Hillary Clinton was supposed to win the Michigan Democratic primary on Tuesday night by roughly 21 percentage points, according to the RealClearPolitics polling average. That average was comprised of any number of surveys that we now know to have been—each and every one of them—“crap polls,” to use the proper political science term. Bernie Sanders won a narrow victory against Clinton in a race that FiveThirtyEight had projected Clinton had a 99 percent chance of winning. “If Sanders winds up winning in Michigan, in fact,” Nate Silver wrote Tuesday night as the surprisingly tight results began rolling in, “it will count as among the greatest polling errors in primary history.” Well, there you go.
Michigan is unquestionably Sanders’ biggest win so far. It also demonstrated that he has made some necessary progress in expanding his coalition. The Sanders campaign had hoped that Northern black voters would be more receptive to his campaign message than black voters have been in the South. They were.
In Mississippi, where exit polls showed 64 percent of Democratic primary voters were black, Clinton cleaned up with a full 90 percent support among black voters en route to a colossal 67-point victory. In Michigan, however—where black voters comprised a smaller but hardly insignificant 24 percent of the Democratic primary turnout—Sanders earned 30 percent of the black vote to Clinton’s 65 percent. Sure, Clinton still led Sanders among black voters by a greater than two-to-one margin. But the results show that if Sanders can make even marginal progress among black voters in states outside the South, as he did in Michigan, this will be a considerably closer nominating race. Expect next week’s Ohio Democratic primary, to conjure just one immediate example, to be a closer race than forecasted.
The problem for Sanders is that even if he’s shown that he can win big states outside the South, he’s not showing that he can win them by much. Hillary Clinton has already won delegate-rich states like Texas and Georgia by large margins, basically wiping out whatever pickups Sanders has had, and she’s run up the score in smaller contests largely based on support from black voters. As a result, she had already amassed a daunting lead of more than 200 pledged delegates heading into Tuesday night. (That doesn’t even count her enormous margin among superdelegates, but Clinton’s goal is and should be to win a majority of pledged delegates.)
Winning the big prize on Tuesday night helps Sanders mostly in terms of horse-race narrative purposes: showing the world that he’s not dead yet. But between Michigan and Mississippi, Sanders on net will lose ground to Clinton in the delegate race Tuesday night.* As of this writing Sanders is projected to take 63 delegates from Michigan to Clinton’s 53. In Mississippi, however, Clinton is projected to take 28 delegates to Sanders’ 1, as Sanders only barely crossed the state’s 15 percent minimum threshold. These numbers will change a tad as the final votes are tallied, but right now they show Clinton netting 17 delegates between the two.
Modest victories, no matter the size of the state, are not going to cut it for Sanders. He needs big victories in big states to cut hard into the pledged delegate lead that Clinton has accumulated. Even if he is able to pull off another squeaker in Ohio next week, Clinton is set to swamp him in the demographically and culturally different—and larger—state of Florida. Winning 30 percent of black voters in Michigan is a move in the right direction, but Sanders needs to start winning majorities of all demographic groups in places like California, New York, and Pennsylvania—all while continuing to win the smaller contests in places like Kansas and Oklahoma that he’s had success with.
Sanders’ Michigan win was impressive and a total surprise. The fact that Sanders gave an impromptu press conference in Miami outside a hotel Tuesday night, rather than a victory speech in Detroit, tells you just how surprised his campaign was with its success, too. But victories like these on their own will not do it for him. He needs to take whatever narrative momentum he gains from tonight’s win and snowball it into something much larger—a complete inversion of the way the contest has already played out, really. Otherwise, Sanders will serve, at best, as the Clinton of 2008: the candidate who got into a big hole early and was never able to dig out.
*Correction, March 9, 2016: This article originally misstated that Idaho and Hawaii were holding their Democratic caucuses on March 8. They will be on March 22 and March 26, respectively. (Return.)