Political junkies are touting this ECU data as a hopeful sign for Biden, which is sort of true. After all, there haven’t been many early-state polls lately in which he’s been leading. Or in second. Or, er, third.

Or, in New Hampshire’s case, fourth.

So, sure, good news for Joe. He hasn’t collapsed in his must-win state — yet. But the trend line is important. The day before the Iowa caucus he stood at 37 percent in South Carolina, nearly twice as large a share as second-place Tom Steyer, who had 19. Bernie was a weak third at 14 percent, in keeping with the rap on him that he can’t win in places with large black populations. Biden’s 37 percent was also in line with his polling in SC for much of the race, which has seen him consistently in the 30s and occasionally even in the 40s. All along, South Carolina was conspicuous as a state poised to deliver not just a win to its first choice but a decisive one, a rare departure in this 40-car pile-up of a primary.

But not anymore. Joe-mentum or no-mentum?

Biden now leads with 28%, followed by Bernie Sanders (20%), Tom Steyer (14%), Pete Buttigieg (8%), Amy Klobuchar (7%), Elizabeth Warren (7%), Michael Bloomberg (6%), and Tulsi Gabbard (1%). The poll was conducted February 12-13, n=703, +/-4.3.

Among likely voters who are African American, Joe Biden leads Bernie Sanders 36% to 20%. Tom Steyer is third (17%), followed by Elizabeth Warren (7%), Michael Bloomberg (5%), Pete Buttigieg (4%), Tulsi Gabbard (1%), and Amy Klobuchar (1%)…

There is a generational divide among likely voters who are African American. Among those 55 years and older, Joe Biden leads with 49%, followed by Tom Steyer (9%), Michael Bloomberg (8%), Bernie Sanders (7%), Pete Buttigieg (4%), Amy Klobuchar (1%), and Tulsi Gabbard (0%). Among African Americans age 54 years and under, Bernie Sanders has a slight lead with 29%, followed by Joe Biden (26%), Tom Steyer (23%), Elizabeth Warren (10%), Pete Buttigieg (3%), Michael Bloomberg (3%), Tulsi Gabbard (1%), and Amy Klobuchar (0%).

The age divide among black voters is fascinating. Four years ago Hillary crushed Bernie 86/14 among black Democrats in South Carolina; there’s no split by race and age in the exit polls from the time but she must have beaten him soundly among voters of all ages in that group to have such an enormous advantage overall. Although Bernie’s strength has always been the youth vote, black South Carolinians were something of an exception to that rule — in 2016. In 2020, the age divide is showing up. Biden destroys the field among older black voters but it’s Sanders who’s ahead among those 54 and under. That’s a key reason, I assume, why Joe leads him by just 16 points overall within that group when Clinton’s margin was 72. (The much bigger field this year is another obvious one.)

The point is, if just eight points separate Biden and Sanders in SC right now, and if the new Nevada poll predicting another Bernie win holds true next Saturday, how confident are we that Bernie won’t overtake Joe in South Carolina? Even if Biden hangs on, at this point it’s a safer bet that the margin here will tighten on Primary Day than expand. Is a three-point Joe Biden victory in his “firewall” state, where he’s supposed to flex his muscles, really going to help much 72 hours later on Super Tuesday?

Relatedly, does Bernie even *want* to win South Carolina?

If Biden winning SC keeps him in the race a few more days, hoarding a few moderate votes from Bloomberg and Buttigieg on Super Tuesday, then Bernie’s probably better off losing there. In fact, this very poll shows how a big, chaotic centrist field is reducing the chances of stopping Bernie. Pete Buttigieg is up four points since the last poll taken of South Carolina, no doubt thanks to his strong showings in Iowa and New Hampshire. Ditto for Amy Klobuchar, who’s up five points. Those votes probably came out of Joe Biden’s pocket, causing his nine-point decline. If Mayor Pete or Klobuchar do well in Nevada next week, that may spark a further bounce in SC that ends up depleting Biden’s vote share further, potentially costing him the race.

One other point about the data: The two billionaires in the race are each affecting the numbers here in key ways. Tom Steyer’s spent a bundle in South Carolina to try to gain traction and it’s paid off with a third-place standing — but I’m not so sure it’s Biden who’s mainly being hurt by his success. Look again at that age divide among black voters and you’ll see Steyer taking 23 percent of the 54-and-under black vote, which is a group leaning towards Bernie, versus just nine percent in the 55+ group, which is a Biden bloc. It may be that Steyer is muscling in on Bernie’s youth vote to some degree. As for Mike Bloomberg, he’s at six percent here, which feels like a poor showing … until you remember that he’s not on the ballot in South Carolina. He’s not contesting this state, in which case I’m, er, not sure why he’s included in the poll. That being so, it’d be nice to know who the second choice for Bloomberg supporters is. If, hypothetically, Biden were uniformly their fallback option, then Joe’s actually at 34 percent in this poll, not 28. Regardless, if you make the safe-ish assumption that more Bloomberg supporters would choose Biden as their second choice than Bernie, then Joe’s actual lead here is bigger than it seems. It’s just a question of how much.

In lieu of an exit question, here’s FiveThirtyEight’s latest breakdown of how likely each candidate is to amass a majority of delegates, thus clinching the nomination, before the convention. Hoooooo boy.