Robert Wiblin: Hi listeners, this is the 80,000 Hours Podcast, the show about the world’s most pressing problems and how you can use your career to solve them. I’m Rob Wiblin, Director of Research at 80,000 Hours.
Today’s episode goes deep into the why and how of voting reform. If you’re interested in politics, economics, collective decision-making or philosophy, then this should be an exciting episode for you. If none of those gets you going, you might find this episode goes further into the details than you’re looking for.
We’ve recently attracted a few thousand new subscirbers, so I just want to point out to everyone that the blog post attached to every episode comes with links we’ve carefully chosen to help you achieve some level of expertise in the topics we cover. I strongly recommend checking them out.
Also, in case you didn’t know, the blog post for almost every episode comes with a full transcript of the conversation. I’ll be honest that there are sometimes transcription errors in there – but if you read much faster than you listen, you might prefer to consume the show that way. It’s also helpful if you want to find a quote later, on or pull something out to share with other people.
Finally, the posts list all the topics we cover, and a list of key points from the episode. There’s a link to the blog post in the text attached to the episode in your podcasting software.
Without further ado here’s Aaron Hamlin.
Robert Wiblin: Today I’m speaking with Aaron Hamlin. He’s the Executive Director of The Center for Election Science; a U.S. non-profit, which aims to fix broken government by helping the world use smarter election systems. CES recently received a $600,000 grant from the Open Philanthropy Project to scale up its efforts.
In 2014, he also co-founded the Male Contraceptive Initiative, where he worked for three years. Before that, he completed graduate degrees in both education and public health, before completing a JD, and briefly practicing law.
He’s written for Deadspin, Independent Voter Network, and The Telegraph, and his work has been featured in Popular Mechanics, National Public Radio, and Scientific American, among many others. Most importantly, he’s a regular listener to the 80,000 Hours Podcast, so thanks for coming on the show Aaron.
Aaron Hamlin: Thank you. It’s a pleasure being here.
Robert Wiblin: We plan to focus on what good electoral reform would look like, what the benefits might be, and whether it’s at all realistic to achieve it, but first, describe for us the problem that you see yourself as trying to solve, Aaron.
Aaron Hamlin: What we’re looking at here is the voting method itself. To get terminology out of the way, when I say voting method, what I mean is the information that you put on the ballot, so say choosing one, choosing as many as you want, ranking, scoring on a scale, and then what that information is done with.
How you’re using that information to calculate a result, and that’s called the voting method. That’s what we’re mainly looking at when we’re looking at these problems.
Robert Wiblin: Okay, so in the U.S. and the U.K., the voting system is basically that everyone picks one name, and one name only, and in Australia where I’m from, you have to number everyone in order of preference. That’s what we’re talking about.
Aaron Hamlin: That’s right.
Robert Wiblin: Why hasn’t everyone just picked the best voting system already? Why is this an interesting and complicated issue?
Aaron Hamlin: It’s a tough issue because it’s a tough problem to analyze. With a lot of things, you might find a nice, easy metric to use and then given that metric you can say, “Oh, well we measured all these things according to this metric and this one came out on top. We’ll just use this one.”
Well, with voting methods, there are a number of considerations to look at. No particular voting method is solid on all these considerations. These are considerations like what kind of winner you get.
If people, on average, seem to be happy with a particular winner, do you get certain anomalies? How consistent is the method? When a method messes up, how badly does it mess up? How frequently does it happen? How does it treat newcomers?
People that lose, even if they lose, are they given an accurate reflection in support? Here, this is looking at single winner method issues, but there are also a number of considerations in terms of metrics that you would consider for multi-member issues.
You might look at, for say a multi-winner election method, does it give proportional outcomes? For a certain percentage of the electorate, if say 10 percent of them choose a particular party, say they choose 10 percent for instance, does the number of elected officials roughly hit at that 10 percent, or is it way off? Do they get no seats, or do they get many more than they should?
Those are the types of considerations that you look at when you’re dealing with whether a voting method is fair or whether it makes sense to use that method in a particular context, and also things like complexity, what ballot design does it have? Is it simple? Is it hard? These are the types of things that we’re looking at.
Robert Wiblin: We’ll come back to a whole lot of different voting systems and the various pros and cons in a minute, but to make this more concrete, can you us an example of where a bad voting system produced a clearly bad outcome?
Aaron Hamlin: Yeah. There’s one fun, and a little bit depressing election that happened and 1991 for the Louisiana Gubernatorial election. Louisiana, they use what’s called a Jungle Primary, which means that you take all the candidates and you have them all run together.
Then you take the top two and they go together in a runoff, head-to-head. In the 1991 election, there were really three main contenders there within the Jungle Primary. You had Buddy Roemer, he was a moderate candidate and who was the incumbent governor.
You also had a couple others, who were a bit interesting characters. One was Edwin Edwards, and Edwin Edwards, he had served in government before, and had had some allegations of bribery, and he was kind of a shady character and that was something that people were pretty aware of, that he was a bit of a shady character.
The other main front-runner was a guy named David Duke and David Duke was known for being a Wizard in the Ku Klux Klan, and so also an interesting character. You had these three candidates that were running in this Jungle Primary, and what the voter is doing is that each voter is choosing just one candidate out of all these candidates, and here we’re talking about three main front-runners.
There were some other candidates, but three that were getting the majority of the attention, and then the two candidates with the most votes would go on to this runoff. The funny thing is, Buddy Roemer, the more moderate of the bunch, did not make it to the runoff.
Instead, Edwin Edwards, the corrupt politician, and David Duke, the Ku Klux Klansman, were the ones that made it to the runoff.
Robert Wiblin: Obviously, you and I would have preferred the compromise character, but why is it clear that that was just an error in the system, due to a bad voting system, rather than just voters having preferences that you and I don’t like?
Aaron Hamlin: One way that we can get an idea that this wasn’t right, aside from just it smells of something that is not right, is that when they took polling information to look at how candidates would do head-to-head against each other, Buddy Roemer against Edwin Edwards, Buddy Roemer wins.
Buddy Roemer versus David Duke, Buddy Roemer wins. If Buddy Roemer ran against either of these two head-to-head, he would have won easily, but the mere fact that it was a three-way race meant that he was knocked out.
Robert Wiblin: That was because of vote splitting, I guess? That some of the people who liked him moderately were taken away in the first round to the other two candidates?
Aaron Hamlin: That’s right. An interesting theme that comes up with a voting method, when you’re only dealing with two candidates, the voting method doesn’t matter so much, but it becomes really important when you have more than two candidates.
Then the voting method that you choose becomes really important. There’s an interesting phenomenon that can occur when you have a three-way race where all three are competitive and there’s a moderate candidate.
There’s this phenomenon called the Center Squeeze Effect, which has the candidate in the middle have their votes split on either side, so here what you’re essentially having is first choice preferences being split.
The person in the middle gets split from the person on the right, and the person on the left, and they can just get a sliver of the vote, even though they appeal to the broadest breadth of the electorate.
Robert Wiblin: They’d win an election against every other candidate one-on-one?
Aaron Hamlin: Yes.
Robert Wiblin: Is this a common occurrence? Is it common for the voting system to produce suboptimal outcomes, or even disastrous outcomes as in this case?
Aaron Hamlin: It tends to occur more often when you have a competitive election, where there are a number of candidates. Primaries are likely one area where you would see that because you have a lot of candidates; that they tend to have a lot of similarities and overlap, so you can have more of this vote-splitting type effect going on.
Robert Wiblin: I guess, an effect of the potential for vote-splitting is that, in these plurality voting systems, where just the candidate with the most first preferences wins, is that you tend to get consolidation to just two parties, because every side wants to make sure that their vote is never split because that greatly weakens their chances.
Aaron Hamlin: There’s a separate reason why you get two parties as a consequence of a particular voting method.
That goes to a political science idea called Duverger’s Law. Duverger’s Law is a predictor for the number of parties based on the voting method. It looks at two factors: one is the amount of support that a particular candidate, or a party need in order to get elected, and when you’re talking about a single winner election, that amount of support is more than anyone else.
That raises the threshold, so that it makes it difficult for third parties or independents that can hit that support. Then the other factor that Duverger’s Law looks at is whether it’s possible for a voter to fearlessly be able to support their honest favorite, even if the person doesn’t look like they’re going to do particularly well.
You really need that in order to give third-party candidates, and these other candidates the ability to increase their support, because if you’re … As a voter, if you’re looking at a third party, you’re an independent, and you’re saying, I don’t think they’re going to do very well; you’re likely not going to vote for them.
You want your vote to be effectual, so you’re likely going to choose amongst the front-runners, but if you have a voting method that doesn’t let you do that, and it forces your hand to say vote for the lesser of two evils, well, now you’ve got a problem and third parties are going to have a problem, too.
If you’re able to address at least one of these two factors, that is being able to have a lower threshold to be able to get in there, which say, proportional voting methods do, or you have a voting method that allows you to work on supporting your honest favorite, that’s also going to help for allowing for another party to thrive.
Duverger’s Law is a little bit of a misnomer in some ways. It really acts as a large predictor, so there are other types of factors that can be predicted in terms of whether you’re going to have more than two parties, but the voting method is, by and large, a predominant factor.
Robert Wiblin: Just before we go into voting theory in detail, which countries and kinds of elections does The Center for Election Science think about? Is it just government elections in the U.S., or is your scope broader than that?
Aaron Hamlin: We think about voting methods all over, and here, we spend a good amount of time looking at single winner voting methods, and right now that’s predominately our focus, although, we also look at, to some degree, multi-winner methods.
There are a number of countries that use better multi-winner methods arguably than the U.S. and some other countries that do elections similar to how we do, so for instance, you have countries like Denmark, Iceland, Sweden, Norway, they use a lot of proportional methods.
They also use some things like leveling seats, which makes it so that you can have a certain threshold, say 5 percent in order to make it so that a party needs to hit that threshold of support in order to get anyone elected for a particular seat.
Some of these countries that use proportional methods can use features like leveling seats to make it so that if the number of people within a party are maybe off compared to what it should be, they can add extra seats to even that out. To make sure that it’s more proportional than it otherwise would be.
You have some countries like that, that use nice features in order to lead towards more proportional outcomes, but when we’re talking in the single winner space, so things like executive offices, like Mayor, Governor, President, those types of things, we see, really, a lot of bad voting methods all around, across the world.
We see first-past-the-post used quite a lot, when you’re only choosing one person and the person with the most votes wins. Sometimes we see a runoff, which was exampled with the 1991 Louisiana Gubernatorial election, and David Duke and Edwin Edwards entered the runoff, which is a spoiler.
If you haven’t read through, Edwin Edwards actually beat David Duke, and interestingly enough is going and running again in other elections. Not successfully, but attempting to run in other elections, so even to the kind of examples that even countries that are using runoffs, there are still some serious issues with using runoffs.
We know that there are voting methods, single winner voting methods, that get around some of these issues, and yet we’re still seeing a lot of governments, even newer governments, that are re-visiting their systems or creating new systems. They’re still using these really bad voting methods, either first-past-the-post or a runoff system.
Even if they’re doing something, say instant runoff voting, that’s still getting into some of the issues that we see with traditional runoff systems. It’s just simulating that process.
Robert Wiblin: All right. Let’s talk about the various reform options. Most jurisdictions in the U.S. and the U.K. just use first-past-the-post voting, so you get to vote for one candidate and the candidate with the most first preferences wins.
What are the various different options that they could have for reducing opportunities for tactical voting and reflecting the actual preferences that the electorate has, rather than having perverse outcomes, like a candidate that would beat every other candidate, one-on-one, failing to win the election.
Aaron Hamlin: A really simple way that we tend to like, and we push as an organization for single winner government elections, is the voting method called ‘approval voting’. With approval voting, what you do is you merely select as many candidates as you want, not just one. If you want to choose just one, you can do that, but you have the option to select as many as you want.
We’re not talking about ranking. We’re talking about merely choosing as many as you want, and then the calculation just being summative like it normally would be. That particular method is really simple, which is nice. When you can use a simple method that gives you the types of outcomes and behavior that you want, it’s normally pretty good to go with that method.
It’s also robust to a lot of anomalies that you wouldn’t like to have occur, so with vote-splitting, it’s very robust at vote-splitting; if you have similar candidates that are running, because people can choose multiple candidates at the same time, it helps to guard against that.
It also has a really nice property to it, which is nice for independents and third parties. That is that with approval voting, when you can choose as many candidates as you want, you can always put your full support behind your honest favorite candidate.
What that means is, say there’s an independent or a third party candidate that you really like, but you’re looking at them and you think: ‘God, this person is never going to win, what I should really be doing is looking just among the top two competitive candidates, maybe the major parties, and choose a lesser of the two evils among them.’ With approval voting, you don’t have to do that; you can have your cake and eat it too.
Under an approval voting election, what you can do is you can support that third party or independent candidate that you really like, and in addition to that, if it looks like the candidate maybe is not going to be able to take on one of the other leaders, and really dominate, and it looks like you maybe want to hedge your bets with one of the other two front-runners, what you can do is you can choose a lesser of two evils in addition to supporting your honest favorite.
That makes a big difference, because now that third party or independent candidate, that candidate can build their support over time. Also, it means that they can’t be marginalized the same way. If we’re using first-past-the-post or party voting, when you can only choose one candidate, third parties and independents get an artificially low amount of support. Not only do they typically lose, but they get an artificially low amount of support, which means that they are marginalized unfairly.
They can bring good ideas to the table that are popular with people, but because they’re perceived to be a losing candidate, they just simply don’t get that support of a party voting, but with approval voting, we see time and again through research and through other polling, that these candidates get a much more accurate reflection of support, and that’s exciting, because that means when they bring these good ideas to the table, they don’t get marginalized unfairly.
Robert Wiblin: Okay, so approval voting avoids the vote-splitting perversity and it means that candidates are more likely to get an accurate reflection of how much the electorate likes them in general in their vote share.
I know that approval voting is your preferred favorite, and it seems to be that the voting system that The Center for Election Science is mostly promoting at the moment. What are the biggest downsides of that system, and are there any alternatives that we should be considering as well?
Aaron Hamlin: Yeah, so every voting method has anomalies that can occur with it. There are theorems out there, like the Gibbard-Satterthwaite Theorem, which says that every voting method can have some kind of tactical voting vulnerability.
Tactical voting, that’s when a voter expresses information on their ballot in a way that isn’t accurate for their actual preferences, but they do it, because voting that way optimizes the likelihood of the outcome that they want to see.
With approval voting, probably one of its heaviest criticisms is that it can tend to cause some people to choose only one candidate, because they don’t want to create extra competition for that candidate, and you can have an approval voting election where a lot of people just choose one candidate, but we would argue that that’s not something that’s fatal.
Really, what’s important here is that when it’s important for a voter to choose more than one candidate, that they had that ability, and that when they don’t have that ability, then it becomes a real issue, so for instance, there may be, within an election, and a lot of people maybe don’t care about some of the non-front-runners. Maybe they just simply aren’t very good candidates.
In that kind of circumstance, maybe they only choose one candidate, and that’s not a big deal, or say your honest favorite candidate is a front-runner, and you don’t want anyone competing with that candidate and there’s really no one else that you have to hedge your bets against.
Then in that case, it makes sense to choose only one candidate. Another interesting thing here is that even when a lot of people say choose only one candidate, having just a portion of people, say 10 or 15 percent, choose more than one candidate, can still make a big difference.
10 to 15 percent within an election, that can easily mean a difference between a win and a loss, or one candidate winning versus another candidate winning, so it can change the outcome of an election, even when a few number of voters decided to choose more than one candidate.
The other issue that comes up is, even when you have a lot of people choosing only one candidate, it can still make a big difference with the people that choose multiple candidates in terms of how that reflection of support is given for third parties and independents.
Robert Wiblin: Okay. I guess, for those who choose to just vote for one candidate, they’re probably no worse off than they would have been, if they could only vote for one candidate; they just also have the option of expressing a wider range of preferences as well.
Aaron Hamlin: That’s right. There’s no harm for them for having that option.
Robert Wiblin: Is that really the greatest weakness of approval voting? Are there no situations where you have a particular kind of allocation of support between different candidates where a Condorcet winner, that is a candidate that would win on a one-on-one fight with every other candidate, could lose in approval voting, or is that just extremely rare?
Aaron Hamlin: With approval voting, one of its nice properties is that it tends to elect, what’s called a Condorcet winner, and as a recap for clarity, a Condorcet winner is a candidate who can beat every other candidate head-to-head, if it was just between those two candidates.
The candidate that can do that, with every other candidate, is a Condorcet winner. Another interesting thing about that, is that a Condorcet winner does not always exist, just like a candidate with majority support does not always exist.
Robert Wiblin: Interesting. Okay, but approval voting will almost always choose one if one does exist?
Aaron Hamlin: Correct. Approval voting is very good at electing a Condorcet winner. It’s not to say that it will do it every single time because, like I said, voting methods aren’t good at being able to fulfill a lot of certain criteria in an absolute sense, but they can perform really well on a particular criterion.
Robert Wiblin: Yeah. You said approval voting is a little bit vulnerable to tactical voting, where someone might only vote for one candidate, kind of for the same strategic reasons that they might vote for a particular candidate in first-past-the-post voting, but does it violate any of the other, or can it violate any of the other ideals of a voting system?
Aaron Hamlin: Part of the rationale for looking at choosing only one candidate would be a criterion called, later-no-harm, which gets back to the idea of you don’t want to create extra competition for your favorite candidate, so for instance, let’s look at the worst-case-scenario for an approval voting election, whereas an individual voter, you see that one of your favorites is in the lead and there a couple other competitors.
One, you really would hate to see get elected and another, you would just be kind of all right with, and so there, to avoid the catastrophic result, you may choose a candidate that has a lot of commonality with what your ideals are and the policies, but maybe doesn’t quite fit the same way that your favorite candidate does, so there, you may choose a candidate, and that candidate may wind up winning as a consequence.
There, at the individual level, you have a failure where you just get a compromise candidate, but the interesting thing about that is that, at the individual level, there may be a little bit of a loss there, but by choosing a compromise candidate, that’s something that, while there may be a little bit of loss at the individual level, at the group level, that tends to be a candidate that appeals to the broadest space.
Another instance here, is that here, when we’re talking about a failure at the individual level with approval voting, the failure is not an enormous one; we’re talking about maybe not getting your favorite candidate, but as a result of the way that you vote, maybe getting a compromise candidate.
Compare that to the type of loss that you might see in another type of voting method where, instead of by the way that you express your ballot, instead of getting maybe your favorite candidate, you get a really terrible outcome, which is possible with other voting methods.
The worst-case-scenario that you see with an approval voting election, is that at the individual level, you may get a compromise candidate, which may not be a terrible deal for the election at the group level.
Robert Wiblin: Okay, so it fails more gracefully. Before we go on to talk about other voting systems, do you want to just respond to the common objection that approval voting, and probably everything else we’re going to discuss here, allows people to vote more than once, which I think is a common complaint among people who are used to first-past-the-post voting where you only get to vote for one candidate?
Aaron Hamlin: Yeah. I think the idea there is that the criticisms that some voting methods may violate, which called, “one person, one vote.” The concept behind one person, one vote is that the weight of one person’s ballot counts as more than others in an unfair way, and as a quick spoiler, they don’t violate one person, one vote.
In fact, the concept of one person, one vote, goes back to some Supreme Court decisions in the U.S., like Baker versus Carr, Reynolds versus Sims, and they were looking at some cases where you’d have a district that has way fewer voters than another district, and yet, the smaller district would get the same number of seats, which was weird, which meant that, if one district had a population that was 10 times smaller than another, essentially, the weight of those people’s vote, within that district, was 10 times greater.
That’s not what is going on with these voting methods, so for instance, say we have six candidates, you really like one candidate, and were using an approval voting election, and you choose that one candidate, and say I like four other candidates that aren’t that one candidate that you like, and I choose all four of those.
Well, my ballot doesn’t count for four times as much as your ballot does, and neither of the people that we voted for, if we just count our two ballots, are in the lead over the other, so neither one of us is getting an extra unfair weight as a result of using a particular voting method.
Robert Wiblin: I suppose one way of explaining would be, in approval voting, if you had an election with six different candidates, if you voted for all six of them, in a sense, you got six votes, but of course you won’t change the outcome at all, because you voted for every single one and you haven’t distinguished between them.
I suppose another way to think about it might be that, everyone gets the same number of votes, they get six votes, they get to decide whether they say yay or nay to six different candidates?
Aaron Hamlin: Yeah, that’s right.
Those are other explanations that also get at the same underlying idea there, and there are a number of elections that we use with current voting methods that allow us to express the number of votes for candidates in the different ways. So, for instance, in a lot of U.S. elections for a city council, for instance, we use what’s called block plurality voting, where there are five people being elected, and say there are 10 candidates or something, a voter would be told that they can vote for no greater than five candidates.
You can have somebody that chooses five candidates out of those 10; you can have another person that chooses three candidates out of those 10, and here you have a different number of candidates being voted for on different ballots, but neither voter is at a disadvantage, merely because of the way that they decided to express their vote, given the voting method, and that’s something that we already use right now.
Robert Wiblin: This makes me think, if I was voting in an election with approval voting, and there were six candidates, how do I decide how many to vote for? Obviously, I don’t want to vote for all six of them. That that’s pointless. I could vote for one, and so I boost the one that I most like, but is there some kind of medium that is optimal from a strategic point of view?
Aaron Hamlin: Yeah, so one simple heuristic that you could use is you could say, okay, well, who’s likely to win here? Like who are the front runners here? Then among the front runners, you would take a compromise. You would take the candidate among the front runners that you like.
Then in addition to choosing that candidate, you could choose everyone else that you liked more than that candidate too, which would take the place of other candidates that you really like, but maybe aren’t as viable.
You could look at it that way, such as choosing your favorite as another way, and then looking at is there anyone else that is likely to win, that you would also find acceptable; who is maybe not competing with your favorite.
You’re also considering things, like is there somebody else that you really don’t like, and is there perhaps someone that you may be able to hedge your bets with to keep that other person from winning, which really is not too far away from the way that we do it in current elections, in terms of the analysis.
Like say that there’s an independent or third party candidate that you like, normally you can’t vote for that person, but you’re looking at ways to hedge your bets against the other person that you don’t like.
You’re really doing the same scenario, only in addition to hedging your bets appropriately, you are also being able to offer that honest expression of support and really choosing the candidates that align with your viewpoint.
Robert Wiblin: Okay. Let’s move on to other voting systems now. I’m from Australia, where we have a system called single transferable vote. I think also called instant runoff, or in the U.K., alternative vote. Do you want to just describe that one and then explain its pros and cons, and maybe why you haven’t chosen to campaign in favor of that voting system?
Aaron Hamlin: Sure. I would offer a quick highlight of a nuance, which is that instant runoff voting is really a different voting method than single transferable vote. When we talk about something like instant runoff voting, we’re talking about a single winner voting method.
Whereas when we’re talking about single transferable vote, we’re talking about a multi-winner voting method; they have a lot of similarities with their algorithm and you can plop one on the other within a particular circumstance and get the same result, but it’s generally a good practice to separate single winner voting methods and multi-winner voting methods.
Robert Wiblin: Okay. Yeah, so I guess we use instant runoff for our Lower House, our House of Representatives, where we have single-member districts, and then I guess single transferable vote for our Upper House, where each state represent, elect seven different candidates.
Aaron Hamlin: Yeah.
Robert Wiblin: Yep, okay, yeah, so what are the pros and cons of that kind of approach in general?
Aaron Hamlin: With instant runoff voting, I think intuitively, it seems like you’re offering more information, and you are offering more information, and part of the issue with that is that, within the algorithm of instant runoff voting, you are only looking at one part of the ballot at any one point. Like the algorithm that is going through is only looking at the first choice preferences at any one point.
To step back for a moment to highlight what’s going on with instant runoff voting, the algorithm is saying, first, the voter is ranking their preferences from their first to their least preferred candidate. The algorithm is saying, does anyone have more than half of the first choice preferences among all these candidates?
If yes, then you have your winner there; if not, what’ll happen is you look at the candidate who has the least number of first choice preferences, that candidate is eliminated, and you look at the ballots that voted for that candidate as their top preference, and then you look to their next preference, and then the next preference is now treated as the first preference, and then you go through, run that algorithm again, and see if anyone has a number of first choice preferences that’s greater than half.
If yes, you have a winner; if not, you keep going through that until you do, and so there, it’s an element of an issue with instant runoff voting to begin with, and that it’s complicated. You have a number of rounds that are going on there, and being able to keep track of all those vote transfers between rounds, can be complicated.
With the expression part, it seems like you’re providing all the information, but the algorithm is only looking at first choice preferences at any one point, and it may not even look at the rest of your ballot, so even though you provided all this information, the algorithm may not even look at the rest of your ballot at all during the election, given what happens in the election.
That’s one part, and that it intuitively feels like you’re providing all this information, but we can’t forget that a voting method, it has multiple components to it. One component is the information that you’re providing. The other component though is, what is the algorithm for that voting method doing with that information?
That’s one element. You can also have things like anomalies that can occur, and to be fair, anomalies can occur with lots of different voting methods, so you can have it so that you rank a candidate as better, and that causes that candidate to lose, you can rank your favorite candidate as first, and that can wind up harming you and leading to a worse outcome.
That’s an issue that we see with first-past-the-post already with when we choose our honest favorite, causing an undesirable outcome to occur. Even though at the moment, highlighting some of the issues with instant runoff voting, it should be clear that first-past-the-post voting, or plurality voting is really awful.
We have our issues with instant runoff voting, but plurality voting is worse than instant runoff voting, so as a caveat, it’s always important to put that out there, in terms of making sure that we always know where the bullseye is, and that the picture of first-past-the-post is always on that bullseye.
Robert Wiblin: Okay. From my experience in Australia, single transferable voting seems to work reasonably well.
I think the case where you can get the most perverse outcome is where you want your second favorite candidate to get removed sooner, so that then their second preferences will be distributed among other candidates, and so you take a candidate who you actually preference second, and put them last, so that they’re more likely to get eliminated early, and then those votes go to another candidate that you might prefer a bit more.
There’s some room for tactical voting, though I think because Australia’s system is pretty two party heavy, that happens fairly rarely. In the case with multi-member districts in the senate, that there’s this quite regular occurrence, where the seventh member that’s elected in each state, is quite random, because often, they’re cobbling together a bunch of votes for tiny parties.
We’ve had members elected to the senate from the motorist enthusiast party that got well under one percent of the vote, because they managed to cobble together preferences from tons of micro parties, before they went to the major parties, but by and large, it seems to work okay.
Although, especially in the senate, often people have to vote. If they want to vote, or actually, in the past, if they wanted their vote to count, and they didn’t want to give their preferences to the parties to decide themselves; they would often have to number like 100 different candidates, which I’ve actually had to do.
Recently, that’s been changed, so you can choose how many candidates you want to vote for, and your vote just expires if there’s no further preferences, so once it’s removed, but yeah, it’s interesting to me that instant runoff and single transferable vote, that they seem to be more popular methods around the world than approval voting, but you seem to think that approval voting, it’s usually a superior system.
Aaron Hamlin: Yeah. Instant runoff voting, and single transferable vote have had some advantages, probably the academic history for them, and initial use was a lot earlier than with modern use of approval voting.
The academic literature on approval voting really begin in the late 1970s with Steven Brams and Peter Fishburn, they were … While there are some other people that looked at approval voting and identified the voting method, independently, the academic research really begin in the late 1970s, and then went from there.
These other voting methods really just really had a headstart, and from that headstart, they’ve been able to see more implementation, but they also had that initial gap too, from when the initial work is done a method, and then you have that gap between when you see the initial use, and then successions from there.
Robert Wiblin: The UK considered switching to single transferable vote, which they called alternative vote, in I think 2012/2013, they had an election about that, and they pretty decisively decided not to switch their voting system. Do you know what the reason for that was, and does this indicate why it might be hard to get any switch in voting systems?
Aaron Hamlin: I’m not quite as familiar with that particular campaign, so it could difficult for me to go on the nuances of that, but I can talk a bit about some of the campaigns in the US for alternative voting methods that look at instant runoff voting, for instance.
Robert Wiblin: Yeah, that would be great.
Aaron Hamlin: In the US, instant runoff voting has been implemented in a number of cities at the local level, and then it’s also been, more recently, pushed at the state level. That’s largely been successful, so there, the idea there, and this has been used previously with other types of alternative voting methods earlier in the 19th century, so this strategy has been taken up again to push instant runoff voting.
The way that’s been done is by using ballot initiatives at the local level, and that makes a lot of sense, and one of the reasons it makes a lot of sense is because if you are elected under a certain voting method, you really don’t want to have another voting method come along that could risk your re-election. You’d really like to have it be the same voting method that got you elected in the first place.
As a consequence, there’s a conflict of interest, with people who are already elected, choosing the method that’s going to be electing them in the future, and so to get around this conflict of interest, a ballot measure is a good way to-
Robert Wiblin: Take it directly the public.
Aaron Hamlin: Right, and circumvent that issue. We’ve seen over a dozen cities use and implement instant runoff voting, and it’s been pretty successful in getting it initiated. Although, there have been some repeals in some cities, and part of that is you have some complexities that come up with that particular voting method.
We can see some anomalies, and when that happens, people can get confused, and so sometimes you just get the wrong result, and people aren’t happy about that. As a consequence, it’s also been repealed in certain cities as well.
Robert Wiblin: Are there any alternatives, alternative systems that would be appropriate under particular circumstances that we should also consider?
Aaron Hamlin: Yeah.
We like approval voting for single-winner government elections, particularly here in the US, because one, it’s very simple, so you don’t get a big discrepancy between the voting method that we’re currently using, and the proposal, and it tends to avoid a lot of anomalies, tends to get you good results, but there are other situations where maybe we want to use a different voting method, and maybe that makes sense, so for instance, voting takes place in a number of different types of context.
If you’re using a voting method where there aren’t a whole lot of voters, say 20 or 30 or something, and you’re making a group decision, maybe something like score voting makes sense, so for clarity, with scoring voting, what you’re doing is you are scoring each candidate on a scale, say zero to five, and the candidate with the highest score, on average, is the candidate that wins.
Robert Wiblin: It’s a bit like how you review things on Amazon or IMDB?
Aaron Hamlin: Yeah, yeah. The nuances within the algorithm may be a little bit different, but that’s the idea. That’s certainly the expression element that’s occurring. One of the reasons why something like score voting is more appropriate when you have a small amount of voters is because you really, when you’re voting and you brought information, it’s a sampling.
When you have not a whole lot of people, or voters, then you have, from a statistical point of view, a lot of random error. With score voting, because you have more iterations of expression along a scale, that sensitivity actually reduces that error, so you’re more likely to hit the result that is good for the group.
Whereas with approval voting, when you have a small number of people, you may have an issue there, because the expression element doesn’t have a lot of sensitivity. With score voting, you’re getting, really, a lot of the features of approval voting, and it’s still an additive voting method, because you’re just taking the average or simply summing the scores.
You get a lot of the same features, but you also get that sensitivity. When you have a large number of voters, the sensitivity isn’t quite as important there. Having a large number of voters tends to compensate for that lack of sensitivity within the voting expression.
Robert Wiblin: Because you have lots of voters, and they have a distribution of how strongly they like something, and it tends to be the case that the thing that people like the most intensely is also the thing that most people are going to prefer.
Aaron Hamlin: That’s right.
Robert Wiblin: Yeah, interesting. Why not use this score voting? Because you were saying approval voting is basically like score voting, except that you only get two scores that you can give everyone, five starts or zero stars; why not allow all these gradients of expression for every candidate?
Aaron Hamlin: Yeah, and so the reason that we push approval voting so hard instead of score voting is that, with approval voting, we see that as being easier to implement as a practical sense. The ballots look virtually the same, except the direction so you can choose more than one, instead of just one, and that’s pretty much it.
Whereas with score voting, you’d have to change the ballot, you may have to tweak the voting machines a bit more, if you’re using a voting machine, but with approval voting, the ability to implement it is very simple, and the education campaign needed for that is not the same as should you be proposing another voting method that’s more complicated, like whether it be score voting, or instant runoff voting, or another voting method.
Robert Wiblin: With score voting, wouldn’t there be a really strong temptation to bury all of the candidates who you don’t like that much? Just to give everyone who dislikes zero stars out of five, to ensure that you’re not giving them any help to get elected?
Aaron Hamlin: Yeah, yeah, so some other tactics can appear under a score voting, because of the way that you’re allowed to express your ballot.
Say even like a candidate say one out of five, but they look like they may be somewhat competitive, and you may bury that candidate, that’s a particular tactic they can use under score voting, but while using that tactic, we would argue that the types of results that would create wouldn’t be as devastating as the types of anomalies that could occur under different voting methods.
Robert Wiblin: Okay, interesting. The reason to go for approval is just that it’s more straightforward, and people can understand it and it’s more likely to get up?
Aaron Hamlin: That’s right.
Robert Wiblin: Cool. Okay, so we’ve been through approval voting, score voting, single transferable vote; are there any other big candidates for voting reform?
Aaron Hamlin: I think those are some of the big ones. We spoke a bit about instant runoff voting. Single transferable vote is a proportional voting method that’s typically used for districts.
Robert Wiblin: Yeah, the multi-member districts.
Aaron Hamlin: Yeah, yeah, and one important thing about … Say we’re talking about using a proportional method within a district that’s electing multiple people at the same time. It’s also important to keep in mind the number of people that you’re electing at the same time.
If the number of people that you’re electing at the same time, say three, well, the proportionality is going to be worse, so there, in order to guarantee a seat, you’re going to need 25% of the vote using a single transferable vote method of when three people are elected, that uses something called the Droop Formula, in order to figure out what that is.
Proportional methods, when you have more people being elected at the same time, you tend to get more proportionality, which is why some countries use methods that elect an entire legislature or parliament at once, so that they can get more proportionality. In fact, some people argue that that gives too much proportionality, and that you get little tiny parties, and maybe we don’t want to see that kind of splintering going on.
As a way to compensate for that, with these huge legislative bodies that are using a proportional method in order to compensate that, they may say, in order to get elected, we need to at least have five percent of support in order to be able to get someone into office, so that’s a way of getting more of that proportionality, while at the same time, avoiding that super splintering effect.
Robert Wiblin: What makes the voting method specifically an important target, as opposed to other reforms that you might have of a political system?
Aaron Hamlin: Well, a voting method is really unique in the way that its role is played within electoral politics. When I say that, we’re always, as citizens, we have all these opinions about what policy should be, who should be in government, and while we’re doing all these things to express ourselves, whether it’s going and sharing things with friends, going to a protest, or lobbying our representatives, they really don’t have to listen to any of that.
The only thing that they are forced to listen to is our vote. That’s one time when they don’t have a say in the matter. If we cast our votes a certain way, and according to the voting method, it produces a particular result; they can’t ignore that, they have to abide by that.
This is our best tool, and because this is our best tool, in terms of what it’s capable of doing, we need to make sure that it’s an effective tool in order to be able to produce the outcomes that we want.
If we’re using a voting method, and that voting method doesn’t allow us to actually express our opinion, accurately or fully, then we’ve got a big issue, because if we’ve never expressed that information, there’s no way that information is going to be considered in the result for an election, and so because of that, we want to make sure that the voting method that we’re using allows us to express information that is capable of being used to produce a meaningful result.
Also, that the information that we’re expressing under this voting method, is accurate enough, so we’re not just using these tactical voting elements that are being pushed on us by the voting method, that that’s not occurring, so that we can present accurate information that is actually being able to be used in a meaningful way, so that we can get results that represent us.
If we’re using a bad voting method, well, we simply don’t get that, and then we just get these other opportunities that aren’t nearly as effective as being able to cast your own vote, and not being able to have that ignored.
Robert Wiblin: Okay, so if I had to summarize your views on this in just one sentence, I think it would be that first-past-the-post voting is terrible, approval voting seems like the best alternative, but other methods, including instant runoff, or score voting are also better, because first-past-the-post is so bad. Is that attitude typical of voting theorists, and is there a consensus, among people who study this topic, on where we ought to move?
Aaron Hamlin: Interestingly, there is. The London School of Economics and Political Science has an organization called, Voting Powers and Procedures.
A while back, they got a bunch of experts together, and they were trying to figure out, we have all these voting methods out there, and you had this collective, this group was a bunch of broad experts on voting methods, and they thought we should probably, since we got all these people together, we should maybe vote on what the best voting method is; that would be a fun thing to do.
Robert Wiblin: How did they choose how to do that?
Aaron Hamlin: Yeah, so they agreed by consensus, initially, just to use approval voting for simplicity. That was the way that they were going to express their votes on all these voting methods, and they had a number of different voting methods, including score voting, instant runoff voting, first-past-the-post, Condorcet methods, Borda count; all these different voting methods.
Here, we’re talking about single winner methods, so that’s the only thing that they were considering. Interestingly, so they were using approval voting, which meant that all these experts, they could choose as many options as they wanted. If they liked a bunch of different voting methods, they could choose all those voting methods.
I think out of the outcome from that, one of the more interesting things was … There was a clear winner, but there was also a clear loser in the bunch, and the interesting part about that, the loser, the losing voting method out of this group of experts, got zero approvals.
Robert Wiblin: Can I guess which one it is?
Aaron Hamlin: Feel free to toss in your guess.
Robert Wiblin: First-past-the-post, right?
Aaron Hamlin: Yeah, it was first-past-the-post.
Robert Wiblin: The top one was approval voting, I’m imaging?
Aaron Hamlin: Yep, it was approval voting.
Robert Wiblin: Yeah, so an approval voting system approves of approval voting, it’s very suspicious.
Aaron Hamlin: Instant runoff voting did do pretty well in that mini informal election, but approval voting was able to beat it pretty solidly, and then everybody just destroyed first-past-the-post, which is telling, and a bit depressing, because first-past-the-post is used all over the world for single-winner elections.
In the US, we talked a bit about multi-winner elections using proportional voting methods, the US doesn’t do that; the US uses winner-take-all, that is they have an opportunity to have multi-winner elections, and use proportional voting methods, but instead, they put people into individual districts, and they elect them one at a time, and the way they elect them one at a time, is typically to use first-past-the-post.
If you get lucky, maybe they use a runoff, but as we discussed already, runoff is not going to solve all of your issues.
Robert Wiblin: Yep, okay. It sounds like the reform you’re pursuing is very popular among people who think about this.
Before we move on and talk more about the potential benefits of voting reform, who’s working on it, and how likely it is to actually happen, I wanted to talk about some other potential changes that you could get in political systems, and get your thoughts on whether they’re also important targets for advocacy or research.
Do you think proportional representation, or changes in how parliaments are structured, how many houses you have, and the size of districts, and gerrymandering is another thing, is that reform attractable and potentially very high impact, and maybe why didn’t you choose to work on that instead?
Aaron Hamlin: The reason that we chose first to look at single winner methods was the tractability. In the US, there’s a bit more modern history there.
Although, interestingly, in the first half of the 1900s in the US, single transferable vote was used in a number of cities, and over a dozen cities, and was repealed in a number of them, except for really one for city council, which is Cambridge, Massachusetts.
Robert Wiblin: Yeah. In quite a lot of European countries, I guess Germany most notably, that they have these proportional representation system that allows for quite a lot of different parties, and these fairly large districts that elect, model for different candidates to the parliament.
Do you have a view on whether that’s a good thing or a bad thing, setting aside whether it’s actually possible to get that in the United States?
Aaron Hamlin: Yeah, I would tend to say that compared to the United States, that it’s better, significantly better. One is, in the United States, we have a big issue with gerrymandering, which is just politicians drawing the lines to figure out who their own voters are, which is on its face, just slaps of nonsense, of something that shouldn’t be an issue.
What’s interesting about that is a lot of people look at different solutions for that, whether it be independent commissions, or a computerized line drawing as a way to have that be fair, but it seems to be, I would argue, a bit intrinsic to single-member districts, the type of bad outcomes that you can get with winner-take-all.
It’s not to say that it can’t be worsened by gerrymandering, and lessened by these independent commissions, but there are some inherent issues with winner-take-all; as an example, Canada has used independent commission since 1964, and the last two elections in Canada, we saw what are called false majorities, which means that a party has less than half of the votes, and yet, it has more than half of the seats, which is something that doesn’t make sense.
If you get fewer than half the votes, you shouldn’t get more than half of the seats, and that happened in the last two elections, and you would think that, if we were doing some fairer line drawing, that shouldn’t happen, but it does happen, and it seems to a consequence of winner-take-all elections.
If you use something like proportional representation, you see that these false majorities, or these manufactured majorities, you see that happen a lot less often when you use a proportional system.
Robert Wiblin: When I wrote on social media that I was going to interview you, quite a lot of people got in touch, asking about increasing voter turnout as an alternative to doing voting system reform. Do you think that that’s a valuable intervention, and something that can be done in practice?
Aaron Hamlin: I think part of it is how you increase better turnout, and you also have to ask why are people not turning out in the first place.
If you’re giving them a tool that’s not very effective, by using what is arguably the worst voting method there is for single winner elections, then you’re probably not going to get a whole lot of people excited about the effectiveness of their vote. Maybe one way to increase voter turnout is to give a real tool that allows them to cast a much more meaningful vote.
Another interesting part about voter turnout is that if you think about this in terms of sampling, so like when you’re doing a statistical sample, for instance, you don’t need to sample the entire population; you can sample a part of the population, and so long as that sample is representative of the population as a whole, it doesn’t matter that you didn’t sample everybody.
The issue becomes when the sample that you have is not representative, and so we have it so that systematically, certain people are not turning out to vote, perhaps because we’re giving them a terrible voting method, and their not feeling empowered, then if we create an incentive for them to turnout and vote, then I think we’re looking at a better scenario.
It’s not so much there that the voter turnout is higher, it’s that the people who turn out are more representative of the electorate as a whole. Another interesting idea here is that with voter turnout, we could also perhaps look at that as a proxy measure for the kind of faith that people have in the political system.
That if people aren’t turning out, and this goes back to the idea maybe they don’t feel like they are empowered as they should be, and even if they’re not recognizing the system main issue, perhaps they just think I don’t feel like I have an effectual vote.
Perhaps if they did have a meaningful way to vote, such as with the meaningful voting method, and they saw that is an issue, perhaps that would change their likelihood of turning out, and also making it so that the people who do vote, are more reflective of the entire electorate.
Robert Wiblin: A number of listeners also wanted me to ask you about possibilities for unequal vote weighting. Like an example of that might be requiring citizens to show an understanding of the electoral system in order to vote, or giving extra weight to voters who, for whatever reason, are considered wiser contributors to the system?
Obviously that has a pretty terrible history in the United States, as being used as a way of excluding people of color from voting, but can you see any way that unequal voter weighting could be used to improve the system?
Aaron Hamlin: Probably not. I would much rather go the route of more meaningful voting methods. The idea of giving a purposefully … Giving people less weight, and there, you’re actually directly violating one person, one vote. Yeah, I can’t see a whole lot of good coming from that.
Robert Wiblin: Yeah. I used to be interested in possibilities for doing this in order to improve the understanding of the electorate, so you would give people who had a better understanding of policy issues more weight, but I think it’s important to explain to some listeners who might sympathetic to this, why I no longer think that that’s a very good idea.
Basically, to begin with, it’s an absolute political non starter, because people look at that and they see opportunities for all kinds of mischief, to basically exclude them, or to exclude a particular groups from being represented. A vote is two things, it’s both people contributing information that they know about, what kind of policies would be effective, and it’s also a whip that voters to have to make sure that politicians actually care about their welfare.
When particular groups are underrepresented, or not able to vote, or their votes aren’t given as much weight, then it tends to be the case that politicians spend much less money on them, and just don’t care about their welfare terribly much, so you can see that with Puerto Rico perhaps after a hurricane that went through there; they don’t get to vote for members of congress, and they don’t get to vote for president, and that seems to be reflected in the low level of interest that politicians have in taking care of their welfare.
The US constitution in the 14th amendment, precludes giving different weight to different voters, and also precludes trying to prevent people from voting, and I think with good reason, because it’s trying to make sure that politicians can’t choose their voters in this way, in order to say people who share my views, it’s because they’re smarter, and they should be given most of the weight, and people who disagree with me, it’s just because they’re misinformed and we should find ways of preventing them from being part of the electorate.
Aaron Hamlin: Yeah. You particularly don’t want people, designing the test, being the same people who are already elected.
Robert Wiblin: Right, but who else is going to do it, right? It faces some serious problems. I’ll put up some links to discussion of that issue, which I think listener might be interested in, but I guess this raises the general issue of voter ignorance, because if you take surveys of the electorate, you find that often they have very poor understanding of the most basic issues of how the system is designed.
I think most people don’t know how many senators their state has, let alone any more complicated information that they might need in order to vote wisely. Does that imply anything about voting reform? Does it suggest that we have to keep the system extremely simple, because just most people are trying to live their lives, and don’t have much time to allocate to thinking about voting?
Aaron Hamlin: I think, in general, if, particularly for single winner methods, if you have a solution that works reasonably well, and avoids a lot of the anomalies that you like to avoid, and it’s simple, then you should probably go with that method, until you get to the point where your electorate is more sophisticated, and then maybe after that point, then maybe you can consider more complicated voting methods.
That’s a bit of a barrier for proportional methods, because proportional methods, the algorithm component of that tends to be a bit more complicated.
I think with proportional methods, there’s a bit more trust involved, so people are just looking at the results, and looking at independent verification in order to have the confidence in the system, so that may be a bit of an exception to the rule, but you need to give people a bit of experience with alternative voting methods.
Letting them see the improved outcomes that can occur, and then after that confidence level is built up, then exposing them to more and more ideal solutions as they become more comfortable with added complexity.
Robert Wiblin: Okay, well, let’s return to assessing how good voting reform is as a cause, from an effective altruist, or an 80,000 Hours standpoint. We evaluate causes, as most of listeners will know, on three criteria; one is scale, that is how large with the benefits be.
Another is neglectedness, so how many people are already working on it. A third is tractability, which is how likely are you to be able to …
How hard is the problem to solve. Let’s take each of these in turn. Is there a way of measuring how large the benefits will be from having voting reform? Is there nowhere that this has been done, and we can see the benefits in terms of improved welfare?
Aaron Hamlin: In terms of improved welfare, the concept that we’re looking at is we have people who are elected, and these people who are elected are in charge of spending vasts amounts of money; more than the richest of the rich can spend.
In addition to being able to spend these vast amounts of money, they also determine policies that affect our day-to-day lives, in very substantial ways, and so the idea of choosing the voting method is that we’re given an effectual way to be able to decide who sits in those seats, that are able to make those types of decisions for us. That’s the underlying concept for why this makes sense as a target.
Robert Wiblin: Do we have a way of measuring the benefits between different countries, or different jurisdictions, where some of them have better voting systems than others today? In some places you have first-past-the-post voting, which is particularly bad.
In other countries, you have instant runoff, or single transferable vote, or multi-member districts. Can we see better governance in the countries that have, what we consider, better voting systems?
Aaron Hamlin: From what I’ve seen, so for instance, I believe it’s in Pippa Norris’ book, Electoral Engineering, when she had looked at, for instance, proportional voting methods compared to non-proportional voting methods, one of the things that stood out was, one, you had more women in government, particularly with closed party-list systems, but another that stood out was more egalitarian systems.
You had less income inequality compared to non-PR countries. That’s one area that I’ve read in terms of different types of social outcomes that have been different.
Robert Wiblin: You think just because government is so influential and the amount of money that it spends, and the amount of harm it can do, if the people in charge will make bad decisions, that having a system that better reflects the preferences of the electorate, is on its face, it’s going to be extremely important, even if we can’t precisely measure its value in dollars, or how much better policy it would be.
Aaron Hamlin: I would say so, yes.
Robert Wiblin: Okay. How big do you think this could get? My understanding is the Center for Election Science is pushing for approval voting to get tested at a local level, and then I guess you’re hoping to get it implemented in a lot more elections. How big could you see it getting? Could it end up being the system that we use to elect governments at the national level around the world?
Aaron Hamlin: I think so. Certainly not immediately, but the simplicity of this, and that there’s so many good things about approval voting, that it’s a difficult thing to put back in the bag, once it’s out there.
The practical approach for this is there’s been decades of research on approval voting in the literature, so it’s gone through the rigors of peer review, and scholarly analysis, and so the idea from here is just to get it implemented at the local level, show proof of concept, and then replicate that in other areas in the general geographic area, and then from there, go to other areas, in addition to larger cities. Then once you have a certain concentration from there, you could start to target states.
The way that government is structured in the US, once you hit the state level, you can start to think about hitting the federal level at the same time. Perhaps at a certain point down the road, once there’s enough saturation, I don’t expect this to be anytime soon, but further down the road, you can potentially look at something like approval voting being able to elect the president, which is due to the nice features about approval voting.
Because approval voting is so simple, it has some nice features called, this one particular feature called, precinct summability, which means that you could take totals from jurisdictions or states, and then take those totals, and then sum them up with other areas, and that makes it so that you can have totals at the national level.
The other nice component of this, is that if you have some states that are using plurality voting, and other states that are using approval voting, for the way that they do their votes, you can add those together in a way that is not really practical with other voting methods, so that if you have some hold-out states during the process of moving towards a system that elects a president using approval voting, you can do that along the way.
There are different approaches for doing that, like one approach right now is of using a national popular vote with priority voting is using, what’s called an interstate compact, which is an agreement between states, that has them sign on to allocate their electoral votes, once enough other states have signed on, so that their electoral votes total 270, which is more than half the total electoral votes.
Then they would just go ahead and have their electoral votes go towards the national popular vote winner, regardless of what their state’s national popular vote winner is. It’s difficult to have that kind of compact when an individual state is using a voting method other than plurality voting, or approval voting.
It becomes more challenging to do that with other voting methods, whereas with approval voting, you can accomplish that and have that same kind of tactic of using that interstate compact. That’s something that’s possible, albeit way down the road, after being able to build up the base of support for this voting method, at the local level, and then building up to the state level, and then federal level.
Robert Wiblin: Okay, so we’ll come back to practical impediments actually getting voting systems changed, but I want to perhaps pause for a moment, and question whether actually this voting reform would be good, and whether it would improve decisions very much in reality.
Why isn’t approval voting already in use, and why don’t lots of people already support this campaign? Are there potentially good reasons that even most people who are very informed in the political system aren’t talking about this issue? Are there institutions that have a self-interested reason to have a better voting system?
Like for example, companies, well, public companies that have to vote for members of the board. Why are almost all of these organizations ignoring approval voting if it’s such a clear win?
Aaron Hamlin: Part of that is the voting method is really an invisible issue. I imagine that there are a number of listeners, who before this podcast, didn’t really think about the voting method in and of itself, or think about how it could really change things at a fundamental level. It’s like being a fish in water and not really recognizing what water is, it’s just this thing that’s always there.
I think that’s really a big part of that, is that the voting method really seems invisible, for instance, if you were to go out on the street and ask people, what’s a voting method? My guess is that a lot of those people would just describe plurality voting to you, and not be able to say a voting method is a way that we make a decision based on information that’s collected, and the way that information is calculated to produce a result.
They’re not going to give you that definition; they’re likely to say, well, you choose one person, and the person with the most votes wins. That’s what a voting method is, and there, they’re making the mistake of actually just giving an example of a voting method, and not describing what a voting method is.
Robert Wiblin: It’s mostly just neglect, you think?
Aaron Hamlin: Yeah, yeah, I think a lot of invisibility there, and then the other part just being that there are so many ways of analyzing voting methods, such as how much utility does it offer voters on average, how often does it avoid these anomalies, how does it treat candidates who don’t win, what kind of ballot complexity does it have, does it encourage more people to run, who otherwise wouldn’t run; there are all these components that are influenced by the voting method.
It’s complicated to look at all of those, and see how they interact with any individual voting method, and looking at all that as complicated, as a consequence, it becomes much more difficult to analyze that and come up with a solution for any particular circumstance. As a result, you can get this analysis paralysis as well, even with people who like to look at these kind of things.
Robert Wiblin: Has anyone ever managed to make an argument to you that gave you doubts about whether switching to approval voting would actually be a good thing?
Aaron Hamlin: I haven’t found one that’s been particularly convincing. I’ve been thinking about voting methods for about 10 years now, is when this issue initially came up on my radar. Part of the reason for not really having a lot of pause for, for instance, like things that might go wrong with using approval voting versus plurality voting, is that the current way that we vote with plurality voting is just so awful.
That it’s such an incredibly low bar, and approval voting just clearly beats it at all these criteria, that it’s very difficult to have pause when the alternative is so bad; it’s difficult to have pause.
Robert Wiblin: It’s a lot easier to make an argument that maybe approval voting is not the very best, but to say that it’s not better is a bit outlandish?
Aaron Hamlin: Yeah. Yeah, not better. You’re looking at the worst voting method there is, and then you’re asking about a really good voting method, and trying to find ways that you shouldn’t use this really good voting method, versus this, what is clearly, a terrible voting method.
Robert Wiblin: Okay, so what about this argument, approval voting and the other options that we considered, do you make government more responsive to the electorates preferences and views, but is that actually such a good thing, given that many people in the electorate have crazy views, and really don’t know what they’re talking about a lot of the time. Maybe we want politicians to have more freedom to do what they think is right, rather than more accountability to the electorate.
Aaron Hamlin: I mean, to the degree that the electorate isn’t responsible for getting people in office, well, then that influence has to come from somewhere, and where do we want that influence to come from?
Do we want it to come from vested interests? Do we want it to be random? I would argue that those two are not great alternatives, having the influence be the national consensus.
Robert Wiblin: Yeah. I guess you can have a system where it’s very easy for politicians to get re-elected, so that they just stick around for a long time, and they’re not very accountable for the things that they vote for and don’t, which is obviously bad in some ways, that they could do what they want, even if voters in fact wouldn’t like it, if they knew what was going on.
On the other hand, I suppose if you think that the electorate is very poorly informed, you might think that the judgment of this career politician, even if they don’t have great incentives, at least they kind of know something about politics, because it’s their profession. I suppose this is one way.
It’s a bit of a perverse argument, but it’s one way that you could imagine that, making politics more representative of the views of the general public might not be so helpful, as if you just don’t trust actually the views of the public.
Aaron Hamlin: I think part of why some of this idea comes up, in terms of, well, maybe the public has some crazy ideas, and we really don’t want them to have as much say; I think part of the reason that that comes up is because we’re using a measurement instrument that is so bad at gaging the public opinion, and we’ve been relying on that measuring instrument to inform our views on what the public thinks.
Going back, I think part of the idea is we have perhaps a bad idea, in terms of what the public thinks, because we’re using plurality voting to measure what the public thinks, and that is not a good measurement tool.
Robert Wiblin: I guess, yeah, in thinking of this question, I was perhaps more influenced by this book, The Myth of The Rational Voter by Bryan Caplan, where he goes through public attitudes on a whole lot of different economic policy issues, and shows that they’re completely at odds with expert consensus within economics.
Now, you could think, well, actually this shows that economists are biased and misinformed, but Bryan Caplan, and I suppose I, as someone who studied economics, I’m inclined to think that, no, it’s actually the experts who are right, and it’s the general public that hasn’t really thought about benefits from trades so much, and a bunch of other areas where there’s quite a lot of consensus among economists.
If you think the public maybe has better incentives to figure out what’s in the interest, but each individual person is only putting in a fairly small amount of time studying these issues, and experts, even though perhaps they’re a bit self interested, they know their particular niche area very well, and might just have a more informed judgment about the consequences of different policies.
If you really buy that, then it could call for all kinds of different reforms. Potentially you could have citizens juries where you call up a small fraction of the population each year, and get them to intensely study the different candidates and the different policy positions that they take, and hear from different experts about their views.
Maybe this is something that, voting, a system reform can’t deal with. I guess I do have some doubts about whether ensuring that politicians are strongly held to account by the electorate is always an unmitigated good.
Aaron Hamlin: Yeah. I have heard of different ideas, like citizen juries before, and I think, certainly, ideas like that certainly have a place in politics, particularly when analyzing complex issues, but I also think it’s important to go back and think of all the ways that a voting method influences the political landscape.
A voting method, it doesn’t just determine the winner; it gives us a flavor for all the candidates involved, and on top of that, it actually influences what your options are to begin with. As an example, in the 2016 election, Michael Bloomberg decided not to run in the election, because he thought he was going to split the vote with Hillary Clinton, and have Trump win, so that’s why he didn’t run.
We have a voting method that causes people not to run, because they’re afraid of mucking up the election, and nobody wants to be accused of being a spoiler as well, and nobody wants to go in and think, well, I don’t have the same kind of name recognition as other people, if people just hear my views, that’s not going to be enough; they’re not going to vote for me.
Well, if you’re using a voting method that gives people a fair shake, that also affects the number and quality of candidates that you’re likely to have as well. It’s also likely to influence other types of policy too, for instance, in the US, the US has some of the worst ballot access laws there are.
You have to, in some cases, collect tens of thousands, in some states, over 100,000 signatures just to get on the ballot for some offices, and that is really just absurd.
Robert Wiblin: It’s just a way of excluding competition.
Aaron Hamlin: It is. Part of the reason that they’re doing that, it’s not accidental, and it’s not just a matter of not wanting competition, albeit, surely that that’s likely one component, to back up again in thinking about when a voting method matters, a voting method matters when you have more than two candidates.
Plurality voting, even if we have a candidate that does really poorly, you can have a really poorly performing candidate change the election outcome, merely by their presence. Major parties are aware of that, and they have a memory for third parties that have mucked up their elections, and they are not happy about that.
They have a couple choices before them; they could, one, change to another voting method, which you get into this issue of people who are elected not wanting to change the way that they’re elected, or they could address the problem by not having more than two candidates, or making it less likely that more than two candidates are going to run.
That’s been the option that they’ve taken in a number of circumstances, and they’ve taken that option by making ballot access really hard to do. If we’re using a better voting method, that’s also likely to change who is able to run, the types of laws that we have, governing who’s able to be on the ballot, and arguably, if we have more choices, I would think that we’re more likely to have people who would run, who otherwise wouldn’t.
Some of those people being people who are experts on particular policy issues, say more scientists, more economists, more academics who are experts on subject matter that are inherent to being able to have a proper governance, or a governance that runs well.
By having those people on the ballot through the incentive of having a proper voting method, perhaps we’d see some of these other candidates winning, and as a consequence, some of the more mature election policies that we would like.
Robert Wiblin: Do you think that approval voting would result in a breakdown of the two party system, or would it consolidate it? Do we have a sense of that?
Aaron Hamlin: I think it would help.
Robert Wiblin: In which direction?
Aaron Hamlin: I think it would help to encourage more independents getting elected, and more third parties. I suspect that it wouldn’t have the same kind of effect as you would have with a proportional voting method, because you’re only getting one of those elements.
If we go back to Duverger’s law, there’s two factors; one being the threshold, it’s a single winner voting method, so the threshold is always going to be hard. Approval voting doesn’t address that factor.
The other factor though that it does address is that you can support your honest favorite now, and that’s a big difference. Now if you’re a third party, you get that accurate reflection of support, and even if you don’t win that first election, if you got good ideas, you can build on that.
That’s something that third parties simply do not have right now, but if you’re using approval voting, people can look at your policies, it doesn’t matter if it looks like you’re not going to win; they can support your ideas, and as a voter, you’re incentivized to do that under approval voting, even if you don’t think your candidate’s going to win, because you want your candidate to have legitimacy.
Right now, you have a number of candidates who are running, or who otherwise would run, that are not able to get that legitimacy, or fear that they can’t have that legitimacy, not because their ideas aren’t good, but because the voting method doesn’t allow that support to be reflected.
Robert Wiblin: Okay. On the other hand though, it seemed like earlier you were thinking that approval voting would encourage the election of centrist consensus candidates who are moderately appealing to everyone, which seems like it might …
I guess you could imagine a third new party, that was a centrist party, that would now do quite well. Is there a tension between the idea that it would encourage a wider range of different parties to actually get members elected, and the idea that would not encourage extremist candidates to get elected?
Aaron Hamlin: Approval voting does tend to elect more moderate candidates, as we’ve seen through polling, and through just technical analysis. We would expect that … Well, one other kind of strategy that comes up, so say you’re a major party, and approval voting is getting traction, and you have some other candidates who are getting to be more competitive.
There are some ways to deal with this if you’re a major party, and this is a tactic that major parties have taken previously, and that is, if you see that a competitor is bringing a popular issue to the table, nothing keeps you from co-opting that issue, and with that issue being taken away as a distinction, well, now you can have that territory back.
From a major party’s perspective, you can always co-opt those ideas, and from a voter’s perspective, not a huge loss. I mean, because maybe this candidate that had this novel idea that was really good, wasn’t elected, but why you cared about that candidate, the idea itself, was able to move on.
That’s another idea and that something like approval voting, because of the way that it more accurately reflects support for candidates, even if the winner’s the same, the winner’s views may not necessarily be the same, because of the way that approval voting can affect the dynamics within the election, including how a policy is discussed.
Robert Wiblin: Okay, so that’s very interesting. I guess where I was going with this question about whether it would breakdown the two party system, is I was looking at another possible objection that you could have to whether approval voting is going to be good, which might be that, yes, at the electorate level, it will encourage each member to be more reflective of the views of that electorate.
Perhaps that will then result in a parliament that has a wider range of different parties, and for the parliament to be more reflective of the views of the country as a whole, but having a parliament that reflects the views of the country more broadly does not necessarily mean that the policy outcomes then from that parliament will be terribly reflective.
An archetypal case where you can get a parliament that’s more representative, but produces perverse policy outcomes, would be one where you have three different parties; one on the center-left party that has 45% of the seats in parliament, a center-right part that has another 45% of the seats in parliament.
Then a far-right, or a far-left party that has 10%, and in order then to govern, because you’ve got this somewhat arbitrary 50% threshold in parliament that you need in order to pass legislation, the far-right or far-left party ends up having a very outsized influence, because they get to choose which party governs, and they can potentially have a lot of influence over that coalition.
Potentially, squeezing the parties so that there’s really only two, and they’re both encouraged to compete over the center, could end up being less representative of the full range of views within the country, but then the policy outcomes from the parliament could end up being more centrist than otherwise. Does that make sense?
Aaron Hamlin: That type of dynamic tends to occur, and kind of as you’re describing, with a multi-party legislature, or parliament. Some scholars have taken a look at this, so for instance, Douglas Amy, who is an expert on proportional representation, and has written a lot about proportional representation in the US, and comparing it to other countries.
Normally when you see these coalitions form with multiple parties, the vast majority of the time, you’re seeing a smaller fraction join a coalition with the largest fraction, rather than the second largest.
Really, and you’re seeing that somewhere around 15% of the time when that’s not the case, so the small fraction going and doing a coalition with the second largest majority, for instance, so most of the time, you’re seeing that coalition happen with the largest majority within the parties. Arguably, that’s what you want in terms of what you would like to see happen the majority of the time, you’d like to see that coalition happen with the majority.
That majority is going to be more likely to work with whatever other faction that that lines up with, its policy views. That coalition may change to some degree, based on the issue, and we see that in US elections all the time, sometimes moderate republicans join democrats on certain issues, and vice-versa, and so this is something that we will likely see breakdown, and some cases, based on the legislation that’s being considered.
Robert Wiblin: Yeah. I guess I am slightly clutching at straws here, because I like approval voting, and I like your ideas, but I guess if I had to envisage a future in which approval voting was widely adopted, and then thinking, how could it go wrong?
I think this case of where you get a breakdown, or you get a very wide range of parties getting elected, something like you have in Italy or Germany, and then that not working out very well. Like the coalitions that get formed, and the policies that get chosen then at that second stage, being not representative of the country as a whole. That seems like one of the more plausible ways that things could go wrong.
I mean, most people who I talk to, in America, and I think the UK and Australia, are very enthusiastic about the idea of having a wider range of parties, and having parliament a bit more diverse, but I think if you look at how that actually works out in countries like Germany, or Italy, or Spain, or Sweden, it’s a mixed bag.
Sometimes it works well, and sometimes you get coalitions that are good, and sometimes you get pretty weird coalitions. In Germany now, we’re seeing a case where there’s sufficiently many far-left and far-right parties, that the only way to form a reasonable government is to just constantly have national unity governments of center-left and center-right parties, which in fact means, that in practice, there’s almost less choice for voters.
Aaron Hamlin: In terms of parties and how approval voting influences that, so here, when we’re talking about approval voting, we’ve been talking about the single winner form. You can really take any kind of single winner voting method, plop a different algorithm on it, and turn it into a multi-winner proportional voting method, but here, we’re talking about single winner approval voting.
Single winner approval voting is not a proportional voting method. I would really not expect it to generate that number of parties. You would see major parties have to be responsive, but I do not expect the higher number of parties. Say, like I mentioned before, for countries that do use proportional methods, they have safeguards against tons of parties, and some of those safeguards are having specific thresholds.
You can use those mechanisms to make sure that you don’t get into a situation where you’ve got a bunch of parties that have less than one percent, and yet they still have seats, for instance.
Robert Wiblin: Yeah, no, that’s a very good point. Okay, so let’s move onto the question of neglectedness. Who else is working on this probably, if anyone in academia, or in government, and nonprofits?
Aaron Hamlin: Within academia, you have … On our advisory board, we have a number of academics, Steven Brams at NYU, basically one of the major pioneers behind approval voting.
We have other folks, such as Jean-François Lisée, also one of our advisors working on approval voting. Marc Kilgour, an academic in Canada who looks at approval voting, as well as multi-winner systems, and then there are just a whole slew of academics that are looking at voting methods in general. I think within the academic sphere, there’s no shortage of contribution within this space.
Robert Wiblin: Okay, so in academia, there’s a lot of research on voting methods. What about actual advocacy for changing electoral systems, or voting systems?
Aaron Hamlin: The other organization that pushes for alternative voting methods, Fair Vote been around for 25 years, and they are the main advocates for instant runoff voting, and they’ve had some success getting that implemented. Other than them, there’s really not much out there in terms of pushing the voting method itself, besides them and the Center for Election Science, us.
Robert Wiblin: Do any people work on this in government, or in parliamentary committees or anything like that?
Aaron Hamlin: You can have some commissions that come up studying voting methods within a locality; we’ve seen that happen before. Sometimes at the state level, but that’s about it. Those organizations tend to look to other experts, or non-profits, like ourselves, but then you run into the issue of has this particular voting method been used before, and that puts us back to why we’re doing what we’re doing.
Robert Wiblin: Okay, so it’s a pretty thin field outside academic research into the theory of different voting methods, and some experiments with it. Is everything overseas? Are there other countries where this is a more of a live issue that there’s lots of groups working on?
Aaron Hamlin: We haven’t seen a whole lot with regard to single winner voting methods. There are a number of different multi-winner methods, in terms of how a government would elect its legislature or parliament, but unfortunately, when we look at other examples, we see a lot of first-past-the-post, which is really saddening for executive offices, otherwise maybe you see runoffs, or instant runoff voting.
With other organizations, there are other international organizations that look at election methods that are non-profits, think-tanks, and organizations that help democracies get off the ground, but even when those groups have been at work, the depressing part is they keep pushing these really objectively bad voting methods, when we’re talking about single winner elections.
Robert Wiblin: That’s really interesting. Yeah, why is that pro-democracy groups in general just don’t really care about this question of whether the voting system then actually causes people to be able to express their views accurately?
Aaron Hamlin: It’s weird, because they’re certainly thinking about it with the legislatures themselves, but there are some situations where you have to use a single winner voting method. If you’re dealing with an executive office, like president, or governor, or mayor, you have to use a single winner voting method, there’s no way around that.
I don’t know if it’s just the debate that’s within the literature about comparing these different voting methods that gives this analysis paralysis, where they just say we’ve been doing first-past-the-post for however long, let’s just keep doing that, even though when you look at it, it’s just objectively a bad voting method.
Robert Wiblin: Yeah. I wonder if it’s … You know, people get very excited about democracy versus authoritarianism, but then these details of exactly how democracy plays out is maybe just harder to get people emotionally invested in.
I suppose you’ve also got the question, like America’s system, were to some extent, designed in the 18th century, and now Americans, maybe just because they’re familiar with them, are exporting them all around the world, just because that’s what they’re familiar with, and that’s what they assume democracy is.
Aaron Hamlin: Yeah, it could be just one of those status quo biases, where it’s not something that we think about a lot, and we just say there’s a model, other people have done it, let’s just go ahead and do that, and not put a whole lot of thought about it.
Just one of those invisible issues, or things where just people don’t put a whole lot of thought about it, because they don’t appreciate the amount of impact that it can have by changing to an alternative system.
Robert Wiblin: Okay, so it sounds like there’s something in the order of hundreds of academics studying this field, and then there’s a couple of non-profit groups that I’m guessing have a budget in the millions working on it. Is that about right?
Aaron Hamlin: Yeah, that’s about right.
Robert Wiblin: Then I suppose if you were looking for other resources that are going into tackling the problem, you might look at how many volunteers are there, or what’s the general support among the public of people who bring up the issue, or contribute in some small way to advocating on it. It sounds like it’s pretty small there as well, just because most people aren’t even aware that it’s an issue.
Aaron Hamlin: That’s right. Yeah, at our organization, the Center for Election Science, one of the big jobs that we have before us is outreach and communication, and letting people know that this is an issue.
We do that through not only general communication, and digital outreach, but I think also to the degree that we can convince organizations, in a number of settings, including informal settings, to say, okay, if we’re making a decision that includes more than two choices, at least not use plurality voting. Let’s at least not do that; it doesn’t have to be that hard.
Robert Wiblin: Yep. Okay, so you’re scoring pretty well on the neglectedness criteria. Let’s move on to tractability or solvability. To start with, you’ve already said in brief, but what does the Center for Election Science, what’s your plan to get approval voting or some other voting system implemented on a wide basis?
Aaron Hamlin: What we do as an organization, we look at different voting methods, and using a number of different metrics, whether it be how it tends to get a good outcome, and the way that it affects election dynamics, in general, in different situations that a voting method may work well, or not work well. With the types of strategy that we’re trying to do, is just using ballot initiatives.
While doing that, we can also be doing outreach, so that other organizations are able to make decisions not using plurality voting, and because this particular voting method is so easy, then can be using this as well, and that while, at the same time of helping other cities enact approval voting, people can at the same time be using better voting methods, including approval voting in their own personal lives, and just realizing that this is not something that’s difficult to do, and yet, you still get all the benefits.
Robert Wiblin: If I read your website correctly, at the moment, one of your campaigns is to try to get approval voting used in the city of Fargo as a trial to show that this can work at the city level?
Aaron Hamlin: That’s right.
Robert Wiblin: Once you’ve got an example there, then you hope to take it to other cities, and then to the state level, and then build up gradually as you can show that it works well?
Aaron Hamlin: Yeah. Yeah, that’s right. We were pretty excited with a … The way that Fargo had started out was, they had started a commission that looked at different voting methods, in terms of how they elect their council and their mayor.
One of the people from that commission had reached out to us, and said we’re doing this, and it looks like you do a lot of research on this, it looks like you’re experts in this space; maybe you could help us out.
Through that dialogue, we were able to bring up approval voting, and see how it was received, and it looks like it’s going to be a good fit for them. Right now, we have folks on the ground in Fargo that are building support locally, and we’re there to help run an educational campaign alongside them.
Really, we see it as we’re providing resources, and then based on who reaches out, and who’s interested within different localities, we can work with them as they go ahead and put the initiative on locally, and then we can run the educational campaign alongside them.
Robert Wiblin: How many staff do you have, and what’s their focus? Is it more on the research or advocacy side, or legal, legal issues?
Aaron Hamlin: From the grant that we received from the Open Philanthropy Project, we only put this different space. Initially, we were a budget of about $50,000 a year, which it’s very difficult to do much on that budget.
Only recently has the organization been able to make it so that I’m working exclusively for them, and being full-time, and consider other positions as reason of this grant. Right now, I’m the only employee, but we are hiring two positions specifically, and we’re also looking at a local campaign coordinator for the educational campaign in Fargo.
Robert Wiblin: It sounds like it’s quite hard to get voting reform passed by the politicians by the present system, so you’re mostly going to go through ballot initiatives, and in what fraction of jurisdictions can you change the voting system with a ballot initiative?
Aaron Hamlin: About half of states have a ballot initiative process, and a portion of those states have what are called Home Rule, which means that a locality can choose its own voting method without having to get permission from the state.
There are differences around that, so you could have those states pass legislation to give permission to localities, but that’s an extra step. Right now, we’re limited to states that offer this ballot initiative process within cities.
Robert Wiblin: What do you think are the odds of succeeding? Of getting this voting reform in at least a few places, and then getting attention elsewhere, and trying to build momentum?
Aaron Hamlin: Surprisingly, really good. I think from an outsider’s perspective, it may look as something like this, so for instance, with Inside Philanthropy, they did an article on this recently, looking at our mission and what our strategy was, and they were a bit dubious about us being able to pass these initiatives, but there’s actually a really good track record in the US for passing ballot initiatives on single winner voting methods, so we expect the likelihood to be pretty high.
The way that we look at it is instant runoff voting has been passed as a ballot initiative in a number of cities, but we see approval voting as producing better outcomes, and having better political dynamics compared to instant runoff voting, and it’s also approval voting is so much easier, and so it avoids a lot of those issues.
We see it as something, well, if instant runoff voting can do this, then surely, a simpler voting method that produces good outcomes and has good dynamics should also be able to do this.
Robert Wiblin: Yeah, for people who are skeptical, can you give some cases where these ballot initiatives were passed, and were they passed easily, or was it close?
Aaron Hamlin: San Francisco is one place, and Minnesota, it’s been passed in Vermont, in places in North Carolina, Santa Fe, New Mexico. Albeit, there have been some places where they’ve also been repealed, because like I mentioned, IRV can have some anomalous results sometimes, and that can produce some anxiety from the electorate.
Although, as a caveat there, so one thing, plurality voting is worse than instant runoff voting, so always keep that in mind, but as a caveat, sometimes people don’t have good reasons for their repeal either; it could just be that they don’t understand the system, or the political party that had a lot of clout didn’t get its way as a result of the voting method, and they pushed up a fit, and used its complexity as a way to leverage their objective.
Just because a voting method gets repealed, doesn’t mean it’s getting repealed for a good reason as well.
Robert Wiblin: If this is so politically practical, then why do most places in the US, and indeed, most places in the world, still use such a terrible old voting system?
Aaron Hamlin: Part of it may just be that some of the more modern analysis, this hasn’t caught on just yet, and it seems like part of it is that status quo bias.
There’s this really annoying argument that, well, we’ve using this method for a really long time, it seems like the world hasn’t burned down; let’s just keep using this method and the mere fact that something has been used a long time is not an argument for it being good.
Robert Wiblin: Also true of bloodletting at one point.
Aaron Hamlin: Yes, you can do terrible things for long periods of time, it doesn’t make them good.
Robert Wiblin: I guess at some point, do you expect to see, like if you started building momentum, wouldn’t the existing interest start mobilizing against you?
My understanding that, in the UKs referendum on alternative voting, the two main parties, labor and the conservatives both organized strongly against it, because they expected to lose out, if the voting system changed. I suppose, could you imagine a system where both democrats and republicans start running active campaigns against approval voting, because they prefer first-past-the-post?
Aaron Hamlin: Yeah, that would be interesting to see both major parties attempt to do that. One interesting thing is major parties, although we’ve talked a bit about empowering voters through approval voting, major parties do stand to have some benefit here as well, particularly through the primary.
The Republicans, are like pretty vocal about not really being happy about the candidate that made it through their primary, and part of that is they had 17 candidates at one point, so that is prime territory for plurality voting to run amok.
Major parties really do have an incentive to use a better voting method themselves, if at the very least for primaries, to make sure that they’re able to advance a better candidate to the general election. They can always adapt from there, but it’s not always to their advantage to use this terrible voting method themselves.
Robert Wiblin: Yeah, so that’s interesting, that the major parties could potentially become allies, at least for using this in the primaries. Are there any other potential allies that you could find, that might have a lot more money or a lot more influence than you do, who could help with the campaign?
Aaron Hamlin: Third parties and independents, they stand to benefit quite a lot.
Other organizations that may be, they have particular issues that haven’t been brought up, because perhaps the flag bearer for those issues are third parties or independents, and not getting the kind of sway, so those types of organizations may also join on for approval voting, being that the rationale, that if they’re using approval voting, and then they have a flag bearer, rather that be an independent, or somewhere else that can support that issue, that that issue is not going to get more support.
People that want to see their issue have more clout than it is under a two party dynamic, they may be supportive of approval voting, just so that their issue can be heard more.
Robert Wiblin: What do you think is the biggest impediment you’re likely to face, or the biggest challenge?
Aaron Hamlin: I would say that it hasn’t been used before. That approval voting hasn’t been used to elect people to government office, and that’s not to say that approval voting hasn’t been used at all.
Approval voting is used in organizations, mostly academic organizations, because they tend to appreciate the positivity of a good voting method, but this is also why we haven’t targeted, really, the state level with any type of aggression, because we’re inevitably going to get the response that we haven’t seen this voting method used before, maybe after it has some experience, then we’ll start to be thinking about that.
The main impediment that we have initially is that we haven’t seen this being used for electing people to government office in the US.
Robert Wiblin: Getting it used for the first time might be the hardest step. I mean, I’m a little bit skeptical about that, because instant runoff voting has been used in quite a lot of places, it’s used in many countries around the world, but it doesn’t seem like people have succeeded in getting that implemented very frequently in the US, or the UK either. It seems like there’s a lot of stickiness in voting systems full-stop.
Aaron Hamlin: Yeah. Perhaps optimistically, I would say that approval voting can have a bit more traction that instant runoff voting with some success, and the reason for that is it’s just so easy to do.
Like with instant runoff voting, you’ve got new ballot designs, in many cases, because the algorithm is more complex, and because of the space it takes on the ballot, they have to truncate it to your first three choices, instead of letting you rank more choices.
There are just a number of advantages that approval voting has, because of its simplicity, that I see that as increasing its likelihood for tractability.
Robert Wiblin: Yeah, I suppose in the UK with the alternative vote referendum, the main tactic that the main parties used to discredit it was just that it was complicated, and different, and confusing, and people often didn’t understand how instant runoff would work.
I think that would be a hard campaign to sell when you’re talking about approval voting, when it’s just so obvious that you just mark as many candidates as you like.
Aaron Hamlin: Yeah, as the naysayer there, your campaign would have to center around convincing people that they can’t add.
Robert Wiblin: Right, yeah. I guess I was a bit surprised that the anti-alternative vote campaign worked so well in the UK, but nonetheless, maybe I’m a fool, but I am hopeful that people could understand approval voting.
You mentioned a bunch of places where ballot initiatives to change voting systems succeeded. Are there a lot of cases where it failed as well, that might indicate that this is going to be a little bit harder than we hope?
Aaron Hamlin: The areas in the US where there’s been a failure has been with multi-winner voting methods, so trying to implement single transferable vote, there are some places where that’s been attempted, and has failed. It was attempted in California, and also in Cincinnati, Ohio, about 10 or 15 years ago, and both of those had failed.
When we are looking at single-winner elections, in the US, I mean, they’ve been pretty good at being able to pass. I’m not aware of a ballot initiative that pushed for single winner voting method, in the US, that wasn’t able to pass.
Robert Wiblin: Is the kind of … An issue with electoral reform, or voting system reform, where it’s stable for a long period of time, and then every so often, it’s thrown; the question just becomes a live one again, and maybe that’s the time to strike?
I think you see this with countries, that you have the political system is often very stable, and then you went to some period of turbulence. Historically, it’s often been in revolutions or wars or something, and then everything is thrown into question again, and potentially, you can rewrite the whole constitution, and change the voting system at that time, but otherwise, it’s quite stable.
I mean, I guess, another model for you could be, once you’ve shown that approval voting works, to just be very opportunistic and go anywhere in the world that is currently reassessing its voting system, or its political system as a whole. Does that seem like a potentially sensible approach?
Aaron Hamlin: Yeah.
At the moment, we want to make sure we’re not biting off more than we can chew, so we’re keeping things within the amount of space that fits our capacity, but in the future, I don’t think there’s anything that is preventing us, say quite some time down the road, in being able to look at some other locality in another city, in another country and say, it looks like you’ve had some really big issues here.
Perhaps they’ve reached out to us, and being able to provide them support with being able to run an education campaign, to be able to change the way that they vote locally. We don’t see this as just a solution for folks in the US; this is really a solution for anywhere where you’re electing an executive office.
Robert Wiblin: Yeah. I suppose also if you’re choosing what movie to see, or where to go for dinner.
Aaron Hamlin: Yeah, yeah.
Robert Wiblin: Yeah, I wonder if that’s … Just off the top of my head, I wonder if that’s a way to get people to be more enthusiastic and understanding of alternative voting, would be to encourage them just to use it in their everyday life when they’re making practical decisions as a group.
Aaron Hamlin: Totally, so there’s that with the small group decision making. Also, contests are another big one, so for instance, I went to a chili cook-off with my girlfriend recently, and they were voting on the best chili within this competition, and they were using a plurality ballot, so they had a meat section, and they had a vegetarian section.
We were competing in the vegetarian section, and for each of the sections, you could only choose one candidate, and so when they were going around and showing the ballot, and one of the organizers was there, I saw the ballot, and instantly, I pulled the person aside a little bit, and I said, “Hey, you know you’re using the worst voting method there is for this chili competition.”
Robert Wiblin: Because there’ll be so many entrants, that I imagine that there’d be an extraordinary amount of vote splitting, right?
Aaron Hamlin: Absolutely. There are seven or eight candidates for each section, and so unless you had one chili that was just absolutely awesome, you’re basically just doing a lottery there when you’re doing that many candidates with all that vote splitting.
Robert Wiblin: Your hope is to get approval voting tried at the city level, but how would an election like that really prove to voters considering a ballot initiative, to change their voting system elsewhere, that approval voting is a better voting system? How would they be able to look at what’s happened in Fargo, and say, yes, I want to repeat that; how would they even tell that it’s a good idea?
Aaron Hamlin: Yeah, so a voting method really proves itself when it behaves in a rational way when there are more than two candidates. One thing to keep in mind when looking at a voting method, is a voting method can choose the same winner as another voting method in some circumstances.
The voting method becomes important when you’re dealing with competitive elections that deal with more than two candidates, and so those are really the elections that are important.
When there’s an opportunity to select someone that wouldn’t be a reasonable winner under plurality voting, and so one way to look at that is, within a particular locality that’s using approval voting, for instance, is to have encouragement, and have good education to say, hey, you can have a bunch of candidates in this election, and it’s okay, it’s not going to muck things up.
To be able to show the reflection of support that all these candidates get, even when there’s a competitive election, and to show how you don’t have these bizarre outcomes from people that nobody likes winning an election.
Robert Wiblin: Yeah. I got to say, I kind of remain skeptical of the solvability of this problem.
I think it’s just the gut reaction that I have, because it’s something that people could have tried to do, and presumably people have thought about changing the voting system, maybe not lots of voters, but people who understand politics well, and it just hasn’t happened in many countries for a long time, which just makes me wonder, maybe there’s some impediments to doing this that we’re not thinking of here, and that I haven’t thought of and asked you about.
Do you have any response to that?
Aaron Hamlin: The way that this is normally done, like I mentioned before, is through ballot initiatives. If there was a hidden impediment, what we would expect to see is a bunch of failed ballot initiatives, and yet, that’s not something that we’ve seen.
Given that, perhaps that should cause us to step back, and say, well, why haven’t people done this as a ballot initiative? Well, then it’s more likely a complexity issue; something that’s just not maybe on people’s minds.
Going back to that invisibility issue as well, and then thinking, okay, are there groups out there that are interested in this, and can do this, and to the degree that there are, and they’re running these ballot initiatives, I think that’s the degree that you can look at its tractability.
Now if there are groups who are out there doing this, and you just see failed ballot initiative after ballot initiative, then perhaps it makes sense to look more down that road.
Robert Wiblin: How hard is it to get a ballot initiative up? Do you have to get lots of signatures of support?
Aaron Hamlin: It varies depending on the city, and the size of the city. With Fargo, it’s a bit easier to get something on the ballot at the local level, but still, even something goes on the ballot, you still have to run an educational campaign alongside that.
Robert Wiblin: Yeah. I’ve put you through your paces here, talking about the solvability of this problem, and whether it would be a good or bad thing. You just received $600,000 or recently received $600,000 from Open Philanthropy Project. Do you have more room for more funding? Is there extra things that you could fund, if listeners potentially wanted to donate more money to the Center for Election Science?
Aaron Hamlin: Yes. I’m sure that was a surprise answer.
Robert Wiblin: Yeah. Well, you could just say you’re going as quickly as you can, but yeah, no, make your pitch.
Aaron Hamlin: One of the nice things about this is that it scales up well. Open Philanthropy Project really helped us deal with our budget for this year, and really helping us create the foundation that we need in order to push forward.
Now, in terms of scalability, we can be doing this in other cities, and so we’re starting at something where we see as low hanging fruit, but once we have a good procedure, and we have good staff experience and institutional memory to be able to replicate this, then we can target other cities, and doing that is not cheap.
Here, we’re talking about doing it at a mid-sized city; Fargo is about 110/115,000 people, but doing that to several other cities next year of a similar size, and then maybe doing it at half a dozen to 10 cities the next year, and then targeting larger cities.
Here, I mean, it takes much more coordination and staff to be able to coordinate campaigns at that kind of level, and to be able to run these educational campaigns at all of these cities. Particularly as the population within cities is larger; you’re talking about a higher cost, because you have to reach out to more people.
Robert Wiblin: If someone suddenly gave you $10 million dollars, do you have a sense of how hard it would be to hire people to run these campaigns in many different locations at once?
Aaron Hamlin: I think we have a pretty good hiring mechanism. There, I think we would just going and targeting those individual locations, and identifying folks who have experience running ballot initiatives, and then having them be paired with people who are already interested within this space on the ground.
One of the ways that helps us with that is we’re a pretty friendly organization, so we coordinate and build relationships with other organizations as well. There are some organizations out there that have been a bit shy about jumping onto approval voting, because of its lack of use in government elections.
Once we’re able to get the ball rolling, that helps us build those relationships even further, and help them become more public, so then it’s not just a matter of us building up our staff, and being able to reach out to these new cities, but now, once that ball is rolling more, we can be using the relationships that we’re having with other people, who are no onboard with joining on with our mission, given that proof of concept has been demonstrated.
Robert Wiblin: Okay, yeah, speaking of hiring, let’s move on to a section where we can actually offer some career advice to listeners who might be interested in working on voting reform.
I think that you’re hiring for a couple of positions at the moment; what are they, and what skills do you most need as an organization in general?
Aaron Hamlin: Right now, we have a rolling deadline for two positions. One is director of operations and programs, the other is director of philanthropy. I’ll start with the director of operations and programs, so that position is looking at doing a lot of digital outreach, so helping us with our website.
At the moment, if you go to our website at electology.org, you’ll see that it’s the work of a lot of volunteers to get it to where it is. This person within this position would be helping us to contract out doing work for our website, helping our digital outreach, as well as being able to help us with coordinating programs, including being able to help coordinate and help with the strategy on these educational campaigns on the ground.
They wouldn’t be doing the educational campaign directly, we would be hiring someone else for that, but they would be helping to coordinate that effort, as well as help building internal strategy to be able to do the type of relationship building that we need to do with other cities, as well as to be able to strategically target what cities we want to be looking at next.
Robert Wiblin: We’ve got the director of operations and programs, and the director of philanthropy. Are you going to be hiring political campaigners as well, or what about researchers, or people who have an academic background in understanding voting, are those potential hires as you scale up?
Aaron Hamlin: Those are potential hires as we scale up. At the moment, we rely pretty heavily on some board members, as well as our board of advisors, so if you look at our Who We Are page on our website, electology.org, you can see that we have some pretty heavy hitters, in terms of our board of advisors.
We rely on them pretty heavily, in terms of figuring out which approaches we want to be taking in terms of particular voting methods, and also, being able to analyze more deeply different types of voting methods, and their attributes.
Robert Wiblin: Do you just want to give a brief description of the kind of person who’d be a good fit for each of those roles, so that someone who’s listening, and would be a good fit, just can’t help but realize that they should think about applying?
Aaron Hamlin: Yeah, so with the director of operations and programs, for both of them really, it really does help to have experience within the nonprofit sector.
Within the operations and programs position, someone who has had experience running complex projects that deal with coordinating a lot of plates at the same time, and being able to have good relationships with people, while coordinating all those tasks.
That’s the type of experience that we want to see, so if you’ve been responsible for coordinating other staff within a project that has a reasonable size budget, that’s the type of experience that we want to see.
Particularly, if you have experience with political campaigns, ballot initiatives in particularly, and understand those logistics, and that’s going to be really helpful as well.
Robert Wiblin: Let’s look beyond just your organization to thinking about the problem area as a whole, and I guess that’s a bit difficult, because it almost doesn’t exist as a coordinated group of people trying to solve a problem.
It’s mostly just academics who are studying it, and might do some advocacy on the side, but I suppose you would hope to scale up, the community that’s trying to solve this problem, to potentially become hundreds or thousands of people with significant budgets.
What would you like people to be studying today, or learning today, so that they’ll be able to contribute as, hopefully, this issue gains traction?
Aaron Hamlin: I think organizational management is really helpful, and also just experience within the nonprofit sector, so reading books, and taking courses on non-profit organization are definitely going to be helpful.
There are some differences within the non-profit sector compared to the for-profit sector that can be material, and absolutely, that’s going to be the case, for instance, with the director of philanthropy spot, for instance, there, we need someone that has experience not only with the grant process, but also experience with fundraising, such as doing our communication and relationship building all the way from prospecting and cultivation, all the way to stewardship with large donors.
It may seem too, so one of the reasons that we’re also looking at that director of philanthropy position is because it’s really exciting when you have a large source of funding coming in, because it means that you have more resources and capacity to be able to fulfill your mission, but just like any kind of sector, you want to make sure that you have diversified funding streams, because you don’t want to have a lot of pressure on an individual funder.
It just makes sense from a business point of view, so having someone come in, that has experience within the nonprofit sector, and really understands that space, is really going to be an advantage.
In addition to that, a side set of skills is good scientific literacy, and understanding how to analyze information, so even things like having a fundamental science background, is going to be able to help with the type of problem solving, and skills that you’re going to need in order to be able to understand this space as well.
You need to be able to, one, be able to be good at dealing with complex tasks and understanding the nonprofit sector, but also, you need to have a reasonable ability to be able to understand the topic area as well.
Robert Wiblin: What about people who’ve been involved in major party political campaigns? Wouldn’t that give them good experience for potentially being involved in ballot initiatives in future, and knowing how to combat misinformation that people might try to put out?
Aaron Hamlin: Yeah. I think so too, I think that that would also be the type of skill while, not direct to ballot initiative, there are a lot of components there that are likely to be able to carry over for sure.
Robert Wiblin: What about having more people in academia, or doing more theoretical trials of voting reform? It sounds like that’s in fact where most of the resources are today, so would there be much point in having even more people doing that? Is there the potential to trial approval voting through getting social science grants or something like that, in order to prove that it does actually work in practice?
Aaron Hamlin: Yeah. When we’re able to implement approval voting, certainly being able to see the types of outcomes that it produces, is something that we’re going to be looking at as well.
To be able to look at an election and see this is what it looked like under approval voting; what would it have looked like under plurality voting as well? There are some things that are more challenging to do, because when you change a voting method, which I alluded to before, you change the complete dynamics of an election too.
It’s really difficult to come along and say, this is how this voting method was going to look; how would it have looked with this other voting method? Well, you can do that, but you can only look at that in terms of that snapshot in space, because had you gone back and said, this election was going to be conducted using a different voting method, well, you’re changing the entire dynamics.
Those are the types of challenges that political scientists and social scientists face when they’re trying to study this topic within a particular setting. That’s just one of the things that makes this difficult.
Robert Wiblin: We have a lot of listeners who are very inclined towards mathematics and philosophy and analysis. Is there any point having more theoretical voting people, or is that an area that’s been studied out of all proportion to the actual application in the real world?
Aaron Hamlin: I think there are perhaps some angles to this that haven’t quite been looked at completely, in the literature, that people could do some gap building on, but really, one of the main things that we see is just not …
Like we have overwhelming evidence that this method that we’re using for single winner elections is awful, so no amount of extra data is going to … We only need our confidence level to be at a certain point before we say, we’ve collected enough data; we need to do something about this.
In that respect, I don’t think we need more analysis, but there are other angles within voting methods, and particularly once we have more elections being used with different voting methods, it’s going to be useful to collect data on those elections, and see how they perform compared to others.
Robert Wiblin: Are there any other potential ways that people could contribute to solving this problem? I mean, maybe they could become journalists or go into think-tanks? Are there any options there? I’m just trying to think of any other thing that we haven’t considered already.
Aaron Hamlin: Well, this space really isn’t talked about much in the media, so certainly, having journalists who are interested in this topic, and being able to communicate that, is going to be an advantage, because the kinds of things that people think about are the kinds of things that they hear about from friends, or on media.
To the degree that that becomes more of a popular media issue, is going to change the degree that it’s on people’s radars. Being on people’s radars is really a huge issue, because like mentioned before, the voting method really is an invisible issue, despite its really extreme importance.
Robert Wiblin: You mentioned to me before the interview that you were looking for board members for the Center for Election Science, do board members make a big contribution, and what do they do for a non-profit?
Aaron Hamlin: Within the non-profit, you’re always looking at perspective board members, and we did just do a recruitment effort, where we got a number of really good board members.
In terms of what people are contributing when they join a board, and the types of things that can help prepare them when joining a non-profit board, so first, the kind of things that non-profits expect when you join the board.
Normally, there’s a giving and fundraising component, which is helpful for diversifying an organization’s unrestricted funds, but also if you’re coming in, and you say you have a really good network or people within the political space, or other non-profits that have similar missions, or you have another relationship with people within the polling industry, those types of relationships are really helpful, because a lot of an organization’s capacity is dependent on its network; the board really plays a large role in expanding what that network is.
The other components that a board brings to the table, in addition to things such as advising and those other responsibilities are being able to understand a non-profit framework, such as being able to go through and evaluate a non-profit’s programs, being able to evaluate the executive director, being able to evaluate the board, and making sure that the organization is staying on task as well.
You’re also an ambassador for the organization too, so when you’re going out to conferences or meetings, you’re really a spokesperson for the organization in that situation.
Robert Wiblin: Ideally, you’re looking for someone who’s a bit senior, who has experience with managing projects and organizations and people, and I guess ideally, some experience with non-profits as well, and I suppose even better than that, someone who also understands the problem area extensively.
Aaron Hamlin: Yeah, that’s right. Within the experience, it’s also helpful, particularly as an organization is growing, to have people who have had experience within a non-profit, that isn’t just a volunteer based non-profit, which is where most non-profits start, but also to have experience within a non-profit that has a budget of say a million dollars or more, where there’s staff, and more complex projects are being run.
Robert Wiblin: Interesting. Yeah, so what kinds of listeners should think about, talking about nonprofits, about potentially joining their boards, and is that the right way to go about contributing, that you would approach the non-profit, or should the non-profit approach you?
Aaron Hamlin: I think in a lot of the situations, approaching the non-profit is appropriate.
If you’re looking in this space, and you’re thinking, well, one way I can contribute is by joining a non-profit board, and whether it’s considering our board, or another board within a topic area that you’re interested in, some things that it can be doing just to start are, one, making sure that you have a good network, particularly a network that expands within the space that would benefit the organization that you’re looking at.
Also, reading pretty heavily on non-profit governance and board structure is going to be really helpful, because it can be really exciting to join onto a board, but then at the same time, look at something and say, well, this isn’t quite as exciting as the type of things, why I would want to join on the board.
Volunteering to join a particular project, for instance, but being able to go and be familiar with the material in terms of best practices for non-profit governance, and best practices for within a particular office or a position for a board, or board structure in general; having that background and to be able to demonstrate that you have that knowledge base, and preferably experience, is going to go a long way in being able to demonstrate that you would be a valuable member for an organizational board.
Robert Wiblin: Yeah. Okay, well, let’s come back to talking about the problem area of voting reform, are there any conferences that people ought to go through if they’re interested in getting more involved in this community, in as much as it exists, and what people should they meet, and how can they find them?
Aaron Hamlin: Yeah, so one really exciting opportunity that we took up recently was our chair, [Janice Dru 02:15:48] and I went to a conference called, Unrig the System, which is put together by Represent US, and it’s also another non-profit I really like.
They do a lot of ballot initiatives that deal with anti-corruption efforts, and so going to that, is that type of conference is really exciting, because you learn about all kinds of political activities that people are doing, and the types of mechanisms that they’re using to address these problems, such as gerrymandering, maybe they’re looking at running in politics, could be looking at voting methods, internal corruption, checks and balances; there’s a number of issues.
There’s another conference that I think, for networking purpose, and also, it’s just really fun to go to in terms of learning about the political space, and you see a lot of exciting people, is Politicon, and that’s held in Pasadena, California, normally in the summer months around June and July. It’s in Los Angeles in October 2018
Robert Wiblin: Excellent. Is there a book you can recommend for listeners who’d like to learn more about this topic?
Aaron Hamlin: Oh, yes. Yeah, there is a really nice introductory book on single winner methods called, Gaming The Vote. Gaming The Vote is by William Poundstone. It does a really nice survey on single winner methods, it also does a lot of storytelling, so you’ll hear him go through the story of the 1991 Louisiana Gubernatorial election, which he puts in a lot of really exciting details, so it’s a really fun one to go through.
That’s a good one in terms of single winner methods. There are some other scholars that are really good within this space. Although, there may be some kind of disagreements on little nuances that are there, but David Farrell, he wrote Electoral Systems, which is a pretty good one.
Douglas Amy, he wrote a book called, Real Choices, New Voices, which is a really good history of proportional representation in the US, and also looks at proportional representation in general. He did another one called, Behind the Ballot Box, and that one is a good one on … Kind of another survey on voting methods in general.
Robert Wiblin: Great. Well, then I guess just to finish, given the challenges that confront voting reform, and the fact that you’re part of a fairly small community, what keeps you really excited about tackling this problem, and waking up early in the morning to send emails and organize everything?
Aaron Hamlin: Well, I think the most exciting thing about this is that it has a high potential gain, and it’s pretty clear that, if we don’t do this, it won’t happen. At least not for a long time, and so because of that, we’re working to produce a gain for people, who otherwise wouldn’t experience the benefits of a better voting method. That’s a lot of people, and that’s a lot of benefit, so it’s pretty exciting.
Robert Wiblin: My guest has been Aaron Hamlin. Thanks for coming on the 80,000 Hours Podcast, Aaron.
Aaron Hamlin: Thank you.
Robert Wiblin: Just a reminder to check out the links that we pick out for every show we put together, to help you become a minor expert on the topics we talked about.
Keiran Harris produced this episode of the 80,000 Hours Podcast.
Thanks for joining – talk to you next week.