As I type this, 88% of Nevada precincts have reported. Close enough to done for us to look under the hood a bit. And we’re finding the conversion from first alignment (i.e. popular vote), to final alignment, to County Caucus Delegates, changes the story. A lot. If it weren’t for Bernie Sanders, we’d lack this information. Nevada would have reported CCDs only. And without the transparency and the extra time it takes to make sure the numbers add up, he’d have been declared the winner with almost half of Nevadans supporting him by late afternoon on Saturday.
I’m sure he’ll be fine the way it turned out. Everyone knows he crushed it. Sanders wound up making the same statement Donald Trump did in 2016, when he ended the idea there was a 30-35% ceiling on his support by getting 47% in Nevada. Bernie is tracking along at 47% himself. But these two numbers are not the same. Republican caucuses don’t have realignment or conversion to state/county delegates. When Trump got 47%, he got 47% of all voters who participated.
Right now, Sanders has 34.3% of the initial popular vote. That’s not a bad number at all in a six person contest. It’s almost exactly double Joe Biden’s 17.9%. But when we move from first to final alignment, Sanders jumps to 40.7%. He then leaps again to 47.1% when assigning the CCDs. A solid win becomes a rout.
Couple things here. There aren’t any caucuses on Super Tuesday. In 2016, several of these states (Colorado, Minnesota, Maine, Oklahoma, Utah) had caucuses. Beyond the math being cleaner, Sanders won’t have the advantage the candidate with the most dedicated supporters gets from the more difficult voting process. He beat Hillary Clinton in almost every caucus last time (though not in the one he just won.)
He recently won a primary in New Hampshire. He may win most/all of these. At this point, you’d need to favor him to. But when benchmarking what to expect, think of Bernie as having a strong Nevada result in the mid-30s, not a hugely dominating one in the upper-40s. And the states mentioned above have somewhat to much smaller percentages of Latinos. Sure, he could get some extra momentum from the perception that he did even better than he did. But so far, we haven’t seen huge momentum-based impacts. Pete Buttigieg didn’t suddenly win over Nevadans of color because he (sorta) won Iowa and almost won New Hampshire. Biden’s support in Nevada didn’t collapse after his dismal fifth place, single-digit New Hampshire debacle.
Because it appears Sanders beat Biden by even more in Nevada than he actually did, a Biden victory in South Carolina will seem like a larger deal than it otherwise would. He’s still leading in the polls there, and while a 4 to 7 point Biden victory there would have seemed weak a month ago, if it happens on the 29th, he’ll appear to have pulled off a miracle.
It’s hard to see how Sanders doesn’t wind up with the largest amount of delegates heading into the convention. At the moment, FiveThirtyEight is showing him with a 45% chance of having an actual majority, plus a 41% chance of no candidate having a majority, but in many of those cases, Bernie has a plurality. It’s way better than 50/50 he’s leading then. In only 1 out of 7 simulations, another candidate has a majority and that’s almost always Biden.
While you’d rather be Bernie than Joe, the ex-Veep has a window. And part of that window is due to Nevada caucus math. Like I mentioned above, he got 17.9% of the first alignment vote. Pete Buttigieg got 15.2%. Elizabeth Warren is at 13.1%. The gap between Biden in second and Amy Klobuchar in sixth is less than 9 points, about half the distance between Biden and Bernie. But by the time of final alignment, Biden moves up to 19.7%, while Klobuchar drops to 4.0%. Buttigieg remains close with 17.1%.
Had Mayor Pete duplicated his conversion skill from Iowa, he could have passed Biden, finishing ahead of him in each of the first three contests, and no worse than second overall in any. There would be real question about who is best positioned to challenge Bernie in South Carolina and after. Buttigieg would have been one more bad Mike Bloomberg debate away from being very important.
But it didn’t go that way. Mayor Pete’s weakness in Clark County (Las Vegas + suburbs) was deadly. The CCD conversion brought Biden up to 21.0% and Buttigieg down to 13.7%. Warren, Klobuchar, and Tom Steyer combined for fewer CCDs than Biden has on his own. This is all very good news for Joe in South Carolina. His non-Bernie opponents were weakened, and the expectations are just about perfect for him. If Biden wins South Carolina by a few points, holds off Bloomberg and Sanders in places like Alabama, Arkansas and Tennessee on Super Tuesday, he’s back to being able to at least take Sanders all the way to Milwaukee.
Looking further down the road, primaries like Florida, Ohio, and Pennsylvania give Biden the chance to argue for the nomination even if Sanders has a few more delegates. Indiana in May would be a key test of strength. Plenty of other candidates would be willing to swap their delegates for a place on Biden’s ticket. Meanwhile, Joe has some work to do. He hasn’t won South Carolina yet. If Bernie does, we’re done. Time to start thinking about November. But if he recovers and at least makes this interesting, Biden will have caucus math to thank.
Side Note: The polls were excellent. Nevada normally has miserable data. Confusing, unreliable numbers in low quantities. Here, the surveys were good enough that I managed to guess first alignment results for each candidate within 3.3% of what really happened. I was within 0.7% on Warren, Steyer, and Klobuchar. After that, the accuracy waned tremendously. Turns out the polls were predictive, but the Iowa delegate conversion results were not. This is the third straight state where the poll averages were dead on. With this many candidates and two of the three being caucuses, that’s remarkable. Remember this next time you hear someone say polls are crap.