For the first narrative: Democrats rejoiced when Rand Paul and Sharron Angle both committed big public gaffes soon after snagging the GOP nomination, and this last Tuesday’s contests gave them even more cause to celebrate:

Elections this week in Colorado and Connecticut yielded a new crop of oddball nominees. Ken Buck, a gaffe-prone prosecutor once ordered to take ethics classes for his handling of an illegal guns case, defeated former Colorado Lt. Gov. Jane Norton in a Republican Senate primary. In Connecticut, Linda McMahon, who founded World Wrestling Entertainment with her husband Vince, was linked to steroid investigations and appeared in numerous violent and sexually suggestive sketches, bested Rob Simmons, a former congressman and decorated veteran, for the GOP’s Senate nomination. [...]

“There’s no way [the National Republican Senatorial Committee] could tell you they’d rather have Ken Buck than Jane Norton, or that they’d rather have Sharron Angle than Sue Lowden, pre-chicken bartering comment,” the strategist said, referring to the GOP candidate in Nevada and her defeated primary opponent.

“Even with Republicans having low numbers, they are the opposition party and are going to benefit from people saying, ‘We’re ticked off and we want a change,”‘ said Republican pollster Bill McInturff, who conducted the survey with Democratic pollster Peter Hart. “The way you vote your discontent is to say you’re going to vote Republican.”

Mr. Hart said the 2010 contest is being pulled by the sentiment associated with the JetBlue flight attendant who fled his plane via the emergency chute after an altercation with a passenger. Calling it the “JetBlue election,” Mr. Hart said: “Everyone’s hurling invective and they’re all taking the emergency exit.”

It seems pretty clear that Americans are mad at incumbents, not Democrats, but it makes little difference in a year when most incumbents are Democrats. It does, however, seem to make the case that Democrats have a better shot at holding their seats by going negative and talking up their opponents’ wackiness rather than their own records.

Take this latest one Sam Stein reports on for The Huffington Post, showing that 77 percent of Americans think it’s important to have a “choice” between government-run health insurance and private coverage.  That poll comes from Survey USA. Meanwhile, Rasmussen Reports  found just 34 percent of Americans support a health care reform plan without a public option. Okay — so far, so good.

So how is it that Rasmussen’s last poll on health care reform found that only 42 percent of voters supported the Democrats’ proposed health reform plan, when that plan still clearly included a “public option”? Back then (last week), the whole Democratic proposal for health reform was sinking, leading President Obama and some members of his administration to start backing away from the government-run option and proclaiming their newfound flexibility.

Well, that Rasmussen poll referred to the Obama plan as simply “the health care reform plan proposed by President Obama and the congressional Democrats.” It seems significant that none of the Rasmussen questions included the word “choice” in them. And as pollsters for NBC found earlier this week, without the word “choice,” only 43 percent of the public favored “creating a public health care plan administered by the federal government that would compete directly with private health insurance companies.”

Meanwhile, another Rasmussen poll earlier this month found that “51% of the nation’s voters fear the federal government more than private insurance companies.” I guess that’s when they’re not offered a “choice” between the two.

If this confirms anything, it may be that Americans just don’t like to commit. MoveOn.org, which commissioned the SurveyUSA poll and supports the public option, was wise to that.

Still, as Stein notes, when read an actual description of the president’s health care plan (when it still included the public option), 51 percent of SurveyUSA respondents said they “favored” the approach; 43 percent opposed it. Asked the same question by NBC and the Wall Street Journal, 53 percent of respondents said they favored the president’s plan, and 43 percent opposed it.

According to a CNN poll released Wednesday, 75 percent of Americans support Obama’s cabinet nominations, including 71 percent who back his choice of Hillary Clinton for secretary of state and 83 percent who approve of his retaining Secretary of Defense Robert Gates. A Gallup poll released Monday showed similar numbers: 69 percent for Clinton and 80 percent for Gates.

Yet today’s Gallup tracking poll shows that Americans’ faith in Obama to be a good president hasn’t budged. In fact, Obama’s overall confidence numbers since November 23 have been remarkably stable, never climbing above 66 percent or dropping below 64 percent:

Of course, the rigidity of these numbers isn’t so surprising when you take a look at President Bush’s approval ratings, which have recently hovered between 26 and 29 percent:

The die-hard Republicans who continue to support Bush unflinchingly may not warm up to Obama no matter what he does. That Bush’s job approval is roughly equivalent to those who are not confident in Obama suggests that it will be tough for Obama to reduce his no-confidence rating to less than 25 percent. That means he’s doing about as well as he can, and barring a cataclysmic event like 9/11 — which boosted Bush’s approval numbers to an unprecedented 92 percent — he may have nowhere to go but down.

Yet as Sherry Bebitch Jeffe points out in a detailed historical analysis, the 1982 electoral loss of black candidate Tom Bradley gives little evidence for the effect that took his name. Instead, Bradley’s loss can be explained by a combination of guns and absentee ballots.

But Jeffe lays it out more articulately than I could, so be sure to take a look at her piece.

The best analysis I’ve seen of the potential Bradley Effect in the 2008 presidential election comes from Nate Silver at fivethirtyeight.com. Looking at the Democratic primary, he argues that there was no such effect, although unlike Jeffe he believes that the phenomenon played a large role in elections in the 1980s and 1990s.