March 9, 2008 - March 15, 2008
We want to update regular readers on two important issues affecting our comments facility.
First, we believe we have, at long last, identified and eliminated the technical glitch that caused long delays and error messages for those posting comments, frequently resulting in inadvertent double and triple posts. As of last night, hopefully, comments post without error messages within 10 to 15 seconds of clicking the "post" button.
I know our various comment bugs have been a source of frustration for just about everyone that has tried to post a comment on Pollster in recent months, and I want to once again apologize for the inconvenience we caused. Hopefully your experience will be far smoother from now on. If not, and in particular if you encounter any error message while commenting, please do not hesitate to email us with full details. We are committed to squashing these bugs once and for all.
Please note that it may take a minute or so for your comment to appear after you post it. To check that your comment has posted successfully, wait a full 60 seconds and click your browsers "refresh" or "reload" button. If you are really impatient, you can also perform a "hard reload" of the page (details here).
Second, now that comments are posting as they are designed to, we want to address another nagging problem resulting from the dramatic increase in traffic over the last few months. On Monday, if all goes well, we will begin requiring a valid Typekey username in order to post on Pollster.com.
Typekey is a free, third-party system that allows you to create a unique username. The registration is fairly painless -- you just need to be a real person (not a SPAM "bot") and leave a working email address with Typekey that can be used to validate your account. However, when you post a comment with a Typekey username, we will not be privy to your personal information, so you can remain completely anonymous.
We are making this change in response to your requests for greater moderation of the increasingly uncivil posts in our comments section in recent weeks. This change will hopefully discourage some of the worst behavior and put us in a position to band the most abusive commenters. It is not a panacea, of course, but it's a start.
If you do not have one already, I recommend that regular commenters go ahead and acquire your desired user name now. Please note that we may hold off on implementing Typekey sign-in if the new procedure creates any new delays or error messages like the ones we just worked to eliminate. We will keep you posted.
Finally, we are committed to maintaining an open comments section that allows for the free and civil exchange of views and dissenting opinion. If you have thoughts about how we might improve the tone of the discussion, please leave a comment below.
(Today's Guest Pollster's contribution comes from Professors Robert S. Erikson and Karl Sigman of Columbia University.)
In late February, SurveyUSA interviewed 600 registered voters in every state for a total of 30,000 interviews, ascertaining preferences in a McCain-Obama and a McCain-Clinton race. The focus was a new set of electoral maps of red and blue states based on who led each state in the survey. Based on who won each state in the SurveyUSA survey, Obama defeats McCain 280 to 258 while Clinton defeats McCain 276 to 262 in the Electoral College.
Of course SurveyUSA's mammoth undertaking at best presents a snapshot of the states at one point in time. And even if all the niceties of polling were perfectly met, the allocation of states as "red" or "blue" is problematic due to sampling error. Here, we take the analysis of the SurveyUSA 50 state polls one step further. Rather than assign states based on who leads in the state surveys, we assign states probabilistically to the Democratic or Republican candidate based on the SurveyUSA state polls. Then, based on these probabilistic estimates, we ask the question, given the SurveyUSA results, what are odds of an Obama or Clinton victory in the Electoral College?
To do this, we conducted one million simulations (in MATLAB) of the Obama-McCain contest and then one million more simulations of the Clinton-McCain matchup. In each case we assume that the state estimates were correct except for sampling error. Using sampling theory and the assumption of simple random sampling, we draw one million estimates of the vote for each state. In each case we draw from a normal distribution with the observed mean (percent Democratic vs. percent Republican) and the standard deviation determined by the number of respondents in the state reporting a preference (always slightly under 600).
What do our results show? First, we pooled the state polls to ascertain the national vote, weighing each state's percent in proportion to the size of its House delegation. We also assign the District of Columbia as a 436th district and assign each Democratic candidate 85 percent of the vote to McCain's 15 percent. With these assumptions, the national popular "vote" is tight as of late February. Obama wins 51.5 percent versus McCain's 48.5 percent. Clinton also wins by an even razor thin margin, 50.7 to 49.3. With 30,000 cases, both estimates are statistically significant. McCain would be in the actual popular vote lead less than one time in 20.
That being said, our simulations yield a 88% chance of Obama beating McCain (with 306 Electoral College votes on average versus 233 for McCain), and a 74% chance of Hillary beating McCain (with 285 Electoral College votes on average versus 253 for McCain). About one percent of our simulated outcomes were Electoral College ties. (We ignored within-state variation in Maine and Nebraska, which divide their electoral votes by district.)
On the one hand, we find the expected numbers of electoral votes (the average from the simulations) for Obama or Clinton to be slightly higher than SurveyUSA reports. On the other hand, there is sufficient variance in the outcomes, so that McCain wins a nontrivial portion of the simulations, even with Obama as the opponent. Our two million simulations remind us that the popular vote winner is not always the Electoral College winner, although probably due mainly to chance -- the lottery aspect of the Electoral College -- and not any identifiable partisan bias in the 2008 Electoral College.
We thank Linda Liu for her technical assistance.
Obama 48, Clinton 46
Clinton 47, McCain 45... Obama 46, McCain 44
Record Level of Pessimism About Economy
NBC News/Wall Street Journal
(NBC story, results; WSJ story, results)
Clinton 47, Obama 43
Obama 47, McCain 44
Clinton 47, McCain 45
Kathy Frankovic considers the Democratic superdelegates in the context of "poll after poll" showing Americans' "desire to have elected officials listen to them."
Frank Newport says "rank and file Democrats" are "on top of what's going on" regarding the controversies over Michigan, Florida and the role of the superdelegates.
Carl Bialik continues to explain the contradictory Democratic delegate tallies with a report on a "little noticed shift" of eight delegates in California to Obama.
Mark Mellman says the "analytic strategy" of projecting "primary performance into the general election" is "fundamentally flawed."
Jay Cost crunches past primary and general election vote data and generally agrees.
Paul Lukasiak mines the SurveyUSA 50-state polls for gender gap data.
The Wisconsin Advertising Project releases data on television ads in Ohio.
Gary Langer looks back and concludes that while Americans "suspected" that Saddam Hussein backed Al Qaeda, they did not affirmatively "believe" it.
My National Journal column, which looks at the Republicans who turned out to support Hillary Clinton in Mississippi is now posted online. It includes some exit poll numbers, courtesy of Sarah Dutton at CBS News, that shed new light on their votes and remind us of the real reason we value exit polling.
The short version: The survey yields evidence that many of these voters like John McCain as much or more than Hillary Clinton. At the same time the Clinton Mississippi Republicans are nearly unanimous in their disdain for Barack Obama. The column has a ton of data, so please go read it all.
Here are some additional results shared by CBS News that did not make it into the column, among the roughly 9% of the Mississippi Democratic primary electorate that voted for Clinton but identify as Republicans (n=147):
- 99% are white
- 56% are female
- 47% have a college degree
- 65% report incomes of more than $50,000 a year; 17% report incomes of more than $100,000 a year
- 68% describe themselves as conservative; only 8% describe themselves as liberal
Some addtional substantive items for the same subgroup of Clinton Republicans:
- 54% said Clinton would be more likely than Obama to beat McCain, 37% said Obama would be more likely to beat McCain
- 67% rate the economy as the most important issue facing the country, 22% name the war in Iraq and 8% name the health care as most important issues
- 38% said "has the right experience" was the candidate quality that mattered most when they voted, 20% said "has the best chance to win in November," 15% said "can bring about needed change" and 14% said "cares about people like me"
- 61% thought Clinton attacked Obama unfairly; 55% thought Obama attacked Clinton unfairly.
- 43% decided how they would vote in the last three days or the last week; 55% decided in the last month or before that
Official exit poll tabulations will be posted momentarily (as the polls close) at these links:
I'll live blog to the extent that news seems relevant in reverse chronological order, all times Eastern:
12:36 - Our comments section featured some speculation early this evening about the early exit poll estimates being wrong. The estimates were off slightly but with 99% of the precincts reporting, it looks like they were off in Clinton's favor. The current count shows Obama winning by a 22-point margin (60% to 38%) and more than 401,000 votes cast.
Two things I notice in the current exit poll tabulation (which may still update again by morning). First, make of it what you will, but Obama's 26% of the white vote was comparable to what he received from white voters in Alabama, Louisiana, South Carolina and Tennessee. Among the Southern states, only Georgia gave Obama more than 30% of the white vote.
Second, the open Mississippi Democratic primary -- the first to be held after John McCain secured his nomination -- included 2-3 times as many Republicans (12%) as the other states. And those Republicans supported Clinton by a 3-to-1 margin, far more than Republicans in any of the other Southern states.
11:07 - An update on turnout: I can't find a better source online, but this MyDD post from earlier today quotes a CNN story that put the total vote in the 1988 Mississippi Democratic presidential primary at "more than 359,000." The current Clinton-Obama vote total, according to CNN, is over 366,000 with 9% of the precincts uncounted (and Thatcher is ahead of me on this in the comments).
10:32 - In the comments, Thatcher notes:
all day long, they were saying that turn-out was light-to-moderate ... about 100K-150K voting ... yet we are over 240,000 now. No, it won't hit 1988 numbers - but what's up with the majorly wrong turnout prediction?
A quick Googling turns up an AP story that had the Mississippi Secretary of State, as of last night, "predicting a light to moderate turnout" of 125,000 to 150,000 today (emphasis added). So we are talking about a pre-election prediction, not an estimate based on actual turnout. What are these sorts of estimates based on? Who knows, but in my experience, this is not the first time that a Secretary of State's early prediction turns out to be wrong.
10:23 - An update in the exit poll tabulations shows the current vote estimate (which should by now be based mostly on random sample of actual results from randomly selected precincts) shows a margin of 56% Obama, 41% Clinton. That's pretty close to the current actual vote count (57% Clinton, 41% Obama with two-thirds of the precincts counted).
8:33 - 6% for Paul, not 62%. My bad.
8:30 - I just want to take a moment to apologize, again, for the comment posting bug that has frustrated everyone over the last month or so with error messages and caused inadvertent double (and sometimes) posts. In an effort to identify the bug, we made one small and very temporary change this afternoon that will allow most posts to go through without an error message (but unfortunately, only after a long delay of 60 seconds or more).
However some of you will get an error message saying your comment was not posted because of "too many comments submitted from you in too short a time." That message is in error - no one has been blocked! However if you see this message you will most likely need to repost your comment. I can't apologize for this enough. Please bear with us for a few more days and we'll get this ironed out (I promise...or your money back).
8:24 - The exit poll tabulations just updated. Mark Lindeman's extrapolations show no overall change in the estimated Obama lead (59% to 41%).
8:19 - MSNBC just called the state for Obama. Note that the networks did not call the race for Obama even though the estimate used to weight the official exit poll tabulations showed Obama with a 19 point lead. Looks like the decision makers at the networks felt they needed to see some actual count (or perhaps a more complete set of exit poll interviews) to increase their confidence before making a call.
8:15 - The initial racial composition numbers in the tabulations are 49% white, 48% black. The latter percentage is a 5-7 points lower in the five pre-election polls released over the last week. The preliminary estimate of vote by race shows Clinton leading 72% to 27%, while Obama leads among African Americans, 91% to 9%.
8:03 - As the first tabulations go up, our friend Mark Lindeman emails the overall estimates extrapolated from the tables: 59% Obama, 41% Clinton for the Democrats and 76% McCain, 13% Huckabee, 6% Paul for the Republicans.
Click here for the usual caveats on how these numbers are derived and how they improve over the course of the evening. Click here from my National Journal column last week on how the estimates at this point in the evening have sometimes been off in Obama's favor (although not in either Georgia or Tennessee).
For those who will be watching results from the Mississippi primary tonight, here is a breakdown of the demographics of recent surveys as well as the tabulations of vote by race. First, the demographic composition and overall results:
Obviously, we have far fewer polls (and pollsters) to consider than for last week's primaries in Texas and Ohio. Only three pollsters have been active in Mississippi -- ARG, Rasmussen and InsiderAdvantage -- and their reported demographic compositions have been reasonably consistent. ARG uses live interviewers, Rasmussen Reports uses an automated (IVR) methodology and InsiderAdvantage has used both in recent months but does not specify their methodology on their last two releases.
The vote results by race are less consistent. All show Clinton with a wide lead among white votes and Obama with a wider lead among African-Americans, but the specific results -- particularly Obama's support among black voters -- have varied. Assuming that the networks conduct an exit poll tonight, we will see in a few hours how the results from that survey compare to those above.
Update: Rasmussen Reports emails with demographic composition numbers (thank you), so I updated the table above.
Update2: As jr886 points out in the comments, the folks at The Page are certainly expecting exit poll results.
Downstairs at The Hotline, my colleague Matt Gottlieb caught a finding in the general election presidential trial heats in the latest Newsweek poll that confirms some analysis in the most recent survey from the Pew Research Center:
In the Newsweek general-election matchup between Clinton and McCain, Clinton leads overall by 2% but leads among Obama primary voters by 51%. In a matchup between Obama and McCain, however, Obama leads overall by 1% but leads Clinton primary voters by just 36%.
The Pew Research Center's analysis on their most recent national survey notes a similar pattern:
The vast majority of Democratic voters say they would support either Obama or Clinton over McCain. But in an Obama-McCain matchup, 14% of Democratic voters say they would support McCain, compared with 8% who would do so if Clinton is the nominee.
While the measurements differ, both show essentially the same thing. In match-ups against McCain, Clinton does slightly better among Democrats than Obama. But keep in mind that on the Pew poll, Clinton's advantage among Democrats appears to be offset by Obama's greater advantage among independents and Republican identifiers, as both Democrats are preferred by 50% of registered voters. The Pew report has full details on the demographics of the independents that Obama attracts and the Democrats that defect. On the Newsweek poll, Clinton's advantage is slightly greater, since she wins 48% of the registered voters compared to 46% for Obama.
Of course, the slight Clinton edge on both the Pew and Newsweek polls is at odds with most of the other recent surveys, although the more recent Newsweek survey is consistent with two automated surveys conducted by SurveyUSA and Rasmussen Reports in recent days.
All of this is a good way to point out that we have just put up tables and charts on the national general election polls (Obama-McCain and Clinton-McCain) and will be rolling out charts for individuals states over the coming days and weeks.
PS: The complete results for the Newsweek poll are not (yet?) posted online, but Newsweek has posted unweighted subgroup sizes something new: A helpful demographic profile of registered voters and the Democrats and Democratic "leaners" that answered primary vote questions. Thank you, Newsweek.
Update: The Pew Research Center's Scott Keeter emails to the observation that "defectors" from Clinton and Obama essentially cancel each other out in their late February survey. In match-ups of Clinton vs. McCain and Obama vs. McCain among registered voters, 8% were for Clinton not Obama, 8% were for Obama not Clinton, 40% favored both Clinton and Obama and 35% were for neither Clinton nor Obama. It would be useful to those debating the issue of "electability" to see the same computation for other surveys.
SurveyUSA reveals Gov. Eliot Spitzer's monthly Job Approval ratings from 1/15/07 through 2/18/08 and shows a drop in his approval from 62% to 35% as well as a rise in his disapproval from 24% to 62%.