Update: A new CNN/ORC poll Wednesday of the Nevada caucuses, right on cue, shows a very close race, with Hillary Clinton at 48 percent and Bernie Sanders at 47 percent. Interestingly, the poll shows "both white and non-white voters about evenly divided between the two candidates."
Here are the possible outcomes of this weekend's Democratic caucus in Nevada:
- Hillary Clinton wins by a wide margin, because of her strength with the state's non-white voters.
- The race is close, with the campaigns of both Clinton and Bernie Sanders fighting to a near-draw.
- Bernie Sanders wins by a wide margin, because of his grass-roots organizing effort and thanks to drawing closer in the polls.
Those are, of course, basically the three options in any state. New Hampshire was Option No. 3 with the twist of Sanders being from nearby Vermont. Iowa was Option No. 2, a hard-fought contest. South Carolina, if polls are to be believed, will be Option No. 1.
Nevada, though? We have no idea.
Normally, we'd have polling to guide us. But, so far, polling in Nevada has been scattered and sketchy. Over the past 12 months, as the Wall Street Journal noted, there have been seven polls conducted among Democratic voters in the state, only one of which was conducted this year. That one, conducted by a Republican firm even as New Hampshire was still voting a week ago, showed a tie.
Even if we had lots of good polling, it's not clear how informative it would be. In 2008, there were a number of polls in January leading up to the caucuses on the 19th. On the Republican side, the Real Clear Politics polling average gave the edge to Mitt Romney, suggesting he'd earn 26 percent of the support at the caucus, followed by John McCain with 21. Romney won -- with 51 percent of the votes. In second was Ron Paul, who RCP had in sixth.
The Democratic side wasn't much better. Clinton outperformed the RCP average by 13 points; John Edwards underperformed by 15.
There are a lot of possible reasons for this, including that the polling itself may have been flawed. Part of it too is probably also that Nevada's system on the Democratic side operates like Iowa's, with horse-trading and negotiating built in. With a large, strong field in 2008, that no doubt shifted things.
A closer look at the only previous recent poll, from TargetPoint, shows an electorate that, on paper, seems similar to what we saw in Iowa. Among those who say they're "definitely" voting, Clinton has an advantage of 48 to 41 percent. Fifty-one percent of Clinton supporters say they've caucused before, versus 41 percent of Sanders backers -- margins of likelihood that favor Clinton as she was favored in Iowa.
The survey didn't ask about ethnicity, which remains a question mark. Sanders was certainly helped in Iowa and New Hampshire by those states' heavily white populations; Sanders has generally fared much worse with non-white voters. But most of those non-white voters are black voters, not Hispanic. (In the eyes of the Census Bureau, as Clinton's campaign tried to leverage last week, white and black are racial identities, while "Hispanic" is an ethnic one. So voters can be both white and Hispanic, though pollsters usually consider Hispanic voters as non-white.) New data from Gallup finds that many Hispanics nationally -- 46 percent -- aren't really familiar with Sanders, even this month. Among those who know him, though, he's viewed favorably.
In 2008, 15 percent of the state's Democratic caucus-voters were Hispanic -- and two-thirds of them voted for Clinton. Whether or not that level of support still exists isn't clear.
So perhaps Clinton's strength with Hispanics (and the large percentage of Nevada Democrats who are black; 15 percent in 2008) will power her to an easy victory. Or perhaps, like Iowa, Sanders has made up a lot of ground and will have an easy win of his own in a caucus battle driven by grassroots energy and organizing.
In other words: As of this moment, it's impossible to tell.