Mumble

There’s no doubt that, as the Rudd government says, this is the most difficult Senate in living memory. But it’s not really the opposition’s fault; Abbott and co are probably no more difficult than Labor was during the Howard years. The reason is the make-up of the rest: for anything to pass it has to be supported by the Greens and Steve Fielding and Nick Xenophon. Is there much these three would agree on?

These equations were anticipated prior to the 2007 election (before Xenophon had announced his run).

If those seven cross-benchers constituted one party, like the Democrats, then a majority of them (ie a partyroom decision) would do it. But it’s very difficult as it stands.

As Antony Green has shown, it is extremely likely that the next election will result in a Labor-Green Senate majority. Doesn’t matter if it’s a half-Senate one (with HoR) or double dissolution. But under a double dissolution there’s a tiny chance that Labor plus independents also = a majority, so the government could have option of dealing with Greens or the independents (much better than and).

And a reminder of the other big difference, that after a double dissolution the fresh Senate would come into effect immediately, while a half-Senate one would have to wait until July 2011.

End of Senate election notes.

Kevin Rudd is threatening the states with a constitutional referendum on health. Anyone with the patience to wade through this impenetrable sludge (my 20 year old honours thesis) will know the following:

 A constitutional referendum needs bipartisan support to have a chance of success.

  If it doesn’t receive federal opposition support, and it is held mid-term (ie not concurrent with an election) it will be walloped.

 If it doesn’t get bipartisan support and is held with an election, it will still fail - but respectably.

One corollary is that the actual content of questions is secondary. Referendums held together on wildly different topics usually get similar levels of support, for example.

There have been 44 referendums overall, 8 of which succeeded. Below are all the ones opposed by federal oppositions, ranked by descending national ‘yes’ vote. None was successful, although three received majority support but not a majority of states.

Constitutional referendums opposed by federal opposition, in decreasing order of national vote

There are patterns. Labor is nearly always in government. (A list of the eight successful referendums, by contrast, contains just one Labor entry.) As mentioned, referendums held together stick together in the list.

But the starkest pattern is that referendums concurrent with elections are at top, mid-term ones at bottom.

You could say that a referendum not supported by the opposition has a better chance of success if held with an election. Support tends to split along party lines, and if the government does very well it might drag the referendum with it as well.

But there is something else. It may be that these  referendums also suppress government votes.

The last ones with elections were in 1984, 1974 and 1946. The 1984 one saw a closer election result than most expected. Perhaps the same with 1974.

It’s not something you could “prove”. But it’s possible that if Thing A lifts up Thing B, Thing B also drags down Thing A.

Perhaps, you think, health is so important that it will transcend all this? Don’t bet on it. Even if Tony Abbott wasn’t dedicated to opposing everything, the Coalition would oppose this. They always* do.

Then the logic of referendums will predominate.

Be careful with that referendum, Mr Rudd.

*The one Labor-initiated referendum that received bipartisan support, in 1946, was successfull.