logged in to comment. Log in or connect with
