There is a lot of nervous buzz today about the national tracking polls. Gallup now gives John McCain a 2-point lead, the first time he has had an advantage of any amount since late May. Rasmussen, meanwhile, has the race converging back into a tie, after having shown Barack Obama ahead by 3 points yesterday.
This tracking polling will NOT reflect any convention bounce (or its absence). These polling firms concluded their interviews by mid-evening, before Michelle Obama's speech and before network coverage of the convention began. So if there is a response to the events of Monday night, it will show up in the field on Tuesday, which means that it will be reflected in polls released on Wednesday. Moreover, our research has concluded that there typically is not any bounce until the third day of the convention. As such, this polling tells us nothing at all about the convention so far, and it probably won't tell us a whole lot until at least Thursday or Friday.
It might tell us something about Joe Biden. I tend to agree with the conventional wisdom that there was liable to be a bit of a near-term backlash whenever Obama announced his VP choice, provided that the VP was not Hillary Clinton. The key phrase in there, however, is "near-term". If Hillary is able to rally her supporters to the Obama-Biden ticket tonight, there could still be a latent/lagged VP bounce for Obama that gets rolled up into his convention bounce.
Besides all that, we also have a number of state polls today which generally look pretty decent for Obama.
Yep, Quinnipiac released its "big three" swing state polling this morning. The results are literally identical to last month in Pennsylvania, where Obama leads by the same 49-42 margin, and essentially identical in Ohio, where Obama's lead is down from 2 points to 1 (although with undecided and Bob Barr improving). The difference is in Florida, which swung from a 2-point Obama lead to a 4-point deficit. Obama's investment in the air wars in Florida does not appear to paying immediate benefits. Still, this result is about where other polling firms had shown Florida, as Quinnipiac had been a modest outlier before. And that Obama's lead is holding relatively steady in Pennsyvalnia should reassure his supporters.
There is also polling out in North Carolina (PPP) and Texas (Rasmussen), which shows North Carolina and Texas at the same margins they have been polling at since antiquity, with McCain holding a 10-point lead among Texans and a 3-point lead among Tarheels.
EDIT: To clarify one thing: how can I imply that some of these state polls are good for Obama when they show his numbers -- at best -- holding steady? Because they beat the model's expectations. Our model had already anticipated about a 2.5-point decline for Obama since the last time these Quinnipiac polls were in the field in late July, and that decline was already priced in to our numbers. That turned out to be prescient in Florida, where his numbers slipped by 6 points. But Obama polled somewhat more strongly than anticipated in both Pennsylvania and Ohio.
Tuesday, August 26, 2008
Today's Polls, 8/26
Thursday, August 14, 2008
Today's Polls, 8/14
The presidential polls have been more or less stable for several weeks now. With Barack Obama in Hawaii and most of the country focused on the Olympics, you'd probably expect them to stay that way until someone holds a convention or names a vice president. However, three new polls released today show significant movement toward John McCain.
In Washington, SurveyUSA has John McCain trailing by 7 points. This hardly moves Washington into competitive territory, however, SurveyUSA had polled Washington no fewer than nine times since Super Tuesday, and had shown Obama ahead an average of 13.4 points, including 16 points in a survey released in mid-July.
A similar pattern manifests itself in Minnesota, where Rasmussen has Barack Obama's lead eroding to 4 points; Obama had held a 13-point lead last month. And in the critical swing state of Colorado, Rasmussen has John McCain edging into a 1-point lead; last month, it had been Obama by 3.
Colorado, Minnesota and Washington are quite similar to one another demographically. There are no overwhelmingly strong hints about what's going in from the cross-tabular results, but it appears that McCain has gained ground with independents (as is almost always the case when the polls move) and also that some Republican voters are moving from undecided to McCain.
Still, it's a little bit perplexing to see movement like this without any obvious proximate cause. As usual, we are simply going to have to wait for more data to find out whether it means anything.
p.s. Although it doesn't quite fit into our storyline above, there is also a University of Texas poll out in the Lone Star State, showing John McCain with a 43-33 lead.
Thursday, July 31, 2008
Today's Polls (PM Edition), 7/31
This is everything from today, including the Quinnipiac numbers that we discussed earlier:
This is generally a pretty good set of polling for Barack Obama. Abstract these numbers for a moment. If you had told a Democrat a year ago that, on the last day of July, their candidate would be ahead in Ohio and Florida, well ahead in Pennsylvania , way ahead in California, tied in Montana, within single digits in a couple of states that went really red in 2000 and 2004, they'd be pretty thrilled with that set of polling.
Generally speaking, the national polls in this cycle have been somewhat more favorable to Barack Obama than the state polling. Our projected popular vote margin, which is based on principally on state rather than national polling, has usually run a point or two behind the national polling averages at RCP or Pollster.com.
That pattern has somewhat reversed itself now. Our model likes these state polls for Obama, even as the national trackers have shown his lead shrinking to 2 points and 1, respectively.
The most interesting result today might be from Kentucky, where Rasmussen has Obama within 9 points once leaners are counted. Obama had trailed by 16 points in Rasmussen's June poll of Kentucky, and 25 points in May. There is no longer a big education/income gap in this election -- Obama has gained ground with lower-income, lower-education voters. That doesn't mean that he's going to win Kentucky. But something like West Virginia, where the candidates are already advertising since its markets overlap with Ohio, Pennsylvania and Virginia, deserves monitoring.
The other interesting result may is Montana, where Rasmussen now shows the race tied after having shown Obama with a 5-point lead four weeks ago. The underlying demographics of the state still probably point to a McCain victory by a few points, but so long as Obama is engaging the state and McCain is not, it has to continue to rate as a toss-up.
Friday, June 27, 2008
Today's Polls, 6/27
We seem to have gotten past some point of no return where we have a half-dozen polls or so to look at almost every day: 
A lot of these polls are good news/bad news for both candidates. In Rasmussen's poll of Mississippi, for instance, Barack Obama hasn't improved his numbers from the May edition of the poll, which had also showed him trailing by a 50-44 margin. On the other hand, that result had looked to me like an outlier before, and now can probably be taken more credibly. While Mississippi remains a longshot for Obama, keep an eye on Louisiana, which has similar demographics but much bluer party identification figures.
The two polls of Texas show the race tightening, but probably not enough to make the race interesting. There's a good rundown here of the pros and cons of Obama investing resources in Texas. The only thing I'd add is that there is far more room at the margins for the Democrats to make up ground with registration among Latino voters than among African-Americans. If Tejanos vote in anything resembling the same proportion that they constitute of Texas's citizenry, the state could be quite competitive.
But it's the SurveyUSA result in Ohio that I want to focus on. Obama leads by 2 here, but had been ahead by 9 in SurveyUSA's may poll of the state. That previous poll had shown a heavily Democratic sample -- 52 percent Democrat, 28 percent Republican, 18 percent independent -- and had triggered a lot of discussion about whether pollsters should be weighting their results by party ID. SurveyUSA does not do so -- although if it had applied the May distribution of party IDs to this poll, it would have shown Obama ahead by 10-11 points rather than by 2. Conversely, if SurveyUSA had applied the June party ID distribution to its May poll, that poll would have shown a dead heat rather than Obama ahead by 9.
I do not mean to be a fair-weather fan on the idea of weighting by party ID. As I implied the other day, I suspect that pollsters are facing something of a trade-off between volatility and potentially introducing bias. Weighting by party ID will almost certainly reduce noise, and perhaps make it easier to perceive trendlines -- but if the pollster's guesses about party ID are wrong, they may be reducing the turbulence but landing at the wrong airport. I'll say this: if a pollster doesn't know what it's doing, I think it should be letting the numbers speak for themselves. On the other hand, if the pollster has a robust and thoughtful method for weighting by party ID, it might be worth the trade-off. It is interesting that, taking two of our three highest-rated pollsters, Scott Rasmussen is a firm believer that you ought to weight by party ID, and Jay Leve at SurveyUSA is a firm believer that you ought not.
Finally, there have been some further refinements to the simulation model based on everyone's feedback, which I'll get around to explaining in a bit.
Monday, June 9, 2008
Today's Polls, 6/9
Three new polls out this afternoon from Rasmussen. In Wisconsin, Barack Obama holds a 45-43 lead over John McCain. Rasmussen's March and May polls had shown McCain with a lead in Wisconsin, so this brings them into line with other polling in the state, all of which shows a small lead for Obama. The McCain team may need to decide relatively early on whether they want to make a serious play in Wisconsin, or concentrate more exclusively on Michigan and Ohio, each of which look somewhat more favorable to them. Both candidates get pretty good favorability scores in Wisconsin, but Obama's support is firmer, so McCain may be playing for a smaller-than-usual number of swing voters.
In New Jersey, Obama leads by 9 points. New Jersey has not been polled as much as it probably should be, but this is another case of the polls coalescing a little bit, as Rasmussen's prior polling in New Jersey had shown the state to be a toss-up. If there's bad news here, it's for Frank Lautenberg, since the poll seems to indicate that many Dems are thinking about splitting their ticket.
Lastly, in Texas, McCain holds a 52-39 lead. This is a "pre-bounce" poll, as the field work was conducted last Monday. Nevertheless, this is yet another case of Rasmussen's numbers gravitating back toward what other pollsters like Research 2000 and Baselice have found in the state. Obama may make some pretense of competing in Texas, but it will be pretense only.
Thursday, May 29, 2008
Today's Polls, 5/29
A strange mix of polling data today:
The most significant result is probably the EPIC-MRA poll in Michigan, which shows the same four-point margin for McCain that SurveyUSA showed yesterday. Also like the SurveyUSA poll, this one had a conspicuously high number of undecideds. My sense is that this probably has something to do with Obama not having campaigned in Michigan during the primary cycle and that the state will probably lean his way in the long run. At the same time, Michigan is a state that has a significant amount of affection for John McCain, and his fundraising has been strong there.
I'm quite honestly at a loss as to how to explain the couple of Rasmussen results in Alabama and Mississippi. Demographically, the states are nearly identical. The Obama campaign has made some overtures about wanting to compete in Mississippi specifically, and it rarely hurts a candidate to call out the importance of a particular state. There might also be some lingering bitterness among Mississippians directed at the Republican Party over Trent Lott. But in the long-run, I don't see how you're going to get a 6-point margin in one state and a 28-point margin in the other.
Finally, for those of you wondering what in the hell that Texas poll is, that survey (from Baselice & Associates, Inc.) can be found here.
Thursday, May 8, 2008
Today's Polls, 5/8
Rasmussen has new surveys out in Missouri and Georgia, two states that have been somewhat underpolled.
In Missouri, John McCain leads Barack Obama by 6 points and Hillary Clinton by 2. These numbers are actually an improvement for both Democrats from Rasmussen's previous poll of the state. Nevertheless, Missouri certainly looks to us like a stronger state for Barack Obama than for Hillary Clinton Hillary Clinton than for Barack Obama, as there are enough of those Appalachian voters in the Southeast portion of the state to take a couple of points off Barack Obama's margin. What's interesting is that you won't hear the Clinton campaign talk much about Missouri, because it somewhat defies the connection between performance in the general election and in the primaries. That is, by trumpeting their superior polling in Missouri, the Clinton campaign would open up the door for Obama to point out that he's polled stronger in states like Michigan, Nevada and New Mexico, in spite of losing (or not competing in) those primaries.
In Georgia, John McCain leads Obama by 14 points and Clinton by 11. Our regression model insists that Barack Obama should be able to do somewhat better than this in Georgia. It's a very young state, and it has higher education levels and more of a white professional class than other states in its region. But so far, those results have not shown up in the polling.
Also, in Texas, a Research 2000 poll conducted on behalf of Daily Kos has John McCain leading Barack Obama by 13 and Hillary Clinton by 15. Overall, it's not a particularly good polling day for the Democrats in the Southern states.
Finally, I'd like to thank Mark Blumenthal of the National Journal and Pollster.com and Meteor Blades at Daily Kos for their kind words today.
Monday, May 5, 2008
Today's Polls, 5/5
In Texas, Rasmussen offers a little bit of a tease to the Democrats. Barack Obama trails by 5 points against John McCain, while Hillary Clinton trails by 6.
Texas now appears on the Swing State List of both Democrats, although our swing state analysis does not account for the amount of resources required -- and in politics, like so much else in Texas, if you're going to invest in Texas, you had better be prepared to do it big.
My hunch in that an investment in Texas would be a bit quixotic for either Democrat -- although for Obama I think you can make a case that he's just as likely to win Texas as Florida, which is more Democratic but also much older.
Or, you could try and appeal to Texans through your VP selection: Bill Richardson would make an interesting choice.
Thursday, April 3, 2008
Polls Now Weighted by Sample Size
Let me show you a catchy new graphic and then I'll explain what's going on.
You'll see that, in addition to listing the date of the poll, we also list the sample size. And not only do we list the sample size, but we also weight the poll in part based on its sample size.
This is something I'd thought about doing all along, but a couple of the commenters had been prodding me on this issue in light of some recent polling that had particularly large sample sizes. The recent PPIC poll in California, for instance, had more than 1,000 respondents, and Quinnipiac's sample sizes have been enormous -- they surveyed 3,484 people in their latest Pennslvania poll!
Now, I don't quite assume a linear relationship between sample size and a poll's reliability. For one thing, different pollsters have different habits about how many respondents they include -- all of Rasmussen's trial heat polls have included exactly 500 people, for example, while Strategic Vision favors sample sizes of either 800 or 1,200. And so, the information about the sample sizes may already be embedded to a certain extent in the pollster's reliability rating, meaning that we might wind up double-counting the impact of sample size if we're not careful. For another thing, sample sizes are just one source of error, with methodological error perhaps be the more important kind. I'm convinced that Zogby or ARG could survey all 325 million American citizens and still manage to fuck things up somehow.
Clearly though, if one poll is going to survey 3,484 voters and another poll is going to survey 348, the bigger survey ought to get some kind of extra credit. So the formula I use is
(sample/600) ^ 0.5
e.g. the poll's sample size, divided by 600, and taken to its square root. The number 600 is chosen because it represents an "average" sample size; it also represents a margin of error of exactly +/- 4. And taking things to the square root has the effect of accounting for diminishing returns. The Quinnipiac poll of Pennsylvania comes out to 2.41 under this formula, for instance, whereas a typical Rasmussen poll of 500 respondents works out to 0.91.
So we now have three factors we use to determine the weight a poll is given in our model:
1. Its sample size, subject to the calculation I described above.
2. Its pollster reliability rating.
3. Its recentness rating, as described in the FAQ. Note that "old" polls from the same polling firm are punished under our recentness formula (FYI, I have validated based on data from previous election cycles that this tends to improve the accuracy of our estimates more than either throwing out "old" polls from the same agency, or not penalizing them at all).
The Quinnipiac poll, for instance, has a sample size rating of 2.41, a reliability rating of 0.95, and a recentness rating of 0.91; we multiply these three numbers together to get its overall weighting, which is 2.10. In the interests of transparency, I'm now explicitly listing the weighting I'm using for each and every poll in our data table (and I'm now calling it 'weight' rather than 'reliability', since the pollster's reliability score is just one of three factors we use in the weighting calculation).
Yes, I know this stuff gets complicated. But that's why you come here, right?
And it does turn out to make some difference. Texas, for instance, had been showing up on Obama's Swing State List in part because of a Survey USA poll that showed him trailing McCain by just 1 in that state. This is the most recent Survey USA poll, and Survey USA is a very reliable polling agency, so this poll does deserve some weight. However, this poll had a sample size of exactly 600, whereas there were a number of Texas polls, taken right around the same time, that have exceptionally large sample sizes. An IVR poll that showed Obama trailing McCain by 22 had more than 2,900 respondents; CNN conducted a poll with 1,500 respondents, and a previous Survey USA poll had 1,725. These large sample size polls now have a comparatively larger weight, and as such Obama's numbers have dropped a couple points in Texas, enough to take it off his Swing State List.
Overall, however, including the sample sizes has turned out to be helpful to both Democrats, particularly on the strength of the large-sample Quinnipiac polls in places like Ohio, Pennsylvania, and New Jersey.
148 comments
Post a Comment