The following questions were sent to Equal Vote Coalition by Kathay Feng, head of Redistricting and Representation at Common Cause. Common Cause has been looking at a host of alternative voting methods and has been active in campaigns in multiple states and localities. The Equal Vote Coalition looks forward to meeting with Common Cause to discuss these answers in depth, go over sources and citation, and work together for better voting and equitable representation.
What problem(s) do you think STAR solves for?
STAR Voting solves a huge variety of problems, but many of them can be traced back to one specific root cause: STAR Voting directly addresses vote-splitting. The simplest example of vote-splitting is under our current Choose-One voting when a majority coalition of similar voters have to decide between two or more similar candidates (because they’re only allowed to support one at a time). Often, this leads to neither of those two similar candidates receiving enough votes to beat a third candidate who was actually opposed by a majority of the voters.
Vote-splitting causes many other problems, including the spoiler effect, where an additional candidate can “steal” votes away from a majority supported candidate, which ultimately can result in a candidate winning who was not representative of the voters. At the end of the day, when voters are only able to support a single candidate at a time, the toxic idea that we can only advocate for one group or one set of ideas at a time is strongly reinforced, and that leads us toward an entrenched two-party system that is not representative of how the people really feel.
For more information on how vote-splitting leads to serious inequities, biases, and barriers to entry in politics and government see http://equal.vote/problem.
Do you think that STAR would assist in situations where there were multiple candidates of color – take Boston Mayoral, for example? Or would it serve to highlight racially polarized voting?
STAR Voting would perform spectacularly in situations with multiple candidates of color! This is the great part about voting methods like STAR that eliminate vote-splitting — voters are able to fully (or partially) support as many candidates as they like!
To take that same concept further, STAR Voting would almost certainly decrease the apparent racial polarity in voting because voters would be able to support multiple diverse candidates. This is fundamental to voting methods that eliminate vote-splitting — they build consensus by highlighting what we agree on rather than how we disagree.
What happens if there is a candidate who gets the overwhelming number of votes (let’s say 70%) but all 2’s, and another who gets a much smaller percentage of votes (say 30%), all 5’s. I think STAR would give the win to the person with more intense voters. In some instances, that might be a “Bernie” win. In others, it could be a “Trump” win. How would you work through that scenario? If I have the numbers wrong in my scenario, stay with me on the scenario – can that happen? I think Americans are so trained to value total number of votes, this would be the case that opposition would raise.
The example you’ve given is actually fantastic for demonstrating why STAR Voting performs better than both Choose-One voting and even other great methods like standard Score Voting.
Let’s say we have an election with 100 voters and many candidates, and the 2 candidates who have the highest scores are named Ava and Bianca. Ava received 2 stars from 70 voters and 0 from the rest, so her total score is 140 stars. Bianca received 5 stars from 30 voters and 0 from the rest, so her total score is 150 stars. Under standard Score Voting, which simply elects the candidate with the highest total score, the more-polarizing Bianca would win, but STAR Voting corrects for that.
Under STAR Voting, the two highest-scoring candidates are declared finalists for the Automatic Runoff round. In the Automatic Runoff, each ballot is counted as one vote for the finalist who that voter preferred. The finalist preferred by more voters is elected.
In this election, 70 voters scored Ava higher than Bianca while 30 voters scored Bianca higher than Ava. In this case, the less-polarizing Ava would be elected at 70 votes to 30 votes, despite having a lower score than Bianca. This is by design to avoid the scenario you’re concerned about.
I’ll note that this election in reality is highly unlikely as most voters will give varying and mixed support across the candidate field, but the important part to realize is that every voter gets exactly one full vote in the runoff regardless of the scores they’ve given on their ballot. STAR Voting elects the majority preferred winner between the two highest scoring candidates, measuring both quality of support (in the scoring round) and quantity of supporters (in the runoff.)
Have you run numbers on how this would play out in jurisdictions that have one or more Majority-minority districts? Would you move to at-large with STAR or keep STAR with districts? How would each play out?
STAR Voting is flexible and can be used in at-large elections in a multi-winner bloc form. However, the Equal Vote Coalition has found through study that at-large elections inherently hurt voters in majority-minority districts within a larger jurisdiction regardless of the voting method used. This is because the overall electorate is able to “overpower” the pockets of minority voters who would have been able to win representation in a smaller district. This fundamental issue is why at-large elections were banned in the Voting Rights Act.
For these reasons, we recommend getting rid of at-large elections in most cases and switching to accurate, equitable, and representative voting methods in single-winner or multi-member districts. Shifting to a proportional voting method like STAR-PR is another option worth considering. The pros and cons for these options depends in large part on to the population demographics and also the style of the election (single-winner, at-large, proportional, etc.) more than the voting method itself, though generally STAR Voting will perform better than other methods, especially our current Choose-One voting method.
You can read more about these options and the pros and cons here: https://www.starvoting.us/multi_winner.
Are there some contexts where you see STAR particularly useful and appealing? (Eg: Urban areas where the distinction may not be based on party, so gradations of preference are helpful).
STAR Voting is a great option in a wide range of scenarios including both partisan and nonpartisan contexts, but STAR is especially attractive for jurisdictions where RCV and/or Approval voting are not allowed, including those with specific restrictions around majority or plurality winners, those that require top two runoffs, and in jurisdictions that would like to eliminate an election, i.e. a primary or a runoff.
For example, in the state of Maine, RCV was passed by the voters but was found to be unconstitutional because their constitution calls for a win by plurality. RCV tallies the top choice votes, finds a plurality winner, and then continues until a (sometimes different) winner is found who has a majority of remaining votes. In contrast, with STAR Voting your vote goes to the finalist you preferred, and the finalist with the most votes wins. STAR Voting is fully constitutional in states like Maine, Massachusetts, and others where RCV will require a constitutional amendment before RCV can be adopted for statewide elections.
Alternately, many other localities are subject to state laws that require two elections to be run for the sake of obtaining a majority. The key here is that there are several different types of majority, and the laws often don’t clearly specify which type it requires. No voting method can guarantee a “true” or “absolute” majority in a single election with more than two candidates, but some voting methods can guarantee weaker types of majorities that can satisfy these state election codes, and allow races to be decided in one election only. This dramatically reduces costs to taxpayers, shortens campaign cycles, and eliminates the effect of turnout biases across multiple elections.
Ranked Choice (Instant Runoff) Voting guarantees a “relative” majority of remaining non-exhausted ballots only. One study found that “In examining 96 ranked-choice voting races from across the country where additional rounds of tabulation were necessary to declare a winner ... the eventual winner failed to receive a true majority 61 percent of the time.” This can be further exacerbated by the fact that most RCV elections need to limit the number of candidates voters are able to rank for logistical reasons.
In STAR Voting, just like in a top-two runoff election, the winner will always be the candidate preferred by a majority of voters who had a preference. It's also worth noting here that STAR Voting always allows voters to weigh in on as many candidates as they like, so for example if a voter gave both finalists 5 stars, or both finalists a 0, we know that that was intentional.
STAR Voting is often already legal in most places, including many jurisdictions that prohibit RCV or Approval, and in many others it would likely have an easier legal battle because every ballot is counted as one vote, because all votes are fully counted, and because it ensures a majority preferred winner by way of an automatic top-two runoff.
STAR Voting excels at allowing voters to fully and honestly express themselves no matter how many candidates are in the race. It also strongly incentivizes voters to be expressive and show their honest preference order. For these reasons, STAR Voting is able to remain highly accurate even in competitive races with large fields of candidates where even RCV can struggle (like the recent NYC primary) or fail due to center-squeeze spoilers (Burlington, VT, 2009).
Whether we’re talking about government elections or use in private organizations, STAR Voting is an incredible choice when voters are well-informed and have nuanced opinions about many of their options. This is actually part of a larger vision of the Equal Vote Coalition to empower voters. In Oregon, for example, voters vote at home with plenty of time to fill out their ballots and state-sponsored brochures with information on all of the candidates. The culture of voting in Oregon is quite different from most of the rest of the country. Many voters will spend lots of time going over the options with their friends and family so they can give deep consideration to all of their choices. STAR Voting can help fuel this vision across the country.
The inverse question: Are there contexts where STAR would work less well?
STAR Voting covers many bases, but in the places where an electorate is particularly resistant to change and unlikely to support reform in general, Approval Voting is another option worth considering which also eliminates vote-splitting, and which is very easy to implement.
Approval Voting is comically simple. It's nothing more than changing the instructions on your ballot from “CHOOSE ONE” to “CHOOSE ONE OR MORE”. This means that instead of bubbling in next to just one candidate, you're free to bubble in as many candidates as you like, or rather all the candidates you “approve” of. Then, the tallying is exactly the same as what you’re familiar with: add up all of the votes, or all of the “approvals”, and the most approved candidate wins.
The real power of Approval Voting is its simplicity. Because it's fully compatible with existing electoral infrastructure throughout the country, the only real financial cost to shifting to Approval Voting is voter education. Additionally, Approval Voting is compatible with some state election codes that STAR Voting may not be.
We foresee Approval Voting being a fantastic choice for small rural communities that want either the most incremental change or a change that doesn’t cost them any money. Of course, Approval Voting does not satisfy any majority clause, so while it is effectively free to implement for taxpayers, it also doesn’t bring any inherent qualities that could save jurisdictions money. This is actually a huge part of the message of some activists in the space: both STAR Voting and Approval Voting are needed to accomplish the reform we need in the US. In fact, most STAR Voting supporters are also Approval Voting supporters and vice-versa.
What voting system vendors can accommodate STAR?
STAR Voting is compatible with most existing vendors, oftentimes only needing a simple software update. Open-source software code is available from STAR Elections or is easy to write from scratch. As always, election vendors do need to get all software updates certified, which does present a one time cost for the first adoption by each vendor. A number of vendors including ClearBallot and HartIntercivic have now confirmed that adding STAR Voting is possible, is something they are willing and able to do as needed, and is something they are considering adding as a default option in the future.
For non-governmental or internal party elections, a number of options exist at little or no cost. For those looking to run paper ballot elections, the STAR Elections project recommends “Remark” from Gravic because it is transparent, open-source software while offering all the bells and whistles of a professional election vendor. Free trial memberships make this an attractive option for those looking to run a professional election on little-to-no budget. https://remarksoftware.com/products/office-omr/try-remark-office-omr/
The Cast Vote Record (CVR) from Remark can then be uploaded to the STAR Voting Spreadsheet Tabulator for tallying.
For online elections, there are a number of options from star.vote for casual polls, to STAR Voting with Google Forms for a more secure, professional option, and the STAR Elections team is available to consult or assist with full scale professional elections such as the 2020 statewide primary for the Independent Party of Oregon, which used STAR in 2020.
There are a handful of case studies going over exactly how STAR Voting was implemented across several organizations and local and state political parties.
For an on-the-ground look at one implementation of scanning procedures, check out this demonstration and Q&A on how to use Remark software to scan and tally STAR Voting ballots automatically with the Democratic Party of Oregon.
Here is more information on hosting STAR Voting elections.
What system is Oregon using?
There are a number of active and ongoing campaigns for STAR Voting around the state, at the city, county and statewide levels. Last Thursday, the Portland Charter Review Commission’s subcommittee on the form of voting recommended STAR Voting to the larger commission. We are hopeful that the commission will refer STAR Voting directly to the 2022 ballot for Portland elections. STAR Voting should also be on the ballot in both Eugene, and Troutdale, OR in 2022. Several local campaigns around the country are promising as well.
Statewide, Oregon still uses the same old Choose-One voting that almost every state in the union uses, but bills to adopt or authorize STAR Voting at the statewide level were introduced in both Oregon (by a Democratic Legislator) and Utah (by a Republican legislator) this year, and those campaigns are ongoing.
What are the critiques of STAR Voting?
The most common and most valid critique of STAR Voting is that it is new and has not been used yet for municipal governmental elections, though STAR Voting has been used at the statewide level in the Independent Party of Oregon 2020 primary, and was used to elect all of Oregon’s presidential delegates to the Democratic National Convention in 2020 despite its newness.
STAR Voting was invented in 2014 and in that time has come a very long way. We believe the reason STAR has developed the traction it has is that it goes further to deliver on the goals of the movement, while addressing legitimate concerns with previous proposals.
No voting method is perfect (which has been mathematically proven), but STAR Voting really is the ultimate culmination of modern voting science and the data is compelling, with new research coming in regularly that further supports the findings that compelled STAR Voting into the spotlight in the first place. STAR is designed to balance and maximize the pillars of a good voting method with a specific focus on Expression, Honesty, Simplicity, Accuracy, and Equality.
Specifically, STAR Voting tops the charts in statistical analysis comparing voting methods for both accuracy and combatting strategic incentives. While some, including FairVote, have raised the concern that STAR Voting does not pass the Later No Harm criterion, (which guarantees that a voter can always support other candidates without any chance that could hurt their favorite.) The key here is that it is impossible to pass this criterion while also ensuring that it's safe to vote for your favorite themself! The “No Favorite Betrayal” and “Later No Harm” criteria are mutually exclusive. Passing the Later No Harm criterion also requires that a voting method not count data which could be relevant for a better, more consensus outcome. Additionally, voting methods which pass Later No Harm are unable to also eliminate vote-splitting and can incentivize the same lesser-evil voting that we see in the current system.
STAR Voting does ensure that in practice it is safe to vote for your favorite and also strongly incentives voters to show their honest preference order and level of support. In fact, every study and simulation to date shows that STAR Voting is more resilient to strategic voting and goes further to empower voters to vote their conscience than RCV. Studies are also clear that even if voters are strategic, STAR voting still outperforms RCV and gets more representative outcomes.
If you have specific concerns we haven't addressed here, please let us know and we'd be happy to address it in depth. That being said, we strongly believe in the robustness of STAR Voting and actively invite others to study it deeply. The further down the rabbit hole you go, the more you’ll likely find that STAR Voting stands up to scrutiny well.
We would like to add a quick note about Ranked Choice (Instant Runoff) Voting. One of our main concerns with RCV is that it doesn’t solve the fundamental problem of vote-splitting. In each distinct runoff round, voters can only express support for one candidate at a time. This leads to RCV suffering from the same vote-splitting as Choose-One voting. RCV is, quite literally, iterated Choose-One voting. This is compounded by the facts that not all rankings in RCV are ever counted, that voters who have their favorites eliminated in a later round will often not have their next choice counted, and that the “majority” which triggers the end of the counting may not be the strongest or largest majority. In elections where voters can support multiple candidates, it’s possible and common for there to be more than one majority-supported winner.
We don’t want to overload you with information at this time, but if you’re interested in an introductory comparison between RCV and STAR Voting, check out these two pages:
By Sara Wolk and Sass