Shoni Field and David Wills: Why the Citizens' Assembly on Electoral Reform chose STV

    1 of 1 2 of 1

      By Shoni Field and David Wills

      Imagine getting a letter in the mail offering you an opportunity to spend almost a year of your life considering how B.C. elects its MLAs? That was the start of a process that in 2004 saw 160 ordinary people chosen at random become the British Columbia Citizens’ Assembly on Electoral Reform.

      Recognizing the conflict of interest that politicians have in selecting an electoral system, the legislature voted unanimously to turn the question over to an independent citizens’ assembly.

      The citizens’ assembly, with a man and a woman from every constituency, spent 11 months studying electoral systems, listening to British Columbians and forming a recommendation for how B.C. should elects its MLAs.

      Voters told the assembly they wanted a fair system, improved local representation, and greater voter choice.

      After considering electoral systems from around the world, the assembly short-listed two systems—single transferable vote and mixed member proportional representation—and looked carefully at which best addressed these three values. Both are equally proportional, but they provide different choices to the voters. The assembly decided that STV, a system where voters expresses their preference among candidates rather than parties, was a better fit with B.C.’s values. Both systems offer local representation, but STV makes constituency elections more competitive and provides greater incentive for responsive local representation.

      The citizens’ assembly voted 95 percent to recommend BC-STV as the best system for our province.

      Single transferable vote is an established voting system designed to ensure that more voters have a voice. It has been tailored by the citizens’ assembly to fit British Columbia. STV uses multi-member districts and a preferential ballot.

      How will STV be different than first-past-the-post?

      Under STV, voters get the representation they vote for. The results of the 2001 election, where 57 percent of the electorate was represented by 77 MLAs and the remaining 43 percent were represented by 2 MLAs, will not happen. With STV, the number of MLAs elected for each party closely reflects the proportion of the votes cast for that party.

      Vote distortion has serious consequences. We do not get the governments or opposition that we vote for. And, so long as it only takes 40 percent of the vote to form a government we will be limited to polarized policy-making, seesaw governments and partisan posturing.

      Under STV, most electoral districts will elect MLAs from both major parties. Further, candidates will be elected from smaller parties, but not fringe parties, providing a voice in Victoria for fresh ideas. Most voters (80 to 90 percent) will have an MLA they voted for.

      Finally, voters, not the parties, will have the final say on which candidates from a party will represent them. If an MLA does not represent a constituency’s interests, voters in the next election can choose another candidate from the same party. Under first-past-the-post, where most of the districts are safe seats, the election that really counts takes place at the party nomination meeting.

      Also, because larger parties can put forward a more diverse set of candidates, and smaller parties or independents may win some seats, the work of the legislature will better reflect new and important ideas shared by many British Columbians.

      Defenders of the status quo like the current system, because attracting 40 percent of the voters requires fewer compromises than those required to attract 50 percent.

      They also try to imply, against the evidence, that fairness to voters and stable government are mutually exclusive and that giving voters a voice is too risky in these troubled times.

      The interests of voters can never be a “luxury” item—democracy hinges on the notion of power coming up from the people. If we start to question whether voters aren’t perhaps just a little inconvenient then we’re heading down a really slippery slope.

      With governments that generally represent 50 percent plus of voters, consensus-based, enduring legislation creates a better investment, business, and labour climate, not to mention greater stability for social programs.

      There are no extenuating circumstances or unique strengths that excuse our current system’s failure to represent the interests of the voters.

      On May 12, when we vote for BC-STV, as recommended by the B.C. Citizens’ Assembly on Electoral Reform, we are raising our voices for the politics of fairness, positive change, and accountability.

      Shoni Field and David Wills both served on the Citizens’ Assembly on Electoral Reform, representing Vancouver-Hastings and Vancouver-Point Grey, respectively. They now sit on the board of Fair Voting B.C. (aka British Columbians for BC-STV).

      See also:
      Bruce Strachan: The single transferable vote—or how to get elected without really winning
      David Schreck: BC-STV is simply no better than our current electoral system
      Damian Kettlewell: Single transferable vote would break down political barriers in B.C.



      David Schreck

      Mar 23, 2009 at 12:59pm

      Ms Field neglected to mention that BC-STV would replace local representation with regional representation since our current constituencies would be replaced with enormous electoral areas. That would make political parties more powerful. STV's voting system is inherently flawed (improving a candidate's ranking for some voters can cause that candidate to lose - check out monotonicity).

      David Schreck

      Wayne Smith

      Mar 23, 2009 at 3:04pm

      The current voting system gives us almost zero local accountability. Most of us live in safe ridings and already know who will be elected in our riding on May 12, whether we vote for them, or vote for somebody else, or don't vote at all. Voters understand the current deck is rigged, and more and more are not bothering to vote.

      Under BC-STV, almost everyone will have a local representative that they actually voted for.

      Antony Hodgson

      Mar 23, 2009 at 5:03pm

      A Dialogue on Monotonicity:

      Me (an ordinary voter hearing Schreck mention the technical term "monotonicity"): "Oh my. STV is non-monotonic. That sounds scary. I sure don't want a non-monotonic voting system. Uh uh. No way." Aside to knowledgeable friend: "Psst! Would non-monotonicity really matter?"

      Well-informed friend: "Not really. All voting systems have flaws - some major, some minor. For example, our current voting system (first past the post, or FPTP) completely fails three of the four main criteria that are usually considered to be markers of a fair and reasonable voting system. These are:

      1. Universality - the complete preferences of all voters should be taken into account. FPTP doesn't let voters say what they really think.

      2. Independence of Irrelevant Alternatives - a voter's opinion about a third candidate should not affect the outcome of a matchup between the first two. With FPTP, if a voter honestly expresses their preference for a third candidate who's not one of the two leaders, their opinion about which of the two leaders should win is ignored.

      3. Citizen sovereignty - it should be possible to figure out which candidate is most preferred by all the citizens. Since FPTP doesn't allow voters to express their full preferences, we simply can't find out who really should have won.

      So FPTP is non-universal, gives results which change with the introduction of irrelevant alternatives, and violates citizen sovereignty. All these flaws are generally considered very serious."

      Me again: "But David Schreck says non-monotonicity is really bad. He's studied political science. Surely he knows what he's talking about and is not trying to misdirect us."

      Well-informed friend again: "In practice, non-monotonicity is considered a minor aspect of STV that is practically impossible to take advantage of. I actually looked up the mathematics paper that Schreck cites to make his case about the potential non-monotonicity of STV. The author states, after showing a highly contrived example to illustrate the theoretical possibility of a non-monotonic result, "We do not know if anything like this has ever happened in Ireland" (see In other words, Schreck has no evidence that this has ever happened, let alone that it's a problem worth being concerned about."

      "In any case, what non-monotonicity really means is that there's a technical possibility for a voter to slightly improve their preferred candidate's chances of winning by trying to control the order of elimination of two other candidates who are expected to be both less popular than the voter's preferred candidate and close to tied at the point where one or the other is to eliminated. To take advantage of non-monotonicity, a voter has to very accurately predict this situation and then lie on their ballot about who their preferred candidate is by elevating one of these other candidates who's likely to lose above their own favourite to make sure the other one loses first. Following me? Didn't think so. However, it doesn't really matter. Since STV is so effective at giving voters the representative they prefer, trying this strategy carries the very real risk of actually electing the person you've falsely declared to be your favourite, so most political scientists recommend simply being honest with your ballot."

      Me, in conclusion: "So, you're saying FPTP violates three of the four criteria in serious and significant ways, while the minor issue with STV is virtually impossible to exploit in practice so you might as well just vote honestly? Seems pretty clear to me that STV is better."

      Antony Hodgson
      Director, Fair Voting BC
      Supporting the recommendation of the Citizens' Assembly on Electoral Reform
      Check for information on the May 12th referendum