Wheat from the Chaff

My Musings on STV

As we approach the upcoming referendum on the multi-member STV (Single Transferable Vote) system, I face a dilemma. Frankly, I favour neither the current First-Past-The-Post system nor the proposed STV method.

FPTP is surely past its time. In an age where deference to authority is on the wane, and everyone rightly expects his or her view to count, our current electoral scheme is obsolete. In election after election, we see most candidates win their ridings with a plurality that’s well below 50%, rendering the majority of votes worthless. Unless regional parties, like the BQ, upset the distribution of votes, FPTP usually leads to artificial majorities that appear to give winning parties more of a mandate than they actually have. The 1988 federal election is a case in point, from which Brian Mulroney claimed to have a strong mandate for free trade. Of course, he had to point to the seats in the House of Commons rather than the popular vote.

The STV electoral system, however, has its own considerable deficiencies.

• To start with, it’s not a proportional electoral system. It is a preferential balloting system that allows a winning candidate to claim significant support, though usually with second and third preferences added to the initial first preference. This gives the candidate a somewhat contrived popular mandate. But let’s be clear: this is about greater inclusiveness and representativeness, not true proportionality.
• Second, as many commentators have noted, multi-member STV systems involve a great deal of complexity. The Droop quota that determines the threshold for victory is the easiest part of the calculation. The many unpredictable variables, like how many people will actually vote, the extent of their preferences, whether candidates will meet the threshold on the first ballot, and the distribution of fractional votes, all make tabulating the final results very difficult. Computers will be required to process the calculations and recalculations. [At least FPTP is easy to figure out: the candidate with the largest number of votes, majority or not, wins the riding.] So, if this exercise in electoral reform is largely about encouraging more participation, a black-box approach to counting will hardly generate renewed enthusiasm.
• Third, I continue to wonder who my MLA would be. Who of the five in my proposed super-riding would actually represent me? What if they all say it's someone else's problem?
• Fourth, we’ll have to deal with an unwieldy ballot. In 5, 6 or 7 member ridings, we will probably face an extremely large list of candidates, longer than a municipal ballot. It will be difficult to know most of the candidates, especially those who live outside your old riding. In the face of this, most people will likely vote for what they know: the political party. In Malta, one of the few places to use STV, few voters rank candidates outside their preferred party. It remains a two-party system. In
  Australia’s senate election, where you can vote one of two ways, most voters choose a party from a list of parties rather than rank individual candidates using STV. To the extent that we see party bloc voting, there will be little difference from our current system.
• Finally, multimember STV systems do not address the issue of party discipline and control by the leader. Riding associations will still need to follow head office rules and potential disallowance of candidates. In fact, these riding associations may become so big - because of the bigger ridings - that even less input from local members would be welcomed. Moreover, big ridings (like the one I would live in) would require more party resources if voters were to hear about distant candidates. This is hardly the recipe for lessening party control over the candidates.

I would prefer a wider referendum, like that in New Zealand in 1992. There, voters were asked to vote on maintaining FPTP, and, if not, which system should replace it. STV was only one of four choices. The voters, in their democratic wisdom, chose a different model - and the one I would pick if I were trusted to make the choice: MMP. Mixed member proportional systems acknowledge that no single electoral system is perfect, and prudently mix two systems to preserve the benefits (and minimize the drawbacks) of each. MMP gives the voter two ballots. In one ballot, a voter casts a party preference, and a directly proportional number of candidates are thus chosen from a descending party list. The list becomes a political artifact: if it has a skewed distribution of gender, region, class, etc., and too many party hacks from the largest city, the party risks harsh condemnation during an election. Since party leadership will be with us in any electoral system, at least MMP creates a form of explicit party accountability. Moreover, if it's done right, creating the list can reinvigorate party activism, and make the party more than a part-time electoral machine. Moreover, quotas are usually introduced to prevent dozens of fringe parties from dominating parliament. Often ignored is that MMP also preserves a large portion of its seats for a traditional FPTP election. The second ballot asks you to choose a constituency representative like we have now. This preserves the benefit of having a locally accountable politician. The degree of proportional compensation and the ratio of riding and proportional list seats can vary, depending on whether the electorate prefers stable majorities with many local representatives, or greater democratic fairness.

In the end, the upcoming referendum offers two unacceptable options. I can therefore boycott the referendum, spoil the ballot, or hold my nose and choose one of the two options. I'm pretty sure I won’t vote for STV. I’d rather campaign for a real choice in electoral reform, and not settle for the lesser of two electoral evils. There should be more than two options available, and we ought to have the right to choose among them.