Over at Core Economics, Stephen King explores the various possible outcomes of the Speakership kerfuffle in the House of Representatives.(1) Assuming that Wilkie and Bandt can be assigned to the Labor camp, and Katter to the LNP's, he concludes:
[A]s a starting point the above analysis suggests three key points:
Given this, it is little wonder that the Coalition are stalling on pairing and pushing for a Labor Speaker. It is unsurprising that Oakeshott wanted the job with pairing but walked away when the possibility of pairing looked lost. It is also obvious that Labor would like pairing – and the Speaker role which it could claim once pairing was agreed to. The only odd thing – why did Windsor want pairing with Oakeshott as Speaker?
- Without pairing, no-one wants to be Speaker. An independent who took the job would have no say in legislation and this would just hand control of the Lower House over to the other conservative independent. One of the major parties will probably take the Speaker role on, but both will want the other side to take the job.
- With pairing, the Speaker role is a bit like a ‘double vote’. Both labor and the Coalition will want the Speaker role.
- With pairing, if a conservative independent gets the Speaker role then that individual controls the outcome for all legislation while the other conservative independent loses all power.
That's the takeaway, but the post is worth reading if the process by which he has reached this conclusion seems unclear.
King's post also reminded me of the fact that, in the 76-seat Senate, the Greens (9 seats from next July) will be just as powerful as Labor (34) or the Coalition (31). The reasoning is this: treating each of Labor, the LNP, and the Greens as one bloc, any two of them can form a majority, and two of them are required to do so; each therefore has a third of the voting power (with the 3 independents being irrelevant, as they can't help any other bloc to a majority).
This voting power can be very neatly quantified by "imagin[ing] each party joining a coalition in random order and count[ing] the proportion of times in which that party’s arrival turned the coalition into a majority." (Tim Harford's explanation refers to Dennis Leech's site where you can calculate the voting power for different combinations of actors, though it's not hugely user-friendly).
Joshua Gans has an example using lower house results (this was before the indies announced their allegiance), but for the three-bloc Senate, the possible random-order coalition combinations are;
| 1) ALP, then Greens | Greens decisive (their arrival creates the majority) |
| 2) Greens, then ALP | Greens irrelevant (they are necessary for the majority, but do not create it) |
| 3) LNP, then Greens | Greens decisive |
| 4) Greens, then LNP | Greens irrelevant |
| 5) ALP, then LNP | Greens irrelevant (they are not even part of the coalition!) |
| 6) LNP, then ALP | Greens irrelevant |
| 7) ALP, then LNP, then Greens | Greens irrelevant |
| 8) LNP, then ALP, then Greens | Greens irrelevant |
| 9) ALP, then Greens, then LNP | Greens decisive |
| 10) LNP, then Greens, then ALP | Greens decisive |
| 11) Greens, then ALP, then LNP | Greens irrelevant |
| 12) Greens, then LNP, then ALP | Greens irrelevant |
Twelve different combinations; four in which the Greens are the decider; ergo, Greens hold 33.3% of the voting power. (Leech's scripts follow a more complex, but in essence similar process.)
So, voting power... cool, huh?! The next time there's a hung parliament in Europe somewhere, you can pop over to Leech's page and impress your mates (well, me, at least) with some algorithms. Yeah baby!
(1) For non-Australians: our recent federal election resulted in a situation in which neither of the traditional ruling sides – Labor, and the Liberal-National coalition – could muster 75 seats in the 150-seat "lower house" (House of Representatives) to form Government. Labor has acquired the support of several independent/minor party politicians to get to 76 votes, but one person from the House of Reps has to act as Speaker, an important procedural role that means they don't vote unless there's a tie.
"Pairing" refers to the convention of compensating for one politician's inability to vote (e.g. through absence from the chamber due to illness, travel, parental leave, etc) by having a second politician, who would usually have voted in the opposite manner to the first, abstain from voting; thus the margin in the Parliament is not affected.
