Interview With President of Center for Election Science

President Center for Election Science Aaron Hamlin

The President of Center for Election Science Aaron Hamlin discusses approval voting’s many benefits for all levels of government and even internationally

Democracy Chronicles founder Adrian Tawfik conducted an exclusive interview series with an international group of election method proponents including prominent signers of the Declaration of Election-Method Reform Advocates. The best place to start off is the Democracy Chronicles introduction and then take a look at each of these interviews:

Also see the DC Interview With Creator of Wooden Models of Voting Methods with artist Peter A. Taylor. 

Democracy Chronicles coverage of every type of election reform ranges from clever election campaign finance proposals to redesigned layout for election ballots.  But one type of election reform that is storming into discussions across the political divide is in regard to election methodology.  Leading the charge are the vocal advocates of Approval voting including long-time Democracy Chronicles writer Michael Ossipoff. Approval voting advocates are in the midst of the biggest concerted push for election method reform ever seen in the United States notching some impressive successes for Approval voting in Arizona and even at the United Nations.

The organization behind much of the recent push for Approval voting is the Center for Election Science in San Francisco.  The Center is in the midst of a huge fundraising campaign, seen in the image here, and are successfully working to redefine the conversation on election reform.  This is what led Democracy Chronicles to invite the President and Co-founder of the Center for Election Science, Aaron Hamlin, to answer a few questions for our readers. The result is our interview below…

Adrian: Thank you very much for agreeing to answer my questions about your work.  First question: what is the state of Approval voting in the world today?

Aaron: Firstly, Approval Voting is a single-winner voting method. Voters pick as many candidates as they want (no ranking). The candidate with the most votes wins. It’s the same as what the U.S. currently uses, the vote-for-one method, Plurality Voting, except you’re not limited to choosing one candidate. Approval becomes useful when more than two candidates run.

Approval Voting  is relatively young (in voting method years). Instant Runoff Voting (IRV), for instance, was invented around 1870. But no one used it for another 50 years.

Approval Voting, on the other hand, was first described in modern times by Guy Ottewell and Robert Weber in 1976. Steven Brams along with Peter Fishburn published it more formally in 1978. But I don’t expect Approval Voting to incubate the same 50 years IRV did. And that shouldn’t be surprising. Today information travels as quickly as electrons can move. Moreover, I suspect Approval’s simplicity (and performance) will cause it to spread quickly once places start trying it out.

Many large organizations already use Approval Voting in some form. This includes the National Academy of Sciences, Mathematical Association of America, American Mathematical Society, Institute for Operations Research and Management Sciences, American Statistical Association, and the Institute of Electrical and Electronics Engineers. Smaller groups include the Society for Judgment and Decision Making, the Social Choice and Welfare Society, the Public choice Society, and many others. Collectively, that’s over half a million people experiencing Approval Voting.

Those are sophisticated groups. They easily could have chosen more complicated methods than Approval. But they didn’t. I think that says something positive about Approval Voting’s effectiveness. Naturally, you have local student groups using it, too: Colorado University’s Student Government, San Francisco State University’s Academic Senate, and Dartmouth’s Student Assembly all use Approval Voting.

Approval Voting has technically been used with voting referenda where there are three or more referenda versions, and the referendum with the highest approval percentage wins. It’s also used to elect the Secretary General of the United Nations. And minor parties like the German Pirate Party, Texas Libertarian Party, and the National Reform Party have started to use Approval Voting internally.

So far Approval Voting hasn’t been used to elect a government official. But it’s only a matter of time there. We’ve seen Approval Voting bills introduced in a number of states including North Dakota, New Hampshire, Colorado, and recently in Arizona where it just passed a vote on the House floor. The Arizona bill would let localities use the method.

Adrian: How did you first get exposed to/ interest in election method reform?

Aaron: My personal interest in voting methods came, unsurprisingly, during an election season. I was an odd voter—I always voted my honest favorites, even if they showed no hope of winning.

At the time, I was in a student group among fellow grad students. It bothered me that my classmates were intending to vote for the very people who opposed the reforms we advocated. And they wrote off candidates that actually supported these reforms. They were “unelectable”. This really got to me. A lot.

Initially I was just angry. I began to think I was wasting my time with this group. And I wondered who was going to vote for these people if my own classmates wouldn’t? But I got past that. I appreciated that my classmates were smart people, and they were just conforming their behavior to what they had to work with. It’s not easy to vote for a favorite who is (in all reality) going to lose. At least it’s not easy under Plurality Voting.

But that troubling experience never left me. It told me there was something fundamentally wrong with the way we vote. So that discomfort motivated me to spend an inordinate amount of time studying voting methods (Gaming the Vote is highly recommended). And along the way I met with some hyper-intelligent, motivated individuals who shared my strife (some of whom now comprise our board). By this time, I was starting my law degree, so I adapted my program to gain the skills needed to build a nonprofit. The Center for Election Science was born shortly thereafter.

Adrian: What is the current status of your fundraising campaign and video project?

Aaron: We’re ambitiously trying to fund the Approval Voting video in just four weeks. This is necessary to get into the Looking@Democracy contest. That’s a contest promoting ideas on bettering government. It’s funded through the MacArthur Foundation. This contest will give us more coverage and a shot at further funding.

We started off well hitting $1,000 in our first 24 hours. Now we’re in our final week. We’ve been getting over 150 views a day on our Indiegogo page. Our promotional video has over 850 views. At our peak, we were the 4th most popular active campaign of over 4,000; and Indiegogo featured us on their homepage.

There’s even some added excitement. As our supporters know, we interviewed Dr. Kenneth Arrow, a Nobel laureate for his Impossibility Theorem. We spoke to Dr. Arrow again recently and he agreed to sign a reprint of the journal where he first published his Impossibility Theorem. We’re offering this as a perk for $3,500 (one of his signed rare books has gone for as much as $5,200). If no one takes the perk, we’re giving it to the person with the most in public donations. Right now that number is $800, which would be quite a deal.

Adrian: Where has your campaign received the most support?  Are there supporters who have surprised you?

Aaron: We have an online presence that’s stemmed from a discussion group. A lot of those individuals have donated. People from technical fields like engineering and computer programming tend to be attracted to our work. Consequently, many who’ve donated have jobs that afford them the ability to be rather generous. Before our audience started expanding, our average donation was over $100. So we have some rather impressive core supporters.

We also have growing support from our Facebook and Twitter pages. Those started building from discussion group members and friends of the board. But it’s been expanding, which we know is essential. It’s good to see people outside technical fields appreciating the importance of this reform. And it really is a simple reform. So there’s no good reason to be out of the loop on this one.  It’s clear that this issue is striking people. Many are donating $100, $200, and even $500 at a time once that light bulb hits.

Our supporters make us proud to do what we do. And we’re excited to deliver for them.

Adrian: What impact could we expect from Approval voting reform for federal offices like President and Congress?

Aaron: Huge impact. Admittedly, however, reforming the election of US President will be quite difficult because of the Electoral College. There are approaches that use interstate compacts which could get around the barrier. That’s the current strategy with the national popular vote bill (which, if successful, would still just be Plurality). But the interstate compact approach can be adapted for Approval Voting.

Competition is (if you will) the Miracle Grow for democracy. And Approval Voting brings it in spades. This competition can occur because Approval Voting more accurately reflects all candidates’ support, win or lose. Approval Voting illuminates these candidates’ support because it always lets you vote your honest favorites—always, regardless of viability. Consequently, candidates with legitimate ideas can’t be marginalized.

Partisan gridlock forever?

But that competition isn’t what happens now. Now you either operate within the status quo of the major parties or else you’re marginalized. It’s very difficult to challenge convention under Plurality Voting. And that’s precisely because it doesn’t let you vote honestly.

When you’re talking about something new, then—by definition—it needs to build. But new ideas can’t build when its supporters foresee Plurality Voting shoving their ballots in the trash. And this distorts those ideas’ support. Voters then find themselves in a political marketplace with few real choices.

You see how easily the media pushes aside minor parties and independents when Plurality (including Plurality polling) gives them artificially low support. Sadly, even candidates from major parties aren’t off limits. If their ideas are too novel, Plurality will nail them to the floor in the primaries. Crazily, this happens even when the ideas themselves poll as popular.

When I spoke with Dr. Arrow in an interview, he recognized this issue with Plurality, too. He called Plurality “stifling” and not something we want as our voting system. I couldn’t agree more.

Adrian: What about more local offices?

Aaron: Local offices are the starting ground for election reform. Historically, this is how it’s been done. So this is where our sights are. We’re ready to see the first Approval Voting-elected mayor. From more competition to flat-out better winners, Approval Voting offers the same benefits  at the local level.

Adrian: There are multiple election methods that have at least some support, how does Approval compare to the most popular alternatives?

Aaron: For simplicity, I’ll limit myself to single-winner methods. There’s a similar method called Score Voting. You score candidates on a scale and the candidate with the highest score wins. It actually performs a little better than Approval when there’s a mix of honest and strategic voters. We’d support Score Voting first if Approval Voting wasn’t so simple. And the gains of moving from Plurality to Approval are just enormous, much more than between Approval and Score Voting.

The two other reforms you see are Plurality with a runoff and IRV. Neither addresses the issues that Approval does. Both can still punish you when you vote your favorite, and both can behave oddly. For instance, with a runoff you can have situations where had you gotten more support from certain voters, then that would cause you to lose; and getting less support from some voters can cause you to win. The 1991 Louisiana gubernatorial election misbehaved this way.

IRV does the same with rankings in unacceptable frequency. Just ask the voters from Burlington, VT’s 2009 mayoral election. Worse, that same election punished conservative voters for voting honestly. Had they insincerely ranked the Democrat first, they could have avoided the Progressive candidate winning (their least favorite). To top it off, both these methods are more complicated than Approval Voting.

To give IRV some credit (the little it deserves), it mitigates the spoiler effect for cases where the spoiler candidate has few votes. But when that candidate becomes more competitive, the spoiler effect comes in with full force.

Some IRV advocates complain that Approval treats all votes the same, so a less preferred candidate has the same weight as one that’s more preferred. What this comes down to, however, is whether you’d prefer this scenario or a more complex ranking method where you can’t even vote your honest favorite. We’re not claiming Approval is flawless. But when it shows imperfection, the utility cost to any particular voter or group is small. Runoffs with Plurality and IRV, on the other hand, give you straight-up bad winners during their episodes.

Finally, IRV proponents tend to claim they provide majority winners. But know that IRV removes many ballots, so any “majority” claimed is not referring to the whole electorate that voted. IRV also does poorly at  electing candidates that are able to beat every other candidate one-on-one. This type of candidate (when she or he exists) is called a Condorcet candidate. Approval Voting, on the other hand, is quite good at electing these Condorcet candidates.

Importantly, reformers and experts reach consensus on one issue for sure: Our current Plurality approach is the deadbeat of voting methods.”

Adrian: With new governments being actively created in places like Egypt and even South Sudan, does Approval voting deserve to be a part of the conversation?”

Aaron: Absolutely. Approval Voting is especially important in polarized environments like the ones you mention. One of Approval Voting’s features is that it’s good at electing candidates with broad-based support. Contrast that to the candidates IRV and Plurality elect. Their elected winners tend to be quite polarized. To see why this is important, imagine how well someone from the middle could govern compared to a candidate entrenched within a  particular faction.

Also, large elections tend to bring many more candidates. And when you have more candidates, you need a method that avoids vote splitting and bizarre behavior. That’s especially true if any candidates are similar or overlap in ideology (inevitable). If you stick with a method like Plurality, Plurality with a runoff, or IRV, then you’re asking for trouble. You increasingly risk bizarre outcomes as more candidates jump in the ring. Such fragile voting methods make terrible tools for democracy. And that’s why Approval Voting needs to be in the conversation.

Adrian: Is there anything else you would like to mention that has not been covered?

Aaron: For whatever reason, many inappropriately brush off voting methods as complex and esoteric. Focus instead goes to other reforms like the right to vote, campaign finance, and so on. Obviously, we’re aware of these issues and more; they’re undeniably important. But we maintain the voting method is the most important. And it indirectly assists with other issues.

Imagine yourself in a world where you have all the reforms you want. But you don’t have a good voting method. You’ve failed. New candidates will still be shamed just for running. Vote splitting will happen. Poor candidates will win. And, despite claims of a brilliant democracy, voters will still be unable to vote their honest favorites (unless they want to see a spoiler effect).

We’re talking about the way we make our most important decisions. And besides a random lottery or dictator, we are quite literally using the worst voting method ever invented. This is appalling. Is it any wonder why we see representation so disparate from our electorate? Approval Voting needs to be everyone’s priority because this height of dysfunction is unacceptable, especially when the solution is so simple.


  1. says

    I hope lots of people read this article. As Adrian and Aaron suggest, Approval voting would be a great choice for use in nations where new governments are being created.

    One additional point worth mentioning. The reason why so few voters are aware of the unfairness in our current single-mark ballot is that no one collects information about what the voters really want. In other words, the use of single-mark ballots, even in public-opinion polls, leaves us without evidence that we, the majority of voters, do not like the special-interest puppets who get elected.

    • Aaron Hamlin says

      Thanks, Richard, for your support.

      You make a good point. Not only does Plurality do a poor job at electing a winner, but it also does a miserable job at estimating support through polling. And so, without the disparity between Plurality and a more accurate support metric, voters have a harder time appreciating when they get shortchanged. Voters’ evidence that something is broken comes from the dysfunctional political process they’re forced to watch. Unfortunately, however, that doesn’t explicitly highlight the voting method.

      It’s unsurprising that the polling method would be equally dysfunctional to the voting method itself. The poll is trying to predict how the Plurality election will turn out rather than how much support the candidates have. Given that, it’s a reasonable prediction that when Approval Voting does become a reality, it will be paired with Approval polling. To do Plurality polling with Approval Voting would just be nonsensical, perhaps more nonsensical than using Plurality Voting at all. Fortunately, Approval, unlike Plurality, does an excellent job at measuring what voters really want.

      I’ll offer what may be a bit of long-term foreshadowing. We’ve talked internally about using a national polling organization to compare Score, Approval, Plurality, and all ranking methods. Doing this for a presidential election would help spark voting method discussion at a national level. Don’t be surprised if we do this down the road. We have lots of ideas brewing. This is just the beginning.

      • says

        Suggestion: If you can convince an online survey service — especially Survey Monkey (.com) — to implement Approval counting in the results they calculate, that will both educate lots of survey-takers and make Approval voting more accessible.

        In contrast, a poll that collects preferences about voting methods will amount to “preaching to the choir” because few people have the time to learn about all the different methods. And it will be flawed because the choice of ballot method will influence the results (because each ballot type encourages a different kind of voting strategy).

        • Aaron Hamlin says

          I think we were considering something more formal than an online survey. Although, Survey Monkey does do other more formal surveys using phone. The issue with traditional online polls is that it’s a convenience sample. So you get systematic error (the worst kind). It’s tough to take meaningful insights from the data. It would be nice, however, if more online polling you see (even though informal) utilized Approval Voting. It’s absurdly easy to implement.

          If we were to do such a project, it would involve a survey provider that offered simple random sampling or stratified sampling. This more sophisticated sampling would allow us to interpret the results more meaningfully. And if people wanted to strategize, then that’d be fine. More realistic. We’d just want to give people an idea of how different methods perform within the same sample looking at the same candidates. To my knowledge, this has never been done in the U.S.

          At least with the survey in consideration, we were just looking to measure candidate preferences with different methods. We wouldn’t necessarily consider preferences for the different methods themselves.

          The discussion on this within our group hasn’t been thorough because it’s not yet timely. If we were to do this, it probably wouldn’t be until 2016. But don’t worry. We’ll have plenty of good stuff in the meantime.

  2. says

    “Some IRV advocates complain that Approval treats all votes the same, so a less preferred candidate has the same weight as one that’s more preferred.” Approval Voting is just low-resolution Score Voting, 0 or 1. If that resolution is too low for your given vote, crank it up to 0..2, or higher.

    • Aaron Hamlin says

      That’s right, Steve. That upgrade (Score Voting as described in the article) would allow voters to further differentiate among the candidates. We’re pushing Approval because it’s so simple and offers such an enormous gain over the current Plurality method.

  3. Abdirahman Mohamed Dirye says

    the interview is fantastic! Aaron has spoken something need to be done. Adrian,too, is determined democracy and hight rights champion. he is always advocate of human rights . However, the US election system is more complicated, that is in my view, thank its counterparts like the UK or Germany. third world countries feel hard to emulate. in my little country, the UK, people turn to

  4. says

    I built some “wooden computer” tabletop voting system demonstrators that may be of interest to you. See my link. These are suitable for teaching children and adults. I cover Borda, Condorcet/Tideman, Plurality, and Approval. I never got around to making a video of my pitch. One nit-pick: with all due respect, Plurality is only the second-worst single-member election system. Borda is worse. I also have a link on the “wooden computer” page to a curriculum for an “Election Systems 101” short course I taught at a church retreat a few years back. Contact me if I can be of any help (info via the link). Good luck!

Leave a Reply

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