After a long drought, there's a new survey in South Carolina from a recognized pollster that takes the temperature of the races in that state. (There have been other polls, like this one, which numbers-thirsty pundits talked up but which had zero explanation of its methodology.)
The American Research Group (ARG) found that the Republican field looks about as you'd expect: Donald Trump in the lead; John Kasich, Marco Rubio, Jeb Bush and Ted Cruz jockeying for second; Ben Carson still running for some reason. But on the Democratic side, the picture is fascinating, especially when compared to the results in New Hampshire.
In New Hampshire, Bernie Sanders won by more than 22 points. According to ARG, Hillary Clinton leads in South Carolina by an even larger margin.
The immediate factor at play -- the one that has already popped into your head and the one that will be tweeted at me after this article goes live from people who haven't read it and who assume I'm an idiot -- is race. There are too few black voters in New Hampshire to be statistically significant. In South Carolina, they were a majority of the primary electorate in 2008. And they overwhelmingly back Clinton.
If this margin holds -- which it very well may not -- reports like this one about Sanders out-organizing Clinton in the state likely won't matter. Turnout is important, but the best field operation in the universe can't make up a 25-point difference. But that's if the margin holds.
Because of that gap among black voters and because black voters are so much of the electorate (ARG estimates just over half), it's not surprising that Clinton dominates among other demographic groups, too. Sanders won women and young people and just about everyone else in New Hampshire. Clinton wins all of those groups in South Carolina, which one might presume is in part because so many of those women and men and young people and old people are also members of the strongly pro-Clinton black population.
And now, we pull the rug out from under you: Clinton also wins white voters in South Carolina.
If this poll is correct -- this one new poll in the past month, from a pollster that uses traditional methods but didn't do great in FiveThirtyEight's pollster ratings -- then Clinton has an advantage in South Carolina that extends beyond race. Perhaps it's the same advantage her husband enjoyed in 1992, when he first started to pile up wins once the voting got to the South (where, of course, he'd been a governor). Perhaps it's because Clinton does better among moderate Democrats, and the Democratic electorate in South Carolina is closer to the middle than that of New Hampshire. Perhaps it's because Clinton generally has a big lead over Sanders, which was masked in Iowa by its white electorate and in New Hampshire by Sanders's home field advantage.
This is why more polls are better. More good polls, that is, not the random ones we seize upon because talking about numbers, whatever their genesis, feels slightly more responsible than just wildly speculating. But even with good numbers in hand, as I've just shown, there's still wild speculation to be done.