In an earlier article, I pointed out several issues I have with the Ranked Choice Voting (RCV) process, specifically transparency and fairness.

I also had a healthy skepticism about the motives for those pushing RCV.  I think I figured out the end game.  (That’s a tease to read to the end.)

Maine 2nd Congressional Ballt 2018

First, let me present RCV in the context of the Maine 2nd Congressional race that was decided using RCV in 2018.

Maine Congressional District 2, 2018 Election

The 2nd Maine district was decided by RCV “logic” in 2018.  The incumbent, Bruce Poliquin, “won” the 1st “round” of voting 46.3% to 45.6% over his key challenger Jared Golden.  Lacking a majority (>50%), RCV kicked in, meaning the ballots for the 2 independents (Tiffany Bond and William Hoar) who polled at the bottom of the 1st round were examined for their 2nd choices.

Jared Golden overwhelmingly won the 2nd place votes from Bond & Hoar getting 68.7% of those votes.  Golden won the election getting 50.62% of the “votes” versus 49.38% for Poliquin.

Let’s dive into the election results.

ME 2nd Congressional District RCV Results, 2018

ME 2nd Congressional District Results, 2018

Results from the Maine 2nd, 2018

Above are the election results taken directly from the Maine Bureau of Elections.  It’s the spreadsheet from the “Certified Results” link.

The proponents of RCV in Maine state:

Ranked Choice Voting is simple, fair and easy. Voters have the freedom to rank candidates in order of individual preference. Ballots are counted in rounds where last-place candidates lose until one candidate reaches a majority and wins.

Let’s see if experience says it’s simple, fair, and easy.

Transparent

First of all, read the 2nd sentence in the above quote.  Understand what it says?  If you do, my hat is off to you.  It’s gobblygook.  If the proponents can’t readily explain it, how simple and transparent can it be?

Now look at the above spreadsheet from the Maine Election Bureau.  Understand it?  Again, if you do, I bow to your preeminent intelligence.  I played with the spreadsheet for about an hour, and I pretty much figured out the math.

However, I didn’t know the definitions for the 3 categories of “Exhausted Ballots” or how they are applied.  I had educated guesses.  A definition guide is not immediately apparent, but a search on the Bureau’s website for “Exhausted Ballots” got me this 7-page guide.

I went to that guide to get definitions for the types of Exhausted Ballots:

  • Overvote
  • Undervote
  • Exhausted Choices

Stunningly, none of those are in the Definitions section of the document!  Overvote is explained in context.  However, Undervote and Exhausted Choices are terms never defined nor used in the document.  So, how did the tabulators know how to determine if ballots fit into those categories?  Even the state law that created RCV does not define or use the term Undervote.

Is this transparent?

~~~

[Side Note: If you understand the RCV rules, you might be wondering why there isn’t a separate round for each eliminated candidate.  Maine uses a “batch elimination.”  Hoars vote was 2.37%.  Apportioning 2nd place votes for those ballots would not put anyone over 50%.  So, both Hoar and Bond were eliminated together.

That’s explained in the above referenced document, but boy, it’s not transparent.  Just trying to explain all this here is difficult!]

Simplicity, Ease, and Fairness

I’m guessing an Undervote is where no candidate was selected by the voter for a given ranking.  Since it’s not defined, that’s a guess.  The Overvotes are where a voter selected more than 1 candidate in a ranking.

The best way to determine the ease of casting RCV votes is to look at the percentage of ballots disqualified for this office.  The nearby table summarizes it for the 2 rounds.

RCV Ballot Disqualifications Maine 2nd, 2018

Ballot Disqualifications Maine 2nd, 2018

My table looks a little different than the previous spreadsheet.  Round 2 focused on the 23,427 votes “Transferred” into Round 2 from ballots cast for Bond and Hoar in Round 1: 16, 552 for Hoar and 6875 for Bond.

When we’re looking at disqualified ballots, 23,427 is the relevant denominator for Round 2.  (I think – again, it’s not transparent…)  The state’s spreadsheet summarize Exhausted Ballots for the rounds combined.

In Round 1, 2.18% of the 296,077 ballots were “Exhausted” due to voter mistakes.

  • 435 people made multiple 1st choice selections (Overvote)
  • 6018 people didn’t cast a 1st choice (Undervote)

That level of disqualification doesn’t seem too bad – a bunch of people just didn’t bother to vote in this election.

But look at Round 235.2% of the 23,427 ballots “Transferred” (or 8253 ballots) were disqualified from Round 2 voting, mostly due to Undervotes.

  • 98 people made multiple 2nd choice selections (Overvote)
  • 7820 people didn’t cast a 2nd choice (Undervote)
  • 335 people voted for Hoar or Bond as a 2nd choice having voted for the other as a 1st choice (Exhausted Choices)

Simple & Easy?

A full third of the ballots that entered the 2nd Round were disqualified.

To me that’s a pretty stunning percentage and throws into doubt simplicity and ease. 

Most of those people (7820) self-disenfranchised from casting a relevant vote.  Many probably did understand what they were doing.  Is that fair?

Note: the Golden’s margin of victory was 3509 votes.  And 8253 ballots were not used in the 2nd round that resulted in Golden’s victory.  Again, fairness?

The Strategy for Winning RCV Elections

I was scanning the votes cast ballot by ballot in the election.  (You’ll see those spreadsheets on the website.)  I was thinking, “What’s the end game here?  Golden won because he won 2/3rd of the 2nd round valid votes.  How does this happen?”

Logically, the 2 independent candidates, Hoar & Bond, were closer ideologically to Golden than Poliquin.  Presumably in a traditional runoff election, those voters would have voted for Golden. But that presumes they would bother to vote in a runoff.

That’s when I figured out the strategy for winning elections in an RCV environment.

Pre-RCV, a candidate would not want 3rd party candidates who might draw votes away.  For example, Perot probably drew more votes from Bush than Clinton electing Clinton in 1992.  Nader probably drew more votes from Gore than Bush in 2000. Et cetera.

With RCV, the incentive is EXACTLY THE OPPOSITE.

You want fringe candidates on your side of the political aisle.

Those fringe candidates will draw their diehard voters to the polls, and hopefully they’ll choose you as a 2nd choice.  The fringe candidates will lose in the 1st round and now you get their votes.

With RCV fringe candidates are a get-out-the-vote strategy. 

This is especially valuable in off-year elections where voter enthusiasm is lower.

Brilliant!

Predictions on the RCV impact on Election Strategy

Strategic voting may be reduced with RCV (as claimed by RCV supporters), but we now will see a wholesale change in election strategy, not by political parties, but by groups on both ends of the ideological spectrum.

We can expect to see:

  • A flurry of fringe candidates in RCV elections driven by ideological groups — with a wink-wink by the major parties once they are educated about the end game.
  • Electioneering geared toward the fringe-candidate voters on how to cast their ballots done by the PACS affiliated with the ideological groups.

An Example of Winning an RCV Election

Let’s take an extreme, but possible example to show how the strategy might work.  Imagine there was a special election for a local state office.  Special elections are notorious low-turnout events.

The district has 5 towns.  Party 1 has one solid candidate with a following in several towns..  Party 2 has no favorite but has very popular potential candidates in each town.

Normally Party 2 would try to get four of the candidates to drop out and support the most electable candidate.  But with RCV, Party 2 would want all 5 candidates to stay in the race with 4 as indepedents.  The 5 candidates would increase the overall turnout for the election because people would get out to vote for their neighbor, friend or colleague.

If the voters for Party 2’s candidates have some ideological alignment with the party, then they would vote for the other Party 2’s candidates for their 2nd, 3rd, etc. choices.  One of those 5 would then be in a far better chance to win with RCV.

Brilliant!