Similar levels of support exist across white, black, blue-collar voters
PRINCETON, NJ -- Barack Obama's current level of support among white voters in a head-to-head matchup against John McCain is no worse than John Kerry's margin of support among whites against George W. Bush in the 2004 presidential election.
Much of the talk following Tuesday's Indiana and North Carolina primaries has focused on just how electable Obama -- now the highly probable nominee -- will be in the general election. The Clinton campaign has argued that Obama's weaknesses among white voters and blue-collar voters will hurt him against McCain in the fall.
But it appears that the way Obama stacks up against McCain at this point is similar to the way in which Kerry performed against Bush in 2004 within several key racial, educational, religious, and gender subgroups. That is, the basic underlying structure of the general-election campaign this year does not appear to be markedly different from that of the 2004 election. This conclusion is based on an analysis of exit-poll data from 2004 compared to the Obama-McCain matchup in 4,000 Gallup Poll Daily tracking interviews conducted during the first five days of May.
Kerry, the Democratic nominee in 2004, lost to the Republican Bush by a 51% to 48% margin in the popular vote. In Gallup Poll Daily tracking data from May 1-5, Obama is losing to McCain among registered voters by a 46% to 45% margin. Although there is a sizable component of undecideds in the Gallup Poll tracking data (and obviously no undecideds in the 2004 exit-poll data), the margins in these two races are quite similar, with Kerry losing by three points, and Obama by one point.
This overall comparison, in and of itself, suggests that Obama, assuming he captures the Democratic nomination, begins the general-election contest in roughly the same position in which Kerry ended his unsuccessful quest in 2004 -- that is, with the prospect of a very close race.
The more specific comparison of Obama's positioning versus Kerry's among various subgroups of the electorate suggests that the basic dynamic of the general presidential election is fairly predictable, based on past patterns. Regardless of Obama's specific characteristics, or the specific environment in which this year's election will take place, Obama will inherit a fairly structured general election environment.
Obama would be the first black candidate to receive a major party nomination in United States history. This could have racial implications in the voting. Some observers have argued that Obama has a particular liability among white voters, whom exit polls showed voted for Clinton in both North Carolina and Indiana.
The general-election exit poll in 2004 showed that Kerry lost the white vote to Bush by a 17-point, 58% to 41%, margin. At the moment, in Gallup's tracking of the general election across the first five days of May (with a sample of more than 4,000 registered voters), Obama is losing the white vote to McCain by a 53% to 37% margin, or 16 points. In short, Obama's relative performance among white voters -- at least at the moment -- is similar to Kerry's in 2004.
Democratic candidates at the presidential level have traditionally received the overwhelming majority of black votes. In 2004, Kerry won over Bush among blacks by 88% to 11%. At this point, Obama is winning over McCain by a 91% to 5% margin. So, there is little difference in how Obama fares among blacks compared to how Kerry did in 2004, in part a result of the already very high, "pre-existing" Democratic tilt of black voters as seen in previous presidential elections.
Another issue that has arisen in the course of the Democratic nomination campaign this year is the so called blue-collar vote. Clearly in the primary campaigns, Obama has done better among those with a college degree than has Clinton, and worse among those without a college degree. But a comparison of Obama's performance against McCain among educational subgroups compared to Kerry's against Bush suggests that Obama is not at an unusual deficit among those with lower levels of education.
Obama loses by a four-point, 47% to 43%, margin among those with no college. Kerry lost to Bush among those with no college by a similar six-point, 53% to 47%, margin.
Kerry and Bush tied among college graduates in 2004. In Gallup Poll Daily tracking, Obama wins by 50% to 44% among college graduates. Among those with postgraduate educations, Kerry won over Bush by 55% to 44%, or 11 points. Obama is beating McCain by 57% to 38%, or 19 points among this group. The comparative data certainly suggest that Obama actually has a relative strength among well-educated voters.
So at the moment, it would appear that Obama's problems with less well-educated voters -- emphasized heavily by the Clinton campaign -- are no worse than were Kerry's in 2004, and are more than made up for by Obama's strength among those with college degrees.
There has been a significant gender gap in presidential voting for at least three decades. As would be predicted, Obama wins among women by four points, 47% to 43%, in the Gallup Poll tracking data. This is very close to the three-point margin by which Kerry beat Bush among women in 2004. Obama loses among men by five points, while Kerry lost among men in 2004 by 11 points. In other words, the gender swing for Obama now is nine points, while Kerry's swing was a slightly larger 14 points.
One of the fundamental verities of American political life is the significant religion gap in party identification and in presidential voting. Republicans typically do much better among those with high self-reported church attendance. In 2004, Bush beat Kerry by 61% to 39% among those who attend church weekly. McCain is beating Obama by 55% to 36% among this group. Among those who never attend church, Kerry won big in 2004, 61% to 36%. Obama wins among this group, 54% to 36%. It appears that the religion gap is slightly more muted this year than it was in 2004.
Each presidential campaign takes place in a new and different environment, with a new cast of characters and issues. If, as expected, Obama wins the Democratic nomination this year, the campaign will have an even more distinctive newness to it, as Obama would represent the first black major-party candidate in the country's history.
Still, the analysis reviewed here suggests that the basic structure of an Obama-McCain campaign is in many ways quite similar to that of the 2004 race between Kerry and Bush.
At the moment, Gallup Poll Daily tracking indicates that this November's election could be close, as has been the popular vote in 2000 and 2004. In other words, just as 2004 was in many ways a replay of 2000, this year's election could be a replay of 2004 with minor changes around the edges.
Certainly the current data show that the patterns of support for Obama when he is pitted against McCain -- among various key racial, educational, religious, and gender groups -- do not look like they have changed dramatically from the 2004 contest between Kerry and Bush.
Results are based on telephone interviews with 4,536 national adults, aged 18 and older, conducted May 1-5, 2008. For results based on the total sample of national adults, one can say with 95% confidence that the maximum margin of sampling error is ±2 percentage points.
Interviews are conducted with respondents on land-line telephones (for respondents with a land-line telephone) and cellular phones (for respondents who are cell-phone only).
In addition to sampling error, question wording and practical difficulties in conducting surveys can introduce error or bias into the findings of public opinion polls.
To provide feedback or suggestions about how to improve Gallup.com, please e-mail firstname.lastname@example.org.