I would never need to blog. Seriously, some of you guys should really consider taking up the sword. It's a responsibility to keep going, but it's so easy to start, and a consistently great blog can really make a dent. Anyway, here's a great comment from poster Cynic:

The problem, I think, is that our punditry is largely innumerate. To a great extent, the stories about Obama's struggles with particular demographics are farcical. They originated in the heat of the primary campaign, when the Democratic primary electorate (DPE) showed clear cleavages along demographic lines. Young people of all races and genders voted for Obama, but older, poorer, white, Hispanic, less-educated, and female voters tended to back Hillary. Those were real divides. It was very clear, during the primary, that within the DPE, Latinos preferred Hillary.

The problem was the conclusions that the punditry drew from these clear demographic facts: that Obama had a 'problem' with Latino and female voters, two traditionally Democratic constituencies. In fact, it was simple to show that this wasn't the case. Obama's negatives remained fairly low among both groups. Among Latinos, his positives were also low - they simply didn't know him, although they knew they could trust the Clintons. Among women, his positives were reasonably high. But both groups were evincing a preference for Clinton in a two-candidate race, not passing judgment on Obama per se. There was, simply put, never much reason to doubt that in the general election both groups would trend strongly to Obama. In fact, the evidence that Obama is now carrying both groups by substantial margins implies that his support among Democrats in both groups must be astronomically high, despite his failure to carry majorities of Democrats in these groups in the primary.


The other mistake that the punditry routinely makes is conflating relative and absolute performance. So we've seen innumerable stories about how Obama has trouble connecting with white working class voters, particularly men. The media is obsessed with this narrative, as if the central responsibility of the presidency is to deal with the problems of plumbers named Joe in Ohio, and not with those of the citizenry as a whole. (Underlining the absurdity of this story-line is the fact that these stories proliferated even as Obama amassed outsized margins in most tracking polls, and white males ceased to matter as swing voters. He's going to win this election whether or not he carries working-class white males. So why are we paying attention to what they think?) But the kicker is that Obama isn't doing badly among this group for a Democrat. It's just that his popularity among most other segments of the electorate is relatively strong for a Democrat, and so his having equalled or only modestly outperformed other Democrats among working class white males seems newsworthy. It's not. This is a group inimically hostile to Obama's agenda, generally suspicious of government programs, and skeptical of calls for transformation. It's also worth noting that the other candidate is, well, a white male. As I noted above, white women of a certain age overwhelmingly backed Hillary. That wasn't surprising. She spoke their language. They identified with her concerns. So when I see that white men are backing McCain 51-36%, or that his support among white males soars with age, I'm hardly surprised. Nor do I find therein evidence of a problem for Obama.

A final thought. The real story of division in this election isn't about race. In general, the racial breakdown is fairly close to what we've seen time after time, and has little to do with these particular candidates. No, what's interesting here is the generational struggle. The much-touted gender-gap has varied between 4-11% since 1980, in favor of Democrats. This year, tracking polls have placed it at or above the high end of that range (Diageo/Hotline, +11; R2K, +10, GWU/Battleground, +10). But that effect is less dramatic than age. The split between voters under 35 and those over 65 is the most dramatic departure in these year's polling from the results of previous cycles, and in every poll is a bigger divide than gender.

Why do I mention this? The Hispanic population is, on average, much younger (36 vs. 27). Pew's research found that 25% of Hispanic registered voters are under the age of 30, while just 24% are over the age of 55. And that accounts for at least a part of Obama's extraordinary performance among the Latino community this year - his support among voters under 30 is 73-16%, but drops among voters over 50 to a still-impressive 59-25%. This split is a stunning development. In 2000, voters under 30 were actually marginally more likely to support George W. Bush, exit polls found. By 2004, a gap had indeed emerged, but mostly because voters under 30 held roughly constant (around 30% for the GOP) while older voters flocked to the GOP in unprecedented numbers, giving it a 20% swing in just four years. So the switch among older voters back to Obama is the more dramatic shift, but it also represents a return to established patterns prior to 2004. The shift among younger voters, however, is titanic and unprecedented - they've gone from being consistently 2-1 in favor of Democrats to almost 5-1. The Republican party is losing the fastest-growing demographic subgroup in America, and losing them because Barack Obama appeals to young Hispanics, just as he appeals to young Americans of all races. That ought to be headline news; it will likely affect elections for a generation.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.