Emory University political scientist Drew Linzer, who created and runs the Votamatic website, paid a visit to Middlebury College last Wednesday to discuss why, based on his forecast model, he believes President Obama is still the heavy favorite to win the presidential election. To construct his model, Drew uses his colleague Alan Abramowitz’ “Time for A Change” forecast model as his baseline starting point. You will recall from one of my earlier posts that Abramowitz’s forecast model uses three factors — the incumbent president’s net approval rating at the end of June, the change in real GDP in the second quarter of the election year and a first-term incumbency advantage — to predict the winner of the national popular vote. However, during the current election cycle Abramowitz updated his traditional model to include a “polarization” variable that, in effect, reduces the advantage enjoyed by a first-term incumbent running for reelection by about half – from a bit more than 5% to closer to 2.5%. Under his “new” model, Abramowitz projects Obama’s share of the two-party vote to be about 50.3%.
As I’ve discussed before, not everyone accepts Abramowitz’s rationale for updating his model. Drew is one of the skeptics, and so his forecast model starts with the “old” Abramowitz model which is decidedly more bullish regarding Obama’s chances. Without the polarization variable, Abramowitz’s structural baseline component has Obama winning 52.2% of the two-party national vote. That’s a much stronger starting advantage for the President than the “new” Abramowitz model suggests.
The second component in Drew’s model is the state-based polls, which he uses to “update” the Abramowitz baseline forecast. As we get closer to Election Day, state-level survey data influences his projection more and more, and Abramowitz’s structural component becomes correspondingly less important. At this point, 10 days out, the state-level polling component is really driving his forecast almost entirely.
So, where does the race stand, according to Drew’s model? As I discussed in my Economist post, as of today, he projects Obama to win 332 Electoral College votes, or 62% of the 538 Electoral College votes, compared to Romney’s 206.
Note that Drew makes several assumptions in his model. First, he makes no effort to adjust for the “house effects” of individual polls in the belief that in the closest states that are polled most frequently, polling biases will largely cancel out. Second, he essentially assumes that the “undecideds” will break in rough proportion to the distribution of the vote, as indicated by the polls, in each state. Third, since he is interested in forecasting the Electoral College vote, he pays no attention to national tracking polls.
It is doubtful that Drew, or anyone, could have constructed such a forecast model even eight years ago. But the proliferation of state polls, particularly in contested states, now allows political scientists to adjust their structural models in light of recent polling on a state-by-state basis. Of course, this type of modeling is in its infancy; Drew only has one previous election cycle, in 2008, to calibrate his assumptions.
I asked Drew what would happen if he changed his baseline starting point by, for example, substituting Doug Hibbs’ Bread and Peace forecast model, which predicts that Obama will win closer to 47% of the two-party vote – or about 5% below the Abramowitz projection. Drew acknowledged that this would shift the baseline parameter enough to move several swing states into Romney’s column. Nonetheless, based on a state-by-state electoral projection and given the current polling, even with this shift Drew does not believe that Romney would gain enough Electoral College votes to overcome Obama’s current projected advantage.
Is Drew right? Remember, for some time now I have been arguing that the state-level polls will gradually align with the national tracking polls, which as of today are showing a much closer race, with some indicating that Romney has pulled into a narrow lead. How much must Romney gain in the key swing states to overcome Obama’s polling lead? Using the RealClearPolitics state polling averages, Owen Witek created the following table showing the state of the race in the 12 closest state races as of today (Electoral College votes are in parentheses). The last column contains an estimate of the number of undecided voters in each state.
|
State |
Obama |
Romney |
Obama Margin |
Undecided |
| Michigan (16) | 48.8 | 44.8 | 4.0 | 6.4 |
| Ohio (18) | 48 | 45.7 | 2.3 | 6.3 |
| Pennsylvania (20) | 49.5 | 44.8 | 4.7 | 5.7 |
| Virginia (13) | 46.8 | 48 | -1.2 | 5.2 |
| New Hampshire (4) | 48.3 | 47.2 | 1.1 | 4.5 |
| Colorado (9) | 47.8 | 47.8 | 0 | 4.4 |
| Iowa (6) | 49 | 46.7 | 2.3 | 4.3 |
| Florida (29) | 47.1 | 48.9 | -1.8 | 4 |
| Wisconsin (10) | 49.3 | 47 | 2.3 | 3.7 |
| North Carolina (15) | 46.5 | 50.3 | -3.8 | 3.2 |
| Nevada (6) | 49.7 | 47.2 | 2.5 | 3.1 |
Source: Real Clear Politics, 10/27/12
Based on the RCP averages (and remember that Drew’s state-based polling formula includes a structural component and thus is different than the RCP simple averaging), Obama right now has 201 Electoral College votes in strong and lean states, compared to Romney’s 191. That leaves 146 electoral votes in the 11 swing states listed in the table still up for grabs. How likely is it that Romney can pick up the additional 79 needed to reach 270? If he holds his “lead” in Virginia, North Carolina and Florida, he picks up 57 more, leaving him with 22 to go. Assuming a small shift in voter sentiment, he might also squeak to victory in Colorado, earning another 9 electoral votes, leaving him 13 to go.
But here is where the math becomes difficult for Romney, and why Drew – as of today – believes Obama will hold on. Obama leads by more than 2% in all the remaining battleground states. His smallest lead is 2.3% in three states: Ohio, Iowa and Wisconsin. Looking only at these three, Romney needs to win either Ohio, or Iowa and Wisconsin, to reach the 270 mark. But, as Witek shows, there are not very many undecideds left in either Iowa or Wisconsin, so they would have to break strongly in Romney’s favor for him to eke out a victory in both states. That means Ohio may still be Romney’s best path to victory, and that assumes Obama loses the other states – no sure thing. Remember, Romney got perhaps a 2.7% boost from the first widely-watched debate that was generally viewed as a convincing win for him. How likely is it that he will be able to almost match that total in the last 10 days among the much small number of undecideds in the absence of a similar focusing event?
Keep in mind that all these calculations are based on polls that are, by nature, very squishy, so we ought not treat the RCP averages as having more precision than they do. In this respect, Drew’s model, which uses a different algorithm to analyze the polls and predict the final outcome in each state, presents a slightly different picture in the battleground states. Here are Drew’ state-based calculations, complete with the 95% confidence interval. (Note the cool color coding!). As you can see by the vertical line in the middle demarcating the 50% threshold, he projects that Obama will do slightly better in key states than a simple reading of the RCP polling averages might suggest. In contrast to my simple RCP average, he has Florida, Virginia and Colorado all leaning toward Obama (although with a confidence interval that suggests Romney might be leading in all three states). That means Romney has that much of a bigger hill to climb among the undecideds.
Of course, if Romney does begin to close the gap in the swing states, Drew’s model will pick that up, and it will adjust the forecast accordingly. Regardless of the outcome, however, there is one overriding reason why you should be visiting his site for the next 10 days: in contrast to other forecasters (you know who I mean), Drew’s methodology is completely transparent. Anyone can see and utilize his modeling equation, which he describes at length in this journal article (which has been peer-reviewed and accepted for publication). Moreover, as an academic, Drew’s rooting interest in this race is to see whether his forecasting tool is validated. He has much less at stake if the model turns out to be wrong. If it is wrong, however, he can tell us why. I call that progress.
P.S If you are coming to this site for the first time (there’s been a lot of traffic of late) I encourage you to follow me on Twitter at @MattDickinson44 – I tweet all new posts there.
Mitt..uh I mean Matt,
Abstracting away from this particular race, if I told you that the GOP was playing strong defense in NC, VA, and let’s assume FL–tied in CO–and losing IA, OH, NH, NV and all of the Gore/Kerry states, then you would predict a victory for the Dems right? I mean c’mon, VA? IA? For the sake of simplicity, lets assume Obama is only as good a candidate as Gore or Kerry, does not expand the electorate beyond likely voters, and does not understand community organizing and GOTV in a way that evokes the ineptitude of the Kerry campaign (all hard test cases in my view). You still think the GOP has a good chance of winning? I know you will say national polls matter most here, but I can’t see how Romney overcomes the basic distribution of swing states on the map unless they have a tidal wave. Tell me what I am missing here.
I’m sorry I can’t pose these questions to you on Monday due to a scheduling conflict.
Orion – Unless I misread you, I think you just repeated my post! (Or at least my interpretation of Linzer’s post!) You tell me what I am missing here! Reread the paragraph I wrote beginning with this sentence: “But here is where the math becomes difficult for Romney, and why Drew – as of today – believes Obama will hold on.”
It is interesting that this model puts South Carolina and Mississippi in play, in a statistical sense (presumably if black voting is very high and while voting is unusually low). By this model, South Carolina is about as likely to go for Obama as Virginia is to go for Romney. That doesn’t really match my statistical intuition, and I would love to know where that comes from.
BTW, You could add the running totals on either side – say left for Obama (0 at the top) and right for Romney (0 at the bottom). That way, someone could just see at a glance what it would mean if (say) Obama won Vermont through Colorado and nothing below that.
Marshall – Good catch. We asked Drew exactly that question – how could the model possibly believe Obama might win South Carolina, or Mississippi? The answer is that because there are so few polls in either state, the “model” doesn’t have very much data to go on. Therefore it has a very wide confidence interval. I think you and I – and Drew! – realize that Obama is not going to win either state. But he’s letting the model “talk”, rather than imposing his own views to determine what it should say. It may be that in a future iteration of the model, he might include a variable that says “How has this state voted in the past?”
I’ll see if I can present Drew’s chart in a more manageable format.
Matt:
Did you look at Dan McGlaughlin’s analysis where he says Obama is toast?
If so, where is he wrong?
Sheldon – Haven’t seen it, but will check now.
I sent it to you this morning. I went through it and cannot find any areas citing facts to which I disagree. I still believe it is Mitt who will ride the wave of a major hell storm to the presidency.
Sheldon – Haven’t seen it, but I’m sure I can find the link. I’ll respond after I’ve read it.
How does Linzer’s “structural model” account for a nearly 8 point swing that occurred after debate #1?
My “American Idol” model can account for it, but I fail to see how a model that is based on “low-frequency” information can predict the outcome of an electorate that goes through wild and bizarre mood swings from week to week.
Matt: Sheldon is talking about a post at RedState here. It’s not a serious analysis, it’s based on handwaving and intuition because a lot of polls show self-described independents favoring Romney. This, of course, after we had to endure all the conservative complaints about these same polls being “skewed” because they supposedly oversampled self-described Democrats. Given that the overall race is not much different from 2008 despite there being about 8-10% more Democrats in polls and 8-10% more independents saying they will vote for the Republican, it looks a lot more likely that more people are identifying as Democrats. Just another case of conservative preference for truthiness over truth.
Matt,
I am a political junkie and have studied many of these forecasts in great detail. No matter how the polling numbers are minced and mashed, I find the conclusions unconvincing as realities unique to this cycle are not considered. Maybe they can’t be quantified into percentages but I would like your response none the less. These are the realities that are not factored.
1. I have never seen a party so disrespectful and hateful toward this president, yet his policies have been fairly moderate. I believe there is a racial component to this election disguised as ideology.
2. The effects of citizens united give these rich middle aged white men the power to affect the racially biased part of the electorate. If Obama were a white man I doubt that Romney would have much of a chance. He has lied and shifted his positions many times. .
3. The republican efforts to suppress voter turnout.
4. The rich and powerful are directly or indirectly vested in electronic polling machines that 30% of the voters will be using to cast their votes. The voting is administered by the republicans in most of the swing states. The hate for this president is so strong. I just don’t trust the numbers. You could be right and Obama could still lose.
Can or should you consider the unique influence (this cycle) of racial hate (man of color) , money (citizens united) or fraud(voting suppression or counting) in your forecast?
All that silly Red State article is really saying is that if Rasmussen and Gallup are right, then yes, Obama is toast. There is no need to pull those two polls apart and compare their composition to previous elections to know that. That analysis is pretty much the opposite of polling aggregation.
I just now looked at McGlaughlin’s article. He has based his models solely on Rasmussen.
Polling analysis using only one pollster seems poor modeling at best.
Also, I wonder at the last time RedState was anything but a near disaster at anything. Bomb-throwers make poor pundits.
Hi Professer
Thanks for your thoughts on the Votamatic model. What do you think of Princeton Election Consortium, model that’s also being mentioned in addition to the 538 and Votamatic models?
http://election.princeton.edu/
Also will your forecast be posted online as well?
Thanks,
Ariel
Matt – My general rule of thumb is not to pin my hopes or fears on any single poll, but instead to rely on a composite of the polling data. By the same token,I do not throw polls out either, unless they are so obviously defective (due to poor wording, or bad sampling, etc.) What these polls show is that turnout among key demographic groups is going to be the difference in this race.
Ariel,
Sam Wang had the most accurate Electoral College state-by-state forecast of anyone I know in 2008. I haven’t been following Sam as closely as I’d like during this cycle but assuming his methodology is the same, I think he’ll get close to the final tally. What Sam does, in the end, is much like what Drew does – he takes an average of state polling. It’s about the most accurate method out there. Of course, “close” might not be close enough in terms of getting the winner right. That’s what is so tricky about this election cycle – in a tight race, even the best forecasting models may not be precise enough to pick the winner.
Thank you for the response!
Rasmussen is hardly a bomb thrower and actually has been the most accurate on a regular basis. Besides, Matt, I think you’d be the first to tell your students not to throw away the message because of the messenger.
If Rasmussen (and Gallup, even more so) are correct, or even mostly correct, this race still appears to be running somewhat like the Reagan-Carter race. Obama has stepped up his attacks and is no longer the loveable, charming, “cool” president. He is damn scared and running like it.
I think the motivated voter indexes we have been seeing will be decisive. We don’t have too long to wait, do we? Even shorter to hear your predication.
Sheldon – No, my general rule of thumb is not throw any data point away unless it is obviously tainted. But neither do I advocate relying on a single poll to the exclusion of others. Gallup and Rasmussen may be closer to “truth” – but we can’t be sure. For now, I’m relying on the composite polls.
Read 538 in the NYT. Rasmussen and Gallup have histories that must be taken into account when relying on them. Also, the Electoral College means that state polls likely are a better measure of where an election can be best viewed.
PEC uses medians, not averages, fwiw.
I do look forward to the dashing to the ground of this right-wing Mitt/Reagan, Obama/Carter fantasy.
Henry – At this point state polls are an increasingly accurate measure of the likely vote, but history suggests that we shouldn’t see a discrepancy between the national popular vote and the Electoral college vote winner. So I don’t dismiss national polls. Because we can’t see what the 538 model is, I prefer to rely on other forecasts that are more transparent. I’d recommend Drew Linzer’s Votamatic site as a start. Simon Jackman at Pollster.com is also good.
Xian – Yes, Sam uses the median of state polls – thanks for correcting me.
“Garbage in garbage out”. The conclusions are only as good as the data.
One of the problems I have in relying on composites is that except for Rasmussen and Gallup, they by and large massage their raw data with turnout from the previous presidential election.
Is there a person who reads this blog that really believes that the Democratic turnout for Obama in 2012 will anywhere equal the 2008 figures?
And, as Wayne Allyn Root points out, not one person in America who voted for McCain will vote for Obama, but lots of disillusioned Obama 2008 voters will vote for Romney.
Plus, the Independents are lining up for Romney; in 2008 they went heavily for Obama.
As you have pointed out, you have to know the questions asked by the pollsters and you need to know the weighting. At this point, the weighting becomes the result.
One wild card here: There is the infusion of Ralph Reed and his return from purgatory. Ralph has raised an estimated $10 to $12 Million to target 17 to 20 million Evangelistic Protestants who did not vote in 2008. These targeted voters are mostly in swing states of Ohio, Colorado, Florida, Iowa, Nevada, North Carolina. He is credited with recently pushing Bob McDonnell over the line by increasing the Evangelical turnout from 28% to 34% in Virginia. This is another factor that changes the modeling based upon 2008.
My last point is the propensity to vote, that has seen Republicans consistently stronger than Democrats in 2012 whereas in 2008 it was the opposite.
I am curious to see your prediction on Monday. I hope you will explain your conclusion in detail.
I have a question concerning the so-called “independent” voters who are making up their mind at the last minute. I see several references in various blogs and articles which suggest that they are breaking for Romney. Yet, I never see any polling data or any other proof that this is actually the truth. Can you speak to this variable that the Republicans are bragging about as the reason they think Romney will win?
Secondly, I am a FL voter. I tried to vote yesterday when early voting began, but here in Lee County, the early voting polling place was swamped with huge lines of voters trying to vote early. I stood in line for a while and then when I learned that folks in the next queue over from me had been in line for FOUR HOURS! and still had to queue through two more iterations of the the line, I decided to give up and try again on a week day (I’m retired and can afford to go during the work week). Question is this: will the long lines due to shorter early voter periods cause even longer lines on election day, thus causing many would-be voters like me (who, with a disability, cannot stand on my feet for four or more hours) to give up and not vote? Will this therefore benefit Romney or Obama? One could see how the hurricane effect and long lines might benefit Obama, since many of his voters have already voted early and would not be affected by the long lines or hurricane disruptions (e.g., power outages, closed polling places, etc). One could also see how it could help Romney as many of his voters will show up to vote no matter and not be dissuaded by inconveniences such as long lines or polling place consolidations, relocations, etc. What say you?
Marvin,
Great questions!
The important point about independents is that most of them are not, in fact, truly independent. Instead, they “lean” consistently in one partisan direction or the other. There are very few “pure” independents. It is true that, based on polls, Romney is doing slightly better than Obama among self-identified independents. So, for example, he leads in the latest several Washington Post/ABC poll among independents by some 16-20%. But as you might expect, his lead is largely due to support from Republican-leaning independents. Democratic-leaning independents favor Obama. Right now, the enthusiasm gap means more Republican-leaning independents are making it through the likely voter screens, giving Romney a polling boost. Will that translate into an actual turnout advantage? I’m not sure and, as you note, Hurricane Sandy is a confounding factor. I’m going to try to post on Sandy later today.
Tell Drew to update his forecast a little more often if he wants regular viewers!
Pat – Although Drew doesn’t comment everyday, I think he’s updating the numbers regularly during these final few days.