City Club of Eugene: STAR voting

On June 1st, 2018, the City Club of Eugene discussed an alternative voting system called STAR voting. First, voters Score the candidates on a 0-5 scale, Then there is an Automatic Runoff. Advocates are gathering signatures to qualify this proposal as a measure on the general election ballot in November of 2018. Panelists included STAR voting advocates Mark Frohnmayer (from the Arcimoto electric vehicle company) and Alan Zundel (retired political science professor), along with Robin Quirke (PolicyInteractive) as the friendly critic.

Listen to full City Club discussion here.

Visual representation of Quirke’s portion of the discussion:

In 2015 and 2016, using electronic mock 2016 presidential election ballots, PI invited online survey respondents to vote for their preferred candidates using the standard U.S. voting system of plurality, along with three championed alternatives: approval voting, score voting, and ranked-choice voting (RCV or IRV). Quirke focused on the following three points, based on study results:

1. Do winners vary when an alternative voting system is used? (Results and methodology)

2. Which voting systems do voters prefer? (Results and methodology)

3. How we might use research to help advise voting system reform decisions?

Each of these three points are discussed below, with bar charts and data tables to help illustrate.

1. Do winners vary when an alternative voting system is used?

Answer: Sometimes

July 2015 (IRV is the same as RCV)

January 2016 (IRV is the same as RCV)

March 2016 (IRV is the same as RCV)

November 2-8, 2016

In a survey conducted along side Center for Election Science (CES), we matched voter turnout in the 2016 general election. Of our 730 survey-takers, 33% were registered as Republicans, 36% as Democrats, and 31% for everything else.

In looking at the above table, you can see that Clinton was the winner with all voting systems and Trump was the second place winner in all accept with score voting, where Johnson was second and Trump third. How did Johnson manage to bump up to second place with score voting? That which plurality voting doesn’t capture, score voting does, explained in more detail directly below.
In the above bar graph representation of the November 2016 mock ballot survey, Trump got more than three times the best and highest score of ‘5,’ but he also got nearly three times the lowest score of ‘0’ as well. With approval voting, those zeros are not captured. See another example with Trump and Cruz in a mock primary here.

2. Which voting systems do voters prefer?

Answer: the current system (plurality voting)

•Every time we asked, a majority of respondents would like to revamp the current plurality voting system, but only 64-66%, which is not very much for reform proposals going to the voters
•And although there was majority support, there was a consistent preference to keep the current system (see more results here)

3. How we might use research to help advise voting system reform decisions?

1.Results can vary depending on which system is used, so that’s heading in the right direction, but we need to choose carefully
2.Getting support for an alternative voting system may be an uphill battle. Testing in a local election is a great start to educate voters. BUT…
  • Exit polls (page iii) in San Francisco found that younger, English-speaking, more formally educated voters were the ones who most liked RCV, so be sure to test it among those who are often left out of political decisions
  • Consider a ‘No opinion’ option to reduce cognitive load
  • Make average score available as well instead of total score (makes the results more intuitive)
  • Consider using simple score voting without the instant runoff feature (there are downsides to only having one election, and instant runoffs make the calculations more complicated)

(Pippa Norris: Electoral Integrity Project)

Quirke referenced Pippa Norris from Harvard, and the Electoral Integrity Project. This project is at: Norris’s book Why elections fail (2015)outlines how 97 countries were assessed regarding electoral integrity (using the Perceptions of Electoral Integrity (PEI) questionnaire), including the U.S., whose electoral integrity score tied with South Africa, Mexico, and Barbados (pg. 54). The U.S. scored lower than many countries that are considered developing nations. Norris writes “Experts expressed concern about problems of US electoral law [“unfair to smaller parties” as one aspect of this (pg. 39)], voter registration, and partisan gerrymandering in the process of drawing district boundaries, as well as the regulation of campaign finance” (pg. 57). The U.S. PEI score (70) is the average of individual state scores, which ranged from the lowest of Arizona (53) to the highest of Vermont (75). It is worth noting that Vermont’s highest score does not reach the scores of Norway (87), Costa Rica (84), or Lithuania (82), as well as several other countries.