BC Electoral Reform

One of the things that makes BC politics interesting, and frustrating, is that the ruling BC Liberal party isn’t really a Liberal party at all; it’s a coalition of Liberals and Conservatives.

Ideally, everyone would fly their own true colours and the Conservatives would re-animate the BC Conservative party. But this means splitting the vote that now goes to the BC Liberal party, leaving the field wide open to the NDP. As long as the BC Liberals keep winning, even with all the friction within, they have no incentive to do this.

The underlying problem is the old first-past-the-post system that determines a winner based on a plurality of votes. Where there are more than two candidates who earn a significant portion of the vote, plurality means we get a result that expresses the intent of the largest minority – not the majority; the incentive is for fewer voices, less choice.

Besides contributing to voter disillusionment, this skews the outcome. And, as people see a need to second-guess this outcome it encourages them to vote “strategically,” which means that rather than voting for their preference, they vote to defeat their least-preference; more skewing.

The failed efforts to introduce Single Transferable Voting (STV) in BC could have removed this dilemma. But, while STV came close, the defeats make its prospects inauspicious for some time to come. Remember, though, that the two key elements in STV are Multiple Representation (MR), and a preferential ballot.

MR systems are not without their problems and detractors (including me), and MR adds complexity, making it harder to communicate how it works, and harder to sell. But the problems of first-past-the-post can be much more directly addressed by means of a simple, single-member, preferential ballot.

The Condorcet Canada Initiative

Since this writing I’ve taken a much harder look at IRV, and while IRV is still an improvement over FPTP, it still shares some of the worst flaws of FPTP, and all in all it sets the bar too low.

Condorcet voting is a far, far, better electoral reform choice. For voters it is easy to understand and to do, while adding some extra effort for election officials. It is nevertheless straightforward and scrupulously fair, and possibly the most robust and reliable way of identifying the candidate who the majority most prefers.

(And – for those for whom Proportional Representation (PR) remains the only answer – there is Condorcet MMPR, which well fits that bill, too.

Let’s just drop the MR part, and keep the preferential ballot. This will re-empower the voter and re-invigorate the process by encouraging more voices and more parties – without the harmful effects of vote splitting.

This doesn’t require big changes. There’s no need to change the size or number of ridings, or even the ballot itself. We just change the way we mark the ballot and the way we count it: count the first-preference, and if a majority is not achieved, eliminate the then-least-preferred candidate reallocating votes per subsequent preferences … until done.

There are many ways of “counting the ballot,” some better than others, and some way better than others. But the worst of them is better than what we’re doing now. Instant Runoff Voting (IRV), which is what we get when we remove MR from STV, for example, would serve well.

Since it isn’t MR, those who dislike PR are happy, and, since many MR systems involve some manner of preferential vote it’s still attractive for those who want MR, being arguably a step in that direction without us all having to buy-in all the way. And, in the end, we will get a decision that much better reflects the will of the majority.

We’ve got an HST referendum (or an early election) coming up; it’s an opportunity, a free throw. Why not throw an IRV referendum on the ballot while we’re at it?

Facebooktwittergoogle_plusredditpinterestlinkedinmail
This entry was posted in No Sidebar and tagged , , , , , , , , , . Bookmark the permalink.