Up: Issue 12 Previous: Paper 6

Voting matters - Issue 12, November 2000

How to ruin STV

I D Hill

To ruin STV by turning it, in effect, into merely a party list system, the following steps may be taken:
  1. Make voting compulsory so that even the laziest have to turn out;
  2. Insist that votes, as given by voter-defined preferences, are not valid unless every candidate (from a long list) is given a preference number, without gaps or repetition;
  3. Allow the voter the alternative option of merely ticking a party box, and take that to indicate an STV vote as specified by the chosen party;
  4. Use traditional STV counting rules, so that it can be guaranteed that, if you choose your own order, either your first choice will not be elected, or if elected but not on the first count, then all your hard work entering later preferences will be totally ignored;
  5. Insist that, as the party box method is optional, this is not taking anything away from the voters.
Since many voters are lazy, most can then be expected (save in very exceptional circumstances) to use the party box method, as to do anything else is a lot of work and almost certainly for no benefit. Is it unimaginable that party politicians would try to pervert STV in this way? Unfortunately not; all these things now happen in Australia, and nearly all the virtues of STV have consequently been lost.

To see the dire effects of this, consider the election of 6 Senators for New South Wales at the 1998 Federal Election, for which there were 69 candidates. In some Australian STV elections not all the candidates have to be given preference numbers, though they usually require a substantial number. In this one all 69 had to be put in strict preference order. Just imagine doing that when the alternative of merely ticking a party box was available.

Probably many voters would not be aware of the effect mentioned in item 4 above, so that may not have much effect on what happens, but it would certainly add to the frustration for anyone who did know about it. The remarkable thing in the circumstances is not that practically everyone used the party option but that 19012 voters, or 0.51%, did not.

The whole output table is much too vast for reproduction here, but the sense of it can be derived by looking at just the party that did best, with candidates A1, A2, A3 and A4 in that order on the party ticket. The first four stages for those candidates were:

Eventually A3 also was elected. It can be seen, just from this small part of the information, how the party listing is totally dominant, and crushes all individualism. In particular, note how the party's preference for A3 over A4 overwhelms the fact that A4 got three times as many first preferences as A3. In fact, after transfers, all the votes ended up pointing at the three candidates highest on the list of the above party that took three seats, the two candidates highest on the list of another party that took two seats, the candidate first on the list of a further party that took one seat, and the candidate first on the list of the runner-up party. For the candidates, it is clear that getting a high place on the party list, rather than being liked by the voters, is what matters, as with party list systems in general.

Is it wise to tell politicians that STV can be perverted like this? Given that it has already happened in Australia, it can hardly be hidden from them anyway. The important thing is to bring the facts to the attention of STV supporters, so that they know that it is something to be ready to fight against.

Up: Issue 12 Previous: Paper 6