When all the votes are counted, Hillary Clinton will easily win a majority of the pledged delegates and would seem to have an outside chance of winning a majority without any super delegates. To reach this conclusion, we stuck to the original Bernie Barrier theory of Democratic delegate math, meaning the Cook Political Report’s estimate of what Sanders needed in each contest to secure the nomination.
For a break-even result, Sanders, at this point in the contest, is supposed to have a 23-delegate lead over Clinton. Instead, Clinton has a 223 delegate lead. This may come as surprise to those who follow the press coverage that tends to focus on wins in individual states--Sanders’ big win in New Hampshire, his upset victory in Michigan--rather than delegates or overall votes case. They might not have realized that Hillary Clinton has received over 1,600,000 votes more than Bernie Sanders, which explains her large delegate accumulation. Saying on Super Tuesday Sanders won four, and Clinton won seven does not explain her large lead. Nor does saying Clinton won Mississippi, Sanders won Michigan. The 13-9 Clinton contest advantage is a mask that obscures the truth. Clinton is wining by a lot. In fact, based on the Bernie Barrier chart, Sanders has only exceeded his delegate take three times in Kansas, Oklahoma, and Vermont. By contrast, Clinton has done better than needed in all the other states besides New Hampshire and Michigan where they both did as expected (according to delegate requirements).
While Sanders’ Michigan win may have been important for the narrative, the reality is that Sanders netted only four delegates out of Michigan. Even if he exceeded the Bernie Barrier’s expectations from here on, it would still be very difficult for him to win. For those interested in the reasons, we explain at the bottom why March 15th states and beyond will be so hard for him and therefore why we are comfortable in this call.
We feel compelled here to point out a bug in the 538 Delegate Tracker that has the effect of making the race look closer than it is. Last weekend’s results are particularly instructive in revealing that bug.
On Super Tuesday, Hillary Clinton ran 77 delegates ahead of her target according to the 538 chart.
Because the Democratic race is proportional any running ahead requires that the chart re-calculate new goals after each contest. Going forward Clinton can do worse and Sanders must do better in order to get to a tie.
This is very important. But 538 doesn’t recalculate. Based on the original chart, Sanders in fact exceeded by one delegate what he needed in the four states that held contests this weekend and thus had a good weekend. What Sanders would have needed in these four states has to change to reflect his being off pace on Super Tuesday. 538’s overall coverage has basically reflected a-Sanders-is-doomed tone but the chart makes things look better for Sanders as they are. Same thing goes for Michigan. As good as expected before is no longer good enough now because of the already built up gap.
The work of re-doing a chart after each contest may be time consuming and annoying but it is essential. This is particularly important because people are making arguments off outdated information. Someone just this week pointed to a 538 what-Sanders-needs-to-get post that was pre-South Carolina to argue that Sanders exceeding the old estimate in Maine was somehow good news for him. The math is hard but 538 needs to keep doing it. The correct story out of Michigan for Sanders is good win, but nowhere near good enough. This needs to be considered going forward on the R side as well. 538’s argument that Trump is on “pace” needs to be read with similar understanding of its flaws.
How we are able to make this call
Our confidence in Clinton’s victory rests upon the outsized role African American voters play in the Democratic primary process. They deliver large numbers of delegates in districts where they are the majority; they provide the winning margin in districts where they are a substantial number of voters; and they also can be exceptionally helpful in districts where they are a minor presence by keeping the margins close. (It is very hard to get to 58% or 62.5%, which is needed to net a delegate, when 10% of voters vote against a candidate by 70% to 30% or worse, as happened in the Nebraska 2nd that includes Omaha.)African Americans’ influence was apparent in Michigan. Even in losing Michigan, Clinton relied upon African Americans to keep the statewide vote exceptionally close and they delivered a considerable number of district level delegates. In Michigan’s 5th district, which includes Flint, African Americans enabled her to win the district and take an extra delegate. In Michigan’s two Congressional Black Caucus district seats (the 13th and 14th), they allowed Clinton to win 6-3 delegate splits in both.
Clinton only won these three districts but they were enough to basically get her to a tie statewide. She lost one delegate statewide because it had an odd-number of delegates. At the Congressional district level, Sanders won 11 out of 14 districts, yet netted just three delegates.
If Sanders replicates similar successes across other states, he also replicates similar problems. Let’s look at how allocations have been made in the contests held so far. The place to start is with Congressional Black Caucus member districts. In previous CBC districts, Clinton has won the delegate count 75-16, including three in which Sanders didn’t get 15% of the vote and thus missed the threshold for getting any delegates. (This excludes Texas where delegates were awarded at the State Senate District level; the pattern held but is impossible to get a precise calculation.)
On March 15th, voting will take place in 12 such districts that collectively hold 109 delegates. Even if one assumes a split similar to the one that took place in Michigan, Clinton would get 74 delegates to Sanders’ 35 delegates, which would net the Clinton campaign 39 delegates. (Clinton is likely to do better in the March 15 CDC districts than in Michigan because there are two southern states on the schedule, Florida and North Carolina, and there is early voting in those states as well as in Ohio. Older African Americans, Clinton’s base, are overrepresented in Southern states and in early voting.)
Even if Sanders won every single remaining odd district allocation (there are 35 across all five states that can be won by winning just a single vote), he would still lose the day by four delegates. Once you get into districts where African Americans have influence (where they make up around 30% of the vote as in the Michigan 5th) or districts with other Clinton-supporting demographics such as downscale Hispanics (see Texas) and Jews (see Newton and Brookline in Massachusetts), Sanders’ problems magnify.
Even if Sanders were able to “win” Ohio, the lack of odd-delegate districts means he could easily “win” the state popular vote but lose the delegates because of bad splits in the Ohio 3rd and 11th, CBC districts. The same goes for Illinois and Missouri. Florida and North Carolina more likely than not will continue the “Southern” trend of favoring Clinton by large margins. (As always, we will be providing a full delegate allocation for all five states before the election.)
CBC districts will continue to have an influence even after March 15th; there are still 14 CBC districts to go. These districts can mitigate bad but not disastrous performances by Clinton elsewhere, just as they did in Michigan.
Overall, Sanders seems unlikely to turn things around enough to make up Clinton’s 223 delegate-lead. To argue Sanders’ best case scenario, even if he wins two-to-one in the 477 delegates remaining in predominantly white states and ties in all remaining states combined, he would gain 159 delegates. Because he is 223 behind, that number would be insufficient. Now let’s assume Sanders does better than tie, and nets the four delegates out of each remaining state that he gained in Michigan. He would then win pledged delegates by eight delegates. But even a tie is extremely unlikely. If Clinton won Puerto Rico with 58% of the vote, she would move back ahead. In addition, there are still 200 CBC district delegates left, which can basically erase all of Sanders’ white district margins. That means he is incapable of making any gains at all, let alone the gains needed to overcome his 223 delegate deficit.
This is why we are confident Clinton will win the pledged delegates.
0 comments:
Post a Comment