We all see the divergence between the national tracking polls, which generally show Romney with a small lead on average, and the state-level polls, which generally show Obama with small leads in most of the key swing states on average.
How could these two things be true at the same time? There's actually two things going on here:
First, it is entirely possible that the President is winning by narrow margins in swing states like Ohio and Colorado but is losing by more in "solid Romney" states than he is winning in "solid Obama" states. That would translate into a national popular vote deficit but, potentially, an Electoral College victory. Most of the sites that create election-prediction models like Sam Wang's Princeton Election Consortium and Drew Linzer's Votamatic use state-level polls either exclusively or more heavily than national polls because the state-level polls provide more information that is useful. They are generally right to do this in terms of predicting what matters - the Electoral College outcome. But those national tracking polls are still unpleasant to look at ... which brings me to my second point ...
Second, it is really striking to me how few "full" national polls there have been in this campaign, especially in the final weeks of the campaign. There are at least 8 different national "tracking" polls that are released just about every day. But these are not the same as a regular poll. I can't claim perfect expertise on the impact (or not) of this phenomenon but it does seem to me there is some important effect of looking at only tracking polls as opposed to regular "full" national polls. Tracking polls take a smaller sample every night for 3 or 4 or 5 or even 7 (in the case of Gallup) nights. They put these together to create a national polling number. Then, when a new night comes on board, the sample from the oldest night drops out. What you end up with is a "rolling average" of the most recent nights. One potentially problematic aspect of this is that, because each night is a distinct sample and because the pollster wants to keep the demographics in the full sample correct from night to night, you have to weight some very small samples night to night before adding them to the sample. This would seem to introduce some error as opposed to doing a regular 2 or 3-night poll and then weighting the whole sample from there. In effect, you'd need to do less weighting as you'd have stronger sub-samples to work with.
Now, I'm just kind of spit-balling on all that and I could be wrong. But I do think that regardless, it is notable that there have been so few full national polls. Indeed, there has not been one full national poll that I can find since the third debate. I'd be interested to see one of those from a good pollster who does live interviews with cell phones included.
UPDATE: In a post today, Nate Silver speaks to the question of the split between the national tracking polls and the state-level polling. The bottom line is that he says the state-level numbers are suggesting leads for Obama in swing states like Ohio that are statistically very meaningful. Mark Blumenthal makes some similar arguments in Pollster.com
Showing posts with label Mark Blumenthal. Show all posts
Showing posts with label Mark Blumenthal. Show all posts
Saturday, October 27, 2012
Sunday, October 21, 2012
Rasmussen's Weighting
As it happens, Mark Blumenthal wrote about Rasmussen's weighting yesterday just around the same time I was writing my blog post. It was a piece about the national polls generally and Blumenthal noted Rasmussen's importance as the one firm that has done state-level snapshots before and after the second debate and the fact that they show slight gains for Romney. But Blumenthal explains these gains are related to the "dynamic" party weighting Rasmussen is using:
Anyway, what this tells us is that Rasmussen is shifting their party ID weights slightly over time but they are always doing so in response to last week's polling. And that's problematic. If you are looking for a snapshot of where the public is today (the point of a poll), using their mood from last week as a determinant is really problematic.
Rasmussen is one of the few pollsters to routinely weight its samples so they match predetermined targets for the percentage of likely voters that identify as Democrats or Republicans. The catch, as Rasmussen Reports confirms to The Huffington Post, is that its weighting targets are now adjusting on a weekly basis to match the average party identification for likely voters measured on their last six weeks of calling (after weighting for demographics, but not for party). So the party weights for the past recent week may be slightly different than the party weights the week before.Now, here's the thing. IF you're going to weight by party, the shift that Rasmussen appears to have made does make some sense. There was reason to believe, in the wake of the first debate, that a poll should have a few more Republicans in it and a few less Democrats than before. But then, this is why weighting by party ID is a bad idea in the first place.
More important, the weight targets for Rasmussen's national samples grew slightly more Republican in mid-October. Although the data are published on pages available to paid subscribers only, Rasmussen indicates that the national interviews for the week of Oct. 8 to 14 gave Democrats a 1-point edge over Republicans (38 to 37 percent). The party balance for the two prior weeks, Oct. 1 to 7 and Sept. 24 to 30, was a 3-point Democratic advantage (39 to 36 percent).
In the past, Scott Rasmussen has explained that the state-level party weighting targets are derived, in part, from national numbers and the "national shifts appear to provide a good indicator" of mid-year changes at the state level.
A 3-point shift toward a more Republican identification would more than explain the one-point shift to Romney in the five states Rasmussen surveyed this week.
Anyway, what this tells us is that Rasmussen is shifting their party ID weights slightly over time but they are always doing so in response to last week's polling. And that's problematic. If you are looking for a snapshot of where the public is today (the point of a poll), using their mood from last week as a determinant is really problematic.
Labels:
debates,
Mark Blumenthal,
Rasmussen
Tuesday, October 9, 2012
Sullivan Unhinged
That's how I describe Andrew Sullivan's most recent lengthy post on the President's campaign. The post was sparked by the latest Pew poll which shows Romney leading by 4 points after Pew had Obama ahead by 8 points a few weeks ago. Here's some of Sullivan's rant:
The Pew poll is upsetting to many in large part because it is a pollster with a good track record and credibility. This top line caused Andrew Sullivan to descend into italics and exclamation points.
I've been arguing that the Romney bounce from the debate appears to have been about 3 points and that it appears it may be receding from that high. That still seems about right. One thing to notice about the Pew poll is that it was in the field on Oct. 4 to Oct. 7. But, as Mark Blumenthal points out, almost all of those interviews were on Oct. 4 to Oct. 6, when the Romney bounce was at its peak:
Also, it is one poll! Rasmussen has the race even (even this morning). Gallup has the race roughly even among likely voters. But beyond the polling, let me make a larger point about Andrew Sullivan's absurd reaction. If you are someone who thinks that 90 minutes of bad debating is going to cancel out 4 years of a record worth defending and months and months of bad campaigning by Romney, then you are someone who believes there isn't sufficient wisdom and virtue among the people for self-governance.
The polling (beyond the Pew Poll) doesn't suggest the race has been "thrown away." Get a grip Andrew.
Look: I'm trying to rally some morale, but I've never seen a candidate this late in the game, so far ahead, just throw in the towel in the way Obama did last week - throw away almost every single advantage he had with voters and manage to enable his opponent to seem as if he cares about the middle class as much as Obama does. How do you erase that imprinted first image from public consciousness: a president incapable of making a single argument or even a halfway decent closing statement? And after Romney's convincing Etch-A-Sketch, convincing because Obama was incapable of exposing it, Romney is now the centrist candidate, even as he is running to head up the most radical party in the modern era.OK, let's start by discussing that Pew poll in more depth.
The Pew poll is upsetting to many in large part because it is a pollster with a good track record and credibility. This top line caused Andrew Sullivan to descend into italics and exclamation points.
Before the debate, Obama had a 51 - 43 lead; now, Romney has a 49 - 45 lead. That's a simply unprecedented reversal for a candidate in October. Before Obama had leads on every policy issue and personal characteristic; now Romney leads in almost all of them. Obama's performance gave Romney a 12 point swing! I repeat: a 12 point swing.Exclamation points!!! Italics too!!!
I've been arguing that the Romney bounce from the debate appears to have been about 3 points and that it appears it may be receding from that high. That still seems about right. One thing to notice about the Pew poll is that it was in the field on Oct. 4 to Oct. 7. But, as Mark Blumenthal points out, almost all of those interviews were on Oct. 4 to Oct. 6, when the Romney bounce was at its peak:
According to Pew Research, 1,046 of its 1,201 interviews were conducted on Thursday through Saturday. Just 155 interviews were conducted on Sunday. The full sample had a margin of error of 2.9 percentage points, with a 3.4 percentage point margin of error among likely voters.So, this Pew poll is likely overstating Romney's actual support at this moment.
Also, it is one poll! Rasmussen has the race even (even this morning). Gallup has the race roughly even among likely voters. But beyond the polling, let me make a larger point about Andrew Sullivan's absurd reaction. If you are someone who thinks that 90 minutes of bad debating is going to cancel out 4 years of a record worth defending and months and months of bad campaigning by Romney, then you are someone who believes there isn't sufficient wisdom and virtue among the people for self-governance.
The polling (beyond the Pew Poll) doesn't suggest the race has been "thrown away." Get a grip Andrew.
Labels:
Andrew Sullivan,
bounce,
debates,
Gallup,
Mark Blumenthal,
Pew,
Rasmussen
Monday, October 8, 2012
Gallup's "Latest"
Yahoo News has a story right on their homepage (updated a little over 2 hours ago) that says
Go to Gallup's site and you'll see that their "latest" polling release has Obama up 50-45, better than the 49-46 margin he had yesterday and the day before. Gallup's tracker does not yet have a full 7 days since the debate so Obama's numbers may well deteriorate from here once we get the full post-debate set of interviews. But it is just not true to say that the story above is "the latest."
Rasmussen's tracker this morning shows Obama and Romney tied 48-48, better than Romney's 2-point lead yesterday. I noted yesterday that Romney's post-debate surge seems to have crested. Both these polls are consistent with that view.
Mark Blumenthal is up with an article this afternoon explaining the Gallup mess for laypeople. Read that if what I've said above is unclear. But importantly, he points out that Romney's bounce may well be fading at this point.
In any case, here's where we are. Obama's approval rating for Gallup AND for Rasmussen is now 51%. Those are good numbers for Barack Obama if they're real.
Now, those are tracking polls and I'll be the first to criticize those two polling organizations for being ... well, not good. But, until we see some actual evidence (good polls) showing Romney with some massive surge, I wish news organizations would stop saying it's happening.
The latest Gallup tracking poll conducted in the three days immediately following last week's first presidential debate between Mitt Romney and Barack Obama shows the race tied - with each candidate currently garnering 47 percent support among registered voters. Tracking conducted in the two days leading up to the debate showed President Obama with a five point lead - 50 percent to Mitt Romney's 45 percent.This would be breaking news if it were actually now ... ya know, breaking. It's actually almost 2 days old.
Go to Gallup's site and you'll see that their "latest" polling release has Obama up 50-45, better than the 49-46 margin he had yesterday and the day before. Gallup's tracker does not yet have a full 7 days since the debate so Obama's numbers may well deteriorate from here once we get the full post-debate set of interviews. But it is just not true to say that the story above is "the latest."
Rasmussen's tracker this morning shows Obama and Romney tied 48-48, better than Romney's 2-point lead yesterday. I noted yesterday that Romney's post-debate surge seems to have crested. Both these polls are consistent with that view.
Mark Blumenthal is up with an article this afternoon explaining the Gallup mess for laypeople. Read that if what I've said above is unclear. But importantly, he points out that Romney's bounce may well be fading at this point.
Gallup's Monday afternoon update suggests the initial bounce that Romney received from last week's presidential debate may already be starting to fade.If only Yahoo news could somehow offer the "latest" by being attached to some kind of worldwide network of computers that sends information around at the speed of light. That would be really cool.
In any case, here's where we are. Obama's approval rating for Gallup AND for Rasmussen is now 51%. Those are good numbers for Barack Obama if they're real.
Now, those are tracking polls and I'll be the first to criticize those two polling organizations for being ... well, not good. But, until we see some actual evidence (good polls) showing Romney with some massive surge, I wish news organizations would stop saying it's happening.
Wednesday, June 20, 2012
We Ask America ...
... you just made my list of pollsters that are suspect.
They have Romney ahead by 2 in Michigan in large part because Obama only leads 45-42 among women. No other demographic data was released to the general public but this was a poll of likely voters and they do indicate they weighted for demographics.
Pollsters pretty much have to weight for demographics these days given the poor and biased response rates they are getting but the devil is in HOW you weight for demographics. Mark Blumenthal made this point the other day in explaining why Gallup tends to underestimate the President's support.
I don't know what it is yet but something is wrong with their sample.
They have Romney ahead by 2 in Michigan in large part because Obama only leads 45-42 among women. No other demographic data was released to the general public but this was a poll of likely voters and they do indicate they weighted for demographics.
Pollsters pretty much have to weight for demographics these days given the poor and biased response rates they are getting but the devil is in HOW you weight for demographics. Mark Blumenthal made this point the other day in explaining why Gallup tends to underestimate the President's support.
I don't know what it is yet but something is wrong with their sample.
Labels:
Gallup,
Mark Blumenthal,
Michigan,
We Ask America
Sunday, May 6, 2012
How Can Obama Be Doing So Well?
A lot of different pieces from a lot of different angles seem to be making similar arguments and asking a similar question.
The argument: Obama shouldn't be in good shape, yet he seems to be ahead.
The question: Why is Obama ahead?
I've been pointing out for some time that Obama seems to be in better shape than the national horserace polls suggest. For instance, almost two weeks ago, I pointed out that swing state polls seem to be better for Obama than his national polls. About a week ago, I highlighted a couple of electoral models, one put together by Nate Silver and one put together by Ezra Klein, John Sides, Lynn Vavreck, and Seth Hill that each suggested Obama was a favorite to win even with some mildly pessimistic assumptions about job approval and economic performance.
But there's more.
1) Two days ago, Nate Silver wrote a sophisticated piece looking at Obama's "magic number" on jobs. Three months earlier, Silver had written an article arguing that the President needed to average about 150,000 jobs created per month in order to win reelection. Now, Silver argues Obama's "magic number" seems slightly lower, closer to 125,000 jobs created per month. In effect, Obama seems to be doing better with less-than-robust jobs numbers than Silver had anticipated.
2) A couple of days before that, in a guest post by John Sides posted on Silver's blog, Sides asked "Is Obama More Popular Than He Should Be?" Sides finds that Obama is indeed more popular than he "should be" using a model of expected job approval based on a bunch of variables including economic performance, scandals, and other significant events that affect job approval. Sides speculates that Obama is "more popular" than he should be because of personal traits (he is likable) and because more voters continue to blame George W. Bush for the weak economy than Obama.
3) Yesterday, Mark Blumenthal posted an article pointing out that Obama seems to be running stronger in the key swing states of Ohio, Florida, and Virginia today than he was at a similar juncture in 2008 when he ... ya know, won those swing states. This is true despite the fact that he is running behind his 2008 numbers in the national horserace polls.
4) Finally, a number of observers have pointed out that the Electoral College math seems to favor the President. Chris Cillizza made this argument well about a week ago. Today, Michael Cooper of the New York Times wrote about nine swing states deemed "critical" to the presidential race. If you put all 9 of these states in the "tossup" category, Obama would have 217 electoral votes that are fairly solidly his. Romney would have about 170. But, as Cooper points out, all 9 of these states are states Obama won in 2012. More importantly, polls currently show Obama with leads (some significant) in a number of these states including Pennsylvania, Wisconsin, Nevada, Virginia, and even Ohio. The inclusion of Pennsylvania strikes a particular nerve with me. Every four years, everyone talks about Pennsylvania being a "swing state." The problem with this argument is that it doesn't "swing." In 2008, Obama won Pennsylvania by 11 points while winning nationally by about 6.5 points. Pollster.com's current polling aggregator shows Obama ahead there by 7.6 points. The last Republican to win Pennsylvania was George H. W. Bush ... 24 years ago and he only beat Mike Dukakis by 2.3 points. So, if we just take Pennsylvania off the list of "swing" states and give it to Obama, he's just 33 electoral votes away from victory with a lot of different paths through the remaining 8 "swing" states to get to 270. Then, add in a couple of other states that are very competitive that Cooper doesn't mention (Arizona, North Carolina) and you see that ...
... Obama seems to be doing surprisingly well for someone with job approval numbers under 50%. Why is this?
One big reason was hit upon by John Sides in pointing out that the President seems more empathetic, etc. than Mitt Romney. While the President's job approval numbers are borderline for an incumbent, people seem to "like" the President, at least more than they "like" Mitt Romney. The President's favorability rating is 47.8%. Romney's is 37.0%. Here's an interesting mental exercise to demonstrate the importance of these numbers. Run back through the last 10 presidential elections and the two major party nominees in each one. Which seemed like the more likable personality? My answers would be Obama, Bush, Bush, Clinton, Clinton, 1988???, Reagan, Reagan, Carter, 1972??? I put question marks for 1988 and 1972 since both candidates "seem" equally unlikable to me in those elections but the winner in every other case is the more likable personality. The one who you can imagine smiling more, the one who you can bear having on your television screen for the next four years, the one you can stand. That's who tends to win. Is that Obama or Romney? I think it is Obama.
Another reason Obama is doing well is that he has paid a lot of attention to key swing states like Ohio since he was elected in 2008. Last November, the Wall Street Journal pointed out that Obama has visited swing states more than any of his predecessors did as president. Yesterday, the President kicked off his campaign officially with stops in two states - Ohio and Virginia. From this perspective, the 2008 campaign never ended for Obama in these states.
Whatever the reason, the President is currently ahead by a variety of measures. Political pundits and journalists appear baffled by this for two reasons: 1) They want it to be a close race and 2) It makes no sense to them that the President could be ahead given that the economy is as weak as it is and given that the 2010 election was as bad as it was for the President and his party.
The election will be close. At a minimum, it will be close in the national horserace numbers. But it is also the case that the President has some built-in advantages described above and I'd rather be holding his cards than Governor Romney's. My one plea to everyone is just this ... can we stop calling Pennsylvania a swing state? Please?
The argument: Obama shouldn't be in good shape, yet he seems to be ahead.
The question: Why is Obama ahead?
I've been pointing out for some time that Obama seems to be in better shape than the national horserace polls suggest. For instance, almost two weeks ago, I pointed out that swing state polls seem to be better for Obama than his national polls. About a week ago, I highlighted a couple of electoral models, one put together by Nate Silver and one put together by Ezra Klein, John Sides, Lynn Vavreck, and Seth Hill that each suggested Obama was a favorite to win even with some mildly pessimistic assumptions about job approval and economic performance.
But there's more.
1) Two days ago, Nate Silver wrote a sophisticated piece looking at Obama's "magic number" on jobs. Three months earlier, Silver had written an article arguing that the President needed to average about 150,000 jobs created per month in order to win reelection. Now, Silver argues Obama's "magic number" seems slightly lower, closer to 125,000 jobs created per month. In effect, Obama seems to be doing better with less-than-robust jobs numbers than Silver had anticipated.
2) A couple of days before that, in a guest post by John Sides posted on Silver's blog, Sides asked "Is Obama More Popular Than He Should Be?" Sides finds that Obama is indeed more popular than he "should be" using a model of expected job approval based on a bunch of variables including economic performance, scandals, and other significant events that affect job approval. Sides speculates that Obama is "more popular" than he should be because of personal traits (he is likable) and because more voters continue to blame George W. Bush for the weak economy than Obama.
3) Yesterday, Mark Blumenthal posted an article pointing out that Obama seems to be running stronger in the key swing states of Ohio, Florida, and Virginia today than he was at a similar juncture in 2008 when he ... ya know, won those swing states. This is true despite the fact that he is running behind his 2008 numbers in the national horserace polls.
4) Finally, a number of observers have pointed out that the Electoral College math seems to favor the President. Chris Cillizza made this argument well about a week ago. Today, Michael Cooper of the New York Times wrote about nine swing states deemed "critical" to the presidential race. If you put all 9 of these states in the "tossup" category, Obama would have 217 electoral votes that are fairly solidly his. Romney would have about 170. But, as Cooper points out, all 9 of these states are states Obama won in 2012. More importantly, polls currently show Obama with leads (some significant) in a number of these states including Pennsylvania, Wisconsin, Nevada, Virginia, and even Ohio. The inclusion of Pennsylvania strikes a particular nerve with me. Every four years, everyone talks about Pennsylvania being a "swing state." The problem with this argument is that it doesn't "swing." In 2008, Obama won Pennsylvania by 11 points while winning nationally by about 6.5 points. Pollster.com's current polling aggregator shows Obama ahead there by 7.6 points. The last Republican to win Pennsylvania was George H. W. Bush ... 24 years ago and he only beat Mike Dukakis by 2.3 points. So, if we just take Pennsylvania off the list of "swing" states and give it to Obama, he's just 33 electoral votes away from victory with a lot of different paths through the remaining 8 "swing" states to get to 270. Then, add in a couple of other states that are very competitive that Cooper doesn't mention (Arizona, North Carolina) and you see that ...
... Obama seems to be doing surprisingly well for someone with job approval numbers under 50%. Why is this?
One big reason was hit upon by John Sides in pointing out that the President seems more empathetic, etc. than Mitt Romney. While the President's job approval numbers are borderline for an incumbent, people seem to "like" the President, at least more than they "like" Mitt Romney. The President's favorability rating is 47.8%. Romney's is 37.0%. Here's an interesting mental exercise to demonstrate the importance of these numbers. Run back through the last 10 presidential elections and the two major party nominees in each one. Which seemed like the more likable personality? My answers would be Obama, Bush, Bush, Clinton, Clinton, 1988???, Reagan, Reagan, Carter, 1972??? I put question marks for 1988 and 1972 since both candidates "seem" equally unlikable to me in those elections but the winner in every other case is the more likable personality. The one who you can imagine smiling more, the one who you can bear having on your television screen for the next four years, the one you can stand. That's who tends to win. Is that Obama or Romney? I think it is Obama.
Another reason Obama is doing well is that he has paid a lot of attention to key swing states like Ohio since he was elected in 2008. Last November, the Wall Street Journal pointed out that Obama has visited swing states more than any of his predecessors did as president. Yesterday, the President kicked off his campaign officially with stops in two states - Ohio and Virginia. From this perspective, the 2008 campaign never ended for Obama in these states.
Whatever the reason, the President is currently ahead by a variety of measures. Political pundits and journalists appear baffled by this for two reasons: 1) They want it to be a close race and 2) It makes no sense to them that the President could be ahead given that the economy is as weak as it is and given that the 2010 election was as bad as it was for the President and his party.
The election will be close. At a minimum, it will be close in the national horserace numbers. But it is also the case that the President has some built-in advantages described above and I'd rather be holding his cards than Governor Romney's. My one plea to everyone is just this ... can we stop calling Pennsylvania a swing state? Please?
Subscribe to:
Posts (Atom)