With both party conventions over and the candidates' bounces having largely subsided, the latest national polls utilizing live callers put President Obama ahead of Mitt Romney by somewhere between 1 percentage point (ABC News/Washington Post) and 6 percentage points (CNN). The Gallup and CBS/New York Times polls have Obama's advantage at 3 points, while Democracy Corps and Fox News put it at 5 points. Private polling on both sides pegs the Obama lead at perhaps 4 or 5 points. The live-caller differential is important because interactive voice-response, or IVR, surveys (so-called “robo polls”) are legally prohibited from calling cell phones; between 30 and 40 percent of voters are now exclusively or primarily served by cell phones. Those polls utilizing cell-phone-only voters reach a constituency disproportionately younger and more likely to be minorities, thus the distinction between IVR and polls by live callers is very important.
Interestingly, among these six polls, the narrowest (ABC/Washington Post) and the widest Obama margins were the two that were conducted soonest after the Democratic convention. Gallup at one point soon after Democrats left Charlotte had Obama up by 7 points, but that lead gradually narrowed and seems to have stabilized at 3 points, 48 to 45 percent for Oct. 10-16. It is the one remaining poll of the group still looking at registered voters; the rest have begun focusing on likely voters. At some point in the next few weeks Gallup is expected to begin reporting trial heats both for all voters and from among the likely-voter subgroup. Some on the GOP side say Romney's choice of Rep. Paul Ryan, R-Wis., as his running mate gained the ticket about 2 points, but the combined effect of the conventions negated the gain, putting the Romney campaign back in the position of being down by 4 or 5 points.
Contrary to what you might read might on the blogosphere, there is very little disagreement between top Democratic and Republican pollsters on where this race is right now, either on the national level or, more importantly, in the swing states. Keep in mind that there are multiple pollsters involved in each of the swing states: a presidential campaign on each side, one or more super PACs with their own pollsters, plus others done for senatorial and gubernatorial races (and their national party campaign committees) or for ballot initiatives. So there is a whole layer of very expensive, high-quality polling in this race, most of which never sees the light of day. But some pollsters or the strategists who commission such polls will agree to characterize the polling in a race on an off-the-record basis, giving, say, a 2- or 3-point range of what their data shows. Among the broadly defined 11 battleground states, Romney is best situated in North Carolina, where the Elon University poll put him ahead by 4 points at the tail end of the Republican convention. More recent private polling is said to be very close, but insiders on both sides expect it will ultimately end up in the Romney column. New Hampshire seems to be about even, give or take a percentage point. Virginia and Wisconsin are the states with the narrowest Obama leads (he is ahead by between 2 and 4 points), while Obama is thought to be up by between 2 and 5 points in Florida and Nevada. Next comes Iowa and Colorado; the Hawkeye State apparently has Obama ahead by between 3 and 6 points, Colorado between 4 and 6 points. Then come the big-ticket items: Obama ahead by between 5 and 8 points in Ohio, by 6 or more in Michigan, and the high single digits in Pennsylvania.
Obviously there are dozens of permutations in the calculus for a Romney path to victory, but as long as Michigan and Pennsylvania are noncompetitive and Ohio continues to look tough for Romney, he would have to come pretty close to running the table to get the 270 electoral votes needed to win. Obama winning Colorado, Iowa, and Nevada would put 276 electoral votes in the Democratic column—six more than necessary for a win—even if Romney carried North Carolina, New Hampshire, Virginia, Wisconsin, and Florida. At this point Romney is only ahead in one, tied in another, and trailing in the other three.
This election is still quite close and could go either way, but Romney badly needs something to happen to change the trajectory of this race. If things remain as they are today, he loses. Presidential debates are scheduled for Oct. 3, 16, and 22. A vice presidential debate is scheduled for Oct. 11, and unemployment numbers are released on Oct. 5 and Nov. 2. These are six events that could prove consequential. Though debates arguably have changed campaign trajectories in 1976 and 1980, in the last seven presidential elections they didn’t materially affect the outcome of the races. Obviously, candidate gaffes or campaign miscues outside the debates can matter as well as external events, domestic or foreign. Anything involving an attack by Israel and/or the United States against Iranian nuclear facilities would certainly create an “all bets are off” situation. Major incidents elsewhere in the Middle East or around the globe—for example, North Korea—could be consequential as well.
Kantar Media’s CMAG recently reported that “this campaign is very developed with few undecided voters in swing states,” pointing to levels of advertising in sample swing-state markets running anywhere from three to 12 times higher than in 2004 or 2008. In one sample week before the conventions, Aug. 15-22, Columbus, Ohio, saw 1,842 presidential ads, three times the 608 ads run in that same week in 2004. Las Vegas voters endured 2,870, three times more than the 867 in 2004, and Orlando saw 1,863 ads, 12 times the 153 ads from that week in 2008.
Adding in the non-presidential campaign ads, CMAG found that as of early September there have been 1.3 million “ad occurrences for all political advertising on local spot TV,” compared with 832,291 at the same point in 2008. CMAG estimates that there will be about 43,000 ads run every day through the election.
If you buy the theory that anyone in a swing state who is still undecided is pretty unlikely to vote, that means that this election becomes pretty much a ground game, coming down to who can get their folks out. It is absolutely impossible to gauge at this point the relative effectiveness of the Obama and Romney campaign operations and their respective party apparatus. At the suggestion of Cook Political Report House Editor David Wasserman, I counted up the state headquarters and field offices for the two campaigns, with the RNC Victory offices counted in the Romney totals. Clearly this is a crude instrument of estimating effort, but in those 11 swing states, the Obama campaign has 526 offices while the Romney campaign and RNC Victory offices total 251. In Colorado, Obama has 55 and Romney 14; it’s Obama 80 and Romney 47 in Florida. In Iowa it’s 65 for Obama and 14 for Romney. Michigan is the one state where there are more Romney than Obama offices (23 to 20). In Nevada, it’s 25 for Obama and 11 for Romney. Obama also holds the edge in New Hampshire (22 to eight), North Carolina (49 to 24), and Ohio (79 to 37). Pennsylvania has 39 Obama and 19 Romney outposts, Virginia has 40 for Obama and 30 for Romney, while in Wisconsin the advantage is for Obama is 52 to 24. This may or may not be a fair way to measure field activity, but I certainly can’t think of any other way to quantify it at this stage.
My advice is to get up every morning and look at the front page and ask yourself, “Do I see anything here that will change the trajectory of this race in a meaningful way that would help Mitt Romney?” Each evening, watch one or more of the three broadcast network evening news programs and ask the same question. If the answer is no both times, then look again the next day and on through the election. But with the saturation levels of advertising that has existed for months and the diminishing numbers of undecided voters, it is increasingly likely that it will take a significant event rather than the normal daily ebb and flow of politics to change the outcome here.