"That's Insulting!"

Originally Posted: April 23rd, 2008

Another blog motivated by the Democratic primary races between Senators Clinton and Obama, challenging the conventional wisdom and spin being generated and perpetuated by spin.

Lol. If you're a Trini, you'll get the reference... If not; well it's still insulting. If I were an American, I must say, I'd be incredibly insulted by much of the media coverage, and Hilary Clinton's current spin on Barack Obama's electablity.

Hilary Clinton's current position is that she is more electable than Barack Obama. Her cited evidence for this, is that she's doing better in the larger states, primarily because she's "doing better with blue collar Democratic voters." (Insert the word "white" there, because it's what the polls actually show.) The media has followed this particular notion - this spin - and it's a strong thread. Anyone from the Times to the Post; CNN to Fox; even a range of international sources, have picked up on this thread. Their reasoning is that this will be the major argument in convincing super delegates to sway towards either candidate.

Why should Americans find this insulting? Any plausible analysis here has centered around one theme: A "disconnect with white working class voters", who Clinton is pulling in in superior numbers. Each candidate has put forth an almost identical policy platform. To say that Barack Obama is not connecting with these folks must therefore hinge on some aspect of campaign strategy, voter identity/disposition, or candidate image.

Please note carefully, that the proportion of white working class voters that Clinton has been pulling in has been almost IDENTICAL over the span of the Presidential Primary race. Since Obama's "bitter" comment, as evidenced from the numbers in Pennsylvania, there hasn't been much change in that distribution.

So why are all of these pundits, and why is the Clinton campaign, assuming that these white, blue collar workers (who will be referred to as WBCWs from here onwards - I refuse to call them "Lunch Bucket Democrats"), would either not vote at all, or vote for John McCain in a general election if Barack Obama were candidate? It is this so called "disconnect" that is being cited. Here's a further question:

Where is the evidence for this so called disconnect? As far as I can tell, there is none whatsoever, especially considering he did not experience a decline in his share of the WBCW vote in this most recent primary. How many unions have turned out in support of the man? How can anyone honestly say that he has no ability to connect with these people? This notion reeks of the Rubin's Fundamental Problem of Causal Inference. For you non-stats geeks out there, that's simply the fact that it is impossible to observe the causal effect on a single unit. To say that these WBCWs vote for Clinton because she connects BETTER with them is fair (A causes B); but to say that because they vote for her instead of him, they aren't connecting at ALL with Obama, not only makes no sense (it is unclear whether, because we know A causes B, that A also causes C, because we can't measure the causal effect on C through B; it must be measured separately if we are to ascertain its value).

Look at the numbers. 46% of WBCWs are still voting for Obama, meaning that 46% still like him better than Clinton. Clinton may be connecting better, but to say that the 54% of democratic WBCWs voting for Clinton instead of Obama would either dessert their party by not voting at all, or turn parties and vote Republican, carries a less than savory assumption.

The underlying assumption - no, I'll just come out and say it - insinuation here, is the WBCWs are not voting for Barack Obama because he is black. One can't claim it's the result of recent campaign developments, such as the "bitter" comment or an alleged elitist, as poll numbers (and you know how much I hate polls) have balanced out. One can't claim it's due to policies towards such workers, because both candidates' policies are too similar.

This is the assumption that I find insulting. The insinuation that there are hundreds of thousands of Americans in Ohio, Pennsylvania, and elsewhere that are vehemently racist and simply refuse to vote for a black candidate. No - even worse - that they vote for a white candidate instead, simply because of race. If I were a voter in one of these states, I would be insulted. Who's to say that these 54% are ALL voting for Clinton in this fashion - because they feel "disconnected" from Obama because he is black?

It's insulting. It might be true, but there's no way of knowing it. I'm sure there are a certain number of Democrats who are effectively voting against Obama by voting for Clinton because they are racists; but to say that all 54% of them in these states are is a gross underestimation of the character of that 54% of WBCWs. Very few people in the media are suggesting that perhaps this 54% aren't completely disconnected from the Obama campaign - most are assuming that they'd rather abandon their party than vote for Obama, by putting weight behind the Clinton campaign's argument that she is more electable in this fashion.

To make matters worse, it also insinuates that they're stupid. That they would vote for a candidate that doesn't represent their values, or that they would allow such a candidate to win by not turning up to vote at all, implies that they'd rather be bitter and suffer than support a candidate who does support their values, but doesn't share their background or skin color.

All the 54-46 distribution tells us, is that on that given day, Clinton would be guaranteed 54% of that particular pool, Obama 46%. We can't simply assume that eithers' supporters will not vote for the other, with any demographic. (The hypocrisy of people negatively criticizing Obama's "bitter" comment, while simultaneously insinuating that all of the same people are racists, is evident here, but let's not get me started on a new rant...) I'm reasonably sure that no one intends to refer to these folks as stupid racists, but that's what is effectively being doing through the aforementioned rhetoric.

Instead of swallowing the bullshit unquestioningly, the question the analysts, pundits, super delegates, and other inquisitive minds such as myself should be asking is, how do we estimate the number of Americans that would vote for either Barack Obama OR Hilary Clinton, within those key battleground states, using the data we have?

It's impossible to know exactly how many people will vote in a racially motivated fashion, especially since people tend lie to the pollsters when asked such sensitive questions; but it is possible to perhaps tease some new information out of proxy data ("Freakonomics" talks about an American Idol study that was used to test people's perceptions via their inclination to vote for or against certain participants based on race - something clever like that maybe).

I'll leave you with one last thought. If there really is a such a "disconnect" between Barack Obama and working class voters, why, when asked whether they felt each candidate connected with them, did they answer an almost identical 70% for both Clinton, and Obama? Do we really think so little of all of these people that we think that they wouldn't vote for Obama based on his race; or that matter, for Clinton based on her sex? If we do, then it really does say something quite daunting about the society we live in.