The Cedar Lounge Revolution

Apologies, but the US Presidential nomination process continues to be a source of fascination, to me at least. And perhaps it should be to everyone, because here to some degree is the vote from the global hegemon… and what way that wind blows can be mighty important to the way the rest of the world goes…

So, West Virginia. Hillary Clinton’s pyrrhic campaign continues, with a significant victory… As John Dickerson said in Slate:

Barack Obama lost West Virginia by 30 points, which looks like an enormous fall. Clinton was favored to win the state, but Obama is the all-but-named nominee. Shouldn’t that have prevented such a rout?

But no, apparently not. Obama did everything bar ignore the state. What is a bit ironic is to hear Clinton now arguing that Florida and Michigan should now be seated in her favour, places Obama and Edwards avoided due to the states breaking Democratic Party rules.

And the only way Clinton can seriously expect to over turn Obama’s lead in ‘the metrics that matter: He’s ahead in pledged delegates, the popular vote, and states won’…is to convince the superdelegates that Obama is unelectable. This is a superficially attractive approach – albeit one rife with danger, particularly that of playing the race card – since as Dickerson notes…

Obama’s problem with white working-class voters does suggest that his powers of persuasion have real limitations. Throughout the campaign, he has touted his ability to reach out to people and to bring them together…he’s been trying to woo working-class whites for months and months—arguably since the start of the campaign—and he can’t get a handle on them. This is not an electoral problem, perhaps, but it’s a governing problem. How can he make the case for his special ability to rally all Americans of diverse backgrounds and interests, yet have such a big problem with one group?

Now therein lies a problem, and indeed a source of discomfort, because it is this argument that Clinton seeks to enunciate without quite saying it (although in recent comments she’s come awfully close). Her only course of action is to intimate that Obama because he is Obama, because he is also black has a limited appeal to a broad demographic necessary to win the race.

But let’s not overstate it… since in polling on the General Election we learn from:

…a recent Los Angeles Times poll, [that] Obama performs only a little worse among white working-class voters than Hillary Clinton does in matchups against McCain. More broadly, in other surveys, such as a recent Washington Post/ABC poll, he is still running as well or better than Democrats have in the past with white voters. Unless white voters are lying to pollsters consistently and in huge numbers, Obama’s problems in the primary don’t seem to translate into the general-election campaign.

And an unlikely source gives at least limited credence to this view, for in a New Scientist from March we read l that:

…for the pollsters, predicting who will win the Democratic Party’s nomination for US president has been a nightmare.

This is because:

Something about the Clinton-Obama tussle – with its overtones of race and gender – has exposed flaws in the science of pre-election polling. The causes are being hotly debated, but the leading contenders are the models used to predict who is likely to get out and vote. “They have no scientific basis,” argues Jon Krosnick, a survey methodologist at Stanford University in California.

This year’s New Hampshire Democratic primary was the biggest embarrassment for US pollsters since the presidential election of 1948 – when Democrat Harry Truman defied predictions to defeat Republican Thomas Dewey. Obama led Clinton by an average of 8.3 percentage points in polls taken over the three days before the primary on 8 January. Yet Clinton won by 2.6 points.

Some attribute this to the “Bradley effect”. This is named after:

the 1982 race for governor of California, when the black Democrat, Tom Bradley, faced white Republican George Deukmejian. Pre-election polls gave Bradley a solid lead, but he was narrowly defeated.

But there are problems with such an analysis. As New Scientist notes, the Bradley effect is usually seen as an aversion on the part of white voters to appear racist by telling pollsters that they won’t vote for a black candidate. But in this instance…

Thomas Guterbock, who led the Virginia research team, finds it hard to believe the polling for this year’s Democratic primaries has been thrown off by voters lying about their true intentions. Obama and Clinton have similar policies and are both popular within their party. This means that prospective voters should be free to voice a preference without worrying about appearing racist or sexist.

So that throws up another possibility, that the polling firms don’t attempt to reflect the reported preferences of actual voters, as against all voters. Instead they make a mathematical adjustment through a formula individual to each firm.

Some polling firms use screening questions, asking people how much attention they are paying to the campaign, how much they care who wins, and so on. They then apply a complex formula to score the answers, and impose a cut-off value to separate likely voters from those who probably won’t bother. Other pollsters pay more attention to the turnout in previous elections. Some even run a variety of models and decide which one to believe based on their gut feeling.

And the result? Well, a multiplicity of results…

…each method can result in a predicted electorate with a different breakdown by race, age and education. For instance, one poll before the 4 March Texas Democratic primary suggested that 14 per cent of voters would be black, while another said 23 per cent. Similarly, estimates of the proportion of Hispanic voters varied from 23 to 39 per cent.

With Obama winning around 85 per cent of the African American vote and Clinton backed by around two-thirds of Hispanic voters, these differences would affect the polling numbers reported by each firm…

Amazingly though we may never know if this is accurate, because as NS reports, some pollsters are “…reluctant to release details of their models. “I understand that people like to see this information as proprietary,” says AAPOR president Nancy Mathiowetz at the University of Wisconsin-Milwaukee. “But you have no right to hide behind that if you release a public number.”" I don’t know. I suspect that is covered by some sort of proprietary rights, although the logic of permitting access is self-evident, particularly in a context where information may have a cause and effect once in the public domain.

New Scientist in an accompanying editorial had some interesting points to make. It was very positive regarding Obama’s speech in March about race, and noted that ‘most importantly, he also argued that it is possible to confront these issues [of race]‘.

It continued…

Overt racial hatred may be in retreat, but there is evidence that bias persists… research suggests that even avowed ‘non-racists’ harbour subtle racial prejudice – apparently a consequence of humanity’s tendency to form a series of ‘in-groups’.

That may sound like a gloomy proscription, but it immediately follows with:

these biases are not immutable however… [research] has shown that subtle interventions, such as showing people videos of discrimination, can reduce subsequent prejudice. The media could do a better job of highlighting such research. Instead, attention tends to focus on controversies such as the simplistic debate over race and intelligence that erupted from the book The Bell Curve in 1994.

[for more on the latter issue look here]

More importantly again NS takes a line familiar to the left when it argues…

Fanning those flames is easy. It gives succour to closet racists whicle letting liberals rage against racism without facing their own prejudices. Far more productive would be a debate over the implications of research by Dalton Conley of New York University, who found the main factor limiting achievement by African Americans to be the low accumulated wealth of the typical black family – just one-eighth that of its white counterpart.

Now that’s a shocking statistic. And isn’t it telling that at this point while we hear an awful lot in this Presidential nomination campaign about class in relation to race the direction is largely one way. And it isn’t focussing on the lack of black capital.

But it is of course in polling data, plus the raw outcome from the primaries, that Clinton bases her last stand defence. This sense that somehow Obama is fatally flawed as a national candidate by the truth that dare not speak its name. And it is the worries and concerns of superdelegates desperate to see a Democrat in the White House in this the ‘Democratic’ year that her campaign plays to. One wonders whether the midnight call TV spot where it is her hand that picks up the telephone to deal with some crisis (although after an unsuitably long time…) also touches on some sort of middle of the night anxiety on the part of the superdelegates.

Worth remembering that no candidate is perfect in any electoral campaign. That somewhere along the line decisions are made and statements put out that alienate one group or another. Worth remembering too that winning the nomination in and of itself is not enough if it divides the Democratic Party election machine and the DP base before November. I can genuinely understand the urgency and sense of opportunity that Clinton may feel as she surveys what appears to be an historic moment for her.

In truth – in any other year it is possible that she would walk away with the nomination with relatively little effort. And that must be galling, as it would for anyone in that circumstance. But not this year.