Crikey subscriber and expat Simon Jackman, the Associate Professor and Director, Political Science Computational Laboratory, Department of Political Science and (by courtesy) Department of Statistics at Stanford University, has been playing with his laptop on his holidays while we’ve been discussing informal ballots and the voting system over the last fortnight.

He’s come up with a swag of preliminary figures and tables on who votes informal – and why. Wonks delight – and civilians can look at the Australian Electoral Commission’s electorate map to guide them through this all.

Jackman has tabulated rates of informal voting in the 150 House of Representatives divisions in 2004. “The distribution has a pronounced skew to the right,” he says. “The minimum rate of informality is just 2.8% (in Higgins, Peter Costello’s seat in inner-metropolitan Melbourne), and most divisions record rates of informality less than 5%, but the maximum rate of informality is 11.8% (in Greenway, in Sydney’s western suburbs). After Greenway, the top ten divisions in terms of informality are (in descending order) Reid, Blaxland, Chifley, Prospect, Fowler, Watson, Parramatta, Kingsford-Smith, and Werriwa. All except Kingsford-Smith are in Western Sydney; all are held by the Labor Party; all have relatively high proportions of voters from non-English speaking households.

He’s mapped informality against the number of candidates on the ballot paper: “The modal ballot length in 2004 was 7 (39 divisions), but 14 candidates stood in Greenway (the seat with the highest rate of informality) and 12 in Dobell. The minimum ballot length was 4, seen in 3 divisions (Riverina, Throsby and Braddon). The scatterplot strongly suggests that has ballot length increases, so too does informality; the correlation between the two variables is a rather healthy.

Jackman’s then looked at the relationship between informal votes and the prevalence of non-English speaking households in electorates. “The NESH variable is quite skewed,” he reports. “Most divisions have no more than 10% of their population speaking a language other than English at home, but five divisions are ‘majority NESH’, with values of NESH in excess of 50%: these are Blaxland, Fowler, Reid, Watson (all in Western Sydney) and Gorton (in Melbourne)… High values of NESH tend to accompany high values of informality, but the relationship is far from perfect.”

He’s considered tertiary qualifications, a degree or higher, or a diploma. In general, as the proportion of people with tertiary degrees increases, the rate of informality decreases. Divisions with extremely high rates of informality almost all have relatively low proportions of the division population holding tertiary qualifications.”

Jackman also ponders the issue of optional preferential voting for the local parliaments in NSW, Queensland and the ACT hand believes his analysis provides “some support for the notion that voters are confused by the fact that two voting systems are in operation at the two levels of government”.

Will any of this be taken into consideration by the Joint Standing Committee on Electoral Matters as it looks at the conduct of the 2004 election? Will the Government take any of this on board; given the speculation it will amend the Electoral Act once it takes control of the Senate?

It all depends. Do our elected representatives want a system that is voter or pollie friendly? Optional preferential seems to be the way to go if we want to reflect punters views – as well as being a pragmatic compromise on the issue of compulsory voting – but some poor pols might get a bit worried if they feel that they might miss out on some crucial fourth, fifth, sixth preferences.

This one is going become an interesting test of just who the voting system serves – electors or politicians.