Sanders’ Big Night

Hillary Clinton led by about 20 points going into Michigan’s primary.  Even if she squeaks out a win, it’s impossible to deny that Bernie Sanders had a huge night there. With 74% of districts reporting, Sanders is up around 4%.  From a delegate math perspective, though, Sanders said it himself “the delegates are going to get split up because of proportional representation.”

Based on the 20-point lead, we projected Sanders would leave with a 30-delegate lead.  If the current numbers hold, Sanders will probably walk away with a 6-delegate lead.  Regardless of what happens, Clinton will widen her gap overall, as she will probably get around 30 more delegates than Sanders in Mississippi. So, in the short term, Sanders win is big and unexpected, but doesn’t matter much.

In the long term, well, Clinton is still has a 200-delegate lead, and the March 15 states look pretty favorable to her.  But if Sanders’ result in Michigan is an indicator that the national numbers are tighter than the current 13-point spread, Bernie could start chipping away, possibly preventing Clinton from getting a majority without having to use Superdelegates.

Dem Race: Clinton Still On Track To Win Without Super Delegates

There’s a lot of talk from Bernie Sanders supporters about the super delegate race, and it’s pretty clear that Clinton currently holds a commanding lead with 460 of the 712 on the board.  They argue, though, that those delegates will start flipping to Sanders when he builds a pledged delegate lead. Based on our projections, however, Clinton is likely to earn the 2,382 delegates she needs to win without using any of her pledged delegates.

Based on our count, Clinton holds a 699 to 458 lead in pledged delegates. Based on current polling, that lead should expand considerably on March 15:

Upcoming states highly favorable for Clinton.
Upcoming states highly favorable for Clinton.

Of the above, only Missouri lacks state-specific data and is therefore based on Clinton’s nationwide 51%-37.6% lead.  If those numbers hold, Clinton will hold a 1244-776 lead following those votes.

Is that impossible to make up? Well, mathematically, no. But the path to get there is highly improbable. To make up a 480-delegate deficit, Sanders would have to win every single delegate in Pennsylvania AND New York, for example.

Does that mean Hillary has it locked up? Well, no. After March 15, we see Sanders basically keeping pace with her.  Our projection sees her building her lead a bit on April 22, but mostly that’s just big-delegate states that are relying on the national poll. Unless things change, she won’t actually wrap it up until the last day of voting on June 7:

Clinton's lead is set to grow March 15 and never substantially narrow.
Clinton’s lead is set to grow March 15 and never substantially narrow.

Sanders is now on record saying he won’t drop out before the convention. That probably makes sense.  There’s not much that Clinton can offer him now that will make him back down, and it probably helps her general-election campaign to have the come-together moment happen at the convention.  That will probably happen:  it’s hard to see Sanders losing the political revolution fight and then letting Trump take the White House.

Projections Update: Post Super Tuesday numbers, what if Rubio drops out?

We’ve updated our spreadsheet that the latest numbers on the delegate projections, based on polling aggregates from and Real Clear Politics.  These are run on a state-by-state basis, with each state’s delegate allocation rules factored into the math.  The poll numbers are up-to-date through March 6, and we’ve added options for Rubio dropping out, as well as forcing new winners in the big winner-take-all states to see what that does to the numbers.

How might our numbers still be off? Well, first off, if the polls are. Some big states, like Florida, haven’t had new poll numbers for nearly two weeks.  National polls don’t yet seem to reflect Super Tuesday (and certainly Super Saturday) voting.

They’re also probably off a few delegates per state, as we use state-wide polling numbers for each congressional district, which favors the state winner by allocating a handful more delegates to them than real life.

A third way it’s probably off is that it apportions undecided voters proportionally among those that make the threshold.  This has no effect on the Democratic side, but does favor Trump on the GOP side for two reasons: it probably undercuts finishers who are polling near the delegate threshold, and the latest results show that late deciders are voting against Trump by large margins.

With those caveats, it’s useful for running scenarios to see what might happen March 15 and beyond.

Run the projections yourself here:

Delegate Math