Roundup: Approval voting and numbers with meaning

While everyone has been enthralled with the electoral reform debate (no, not really), and been gripped with substance over process (no, not really), there was an op-ed in the Citizen last week that I never really had a chance to talk about amidst a number of other things going on, so I thought I’d take a moment now to address it. The issue: the electoral system known as “approval voting.”

So what is it? Basically you take the same ballot you have now, and you mark it for as many people as you want to. Supposedly this discourages strategic voting because you can vote more than once and can vote for both the person your heart wants to vote for, as well as the one you hope to defeat the person in there now. And okay, sure, it’s simple, and sure, it gives you that emotional thrill about being able to vote for more than one person (which I don’t think is that big of a concern for most people, but maybe I’m wrong), and if you do something silly like vote for everyone on the ballot (because they’re all winners for participating?), then it basically cancels out the vote and doesn’t come out any worse off. But I keep going back to the basic question: what problem is this trying to solve?

If that problem is the emotional dissatisfaction with electoral outcomes, then I’m not sure that this is the problem that we should be addressing, and I also have to wonder about the unintended consequences of picking such a system. And what could those be? Really, the quality of the data that an election produces, and what that data tells us about the election. Because believe it or not, that actually matters. What percentage a candidate received matters a lot. It gauges support, it sends a message about how solid or tenuous their support is, and about how much support their rivals have, which could mean clues for them as to how to better organise in the following election, and who to target. If the number of votes cast is divorced from the number of electors, what kind of message are we able to send? That would seem to be a pretty important consideration to me, and to a lot of people running, I would imagine.

I also have an issue with how this portrays what a vote means. In our system currently – and yes, this electoral system purports to keep the system otherwise intact, along the lines of “one simple trick to make the system more emotionally satisfying!” – when you cast a ballot it is to decide who will sit in the seat that represents your geographic area. And this is where a lot of electoral reform nonsense falls apart – it becomes about feelings rather than the fact that there is one seat and you have to help decide who fills it. How casting votes for multiple people to fill that one seat seems to defeat the purpose in many ways, and admission that it’s too difficult to make a decision so let’s cop out and muddle it so that I don’t feel so bad when I do it. But democracy is about making choices, and we should make it clear that it’s what it is, and just what that choice is (i.e.: Who is filling this one seat, rather than who is going to form a government, because that is decided once a parliament has been assembled). We’re not making that clear, and we’re constantly talking in terms of horse race numbers and leadership politics, and not about the actual choice that faces people, and I think this is something we should be paying more attention to, and being more vocal and precise about, so that we don’t wind up with yet more pie-eyed schemes that are designed to make us feel better while not actually doing what we’re supposed to. And this isn’t something that I’m seeing in the discussions on electoral reform – just a lot of pouting about “fairness” based on made-up numbers that don’t actually mean anything, and approval voting would make the numbers that do mean something, mean even less.

Good reads:

  • From the NATO summit in Warsaw, Justin Trudeau talked tough about Russia’s illegitimate actions in the region while pledging forces to Latvia.
  • Post-NATO summit, but before his visit to Ukraine, Trudeau toured the Auschwitz concentration camp.
  • Stephen Harper gave what was likely his final speech at the Stampede, where he didn’t say he was resigning but he did endorse Jason Kenney’s provincial bid.
  • Tristin Hopper gives a fantastic account of that Stampede barbecue and the era of the magic being gone now that the Conservatives are in opposition.
  • Talk at Stampede was also over Northern Gateway, where Rona Ambrose blamed “officials” for not properly consulting First Nations, which is the Crown’s job.
  • As a demoralized RCMP is looking to unionization to force change in the ranks, a fight is brewing over their Legal Fund loaning money to a union drive.
  • The government is looking into re-opening the prison farms that their last government shuttered.
  • The Canada Post dispute and bills that haven’t yet passed are straining the government’s relationship with labour.
  • Another female inmate’s death has Louise Arbour calling for an end to segregation in prisons.
  • Paul Wells notes Trudeau’s good summer (so far) on the global stage.
  • Robyn Urback wonderfully skewers the democratic reform event “toolkit.”

Odds and ends:

Here’s a look at how more women in politics is changing the discussion on topics like abortion and domestic violence.

One thought on “Roundup: Approval voting and numbers with meaning

  1. > But I keep going back to the basic question: what problem is this trying to solve?

    A huge, massive one: getting the wrong electoral outcome because of a simple mathematical anomaly called vote splitting, aka “the spoiler effect”.

    Among election systems experts, the vast majority of them agree that Canada’s current Plurality Voting (aka “First-past-the-post”) is the worst voting system on Earth. For instance, a group of 22 experts convened as part of the London School of Economics “Voting Power and Procedures” department in 2010, and Plurality Voting came in dead last out of the 18 systems they analyzed. You know which method came in first? Approval Voting.

    A simple and dramatic example of the impact of Approval Voting would be the past two gubernatorial elections in the state of Maine. Both times, Maine got an unpopular extremist Republican named Paul LePage, whom all evidence shows they did not want. He won simply because his two opponents split the vote. A moderate liberal attorney who used to work in Jimmy Carter’s administration ran as an independent and split the vote with the Democratic nominee.

    LePage is one of the least popular governors in the US, based on approval ratings. And polls show that the independent candidate (and the Democratic nominees) would have beaten him in a two-person head-to-head race by a landslide. He won purely because of a flawed voting method.

    But the Center for Election Science conducted an exit poll in three Maine cities, and found that the finish order would have _reversed_ under Approval Voting. In our poll, the independent came in first and the Republican came in dead last.

    http://scorevoting.net/Maine2014Exit.html

    Another obvious problem that Approval Voting fixes is that voters often feel reluctant to vote for their favorite candidate if they don’t think he can win. Instead of focusing purely on the merit of his ideas and experience, they concern themselves with polls and other “indicators of electability”.

    I would recommend reading “Gaming the Vote”, a tour de force of social choice theory which deeply analyzes the commonly discussed alternative voting methods and comes out advocating Score Voting (aka Range Voting). Approval Voting is effectively just Score Voting on a 0-1 scale.

    Clay Shentrup
    Co-founder, The Center for Election Science

Comments are closed.