More on the Mythical 'Obamacan' Masses

It's nice to know you care. 

Yesterday, I posted an item called "Chasing the Mythical 'Obamacan' Masses" in which I wrote that "there's little statistical evidence to support the claim that the number of Republicans who favor this year's Democrat is substantially larger than the number of Republicans who favored his predecessors"--despite what many Obama fans, including his promiment Republican endorsers, seem to believe. As evidence, I cited new numbers from the Pew Research Center--typically considered one of the nation's most reliable polling firms--that show the Republican portion of Obama's support base declining from 10 percent to seven percent from June to August, as McCain's Democratic backing grew slightly from 10 to 11 percent (along with a slew of other, similar surveys).

Since then, I've received an avalanche of responses from readers questioning my analysis. Many of these messages qualify as what I like to call "'Your an Idiot' Emails," which invariably convey their author's impression of me as either a) an incurable moron or b) a Sean Hannity clone (assuming, that is, that they count as two different things). "Did you get wind of an opening at Faux news or something?" wrote reader W.C., for example. "After all, they hire all the other GOP hatchets, why not apply?" That said, a refreshing number of readers were actually thoughtful, picking up on a factor I didn't mention in my original post--a factor that they think could be working in Obama's favor--and requesting I explore it in a future Stumper entry.

Here's reader C.M. explaining it better than I ever could:

I think you're ignoring the fact that there has been significant (anecdotal) bleed in the Republican party, where people who formerly have identified as Republican are now identifying as Independent.  Normally, one would expect that this 'bleed' would bias the Independent more towards the Republican candidate than should otherwise be expected, as these individuals by-and-large agree with the Republican philosophy but have drifted away from the party for a variety of reasons.  However, again through anecdotal evidence, I have heard that the Independent pool looks quite similar to the breakdown in '06.  This would imply a non-negligible portion of the Republican Party from 2004 is today voting for Obama, though these people no longer identify as Republicans.

It's a terrific point, and one worthy of further analysis. So here goes. The best way to measure whether Obama is winning over large numbers of people who voted Republican in 2004 but now identify as Independents--which, by the way, is a pretty common occurrence these days--is to ask them. Imagine a poll that inquires whether its respondents voted for Bush or Kerry last time around, then shows how those same voters are planning to vote in November. Such a survey would allow us to see how many former Bushies, regardless of whether they currently self-identify as Independents or Republicans, are now backing Obama--a more inclusive measure of "Obamacan" strength than the usual Republicans for Obama statistics. Unfortunately, I haven't found any polls that actually do this (although I'm still looking, and may ask NEWSWEEK's pollster, Larry Hugick, whether he'd be willing to give it a try in the near future).

In the meantime, we'll have to look at how Obama is faring among Independents to get a sense of whether "a non-negligible portion of the Republican Party from 2004 is today voting for Obama, [even] though these people no longer identify as Republicans." The theory here is that if a lot of 2004 Republicans are now calling themselves Independents and voting for Obama, he should be performing better against McCain with Independents overall than John Kerry performed against George W. Bush in 2004. Sadly, that doesn't seem to be the case. In August 2004, Pew found Kerry leading Bush by eight points among self-identified Independents, 48 to 40 percent. Currently, Obama's edge within the same subgroup is half that size, and when you include Pew's May and June results (44-44, 42-41) any statistically significant lead disappears.

Sure, it's impossible to discern whether Obama is outperforming Kerry among the "former Republican" segment of the Independent electorate without stats showing how these folks voted in 2004. He very well could be. That said, Obama's tie with McCain among Independents overall indicates that either a) his performance mirrors or lags behind Kerry's or b) McCain is compensating for Obama's gains by besting Bush among other types of Independents. (It's possible that his support from former Clintonites offsets Obama's support from former Bushies, for example.) Either way, the current stats suggest that 2008 will look a whole lot more like 2004 (when Kerry and Bush split the Independent vote 49-48 percent on Election Day) than 2006--when Independents told ABC News that they would support Democratic candidates for Congress over their Republican rivals by a two-to-one margin, 59 to 31. In other words, "there's little statistical evidence to support the claim that the number of Republicans who favor this year's Democrat is substantially larger than the number of Republicans who favored his predecessors"--or, if so, that they'll make much of a difference on Election Day.

And so our hunt for the mythical 'Obamacan' masses goes on.