As we approach the "convention season" in the race for the presidency, it behooves us to take another look at how the electoral math currently stands. Mitt Romney chose to announce his running mate rather early, which is just beginning to be reflected in the polling. But, starting next week, each party will likely get a noticeable "convention bump" in the polls. Because the two conventions are happening right after one another, this should stir the big data pot well into September. Which is why now is a good time to look at the state of the race, to establish a baseline to measure all this expected frenzied movement.
We only have time for a short column today, so we're only going to look at two charts this time around, and then quickly wrap up with my picks. I promise next time around we'll take a deeper look at what is going on. Let's begin with Mitt Romney's chart (click on either of these charts to see a bigger copy):
[Definition of terms: Strong means 10% or better in the polls, Weak means 5% or better, and Barely is under five percent.]
The last time we took a look at the math was two weeks ago (the vertical lines in the charts show the dates of these columns). In that time period, Romney showed some positive movement, which was dampened by some weakening. How much of this was due to the selection of Paul Ryan is anyone's guess, really.
Romney gained two pickups this time around, Iowa and Florida. This put his overall total an impressive 35 Electoral Votes ("EV") higher, but Romney also saw Missouri weaken from "Weak" to "Barely" during this period. Importantly, this was before the Missouri Republican candidate for Senate caused such a ruckus a few days ago, meaning Romney's hold on Missouri was already heading in the wrong direction before the news broke.
Because Florida is so large, this all means Romney rose from a total of 206 EV to a whopping 241. That's a pretty big jump for two weeks. The critical "Strong Plus Weak" line on that chart, however, moved downwards from 191 EV to only 181 EV, while his "Strong" number held steady at 139 EV. So far, at least, Romney's choice of Ryan has not hurt him in Florida, at least yet. Florida is crucial for Romney's chances to win, so this is good news for him.
Let's take a look at Barack Obama's chart:
The news for Obama was mostly bad this time around. Again, this could be the bump from Romney naming his vice presidential pick, or it could be unrelated. The good news for Obama is that he managed to move Colorado back from being a tie into Barely, and New Hampshire improved from Barely to Weak. Virginia started off with bad news, moving to a tie, but then migrated back to Barely Obama again, for no net gain or loss. The worst news for Obama was losing Florida, but Florida can be expected to bounce back and forth during the whole election, and not really in either candidate's column as of yet. The loss of Iowa to Romney hurt Obama as well. Obama weakened his hold on two states, Wisconsin and Ohio, although they still are in the Barely column for him at this point. Wisconsin could become a much bigger fight than the Obama team was likely expecting, now that they've got a "favorite son" on the Republican ticket. It will be very interesting to watch the polling coming out of the state over the next month.
Obama's numbers fell overall, with the loss of Florida's 29 EV leading the trend. Obama started two weeks ago with a total of 332 EV, fell to 284 EV until Virginia firmed back up to leave him at the end with 297 EV overall. This was a net loss of the same 35 EV that Romney picked up. More importantly, while Obama's Strong number held steady at 172 EV, his Strong Plus Weak line went up slightly to 277 EV at the start, but then headed down to 247 -- which translated into a net loss of 20 EV.
While both candidates' Strong Plus Weak numbers fell this time around, Obama's lead over Romney in this category has fallen from a 76 EV margin down to 66 EV -- which is still a healthy lead, at this point in the race. However, Obama has to hope that the Democratic convention will turn the trendline around for him in some key states.
My Picks
This is the part where I toss darts at the wall and try to classify states on more of a gut feeling than just raw poll numbers. To do this, I use different labels to avoid confusion. At the end of this article are full lists of states for each category (with their EV numbers), as well as a list of states which haven't been polling in a long time (so you can see which polling data is getting stale).
First we look at "Likely" states for each candidate, broken down into "Safe" and "Probable" for each. Then we get to the "Tossup" states, which are divided into "Leaning" for each candidate and then a "Too Close To Call" category, before adding it all up in conclusion.
Likely States -- Obama
Safe Obama (14 states, 179 EV)
I'm moving Oregon up to Safe Obama, just because I really don't think Romney's got a chance to win the state. The polling data isn't very recent, but my gut tells me this state is pretty safely in the Obama column.
Probable Obama (7 states, 68 EV)
Oregon moves up from Probable to Safe for Obama, and New Hampshire moves in to take its place. Now, New Hampshire may wind up being closer than I now think, but as of now it seems to be getting pretty solidly behind Obama. Perhaps I'm being too optimistic, to put it another way. If New Hampshire shows signs of weakening, then I'll move it back down, but for now my gut tells me it's at least Probable Obama.
Likely States -- Romney
Safe Romney (18 states, 148 EV)
Romney's Safe states showed no change this time around. He can rest assured that these 18 states will be showing up in red on election night.
Probable Romney (4 states, 33 EV)
Romney does lose one state here downwards, as Missouri heads into a storm of uncertainty. Missouri has been, demographically, getting redder and redder over time, so it really should be an easy state for Romney to claim, but it is showing signs of weakness already -- which happened before the Todd Akin "legitimate rape" comment controversy. Now that this has become the hot-button issue of the week, Romney's standing in Missouri could be affected by the fallout (especially since Akin appears to be staying in the race, at least for now). Either way, the state can't even be seen as Probable Romney at this point.
Tossup States
Lean Obama (2 states, 27 EV)
Obama loses three states here, one of which was good news (as New Hampshire moved up), and two of which were bad news. Both Iowa and Wisconsin have to be seen as Too Close to Call at this point, at least until Obama gets some better polling in each. In the last of these columns I wrote, I moved Iowa up to Lean Obama because I thought it was firming up for Obama, but I was wrong as it moved all the way to Romney's column this time around. Wisconsin is going to need some polling to see what effect putting Ryan on the Republican ticket is going to have. Colorado and Ohio are the only states which remain Lean Obama this time around, for a total of 27 EV.
Lean Romney (2 states, 25 EV)
North Carolina stayed in the Lean Romney column, but Missouri moved down here as Romney's numbers weakened. I almost moved the state into Too Close To Call, but will give Romney the benefit of the doubt here, at least until I see polling which indicates otherwise. Even if Akin stays on the ticket through November, it may not harm Romney's chances in Missouri. We'll see. Taken together, these two states add up to 25 EV.
Too Close To Call (4 states, 58 EV)
Florida and Virginia both remain Too Close To Call, and at least Florida may stay in this category until the very end of the race. Two states moved down to Too Close To Call this week, both bad news for Obama. Iowa and Wisconsin join the battlegroundiest of the battleground states, for a grand total of 58 EV in the category.
Final Tally
Counting up all my picks without any leaning or tossup states, Barack Obama improved by 4 EV this time around, with the addition of New Hampshire, for a total of 247 EV that seem safe for him. Mitt Romney dropped 10 EV with the loss of Missouri to close the period with just 181 EV safely in the bag for him. This widened the gap between the two to a rather impressive 66 EV lead for Obama. Obama needs 23 EV out of the remaining tossup states, whereas Romney needs to win 89 EV -- which would count as "running the table," since there are only 110 EV in the category. Romney would have to win Missouri, North Carolina, Florida, Iowa, Virginia, and Wisconsin -- and then he'd still have to take away either Colorado or Ohio from Obama in order to get over the finish line. Obama, at this point, could win his safe states plus just Colorado and Ohio and still win the race.
Of course, this could all change quite drastically in the next two weeks. It's hard for political wonks to realize, at times, but most of America simply doesn't pay much attention to political news. This begins to change during convention season. More and more of the public will start paying attention, and polling numbers can often swing wildly when this does happen. So while Obama looks in pretty good shape at this point, in two weeks he could be in a very different spot indeed. We'll be able to look back, after the conventions, and see how much has changed from today's baseline.
[Electoral Vote Data:]
(State electoral votes are in parenthesis following each state's name. Washington D.C. is counted as a state)
Barack Obama Likely Easy Wins -- 21 States -- 247 Electoral Votes:
Safe States -- 14 States -- 179 Electoral Votes
California (55), Delaware (3), Hawaii (4), Illinois (20), Maine (4), Maryland (10), Massachusetts (11), New Jersey (14), New York (29), Oregon (7), Rhode Island (4), Vermont (3), Washington D.C. (3), Washington (12)
Probable States -- 7 States -- 68 Electoral Votes
Connecticut (7), Michigan (16), Minnesota (10), Nevada (6), New Hampshire (4), New Mexico (5), Pennsylvania (20)
Mitt Romney Likely Easy Wins -- 22 States -- 181 Electoral Votes:
Safe States -- 18 States -- 148 Electoral Votes
Alabama (9), Alaska (3), Arizona (11), Arkansas (6), Idaho (4), Kansas (6), Kentucky (8), Louisiana (8), Mississippi (6), Nebraska (5), North Dakota (3), Oklahoma (7), South Carolina (9), Tennessee (11), Texas (38), Utah (6), West Virginia (5), Wyoming (3)
Probable States -- 4 States -- 33 Electoral Votes
Georgia (16), Indiana (11), Montana (3), South Dakota (3)
Tossup States -- 8 States -- 110 Electoral Votes:
Tossup States Leaning Obama -- 2 States -- 27 Electoral Votes
Colorado (9), Ohio (18)
Tossup States Leaning Romney -- 2 States -- 25 Electoral Votes
Missouri (10), North Carolina (15)
Too Close To Call -- 4 States -- 58 Electoral Votes
Florida (29), Iowa (6), Virginia (13), Wisconsin (10)
Polled, but no polling data since the primaries:
(States which have not been polled since the beginning of June, with the dates of their last poll)
Maryland (5/21), Nebraska (5/16), Rhode Island (2/22), South Carolina (1/13), Tennessee (5/9), Texas (5/13), Vermont (2/22), and West Virginia (4/28).
No polling data at all, yet:
(States which have not been polled so far this year):
Alaska, Arkansas, Delaware, Hawaii, Idaho, Kansas, Kentucky, Louisiana, Mississippi, Washington D.C., Wyoming
Electoral Math Column Series Archive:
Chris Weigant blogs at: 
Follow Chris on Twitter: @ChrisWeigant
Become a fan of Chris on The Huffington Post
Follow Chris Weigant on Twitter: www.twitter.com/ChrisWeigant
Mitchell Bard: Why Akin's Crazy Claim of "Legitimate Rape" Matters
Al Eisele: A Legitimate Case for the Teddy White Rule
 |
 |
|
| Obama | Romney | |
|---|---|---|
| Electoral Votes (270 to win) |
332 | 206 |
| Obama | Romney | |
|---|---|---|
| Total | 65,899,660 | 60,932,152 |
| Percent | 51.1% | 47.2% |
| Democrats* | Republicans | |
|---|---|---|
| Current Senate | 53 | 47 |
| Seats gained or lost | +2 | -2 |
| New Total | 55 | 45 |
| Democrats | Republicans | |
|---|---|---|
| Seats won | 201 | 234 |
Two things I think you should factor into the argument here. The first being the gluttonous largess of super PAC money that is going to be dumped into the election on the right and the fact that a lot of people still aren't paying attention to the election. Come debate time, I think we shall see a greater picture to what the general population thinks of this race, but my gut feeling is that Obama should clear 270 no problem.
Reason being that even if these voter suppression laws come to pass there's not nearly enough time to implement enforcing them, they clearly are in violation of the voting rights Act, and democratic get out the vote is a well oiled machine.
Barring massive voter suppression and fraud I think Obama has this one.
So the economy is not the pivotal issue some folks claim it to be....for Obama is at least even with Romney despite many months of poor economic growth.
Other issues will then resonate....health care affordability and access, social issues agendas, etc.
That will help Obama.
Florida is slightly more to the political right than is the nation at large.
Unless Obama can establish a lead of say 5 points or more in Florida, it probably goes for Romney.
I think Ohio goes for Obama...and either Colorado or Virginia, but not both.
Obama is slightly more likely to win in November than Romney.
But I'm confused. From everything I've read this week, I assumed that Obama is running against the Ryan/Akin ticket. Who's this Romney guy?
Ah, Grasshopper, you have asked the key question: "Who IS this Romney guy?"
Heh.
-CW
Think Akin will win, or will McCaskill get another term? I'd be interested to hear what folks on the ground in Missouri are saying right about now...
-CW
If you establish a baseline now, Romney's inevitable rise will appear far more dramatic than it actually will be.
The summer polling is more often than not based on registered rather than likely voters and overcounts Democrats and undercounts the GOP by between 6% and 13%. If you apply a realistic partisan weighting (roughly thirds between Dems, Reps and Indis with maybe a Dem+2 advantage), then polls like the latest AP poll with Obama +1 under a Dem +9 weighting turn into current Romney leads.
Once realistic likely voter screens are applied after the conventions, the current Romney lead among decided voters will become clear.
The actual post convention shift will occur during the debates when the undecided shift to the challenger Romney. All this talk about statistically tied horserace polls consistently ignores the 6% to 10% undecided. When you look into the internals, you find the undecided are largely white, male, church going and GOP leaning folks who overwhelmingly disapprove of Obama. https://thecitizenpamphleteer.wordpress.com/2012/07/14/who-are-the-undecided-voters/
Actually, a lot of the pollsters are switching over to Likely Voter counts. But I take your point -- it's early days still, and a lot could happen.
-CW
http://fivethirtyeight.blogs.nytimes.com/2012/07/22/do-presidential-polls-break-toward-challengers/
Nate's analysis is inapt because it uses horserace (not approval) polling over time (whose methodologies also change over time) and applies it to situations where the incumbent president is not running for reelection.
OH 1.8 Obama (within margin of error)
VA 2.0 Obama (within margin of error)
FL Tie
IA 1.0 Obama (within margin of error)
NC 1.0 Romney (within margin of error)
CO 1.0 Obama (within margin of error)
NV 4.2 Obama (barely out of margin of error)
Missouri 6.3 Romney safe
WI 1.4 Obama (within margin of error)
MI 3.3 Obama (within margin of error)
NH 3.5 Obama (within margin of error)
That is a lot of states and electoral votes in play. Don't forget the "Bradley effect" and poll sampling that has polls with D+9 voter samples and I feel good at this point with Romney's polling.
Bradley effect is just one of many types of sampling bias. Bias is unavoidable and hard to quantify. That's one reason why it's very helpful to look for consistency in polling trends among polls known to favor Red or Blue to varying degrees. Curve fitting across different time periods is also very useful since it tends to filter out statistical noise.
One beef I have with RCP is the category called tossup. To me, tossup is a tie in the metrics, anything else is a degree of lean. Most RCP "tossups" actually favor Obama, if fairly weakly.
The best way to feel good about the current polls is to be ignorant of how they work. The more know, the more churn you'll have in your stomach, be that stomach D or R.
If all the polls have a margin of error of about 3 to 4% you can't say the average is exactly 3 to 4% margin of error, that is true. But you can not completely discount the margin of error either.
Example a poll that is 45 to 45% with 3 % margin of error could really be 48 to 42% or any grading in between. But your average of the polls would give a much different result if the numbers used were 48 to 42 rather then 45 to 45.
Statisitics dont carry over exactly with margin of error, because the margin of one poll may favor one candidate and the other poll another candidate cancelling out in the average. However, if the errors are both for the same candidates favor then the margin of error is compounded not erased.
Further sampling pool is also key in polls. What is the ratio of Democrat, independent and Republican Voters that actually will vote? If the poll questions 9% more Democrats then Republicans is that really reflective of National or more specifically state averages?
Example a recent Quinnipak poll in OH shows Obama up by 6. But the sample has 34% Dem, to 26% Republican. Yet voter registration is 37% R to 36% D in OH and that clearly must be known when analyzing results. Since most R will vote Romney that 10% difference in sampling puts OH back as toss up.
You're right that's a lot of states in play, but the difference is in how many of them each candidate will need. Romney will have to almost run the table, whereas Obama has more paths to victory.
But, hey, it's early days. We'll see what happens after the conventions, when people start paying closer attention.
-CW
All the same, I've no disputes with your polling-based analysis, I'm just still surprised about Virginia.
rbenjamin (below) is right. Also, Ohio's unemployment is better than the national average -- one result of the auto bailout. One more thing -- Ohio Republicans (some of them) are a little annoyed that Rob Portman didn't get the nod for veep.
But Ohio could change. It's definitely a key state for both candidates.
-CW
Stranger things have happened. I'm just sayin'....
-CW
My own gut reaction is that fluctuations in both the RED and BLUE EV totals are mostly random noise which can be attributed primarily to different methodologies among the various polling organizations, mixed in with some sampling error.
In other words, I don't see much evidence of significant change in the electoral landscape since May. Obama remains the favorite, but not anything close to an overwhelming favorite.
It's not entirely clear how the Ryan choice will pan out with those critical undecided voters in swing states. For the moment, Ryan has shifted the debate towards social values, rather than the economy, which seems like a bad strategy to me. The Akin abortion flap has tilted the argument even further away from the economy, and entangled Ryan in some very extreme rhetoric. The Obama camp is beginning to hit these weaknesses with some effective ads. Romney still sucks at retail politics.
Maybe things will shift a bit one way or the other during the debates, but maybe not. On election day we may still face a highly polarized electorate, with few persuadable voters in the few states up for grabs.
Boy, I have to say "admirably concise" is a compliment I've never actually received before!
I think that the conventions -- especially because they're so close together -- will tend to cancel each other out (in terms of the "bump" they traditionally provide), and that while we may see a little fluctuation, the polls will stay remarkably steady through the next three weeks. But that, too, is just a gut feeling.
-CW