Florida: the most polled state in the 2012 cycle
If you are a reader of our Daily Kos Elections Polling Wrap, you already know that
a common theme recently has been a marked divide between national polling on the presidential contest between Barack Obama and Mitt Romney, and the state-by-state analysis of the race.
Today we look at the state polling in a slightly different way, in order to see how Mitt Romney can assemble a coalition to get to 270 electoral votes. This week, I counted up every state-by-state poll that has crossed my eyes during the entirety of the 2011-2012 campaign cycle. For those scoring at home, that adds up to 487 polls in all except seven of the 51 locales where the race for the White House will play out.
(Trivia question for the true junkies, with the answer below the fold: Name as many of the seven locales that have not seen a single presidential poll to-date. The fact I used the word "locales" instead of "states" should give you a hint!)
Then, a very simple metric was used: In what percentage of the polls conducted in a state did Barack Obama have a lead? In what percentage of the polls conducted in a state did Mitt Romney have a lead? How often did they tie? Those states where Barack Obama led a plurality of the polls went into his column, and the same for Romney. Furthermore, to get a feel for the "true base" for each candidate, a separate calculation was made of the states where one candidate led in every single poll conducted in a state.
The bottom line: No matter how the analysis is conducted, it still remains easier for Barack Obama to cobble together that collection of states to make it to the magic number.
For the details, head below the fold to check out the analysis ... and the answer to our trivia question!
(Trivia Answer: The 7 states with no polling to date are—Alaska, Delaware, Idaho, Maryland, Rhode Island, Washington D.C. and Wyoming)
Before we begin, a couple of caveats.
First, the polls analyzed were taken from a database I created culled from our Polling Wraps and the old Weekend Digests that existed for about 15 months. This includes a ton of polls, but I freely concede that I may not have every single poll ever conducted on the Obama-Romney race. I feel pretty confident that I have damned near 100 percent of them, but there is no reasonable way to guarantee that.
Second, utilizing this method is an occasionally imperfect indicator, because 18 months is a pretty big window of time, and feelings do change. By this metric, Barack Obama looks very comfortable in Iowa, because he hammered Romney there in 2011. Recent polling, by and large, paints a different story there. Conversely, by this metric, Mitt Romney looks way better in New Hampshire than current polling would seem to indicate. That those are the two states that have shifted the most is uniquely puzzling, since it could be fairly argued that Mitt Romney has spent a disproportionate share of his time this cycle in those traditional "first two" states. Perhaps he wore out his welcome in one, but not the other? Who knows?
With those two caveats out of the way, let's go ahead and check out the numbers:
THE "TRUE BASE" FOR PRESIDENT OBAMA AND MITT ROMNEY
Let's start with the bases. As those who follow presidential electoral politics already likely knew, the majority of states in this union tilt clearly to one party or another. Indeed, in addition to the seven states listed above (all of whom have clear partisan preferences), there were another 28 states where all of the polls conducted to date went unanimously to one candidate.
And, right away, we see part of the dilemma for Mittens. He actually runs the table in more states than Barack Obama (18-17). But when those electoral votes are tallied up, as you can see, it is not even close.
ROMNEY "BASE STATES" (134 Electoral Votes): ALABAMA, ALASKA, ARKANSAS, IDAHO, INDIANA, KANSAS, KENTUCKY, LOUISIANA, MISSISSIPPI, MONTANA, NEBRASKA, NORTH DAKOTA, OKLAHOMA, SOUTH DAKOTA, TEXAS, UTAH, WEST VIRGINIA, WYOMING
OBAMA "BASE STATES" (201 Electoral Votes: CALIFORNIA, CONNECTICUT, DELAWARE, HAWAII, ILLINOIS, MAINE, MARYLAND, MASSACHUSETTS, MINNESOTA, NEW JERSEY, NEW MEXICO, NEW YORK, OREGON, RHODE ISLAND, VERMONT, WASHINGTON, WASHINGTON DC
Looking at that list of 35 states, only the most ardent cheerleaders would suggest that their candidate has a better-than-average shot at a state in the other guys column. Democrats might quibble that Obama could snare Indiana, while Republicans could argue that New Mexico and Oregon are within the realm of possibility. But, with less than four months until Election Day, let's face it—the outcome in these three dozen states are essentially baked in.
This doesn't tell us much definitive about the election, though. All this tells us, in essence, is that the outcome will be somewhere between a 404-134 electoral college landslide for Barack Obama and a 337-201 electoral college victory for the challenger. The margin of error, so to speak, is pretty damned high.
But if you look at the sixteen contests where both candidates have claimed a polling lead at least once, you start to see a very real institutional edge for the president in this election.
What it boils down to is this: If Barack Obama can maintain the polling leads he has already enjoyed, he wins reelection rather easily. He can even absorb some slippage, actually. To lose reelection, his performance with the electorate has to regress to the point that he loses states where Mitt Romney has only led in 20 percent of the polls conducted to-date.
Case in point: If you change the calculus to label a state as "Safe or Likely" if one candidate has led in at least 80 percent of the polls in that state, the election comes very close to being decided.
By that metric, Barack Obama locks down an additional four states: Pennsylvania, Ohio, Colorado and Wisconsin. That gets him to 258 electoral votes—just 12 shy of reelection.
By dropping the "safe zone" down to a Romney lead in 80 percent of the polls to date, Mitt Romney only picks up two more states (Tennessee and South Carolina). That nets him 154 electoral votes.
Examining the 16 "contested" states, you can see the challenges facing Mitt Romney:
Contested States, with percentage of polls in which each candidate has led, with number of polls conducted in parentheses (percentages do not add up to 100 because some polls were tied):
Wisconsin (27): Obama 89 Romney 8
Colorado (12): Obama 83 Romney 0
Ohio (33): Obama 82 Romney 15
Pennsylvania (31): Obama 80 Romney 10
Michigan (20): Obama 70 Romney 20
Nevada (10): Obama 70 Romney 20
Virginia (33): Obama 64 Romney 33
Iowa (13): Obama 61 Romney 31
North Carolina (30): Obama 57 Romney 37
Florida (45): Obama 49 Romney 44
New Hampshire (18): Romney 56 Obama 39
South Carolina (5): Romney 80 Obama 20
Tennessee (6): Romney 83 Obama 17
Arizona (12): Romney 75 Obama 17
Missouri (8): Romney 75 Obama 13
Georgia (9): Romney 78 Obama 11
Does this mean that Democrats should breathe a sigh of relief, content in the knowledge that all is well? Hell. No.
Some of those polling leads, as you would expect, were extraordinarily narrow. It would not take a very strong tailwind at Mitt Romney's back to turn some of those states where Obama has had a consistent 2-5 point lead into state that the Republicans could carry in November. That would certainly be true of Florida and Iowa, but could also be true of some of those states where Romney has scarcely enjoyed a lead.
However, it would not be inaccurate to say, if the current state of the race holds through November, the path to 270 electoral votes for Mitt Romney is quite a bit harder to assemble than the one for Barack Obama. And this generic edge has been accomplished, furthermore, with the president languishing with some very middling job approval numbers. If he gets even a slight boost in his numbers, he suddenly looks pretty tough to beat, given the state of the states, which do look better than the national numbers which the national news media tend to be more likely to flog mercilessly.