It’s a long time since political parties used to deliberately preselect people whose surnames came near the front of the alphabet, so they would appear at the top of the ballot paper. Since 1984, ballot paper positions have been determined by lot, and last Friday the AEC conducted the draw for this year’s election (complete list here).

There’s general agreement that the advantage of the “donkey vote” – votes that just flow straight down the ballot paper, regardless of who the candidates are – is small but not insignificant. In a paper published earlier this year, Andrew Leigh and Amy King found that being at the top of the ballot paper increased a candidate’s vote by an average of one percentage point.

Sure enough, in a year when it seems Labor can do nothing wrong, the luck of the ballot draw has gone Kevin Rudd’s way as well.

What matters, of course, is not the ballot paper position itself, but whether it has changed since the last election – any donkey vote that a candidate received last time is already factored into their margin.

I looked at the 41 Coalition lower house seats that Adam Carr classifies as being at “acute” or “significant” risk. In 16 of them, Labor was drawn below the Coalition in 2004 but will be above this year; only six have moved the other way.

If we remove from the sample the ten very marginal seats where the betting markets have written off the Coalition’s chances, Labor’s advantage comes down to eleven to four.

Adding in Cowan, the one Labor seat in any real danger, makes it eleven to five:

Better for Labor: Boothby, Corangamite, Deakin, Forde, Hinkler, Kalgoorlie, Leichhardt, Page, Paterson, Petrie and Wentworth.

Better for Coalition: Bennelong, Cowan, La Trobe, McPherson and Robertson.

It’s unlikely that more than a couple of seats will be close enough for the change in ballot paper position to make a difference, and the evidence is that this election just isn’t going to be that close. Even so, it’s an added handicap that the government could well do without.