What if the Democrats had used ranked-choice voting in 2020? - The Washington Post

riariaga.blogspot.com

Political nerds being political nerds, a colleague asked me this week if it was possible to figure out how the 2020 Democratic presidential primary would have turned out had the party used a ranked-choice voting system.

You’re no doubt familiar with the concept by now, given all of the attention paid to the mayoral contest in New York City which, for the first time, is using such a system to determine a winner. Voters there pick up to five candidates in order of preference. The first preferences of each voter are counted, with the votes for the least-popular candidate being redistributed to the second-ranked person on those ballots. And so on, until either a candidate has 50 percent of the vote or there are no more votes to redistribute. Not simple, but not that complicated.

The problem in applying it retroactively to a presidential primary, though, is that normal polling — who do you want to win? — isn’t great for determining who a respondent’s second-, third- and fourth-place candidates might be. What’s more, the primary is held over the course of multiple weeks, meaning that, to model the entire primary season, you’d need robust polls taken close to the election date itself for each individual state. Prohibitive.

But we at The Washington Post are not the types to give up easily, unlike people at the New York Times, probably. So I dug around until I found a Quinnipiac University poll from January 2020 that gets us sort of close to our goal. It included first choice selections for voters but also broke out who the second choice would be for those who preferred any individual candidate. It’s not five layers of selections, but it’s a start.

Of course, we now have to layer another caveat onto this thing. After all, divvying up the second-choice candidates of supporters of, say, former congresswoman Tulsi Gabbard (Hawaii) is marred by the microscopic sample size of Gabbard supporters. If Quinnipiac only spoke to 10 Gabbard supporters (to pull a number out of a hat), that two of them chose Andrew Yang as their second pick is not statistically significant. But the nice thing about a speculative examination of a national poll is that you can just sort of shrug at things like accuracy. We’re making all of this up! Please don’t write academic papers based on the calculations below.

That said, after receiving the full breakdown of results from Quinnipiac (thank you to them), the results of the ranked-choice process were interesting and seemingly informative.

Here’s how I did it. As with a normal ranked-choice process, I dropped out the candidates with the least support after each of what ended up being eight rounds. Then I redistributed the support those candidates had earned just among the people who were still left in the race.

The result? Using this poll and this process, the 2020 Democratic nomination would have been won by … Joe Biden.

Here’s what happened in each round.

Round one: There was a three-way tie for last, with three candidates getting zero percent of the vote. It was easy to distribute that support, since anything times zero is zero.

Round two: Gabbard was next to go. The biggest beneficiary of her support was Yang (as indicated by the dashed line) but even a large percentage of a tiny number is itself a tiny number. Nothing changed much.

Round three: Executive Tom Steyer is out. The candidate who benefited the most was Biden. But, again, he didn’t have much to redistribute.

Round four: Yang loses the first of as many as two ranked-vote contests. Again, Biden is the biggest beneficiary. But the most interesting shift is that former South Bend, Ind., mayor Pete Buttigieg got the second most support from Yang voters, allowing Buttigieg to almost catch his sort-of nemesis Sen. Amy Klobuchar (D-Minn.). Almost. Not quite.

Round five: So Buttigieg was next to go. By now, the amount of support being redistributed is fairly significant. Once again, Biden is the biggest beneficiary but, this time, Sen. Elizabeth Warren (D-Mass.) picks up more support than Sen. Bernie Sanders (I-Vt.), gaining a point on her progressive rival.

Round six: Klobuchar gained a bunch of support when Buttigieg was eliminated but it wasn’t enough to allow her to catch up to the person she trailed, former New York mayor Mike Bloomberg. (Remember when he ran for president? He spent the annual gross domestic product of Brazil or thereabouts.) Who benefited the most? Biden.

Round seven: Bloomberg is out. Biden gains. But, as with the two prior rounds, Warren picks up more support than Sanders. In rounds five through seven, she gained 7.4 points to Sanders’s 3.5 — more than twice as much.

See the pattern here? Sanders had a base of support but was often not the preferred second pick of Democratic voters — or candidates. That latter factor is why a number of candidates dropped out of the real race to back Biden before Super Tuesday. Biden, meanwhile, was a common second choice for Democratic voters.

Round eight: Warren is out. She has a lot of support to redistribute and most of it goes to Sanders. But it’s too late. The fraction that goes to Biden puts him over 50 percent.

Reader, this is not scientific. It uses one poll with incomplete data and layers on a metric ton of unjustifiable assumptions, like a chef at greasy-spoon diner coating everything with a quarter-inch of salt.

But someone asked if it could be done and so I did it. And we now know that Joe Biden will be the 2020 Democratic nominee for president. My prediction is that he flips five states, winning the presidency by millions of votes.

Adblock test (Why?)

"had" - Google News
June 23, 2021 at 05:21AM
https://ift.tt/2TWSag8

What if the Democrats had used ranked-choice voting in 2020? - The Washington Post
"had" - Google News
https://ift.tt/2KUBsq7
https://ift.tt/3c5pd6c

Bagikan Berita Ini