Pre-Caucus Thoughts: The Iowa Outcome and Delegate Math

 (source)

by Thomas Neuburger

The Iowa caucuses are today, and the results will be known either tonight or tomorrow. Iowa has 41 pledged delegates available (along with eight superdelegates, who won't vote at the convention until after the first round).

So let's take a quick look ahead. As Bobby Fisher reminded us in Bobby Fisher Teaches Chess, the goal of the game is to checkmate. Anything that gets you there faster is a plus. Anything that slows the process — like oh-so-clever pins and forks — isn't.

It's the same here. The goal is to acquire 1990 delegates to the Democratic National Convention and win on the first round. The faster you get to that point, the better. Anything that slows you down is worse.

So here's the state of play, using projections and delegate math, in two states — Iowa and California.

Iowa Projections and Delegate Math

Predictions for Iowa vary widely, from it being a virtual Sanders-Biden tie to Sanders winning by a wide margin. (For some background on delegate allocation in Iowa and elsewhere, see "How Iowa Chooses National Convention Delegate and What That Means For Us.")

It's true that as of this writing Sanders leads, Biden trails but is viable, and both Buttigieg and Warren hover just at the statewide viability point:

At the top of this piece are what one person claims are the results from the now never-to-be-published Des Moines Register poll, one of the most widely respected polls in Iowa. So for the sake of  discussion, let's take those polling numbers as our starting point and assume this is the actual outcome:

• Sanders: 33%
• Biden: 15%
• Buttigieg: 12%
• Klobuchar: 13%
• Warren: 11%
• Yang: 7%
• Steyer: 1%
• Gabbard: 3%
• Bloomberg: 0%

Notice that, despite the mass of votes garnered by the bottom seven candidates — 52% of all votes tallied — only the top two, Sanders and Biden, are viable in this distribution ("viable" means a candidate got at least 15% of the vote). This in turn means that, in the pure case, only Sanders and Biden are awarded delegates.

If only Sanders and Biden are viable, the pure delegate split, even if they get only 48% of the vote between them, is this. (Some candidates who are unviable statewide may get a delegate or two in districts where they are stronger, but we'll ignore those small differences for now.)

• Sanders: 28 dels. (69% of available)
• Biden: 13 dels. (31% of available)

But if you include Warren as viable (at 15%) and leave Sanders' and Biden's totals as they are, the delegate split changes dramatically:

• Sanders: 21 dels. (52% of available)
• Biden: 10 dels. (24% of available)
• Warren: 10 dels. (24% of available)

These differences between a two-person race and a three-person race are entirely independent of whose vote goes to whom on the second ballot after supporters of unviable candidates realign themselves to candidates who are viable. It's just the math — if two candidates, with 48% of the vote between them, split all delegates, each gets more delegates than if three candidates, with 63% of the vote between them, split the same number of delegates.

I've seen projections that have Sanders winning as much as 36% on the first ballot in Iowa and 40% on the second ballot. That second scenario was modeled by Doug Hatlem here:

• Sanders
  1st ballot: 25.6%
  2nd ballot: 40.4% 
  Delegates: 17

• Biden
  1st ballot: 20.7%
  2nd ballot: 32.7%
  Delegates: 14

• Warren
  1st ballot: 17.0% 
  2nd ballot: 26.9%
  Delegates: 10

• Buttigieg
  1st ballot: 12.6% (unviable)

• Klobuchar
  1st ballot: 8.7% (unviable)

Notice the delegate split. Even with Sanders winning over 40% of the votes on the final ballot, his delegate total — and his separation from Biden, the runner-up — is much smaller than it would be in a two-person race with the same percentage outcomes. (I doubt, by the way, that Biden will get 33% of the Iowa vote on the second ballot. I think the leaked DMR numbers, whatever their source, are closer to the mark.)

So the real number to watch as you watch the contest is Warren's number — Will she be viable statewide or not?

California Projections and Delegate Math

Now let's do the same exercise with California, which votes on Super Tuesday, immediately after the four early states — Iowa, New Hampshire, Nevada and South Carolina — weigh in. California offers a whopping 415 pledged delegates to be divided up among the candidates, with Sanders as the front-runner as of this writing.

For the sake of this discussion, let's say the voting split is the same shown in the recent KQED/NPR poll (I've highlighted the winners totals below):

• Sanders: 33%
• Warren: 16%
• Biden: 15%

In this poll, the rest of the field is below 10% and thus unviable. At this voting split, the pure delegate allocation is:

• Sanders: 204
• Warren: 109
• Biden: 102

But let just one of the lower two, Biden or Warren, become unviable — turning the contest into a two-person race — and the split changes to this:

• Sanders: 271
• Warren or Biden: 144

That is, Sanders, or whoever else leads with 33% of the vote, goes from 49% of delegates available in a three-person race, to 65% of delegates available in a two-person race — with no change in his actual vote totals. 

The conclusion is obvious. The drama around who's viable and who's not — or who has dropped out to support someone else — is almost more important for the winner of a given primary or caucus than his or her actual vote totals. Keep that in mind as you watch the Iowa results come in.