[EL] Rick Pildes Guest Post

[Pildes, Rick]

I’m open to persuasion on MMDs with RCV for Congress, I just have various concerns I am still unresolved about.

Steve says voters only need to be able to rank 3 candidates, which political science tells us is probably the limit for informed voting.  But notice this is in tension with one of the major arguments for RCV, which is that it ensures candidates are not elected without majority support.  If a region is electing 5 members to Congress, and most voters only rank 3, some candidates will be elected without majority vote.

I agree the Australia Senate is our best practice experience with this.  But members to the lower chamber there are elected from traditional single-member districts.  That means residents do have a local representative they know is “theirs.”  Particularly in the US, it is less well-off residents who need a congressional office they can to for help navigating the government bureaucracy, to help with their concerns/needs about benefit programs of various sorts and other services.  If we talk about MMDs for the US House, we are talking about 5 representatives for a region of 3.5 M people, in a place like NC.  In regions with a major urban center, I wonder whether most of those elected will come from that center (and exacerbate the rural/urban divide that is already among the major issues in many democracies today).  I’d be curious to learn if that’s been an issue in certain states in Australia.  Even if that’s not a problem, though, I wonder if voters will actually feel they have a representative they can turn to, because they voted for that person, even though that representative lives 400 miles away.  How many local offices will these members of Congress have to open across 1/3rd of NC to be able to provide these direct connections to constituents?  Where are they going to get the resources to staff these offices, when those who voted for them might be scattered across a third of the State?

Lastly, on Mark’s question of whether we could use this system to get rid of primaries:  if we think about that concretely, with 5 seats to fill, the two major parties will likely field 5 candidates each, bringing us to 10.  Impossible to know how many primary challenges from within the major parties, but let’s say 2 in each party, so we are up to 14.  Of course, one of the arguments for MMDs with RCV is also that it will allow other parties to flourish, so let’s imagine 5 candidates from other parties.  We are now talking about an election with nearly 20 candidates, with voters willing and able to rank maybe 3 of them.  If you want to retain primary elections and fold them into a single election day, that’s the scenario I see.

I’m sure Rob Riche will have thorough answers for all this, including explaining why it’s a good thing that a candidate from the “Motoring Enthusiast Party” got one of six seats in the Australian Senate from Victoria, even though only 0.5% of voters ranked him as their first choice candidate.☺

I’ll read any responses with interest but that’s it from me on this issue for a while, since work beckons.

Best,
Rick

Richard H. Pildes
Sudler Family Professor of Constitutional Law
NYU School of Law
40 Washington Sq. So.
NYC, NY 10012
212 998-6377

From: Steven John Mulroy (smulroy) [mailto:smulroy at memphis.edu]
Sent: Friday, July 5, 2019 12:10 AM
To: Rob Richie <rr at fairvote.org>
Cc: Pildes, Rick <rick.pildes at nyu.edu>; law-election at department-lists.uci.edu
Subject: Re: [EL] Rick Pildes Guest Post

I think Rob Richie has it right.  As to Rick Pildes' concerns, let me respond.
"My concerns are that this would raise the cost of elections; demand more of voters regarding information than is reasonable to expect; would leave voters without feeling they had a representative who was "their" person."

1. Election costs would not go up as much as you might think. A candidate in such a multimember district would only need 17% to 25% of the vote, depending on the size of the multimember district. She would not need to do advertising and outreach to the entire district, but instead could target the geographic or demographic subsets that would work best for her.  In an age of micro targeting, targeted direct mail, and social media, this would be eminently doable.

2.  Voters in such districts would only have to gather enough info to weed out obvious disliked candidates (usually pretty easy to do) and then decide on their 1sr, 2nd, and 3rd choice. Most voters do not rank more than 3 even in the single winner RCV elections which Rick (I'm glad to hear) likes. They can use endorsers, party labels, media endorsements, and other heuristics to do the weeding and the ranking just as they do nowadays. And you can choose to rank as few or as many as you're comfortable with.

3. Actually, there would be MORE voters who felt like they had a rep who was "their person." Under single member district winner take all, usually 40% or so in each district end up feeling like "I didn't vote for THAT guy- he doesnt represent ME." Usually that same  40% feels that way time and again, leading to alienation. Under Proportional Representation, almost all voters can point to at least one representative and say "I voted for that person- she represents ME."

Note that none of these concerns seemed fatal in Cambridge Mass or Minneapolis, where multimember district proportional representation has been used for decades. Ditto Australia, which has multimember PR Senate districts just as big as those in Rick's NC hypothetical.

" But none of these changes are going to happen, certainly not for the 2020 round. "  True only as long as we all tell ourselves that. And  they said the same in Cambridge Mass and Minneapolis too.

"So for that, I still will focus on IRCs."
IRCs are certainly part of the solution. We should focus on them. Just not exclusively.

On Thu, Jul 4, 2019 at 4:28 PM Pildes, Rick <rick.pildes at nyu.edu<mailto:rick.pildes at nyu.edu>> wrote:
I support ranked-choice voting in certain kinds of elections, particularly for a single officeholder (Ned Foley's idea that States should use it in presidential elections is a good one).

I have been more skeptical about using MMDs and RCV when electing members of Congress.  Take a State like NC, which might be carved up into three regions for MMDs; the regions would elect 4, 4, and 5 candidates.  So we would have 8-10 candidates running in a general election, across a third of the State, with voters having to rank them.  My concerns are that this would raise the cost of elections; demand more of voters regarding information than is reasonable to expect; would leave voters without feeling they had a representative who was "their" person.  I think a better approach, if we are going to think outside the box this far, is the hybrid German system, in which voter elect individual representatives from districts but there is also a party-based vote, and the overall representation reflects the correct PR outcome.

But none of these changes are going to happen, certainly not for the 2020 round.  So for that, I still will focus on IRCs.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://webshare.law.ucla.edu/Listservs/law-election/attachments/20190705/ed03e471/attachment.html>