Presidential Race Maps Writing on the Wall

Variable #3: The Bradley Factor

Posted Nov 03, 2008 at 6:55 AM by Maurice Berger

The so-called "Bradley Factor" in contests with black candidates--in which white voters tell pollsters they are undecided or voting for the African-American candidate out of embarrassment or fear of being judged as racist, only to vote for the white challenger in the privacy of the voting booth--is the greatest variable in this presidential cycle. Since no African-American has ever served as the presidential nominee of a major party, there are no national models on which to gauge and understand the Bradley factor. As of this morning, there are enough very close battleground states--at this stage containing large, even unprecedented blocs of undecided and persuadable voters--to make this effect meaningful (if it were to occur). In Ohio, where a number of polls out this morning report only a +2% lead for Obama, any sharp movement of remaining wavering or undecided voters could throw the state to McCain. Ditto other races that are exceedingly close as of this afternoon: Florida, Missouri, North Carolina, and Indiana (though Obama could lose all five states and still win). The good news for Obama is that his lead in nearly all Kerry-blue and some swing states is by sufficient margins (and over the 50% mark) to offset any potential McCain advantage vis-a-vis the Bradley effect. BUT, there are signs out there that the ghost of Bradley is lurking, exemplified by Pennsylvania Governor Ed Rendell's publicly stated fear that PA is not a done deal for Obama (or Democratic Congressman Murtha's impolitic musings on the "racism" of western Pennsylvanians). Even though Obama holds a healthy aggregate lead in PA of +7.6% (a lead that is increasing as of this morning)--requiring at least an 8% swing to reverse the Democrat's numbers--a swing of a far greater magnitude, and with a within a much more liberal voting base, took place in the New Hampshire Democratic primary this January, when Obama entered Election Day with a +8.3% lead, but lost to Hillary Clinton by +2.6%. That a number of battleground states have drawn very close within the past 48 hours may, in fact, suggest that undecided voters (who now are predominantly center-right, older, and demographically disinclined to vote for Obama) may already be breaking for the Republican. If a substantial shift were to occur towards McCain, another question arises: will Obama's enormous advantage in early voting (and new voter registration) offset any of McCain's gains in the now surprisingly large bloc of voters who now call themselves undecided or still persuadable? And has the dramatic tightening in a few key swing states in recent days made the Bradley Effect more of a factor?

What's Going On With Pennsylvania?

Posted Nov 01, 2008 at 3:36 AM by Maurice Berger

With several new polls showing the race in Pennsylvania drawing closer, the question of the day is why. Rasmussen this morning reports on the narrowing trajectory over the past month: the latest "survey of voters in the state shows Obama with 51% of the vote while McCain picks up 47%. That four-point advantage for Obama is down from a seven-point margin earlier in the week and a 13-point advantage for Obama earlier in the month." First it is important to note that the Democrat does lead, has held this advantage for more than a month, and passes the 50% threshold. The competitiveness of the race, however, may relate to the state's demographics, which tend to be evenly divided among Democrat- and Republican-leaning voters (the old joke about PA: it's New York in the big cities at either end of the state; Alabama in the middle). Kerry won the state very narrowly four years ago. The state has one of the oldest populations in the nation (+65 voters tend to favor McCain), a large and politically active bloc of gun owners, and a large contingent of conservative, working class white voters. Even among Democratic primary voters in April--who trended considerably more progressive than the statewide electorate at large--Hillary Clinton defeated Obama by nearly a +10% margin. Indeed, during the fall campaign, PA has been the one Kerry/Gore blue State that has given Obama the most trouble. PollTrack was the first polling website to note Obama's problem in the state, writing on 11 September: "With three new polls all showing the race in Pennsylvania drawing down to a statistical tie --Obama now leads by an average of just over 2%--PollTrack moves the state on Today's Map from 'Leaning Democrat' to 'Too Close to Call.' It is quite possible that the RNC and Palin are helping McCain in the more conservative middle section of the state--an area rich in small towns, Evangelical and Christian conservative voters, and gun owners." (NOTE: the previous passage is a quotation from PollTrack's post in early September: PA remains "Safe Democrat" as of 1 November and until further evaluation on Today's and Tomorrow's Map) If Pennsylvania, and its unique demographics, represents an isolated example of a narrowing race, Obama may hold enough of a structural advantage in the electoral college to win handily on Tuesday. But if PA is a harbinger of a broader national pattern--say, for example, indicating a tendency of white working class and Reagan Democrats to vote for McCain, whether they are admitting this to pollsters or not--certain statewide contests could draw closer as well, particularly Ohio and Missouri.

Is Pennsylvania Narrowing?

Posted Oct 29, 2008 at 4:32 AM by Maurice Berger

While Obama's polling average lead in Pennsylvania is over +10%, two new polls are actually reporting the race narrowing in the Keystone State. Rasmussen reports that Obama's advantage is down to +7%, 53% to 46%. Strategic Vision also reports a +7% lead, 50% to 43%. Still, the Democrat's lead is these polls suggests they "Lean Democrat," at the very least, though the state continues to be listed as "Safe Democrat" on Today's and Tomorrow's Maps. (The freshest polling from the state--Associated Press/GfK, Quinnipiac, and Franklin and Marshall--this morning also suggests that the fundamentals in the state favor Obama, he leads by +12% in the first two polls, 13% in the latter.) Yet, as the Boston Globe reports, "Obama's repeated visits here--he held rallies in Chester, outside Philadelphia, yesterday, and in Pittsburgh the night before-- suggest that his campaign is worried enough about the state, which he lost handily in the primary to Senator Hillary Clinton, to maintain a major presence this close to Election Day. One of Obama's top surrogates here, Governor Ed Rendell, said yesterday that McCain's heavy campaigning in the state, especially in southwestern counties around Pittsburgh, was whittling away Obama's lead. 'I never thought it was a 10-plus lead to begin with," Rendell said in an interview. "This is still not a given.'" The big question: are the campaigns' internal polls indicating a much closer race, as some suspect, or is the Obama campaign faking out McCain into pouring money, time, and resources into a state he cannot win? Even more dramatically: are the internal numbers in PA early harbingers of a "Bradley effect," in which white voters tell pollsters they are voting for the black candidate out of embarrassment or a sense of duty, even though they intend to vote for his white opponent in the privacy of the voting booth?