Wednesday, April 22, 2009

STV: An example of electoral system abuse (bad example)

The contention brought up here only pertains to a situation in which a party would be able to present only one candidate in a two candidate riding. This of course is ridiculous. Sorry, I didn't think out my example well enough


STV has been touted as a form of proportional representation, in which less votes are lost than in the "first past the post system". It is designed to restore electors' confidence in the democratic process. However, STV can easily lend itself to quite the opposite of proportional representation.

STV allows for abuse of the electoral system due to its "Transferable" quality. The transferable quality is designed so that no vote is wasted. If the first choice of a voter does not have enough votes to be elected, that vote goes to the second choice.
This is also the case when a candidate reaches the quota for being elected. The second choices of the voters supporting that candidate determine the proportion of surplus votes that goes to each of the other candidates.

This is where the true problem with STV lies. A vote can be reused multiple times. This may seem fair. To me it doesn't, but it may. However, what is far more important, IT CAN COMPLETELY COUNTER PROPORTIONAL REPRESENTATION.

Say a BC Liberal has 60% of the vote in a riding of 10000. The required quota for election is 3334 (votes/seats+1)+1. This means that there are 2666 extra votes. Now suppose that the NDP received 3332 votes and the Greens the remaining 668 votes. Since the all the BC Liberal voters are vocally anti NDP, and also believe the Greens are an easier party to deal with in the Opposition, 100% of their second choices are for the Greens.

This means that all of the surplus Liberal votes transfer to the Greens bringing 668 to 3334, the quota for election. The Green candidate is elected.

Does this still seem fair to you?

Think about it. In a properly proportional system, the Liberals would have received 60% of the seats, the NDP 33% and the Greens 7%. But in this scenario with STV, the Greens get elected while the NDP is shut out.

The reason this can happen is the retention of the ridings system in STV designed so as to not entirely lose regional representation. In other words, the riding system is still used so as to make sure that regional interests are still represented. In addition, the system of ridings ensures that there is a greater degree of accountability that MPs or MLAs have to adhere to. This compromise leads to the problem I just illustrated above.
Recommend this Post at Progressive Bloggers If you liked this post, please vote for my blog at Canadian Blogosphere Canadian Blogosphere

18 comments:

  1. You're comparing apples and oranges. You raise an example of a BC Liberal and a Green candidate being elected, while an NDP candidate is not. You then compare that riding total to the provincial total, and conclude that the provincial total will not be realized at the riding level. That doesn't follow, though. The example you give would have to be repeated across every riding in the province, which stretches the boundaries of credibility.

    That said, you're also missing the fact that the alternative on the table is SMP/FPTP. Which is far less proportional than STV.

    ReplyDelete
  2. STV is very proportional, but don't take my word for it there is a measurement called the Gallagher's Index that measures disproportionality of voting systems. Anything with less than 5% disproportionality is considered a proportional system.

    STV nicely balances proportionality and local representation and scores between 3.5-4.5% so it is considered a proportional system.

    That means if a party gets 40% of the vote it will get around 40% of the seats (the higher the vote a party gets the closer they get to a proportional representation. A party getting 5% under STV will have a tough time getting that many seats, while a party with 10% will likely have a more proportional representation)

    Our current disproportional system gives a party with 58% of the vote 97% of the seats, or can give a party with 20% of the vote 3% of the seats ... it can be wildly unproportional.

    So why would you choose First Past the Post over STV?

    Also - our own Liberal Party will use STV if we pass the OMOV motion at convention.

    ReplyDelete
  3. This comment has been removed by the author.

    ReplyDelete
  4. First of all, sorry I should have checked my self-manufactured example. It only applies if a party can only put forward one candidate in a two candidate riding. Sorry, it doesn't really apply then to the suggestion in BC.

    Also, this is the first I heard of STV being included in OMOV. How is that going to work?

    ReplyDelete
  5. Your example is perfectly accurate. The odds of either major party running more than one candidate in Northern BC is no more than 50/50. Parties will run the largest number of candidates they think they can get elected and not one more.

    The Liberal Party will of course be using IRV and not STV. We have no need to elect more than one leader.

    ReplyDelete
  6. Thanks for pointing that out. That makes me feel better about the example. So it is significant in Northern BC. It would be best to have a system that could work around that.

    ReplyDelete
  7. Mr. Ginsberg,
    Your info is incorrect.

    No, we wont choose multiple leaders but awe will choose a percentage of each riding to award to each leadership candidate and we will use STV.

    Also, in Northern BC there are plenty of areas where the Liberals could win two seats or more and will run, typically, one more candidate than they think they can win.

    ReplyDelete
  8. Partisan non-partisanApril 23, 2009 at 7:47 PM

    Young Liberal, you clearly don't understand the logic of STV. If a party thinks they can get around 60% of the vote in a two-member riding, it's assured they are going to run two candidates.

    ReplyDelete
  9. I realize that. I said it's a bad example in previous posts.

    ReplyDelete
  10. "A vote can be reused multiple times."

    Votes don't count more than once under STV. On the contrary, the transfers of surplus votes for candidates who are elected and votes for candidates who are elected are done to make sure that every vote counts fully, instead of going into the garbage can, as under first-past-the-post.

    That's the point.

    ReplyDelete
  11. Actually, it's quite a good example of the way a proportional system works when we have minimum number of seats. In your example, every party had a vote reminder and (unlike MMP or any other list PR system) it was not a mathematical formula but voters' second choices that determined whether that reminder should be rounded up or down.

    The problem with your scenario is that the Liberals and the Greens effectively act as one party or as a coalition. Instead of running another candidate to absorb the surplus, Liberals transferred all their surplus votes to the Green party candidate, effectively making him a second Liberal. If even a handful of Liberal voters objected the deal - the Green candidate would fall short of meeting the quota. (In your example - as few as 2 ballots with no second choice specified would be enough for the NDP to win.)

    If the three parties functioned separately, then the Liberals (with their support level) would be running two candidates and then it would be:

    NDP: 3332 votes,
    Liberal 1: ~3200 votes,
    Liberal 2: ~2800 votes,
    Green: 668 votes.

    The Green candidate would be the first to be eliminated and then again, it would be up to the Green voters' second choices to determine whether they put the NDP or one of the Liberals over the top. Unless all the Green voters unite behind the Liberals (in which case we're back to Liberal Green Coalition with enough support to win both seats), it will be 1 Liberal and 1 NDP elected.

    P.S. According to the recent redistribution, if STV is adopted, there will be only 1 2-seat riding in BC. All other ridings will have 3 seats or more, (up to 7,) so the share of leftover votes will be between 12.5% and 25% per constituency. That's a great improvement compared to 50%-65% of wasted votes under FPTP.

    ReplyDelete
  12. The vote is reused though. At a lower value, but it is still reused. In fact, some votes count more under STV than under FPTP. The votes that elect candidates on first preference are used to determine how much surplus goes to the other candidates.

    ReplyDelete
  13. No. The whole notion of "reusing" a vote is incoherent under STV vs. SMP. A vote under STV includes multiple preferences; counting those preferences doesn't imply "reusing" -- it's how you use a vote under STV. SMP doesn't register multiple preferences. It gives us absolutely no information beyond first preferences, and holds that whoever gets the most first preference votes wins -- even if someone else would get more first+second preference votes.

    You can't, in defense of SMP, just assume the way it counts votes is a neutral or default state of counting votes. The method of counting votes is exactly what's in contention.

    ReplyDelete
  14. You're talking about the BC referendum. The choice is STV vs. SMP. If you oppose STV, you are necessarily defending SMP. This maybe in a weak way, i.e., it's better than STV. But, nonetheless, you are defending it.

    ReplyDelete
  15. No, if I could vote in the referendum, I would choose STV over SMP. I'm pointing out how there are still odd parts of STV.

    ReplyDelete
  16. Given that your original post closes with "Single Transferable Vote is a sitting duck to electoral system unfairness and abuse.", you'll forgive the misunderstanding.

    ReplyDelete
  17. Sorry, that's from the original post when I hadn't thought out the example properly. It's certainly not a sitting duck.

    ReplyDelete

Progressive bloggers