DC Initiative 83

Students are asking, so I feel obligated to compile my thoughts. Voting on this begins soon.

I-83 calls for open-to-unaffiliated-voters primaries (with RCV) and then RCV (Alternative Vote) for all public offices and special elections. I am told that the ‘bottom-up’ variant would be used for elections to at-large seats.

I-83 would not set up jungle primaries. It would not set up decisive-round RCV elections with multiple candidates carrying the same (or no) party endorsement.1 I have written a bit about that configuration. That is not what’s on the ballot.

There are real concerns about ballot-invaliding error and voter confusion (which probably are correlated but conceptually distinct). My guess is that the confusion issue would apply more in primary than general elections because, in the primary, candidates are undifferentiated by party.2

Yet I also have these Australian data in mind, from a variant in which not more than one candidate can have a party’s endorsement. No problem there with muddy cues.3 Why the high error rate? One explanation is refusal to go along with compulsory-ranking instructions (see page). However, this article points to confusion issues in what seem to be immigrant-heavy areas.

I am reading the Wikipedia page on Initiative 83. It currently claims: “Independent voters are functionally disenfranchised by not being allowed to participate in the primary that chooses the Democratic Party’s mayoral candidate and other important races.” I find this argument unpersuasive because those who don’t want a Democrat could come together behind someone else. However, I see how this argument might be compelling in the moment.

As for the open-to-unaffiliated primary, I go back and forth. The research I can think of offhand concerns NOMINATE scores in Congress. I don’t think we know enough about this institution as (it might be) used in local elections. Feel free to send me a paper on it if you know if one.

The official position of the local Democratic Party is that RCV should not have been combined with what it is calling “semi-closed primaries”4 because those could lead to candidates who do not “align” with the party’s “core values.” This paragraph exists for information purposes. I have no insider knowledge.

That is all. Good luck.

  1. Technically >M candidates, due to the two at-large seats.
  2. Here is my favorite paper on this so far using U.S. data. Here is another with experiments, focused on who wins. Here was my contribution to the confusion debate. Here is a paper on error in the 2021 NYC primaries.
  3. Yet this page depicts a real ballot with 22 candidates and numerous independents — not fun if all must be ranked, which would not apply in DC.
  4. This term is more meaningful in my view than “open primary.”

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.