The DNC debate qualifying rules made one huge mistake

[cos]

Democratic candidates for president this year had to qualify for the official debates based on a set of simple, neutral rules published in advance - standards that got progressively a bit harder to qualify for with each debate. Each candidate had to get a certain percentage of support in a certain number of qualifying polls, and a certain number of donors. For the first debate, it was 1% in three polls or65,000 donors, and with each debate the standard got a bit higher.

Overall a pretty clever system, and a good way to solve the problem of choosing who the important candidates are for people to see, without bias or too much public perception that it's rigged in favor of or against particular candidates because of their views. But I think they made a huge mistake with the rule about polls!

Especially in the beginning, this summer when the debates began, lots of potential voters were undecided and most of them were considering most of the candidates. Even after the first few debates, there were still a lot of undecided voters, and most voters who did have a favorite were still considering others. The point of rules like this should be to show you the candidates you may still be considering, not the candidates you or someone else have already chosen. With these rules, it was quite possible for a candidate who nearly everyone was still seriously considering to be exluded from the debates, while someone else (such as Tulsi Gabbard) who very few people were considering, and who most voters had already decided against, would still be included.

What the DNC should have done is announced in advance that they would *only* consider polls that ask who you're seriously considering voting for. They could've set much higher thresholds, starting with 10% rather than 1%. Debates would then have emphasized showing us all the candidates people actually want to learn more about.

There weren't very many polls like that, but you can bet that if the DNC had announced a rule like that in late spring, there would have been plenty more. And the DNC could have commissioned a few national polls themselves to add to the mix.

Comments

[elusiveat]

This is an interesting thought.

I noticed that somewhere along the line FiveThirtyEight shifted to talking much more about which candidates were under consideration, which made a lot more sense to me than just looking at who everybody's first choice was.

I'm a little concerned about the part on setting a threshold more like 10%. That essentially cuts out candidates who don't already have pretty high name recognition.

Going into the first debate, less than 50% of likely Democratic voters had even formed a favorable-unfavorable opinion about Klobuchar, Castro, de Blasio, Ryan, Gabbard, Inslee, Delaney, Yang, Gillibrand, Swalwell, Hickenlooper, Bennet, or Williamson (https://projects.fivethirtyeight.com/democratic-debate-poll/). I think it was correct to take the candidacy of people like Inslee, Hickenlooper, Bennet, and Swalwell seriously at that point, but with only 20-30% of Democrats having *any* opinion about them at that stage, I'm not sure how easy it would be to meet the 10%-seriously-considering threshold.

What if the metric were something more like >10% (or perhaps higher still... could we go as high as 20%?) for people who could demonstrate familiarity with each candidate?

I also think you've overestimated how many Democrats had already decided against voting for Gabbard by the start of the summer. The only candidate with *really* bad favorables going into the first debate was de Blasio.

[barking_iguana]

I think for the first couple of debates, it made sense to include candidates who had a small but enthusiastic and not tiny base of supporters. Even if most folks had barely heard of the candidate.

I think the reason for that gets weaker as time goes by, because more voters will have been exposed, at least to the candidates who made the previous debates.

[netmouse]

I don’t know about in general, But I was certainly turned off by Gabbard really early.

[netmouse]

I still think we need a national platform that gives all candidates air time to let people know what they are about.

PBS. and NPR would logically do this.

As it is, I feel I have gotten the best impressions of candidates by reading policy proposals, watching speeches and, especially, watching interviews on “late night” tv. Colbert and Trevor Noah are both excellent interviewers.

[elusiveat]

I know a lot of people feel that way. And even more people dislike her *now*.

I also think that it's important to use polls as a reality check on whether our own preferences are reflected in the views of other voters, and in this case I'm not seeing the evidence. If someone has statistics to show otherwise, I'm open to being convinced.