2012 Presidential Race Is Its Own 'Perfect Storm'

Eric Gay, David Goldman/AP Photo

As it enters its frenetic final week the presidential race is drawing more close attention than any such contest in 28 years, a testament to a campaign that has been the closest by some measures in pre-election polls dating back even further - to 1960, or even to the early days of polling in 1936.

Support for Barack Obama and Mitt Romney has averaged 48-48 percent since September, the closest in ABC/Post polls, or Gallup polls before them, in comparable periods dating back 76 years. It's also the first contest since 1960 in which neither candidate, in this period, has exceeded 50 percent support (adjusting for third-party vote where applicable). See PDF with full results, charts and tables here.

Steady the past three days, the gap is now a single point, with 49 percent support for Romney among likely voters, 48 percent for Obama in this poll, produced for ABC by Langer Research Associates. At this stage in 2008, Obama had held majority support since Oct. 11.

The comparison of the average gap between the leading candidates was done using results from September through this point before the election in ABC/Post polls since 1984 and available Gallup polls before them dating to 1936. The data are thin in some years - just one poll from this period in 1944, and just three in 1940, 1948, 1956, 1960, 1964, 1972 and 1980.

The tight race is making it a closely followed one: Sixty-eight percent of likely voters say they're following the election "very" closely - a non-significant 2 points more than at this time four years ago, and the most in ABC/Post or Gallup polls dating to the second Reagan election.

Possible effects of Hurricane Sandy, bearing down on the Northeast, are unclear. But high attention to the election could presage high voter turnout. And marking Romney's competitiveness compared with John McCain's in 2008, close attention is up especially among conservatives and Republicans, two of his strongest support groups.

TWO ISSUES - Two particular findings also indicate the unusual dynamics of this year's election. One is taxes: Republicans customarily prevail on the issue, but Obama and Romney instead run essentially evenly in trust to handle it, 46-48 percent.

That result - possibly indicating continued suspicion of Romney's economic priorities - is useful for Obama. At the same time, it's been better for him; trust on taxes has tightened from an unusual 11-point advantage for the Democrat nearly a week ago - a change that occurred almost entirely among political independents, the most moveable voters in national elections.

The other issue is health care, usually a strong point for Democrats, but now one on which preferences again divide very closely, 49-46 percent Obama-Romney, likely reflecting doubts about elements of Obama's health care reform law.

APPROVAL and ECONOMY - These are hardly the only tight measures. In another fundamental gauge in an incumbent election, Obama's job approval rating stands at 50 percent among likely voters; it's been 48 to 50 percent since before the conventions.

George W. Bush saw 50 percent approval among likely voters in an ABC/Post poll on Oct. 20, 2004, and he was at 48 percent among all adults in a Gallup poll Oct. 31 that year. The only other incumbent to win re-election with approval this low was, possibly, Harry S. Truman, with 40 percent approval in a Gallup poll in June 1948. No result closer to that election is available.

A continued challenge for Obama is the intensity of sentiment about his work in office to date - a 10-point difference in the number who "strongly" disapprove, 40 percent, vs. strongly approve, 30 percent.

The economy has been Obama's millstone; it could be asked why he's competitive at all, given the length and depth of the downturn. Poor economic conditions usually doom a president, as with the first President Bush in 1992. One answer is that politics are comparative, and Romney's not sealed the deal in persuading voters he could do better.

Likely voters now divide 50-45 percent between Romney and Obama on who would better handle the economy, back to within the margin of sampling error after a foray by Romney beyond it, 52-43 percent, last week.

It's also a non-significant 5-point gap on another important measure, economic empathy, defined as who better understands people's economic problems - on this 49 percent pick Obama, 44 percent Romney. Last week Romney had closed to within 2 points on empathy.

VOTING and GROUPS - Another key dynamic in the age of early voting is who casts their ballot, and when. Twelve percent of likely voters say in fact they've already voted, a number that's more than doubled in the past week and likely will rise very steeply this week.

Early voters report having supported Obama over Romney by 57-40 percent, almost exactly Obama's advantage among early voters against McCain in 2008. The difference this year is that Romney leads Obama by 10 points among likely voters who plan to cast their ballots on Election Day - a group in which Obama and McCain were essentially even in 2008 polling.

It's a result that suggests the outcome of this contest will be hard to anticipate; it likely will come down to Election-Day turnout. And - again unlike in 2008 - the campaigns' outreach to their potential voters is evenly matched this year, as detailed in Friday's analysis.

Finally, divisions among groups match their recent patterns - a wide lead for Romney among whites (now 17 points, 57-40 percent), especially white men (61-36 percent); and a wider lead for Obama among nonwhites (75-23 percent). Independents divide by 55-40 percent in Romney's favor, moderates by 57-39 percent in Obama's. And while Obama holds a nearly 2-1 advantage among likely voters younger than 30, the contest has tightened among those in their 30s - potentially a new battleground group, among the many in the 2012 contest.

METHODOLOGY - This ABC News/Washington Post poll was conducted by telephone Oct. 24-27, 2012, among a random national sample of 1,278 likely voters, including landline and cell-phone-only respondents. Results have a margin of sampling error of 3.5 points, including design effect. The survey was produced for ABC News by Langer Research Associates of New York, N.Y., with sampling, data collection and tabulation by Abt-SRBI of New York, N.Y.

Partisan divisions in this survey, Democrats-Republicans-independents, are 35-28-34 percent among likely voters. Partisan divisions in the 2008 exit poll were 39-32-29 percent.