Podcast 2012-08-20: Interview with Occupy Wall Street’s TJ Frawls on Electoral System Pilot Poll
TJ Frawls from an OWS working group shares his group's poll in New York City, which used alternative voting methods.
TJ Frawls: TJ is the founder of the Occupy Wall Street Politics and Electoral Reform Working Group (PAER). [Now Make Voting Count]
Aaron Hamlin: Executive Director of The Center for Election Science
CES: Welcome to The Center for Election Science podcast. The Center for Election Science is a nonpartisan nonprofit dedicated to informing the public about voting systems. You can find more of our work on the web at electology.org as well as on Facebook and Twitter. I’m Aaron Hamlin, co-founder and president of The Center for Election Science. And today, we’re talking with TJ Frawls. TJ is the founder of the Occupy Wall Street Politics and Election Reform Working Group. He’ll be sharing the results of a voting methods experiment done at Occupy Wall Street in New York City.
CES: Yup, this is Aaron. How are you doing?
TJ: Hey, hi. How are you?
CES: Would you like to go ahead and start by explaining a little bit about the Occupy Wall Street Politics Electoral Reform Working Group and what your role is with that?
TJ: I actually started up the Politics and Electoral Reform Working Group myself about in late September, 2011. The reason I kind of put out the call for the formation of the group was because there were a lot of people at Occupy Wall Street who were concerned about the potential cooptation of the movement by the usual suspects, you know, for instance, the Democratic Party and whatnot. There’s obviously a lot of precedent for that. It happened with the Tea Party and the Republican Party. It has happened to some extent with Occupy and the Democrats. The original idea was to brainstorm proposals for electoral reform to open up our political system to everyone who is not represented by the Democrat and Republican parties, which is a significant plurality of the American public. So that was the idea of the group from the beginning. And then fairly on for a few weeks, we were having near daily meetings. What originally came of that effort was a document called “People before Parties,” which identified eleven or twelve different areas for potential reform: everything from alternative voting methods to independent redistricting, smaller districts, proportional representation, ballot access reform, things of that nature.
CES: Now, I think normally when people are thinking about election reform, they’re thinking about things like barrier to be able to vote. Maybe in the back of their mind they’re thinking ballot access. I don’t know that too many times people are thinking about the voting method itself. So, what really drew that to your attention?
TJ: There was someone in the group who came up with the idea about this voting experiment. And there was significant support for the idea. A lot of people aren’t aware, you know, our voting methods and voting systems. But there is actually a significant amount of awareness of it as a problem. Every once in a while you’ll see an op-ed in the newspaper calling for alternative voting methods. And, as you’re probably well aware, alternative methods have been implemented at the local level in San Francisco and a couple other cities here and there in the North-West and in the Mid-West. And in our group, what we decided to do, we thought the voting experiment was a good idea because it would allow us to gather data on how a specific set of individuals in the aggregate and at the individual level across a variety of voting methods. But it would also give people practical experience with voting methods. And we also hoped that if we did the experiment it would also draw some attention to the issue and get some dialogue going on in the media about this possibility.
CES: Now in this experiment, you took just over three hundred people and gave them i-Pads and asked them to go ahead and participate in a poll where it used different methods to measure their expression about different candidates. Would you be able to tell us a little bit more about what exactly this study was and what you did.
TJ: Yes, we developed a model for the experiment and then one of the members of the team programmed the code for the application. And then we loaded it onto an i-Pad, and then we stood out and asked people if they wanted to participate in the experiment. Over the course of April and May, we went to various Occupy Wall Street events. We just kind of walked right up to people here in New York City. I think the polling, the places where we conducted it were Union Square, Times Square, Central Park, Liberty Plaza, downtown. Basically at Occupy Wall Street events. With this test we had no intention of getting a representative sample of the entire American voting public like you would find in a standard polling organization survey. Although we probably did have enough of a sample to draw those kinds of conclusions about Occupy Wall Street protesters, for us this was the first big test of our model and the program. And what the experiment did was it asked one question under four different voting methods. So the question we used for this was what pollsters and survey organizations call the generic ballot question: “If this year’s election was held today, what parties or candidates would you favor?” And then we offered up a series of six choices, the kind of usual suspect party choices–Democrats, Republicans–but also Greens, Libertarians, Socialists, and independent candidates. And there was also a write-in option under every method. So, everyone who participated in the experiment, they answered that question under Plurality. And then they answered the same question under two of the three alternative voting methods.
CES: When you say, “Plurality,” I think a lot of people, they know what their current voting method is, but they don’t always know what the name is.
CES: So when you say, “Plurality,” what do you mean?
TJ: Plurality is the voting method or the voting system that is commonly used in the United States. It’s also known as First-Past-The Post. And anyone who casts the ballot in a Plurality election, they pick one candidate from the list, and then the candidate with the most votes wins. It’s called Plurality because you don’t actually need a majority to win that kind of an election. For instance, if there were three or four candidates in a given election, then the person with the most votes might only have 38% support, for instance. There was an election (I think in 1996 [1992, particularly]), Clinton won with less than 50% support. That’s our traditional method. It has a specific logic to it. It’s well known among political scientists Plurality, the voting method that we use, tends to result in a two-party system, which a lot of people feel problematic today as the domination of our politics by the Democratic and Republican parties. But there are other voting methods that have a different logic to them and that don’t lock you into what’s called strategic voting. This is one of the big problems with the Plurality system, with the system that we use today, is that even (let’s say): You have someone that’s kind of Libertarian on a lot of issues. Instead of voting for the Libertarian candidate, they might cast their vote for a Republican who they don’t really like because they’re voting against the Democrat–who they like even less. So, we basically kind of have a race to the bottom. This causes us to choose candidates who don’t represent us very well because people are voting defensively, and some people argue dishonestly. And that’s a product not of the political system but just of the logic of our voting method that we employ.
CES: So in giving people other options with the voting methods and giving them exposure to these methods, is that your way of allowing them to express in a way that doesn’t force them into these corners?
TJ: Yeah, we wanted to accomplish a number of different things with the experiment. We wanted to actually gather some actual data and see how the different methods compare with one another. We also wanted to introduce people to these alternative methods. And a lot of people, even people who know about them, might not have a practical experience of actually having to make a decision under an alternative method. And we’re hoping that once we’ve completed the experiment that it would create some dialogue around these issues and potentially create momentum to implement reforms and changes. One of the reasons we decided on the voting methods issue is because it’s actually locally controlled. If a town anywhere in the United States wanted to experiment with alternative voting methods, they don’t have to go through their state legislature. They don’t have to go through Congress. That town can implement this kind of a reform all by itself for its local elections. This the idea of the state as the laboratory of democracy. You don’t need a national movement to do this. This can actually be implemented by everyday people at the grass-roots level in their local government.
CES: Now you looked at a few other different methods besides our traditional method of Plurality Voting. Would you like to talk a bit about those?
TJ: Yes. the three alternatives methods that we tested were Approval Voting, Range [Score] Voting, and Instant Runoff Voting. Instant Runoff might be the kind of alternative method that people are most familiar with. So the traditional method of Plurality Voting, it’s pick one from this list and the person with the most votes wins. Under Ranked Choice (or Instant Runoff Voting as it’s called), an individual ranks their, say, top three choices in their order of preference. And then their votes are counted like an instant runoff. That’s Instant Runoff.
CES: You also mentioned two other methods, Approval and Score Voting.
CES: Would you like to talk about Approval first. I think that’s maybe a little bit simpler.
TJ: Sure. Under Approval Voting, well, let’s compare it to Plurality. Under our traditional method you choose one from the list and that’s it. And the person with the most votes wins. Under the Approval method, each individual [voter] indicates what candidate they approve of. So they just tick off an Approval or as many as they’d be happy with. And then the candidate which was approved of by the most people wins. So that’s Approval. And Range Voting (or Score Voting), the last method that we tested. Approval is basically a simplified form of Range [Score] Voting. Under Range [Score] Voting, each voter rates each individual candidate on a given scale. Say from 0-100 or 1-10. In our experiment, we did it from 0-5. You would give your rating to each candidate, and at the end the candidate with the most points cumulatively wins.
CES: Now, as you did this, were you noticing any kind of differences between these voting methods as some of these results were coming in? For instance, did some voting methods choose winners that other voting methods didn’t, or were there differences in order or support?
TJ: Ah, yes. Actually, it was very interesting. What we found that (as we said in the report), the Plurality method results in anomalous outcomes. And the basis for that conclusion is that the results under the three alternative methods converged with and substantiated one another. Whereas, the results of the Plurality Vote returned anomalous outcome, which did not correspond with anything. It wasn’t even close with the outcome of the three alternative methods. Although the outcome of the three alternative methods were very, very similar results.
CES: What in your mind is telling that the Plurality results were the results that were off rather than the alternative voting methods.
TJ: [laughing] It’s well known that New York is a liberal-leaning, democratic, party town. Although, it has a very strong independent streak in New York City, too. And Occupy Wall Street is very liberal, left leaning. But Occupy Wall Street likes to keep the Democrats kind of at arm’s length. And, as I said earlier, under the Plurality method, people feel compelled to engage in what we call strategic voting, right? You vote for the lesser-of-two-evils rather than someone who you might actually positively support just because you want to ward off the greater-of-the-two-evils. What we found was that under Plurality the Democrats came out on top but with very, very weak support. Almost 70% of the people didn’t vote for the Democrats under Plurality. But the Democrats (if this were an official result or something) would have won, even though 70% of people did not support them. Under the other three methods, the Democrats were; under Approval and Range [Score] Voting, the Democrats came in a distant fourth place behind Greens, Independents, and Socialists.
CES: I’m looking at the Approval results. And for Approval it shows the Green Party at 74% approval. And the Democrats, here they’re showing up in fourth with about 48% approval.
TJ: Yeah, that’s right. That’s right. And if you look at the Plurality results, the Democrats came in first with something like 33 or 34%. [laughing] I mean, they would have won. Quote-Unquote, they would have won if this was somehow official even though the vast majority of people didn’t cast their Plurality vote for them. I think the closest we come to the Plurality results was under IRV [Instant Runoff/Ranked Choice Voting]. When you count the votes, you count it up like an instant runoff. In the first round of voting, the Democrats, they were the largest top-line choice among people in the sample. But in the end, the Green Party still won after you counted up second and third place votes under IRV [Instant Runoff/Ranked Choice Voting]. So even under IRV, the winner of the Plurality test was not the winner of the alternative method.
CES: Did you find any issues with the complexity within the Instant Runoff Voting?
TJ: A member of our team who coded the program for us (we didn’t have an automatic tabulation for IRV). So we had to do that by hand. Whereas, it was much more simple to add up the Approval and Range [Score] votes than IRV. Because under IRV, the ones who come in last place in the first round, you drop them off the list. You take those voters’ second choice votes, and you add them to the top. And then you continue that process until you have a winner. It’s more complicated.
CES: How many rounds did that take for your study?
TJ: It took thirteen rounds to get to the final result in the Instant Runoff.
CES: Are you seeing from this that there are some methods that are able to reflect the will of the voters better than Plurality, even among these alternatives? Or do any of these stand out for you?
TJ: I really think that any of the methods that we tested is superior to Plurality. I really think that it’s the worst method of all the ones we tested.
CES: Do you think that voters, just given that they’re used to Plurality, that any of the complexity within the other methods would create an obstacle?
TJ: I think opponents of alternative methods, they want to make things seem like they’re a lot more complicated than they really are. Because often the opponents of alternative methods are the people who, they remain in the positions where they are because of the logic of the Plurality system.
CES: I also noticed that with your study, looking at the Instant Runoff Voting data, in the 13th round when the Green Party came on top, it only had 46% of the first choice that were transferred over to it, and not greater than 50%.
TJ: I believe that is true. And the reason why that happened was because there was a non-trivial percentage of people who only answered their top choice. In other words, they didn’t enter second and third choice options on their ballot. So the final tally between the Democrats and Greens does not add up to 100%.
CES: Looking also at the data, there’s something that’s interesting. In that, you would think that with the Occupy group that there would be less people supportive of, let’s say, Libertarians. And yet, when you look at, for instance with Approval Voting (which happens to be a system where people can always vote their favorite and you tend to get results that are little bit easier to read and understand), even among this sample it seems to show they’re getting more support. Say, here [Approval Voting] you see 28% whereas with the Plurality they were getting 3% support.
TJ: Yes, exactly! Totally true. You might be underestimating a little bit Libertarian supporters among Occupy Wall Street folks. There’s a significant Libertarian contingent there. But it’s true. There’s a very significant jump in support. As you said, from 3% [Plurality] to 28% [Approval]. Comparing that with the Republicans, for instance, under Plurality the Republicans had 3% and under Approval they had 4.3%. They’re about, pretty much the same. Whereas, the Libertarian number jumped from 3% [Plurality] to 28% [Approval]. Significant, definitely significant.
CES: Now seeing this outcome from the data that you were able to collect, Plurality Voting, how do you see that playing out within our current election? Do you see that creating any kind of clarity within the will of voters or do you see that distorting it all?
TJ: The Plurality system specifically?
TJ: I think it’s definitely distorts it. For instance, just think about the way polling organizations inquire about support for presidential candidates. The question is always, “If elections were held today, would you support or Democrat Barack Obama or Republican Mitt Romney?” Even though, when we go to the polls in November, there are going to be, there might be as many as five or even candidates on a given state’s ballot. The Libertarian Party candidate Gary Johnson, he’s on, he’s looking to be on the ballot in all fifty states. The Green Party candidate, Jill Stein, is going to be on a lot of states, enough to get (theoretically) 270 electoral votes. But as I said before, the Plurality system tends to result in a two-party system. And that’s reinforced by the media and by polling organizations who pretend that the Democrats and the Republicans, that they’re the only political positions out there. And the numbers change significantly in media polls when third party and independent options are provided. But the Plurality system distorts that by making everyone, by creating this false illusion that there are only two choices. And the Plurality system tends to devolve into a two-party state. And my own assessment is that it’s even worse. The system has devolved into a one-party state. In a lot of states across the country, either the Republicans or the Democrats are kind of the only game in town in terms of political choices. So it’s even worse than it being a two-party system. In a lot of places it’s just a one-party system. And that’s not choice at all.
CES: So are you saying that merely giving people a choice on the ballot and giving them other options, that that in itself isn’t enough, that we need to go further and actually change the voting method?
TJ: I think so, yes. In media and polling positions, it’s even reflected the reality of the fact that there are more options even today other than the Democrats and Republicans. There are actually, there are more independents in this country than there are Democrats and Republicans. It’s not going to be very long until independents are the absolute majority [>50%] of the public. So the real question is why would we continue to maintain this artificial two-party system when it no longer corresponds to the political reality of the American electorate. Simple reforms like, for instance, voting methods can now bring about that kind of reasonable change.
CES: Now, I’m looking on your results now for within the Plurality method. And I’m seeing some of the different parties. It looks like they’re getting just very little support. And yet, when I’m looking at something like Approval or Score Voting, it seems like they get a much large support. When we’re seeing that, and we’re seeing Plurality polling done, do you think that plays a role as far as how these parties are treated? Even if they’re not winning, just the fact that they’re given less support with Plurality Voting, do you think that plays out as far as the way their ideas are heard?
TJ: Oh, definitely! The Libertarian and Green Party candidates for president are lucky if they get any coverage whatsoever. Their ideas are basically, they’re not heard. They’re excluded from mainstream discourse. I think they should be included in presidential debates. But they’re very likely to be excluded. And the reason for that is simple. It’s because The Commission on Presidential Debates is literally run by the former head of the Democratic and Republican National Committee[s]. It’s not a disinterested, nonpartisan group; it’s an interested bipartisan group. And they have a vested interest in excluding alternative ideas, positions, and parties, and candidates from the national discourse. And that is reinforced by the media and polling organizations.
CES: After you’ve collected all this data and you’re looking at it, how does that affect your decision making? How are you going to be using this data?
TJ: Our full report on the experiment is online for anyone to read. They can find a full report at our website paercom.net. P-A-E-R-C-O-M dot net. And our hope was that we could create some dialogue around these issues, simple reforms that can be implemented that can open up our political system and result in having a more representative government in the United States. That was our hope. You know, as I said earlier, this is something that’s controlled at the local level. Because if there’s a town out there where people want to implement reform, it can be done. You don’t have to go through your state legislature. You don’t have to go through the Congress. You can go around the parties at the local level.
CES: And what are your group’s plans now moving forward?
TJ: Our plan now is on election day, our group is going to be conducting an exit-poll-style survey in a strategically chosen district in New York City. So we’re hoping to get some more tablets to conduct the experiment outside polling places. You know, exit poll style, on election day. And then that will allow us to take our results from under the alternative methods and compare them with official results from the specific polling places and to see what happens. We’re very interested an excited to see what happens. We’ll be able to compare. Let’s say, official results would show that (I don’t know) X Candidate won. Well, would X Candidate still win using a large sample of the same voters under an alternative method? That’s what we want to find out. So we’re very excited. We’re changing the interface and the program a little bit. We’re tweaking it to make it more efficient and to gather a little bit more data. For instance, we’re planning on collecting some demographic information that way we can be certain that we’re having a representative sample. We want to have number of different teams out there on election day in front of polling places gathering this data that way we can have the largest data set that we can collect that will make our data more robust. We’ll be able to make more precise conclusion on the basis of it.
CES: I think we look forward to hearing more of your work.
TJ: Thanks! Thanks a lot. If anyone out there wants to get in contact with us, they can contact us through the website I mentioned earlier [paercom.net]. We’re going to be doing fundraising over the next few months to save up some money to get some extra tablets. And we’re also going to be looking for volunteers on election day for people to conduct our experiment outside the polling places. And if anyone has any interest in that, please get in contact with us.
CES: Awesome. Thank you, TJ.
TJ: Thanks a lot, Aaron.[music]
CES: This podcast was brought to you by The Center for Election Science. You can find a transcript of this podcast and more of our work at our website, electology.org. If you enjoyed this podcast, you can support us by donating on our webpage and by sharing our work on Facebook and Twitter. You can also follow us at facebook/electology and on Twitter under Election Science. Until next time, thanks for listening.[music]