The Obama campaign made two major mistakes: failing to tout the president's successes in the early stage of the campaign when Romney wasn't advertising, instead focusing on defining Romney negatively, and failing to prepare adequately for the first debate. The first failure meant the second failure left the president with a very weak message: Romney is a flake, but I'm not very convincing in making that case. It also allowed the Republicans to run riot with defining him negatively. According to some of the propaganda, he's a fundamentalist Islamist Europe-loving Marxist. If you can believe all of that simultaneously, you have a very weird head. Or you're a tea party Republican. (Did I say "or"?)
That takes me to the real significance of the campaign.
Obama won despite these very significant mistakes in a year when the economy has been doing poorly. Had he defined himself positively, taking credit for health insurance reform (emphasizing it's essentially the same model as Romney's in Massachusetts), stopping the Bush-era slide, saving Detroit and putting an end to expensive and damaging foreign wars, he should have won easily. So why did Romney still contrive to lose? Part of if was an astonishing succession of mistakes, including:
- having aligned himself with the tea party in the primaries, he compounded the mistake by choosing a running mate notable only for being the poster child of that side of politics (without e.g. the sort of state-wide profile that could deliver his home state... did Romney really think he needed to pander so comprehensively to one niche of voters who would never vote Obama?)
- letting Clint Eastwood do an unscripted monologue at his nominating Convention, shifting the news coverage from the candidates
- being caught on camera dissing 47% of the electorate (incorrectly claiming that non-taxpayers are all Obama supporters: retirees for example disproportionately vote Republican)
- taking positions to win over the tea party, then reversing himself when those turned out to be unpopular, yet failing to repudiate in no uncertain terms weird views of Republican senate candidates on rape and abortion
- an advertising campaign in Ohio that earned the ire of auto executives for inaccurately claiming that Jeep production was to be moved to China
- a failure to offer a clear economic alternative: I'm a businessman and I know best is not exactly compelling
- tying himself in a knot in the foreign policy debate over what was or wasn't said about the death of the ambassador in Libya
- a chaotic attempt at converting a rally to a hurricane relief event
- a last-minute dash to win over Pennsylvania, when the vote in Ohio, Virginia and Florida was much closer (the only rational explanation I can think of for this was an attempt at diverting Obama resources from tighter races, but it didn't work)
Aside from demographics, there are other interesting breakdowns. Let's look at one, ideology:
- liberal (25% of voters) voted Democrat overwhelmingly (86%)
- moderates (41% of voters) voted Democrat by a significant margin (56%)
- only 35% of voters identify as "conservative" and the Republicans win 82% of those
As long as tea party activists retain their enthusiasm for voting in primaries that gives them a disproportionate say in choice of candidates, the Republicans will battle to nominate electable candidates. About the best they can do is find one who is willing to say different things to different audiences. And that worked, didn't it?
What about finding another Reagan, someone who can seem friendly, fun and caring, while promoting these positions that, objectively, most voters see as completely wacko? There's little chance of that. They've been trying since and not found anyone. And anyway, it's a hard act to follow. If you've done it once, the other side sees through it and attacks you on substance.
Finally, a word on polls. Various right-wing pollsters claimed that the mainstream polls were all wrong, and created their own. The narrative was that the polls were "skewed" so they "unskewed" them. This is all very reminiscent of attacks on climate science. Real-world data is noisy, and any proper analysis of it should eliminate sources of bias, like artificial warming in urban areas, and artifacts like moving the location of a weather station. Properly done, these measures produce reliable data. The anti-science position is it's all just data massaging, and they can do it just as well to produce the result they want. With electoral polling, we've seen how well that works. Good scientific analysis requires honest unbiased processing of data. That applies as much to polling as it does to climate science.
BONUS: Here's another view of the demographic time bomb facing the Republicans.