More on the referendum

Due to the dramatic differences in informal vote numbers across electorates - and hence a skewing of the distribution of Part B preferences - it's hazardous to draw exact conclusions from correlations across the different votes. Also, the figures I have are aggregates of electorates, not ballot-papers individually. I've done the maths anyway, but I won't cite figures, as they are not really precise. Keep in mind that you could vote, for instance, for FPP in Part B, and yet vote either for MMP or for change in Part A.

Overall, the percentage of an electorate's pro-MMP vote in Part A correlates almost one-to-one with informal votes in Part B. So those informal votes must be mostly pro-MMP voters.

Other correlations are weaker, but they suggest that:

1. The more pro-change votes in an electorate, the more votes in favour of both FPP and SM in Part B.

2. The more pro-MMP votes in an electorate, the more votes for STV in Part B.

3. PV seems to be evenly split between pro-MMP and pro-change voters who made any choice in Part B (once one takes out the Maori electorates which showed a strong preference for MMP but also gave PV the second place in Part B).

Points 1 & 2 are unsurprising. One might have expected that PV would correlate more strongly with the pro-change vote, but it was, after all, the least favored of the 4 alternatives.

As I argued in my previous post, the pro-change voters' preferences seem to be divided largely between the familiar (FPP) and the option pushed by the 'Vote for Change' lobby-group with Mr Key's support (SM).

It makes sense that pro-MMP voters, if they chose any alternative, would go for STV, as that is the most proportional of the four alternative systems.

Maori electorates were more in favour of PV than the nation as a whole, probably because it delivers 12 Maori seats. If they chose to go on the Maori roll, then presumably they like the idea of more Maori seats.