As Latika Bourke reports in this morning’s Age, the crossbench senators are stepping up their efforts to ward off the advent of democracy — with the honorable exception of South Australia’s Nick Xenophon (the only one who was unmistakably shafted by the present system), who has been frozen out from their discussions as a result.

William Bowe explained the issues in Crikey yesterday, but basically Xenophon has proposed a small refinement to the existing proposal from the Joint Standing Committee on Electoral Matters (JSCEM) to abolish automatic group ticket preferencing. The others, however, led by Liberal Democrat Senator David Leyonhjelm, are determined to defend the system that elected them in defiance of the voters’ wishes.

On one level, this shouldn’t matter: if the major parties agree on reform (as they did at the JSCEM), there’s nothing the crossbenchers can do about it. Even with Labor getting cold feet, the Coalition and the Greens between them have a clear Senate majority. And the threat of crossbench retaliation on other legislation could be met by leaving voting reform until the last months prior to an election.

But there’s a further hiccup. As the ABC’s election analyst Antony Green has pointed out in two recent posts, the Australian Electoral Commission needs time to implement any changes prior to an election. The more drastic the reform, and, in particular, the more it results in voters expressing their own preferences on ballot papers, the more time it needs.

To implement the JSCEM proposal, Green says the AEC has asked for a 12-month lead time, otherwise it will have to conduct the count manually.

And if large numbers of people express their own preferences, then a manual count of, say, the New South Wales Senate, with over 4.5 million votes, would be a logistical nightmare.

So if this has stopped being a matter of high principle (since no one but the self-interested micro-parties defends group ticket voting) and become just a matter of practicality, why not go the whole way with simplicity, and get rid of preferences entirely for Senate voting?

We know that the vast majority of voters (96.5% in 2013) are comfortable just voting “1” for a party above the line, without worrying about its preferences or even who its candidates are. Why not just make the party votes the whole story, and elect senators in proportion to their party’s vote?

The truth is that leaving aside the bizarre deals that have elected the micro-party senators, preferences make very little difference to our Senate results. I’ve redone the calculations from the 2013 half-Senate election, on the assumption that they were a simple proportional election, using a Sainte-Lague system (as used, for example, in Germany and New Zealand). Here are the results. 

There’s very little difference: three of the micro-parties would miss out, with two extra seats going to Labor and the third to Xenophon’s running mate. The fundamentals of the balance of power would not have changed at all.

That confirms what’s already been learned from New South Wales, where the last four upper house elections have been held using essentially the proposed JSCEM method. On only two occasions has a candidate leapfrogged another to win on preferences: for 82 of the 84 members elected, preferences made no difference at all to the outcome.

With this change, the complexity that worries the AEC would disappear.

In effect, all votes would be “above-the-line” (although a below-the-line option could be retained for those who want to re-order the candidates in their party’s ticket, as is allowed in many European elections); they would only need to be counted once, and anyone with a spreadsheet can do a Sainte-Lague calculation in 30 seconds.

I’ve got nothing against preferential voting, and for single-member electorates or even small multi-member electorates (as in Tasmania or Ireland) it works well. But for an election with millions of votes where the preferences are unlikely to change anything, it seems an expensive and unnecessary indulgence.

It’s especially interesting to note that Leyonhjelm, with 9.5% of the vote, would still have won his seat without preferences. His opposition to reform is more subtly based. He knows that his election was a product of voter confusion: faced with an absurdly long ballot paper, many voters who were looking for the Liberal/National ticket lighted instead on the Liberal Democrats, who had drawn the lead position. And that in turn was a consequence of the lottery of group ticket voting.

With no automatic preferencing, the incentive for most of the micro-parties to stand would disappear, meaning simpler ballot papers and less confusion. Not good for some people’s self-interest, but an unequivocal gain for democracy.