All of this writing and data analysis is a lot of work! So after more than five years of posting, I've finally launched a Patreon to help pay the bills.

1/9/20

Another very bad argument for Sanders-Warren "unity"

I have a broader article about the Sander-Warren "unity" discourse on the way, but here I want to make a narrow point. Sam Adler-Bell, with a rationale for keeping Warren in the race:
According to Morning Consult, Biden is the second choice of the most Buttigieg supporters, at 27 percent, followed by Warren at 20 and Sanders at 12 percent. So if both Warren and Sanders stay in the race, the left rises if Buttigieg falls. If only Sanders is in the race when Buttigieg begs off, Sanders would fall significantly relative to Biden.
Let's do some math. Right now, the polls look something like this:

If Warren drops off, her voters get spread among the remaining three candidates like this:
Here, we see that Warren's drop-out disproportionately helps Sanders - he gains 2.1 points on Biden. Now, suppose that Buttigieg drops out:
Yes, it's true that a Buttigieg dropout disproportionately helps Biden. But clearly, the margin he lost from a Warren dropout (2.1 points) is larger than the margin he he gained from a subsequent Buttigieg dropout (1.2 points).

Of course, if (like Adler-Bell does here) you begin by stipulating that the left's advantage is measured by adding Warren's numbers to Bernie's, then it follows trivially that Warren shouldn't drop out. That's just true because S + W > S, it has absolutely nothing to do with Buttigieg.

This may seem like a small point, but it's a good illustration of how flimsy and simplistic a lot of these strategic rationales for Sanders-Warren "unity" really are. We don't know how any of these numbers will change once candidates start winning states, we don't know how candidate drop-outs will change the distribution of second-choice preferences among different coalitions - and as we've seen here, people are having trouble understanding the direct implications of the numbers that we do know.