As Australian readers will almost certainly be aware, due to the obsessive media attention, there was a by-election last Saturday in Eden-Monaro, a federal seat in southern New South Wales.
Political journalists love by-elections, because they offer lots of material for speculation and the promise (rarely delivered) of upsets. And they also love Eden-Monaro, partly because it has long had a reputation as a bellwether seat, but equally because it’s scenic and close to Canberra, so press gallery journalists can do fieldwork without having to stray far from home.
So the combination of the two was irresistible, resulting in blanket coverage. To put it in context, let’s run through some of what we know about by-elections.
Firstly, contested by-elections* tend to produce a swing against the government, because they offer the opportunity for a protest vote without consequences. Secondly, the party holding the seat tends to suffer from the loss of whatever personal vote its sitting member had enjoyed (except in the unusual case where a sitting member recontests).
The latter effect, of course, is stronger when the sitting member was of long standing or particularly well-regarded. And it tends to be more pronounced in rural and regional seats (which Eden-Monaro is), where MPs have more opportunity for personal publicity.
Most contested by-elections are in government-held seats: government MPs are more likely to be given new jobs, and oppositions are more likely to contest government-held seats than vice-versa. That accounts for some, but not all, of the average anti-government swing.
So in a seat like Eden-Monaro, where a popular opposition member retires in a regional seat, you’d expect the result to be pretty much a wash – absent any one-off factors or something special happening in the general political climate.
There’s one other very important thing we know about by-elections: they have almost no predictive power. Whether a government does well or badly in by-elections tells you very little about its prospects at the following election.
Obviously a huge by-election swing is a sign that something is wrong, and those (rare) examples tend to be the ones we remember, like Bass in 1975 or Canberra in 1995. But the message for either side in the difference between, say, a 2% swing and a 4% swing is essentially nil. (For much more on what by-elections don’t tell us, with figures, don’t go past Peter Brent.)
With all that in mind, the result in Eden-Monaro was utterly underwhelming. Labor was sitting on a margin from the last election of 0.9%. It retained the seat with a small swing against it, cutting the margin to about 0.4% (a few postal votes are still outstanding).
Almost everyone can find something to be pleased about in this. Labor is pleased because it held the seat; the Liberals are pleased because the swing was in their favor. The Nationals and the Shooters are pleased because they demonstrated some relevance, with their preferences (leaked from the former, directed from the latter) getting Labor over the line.
The array of minor parties and independents are pleased because they got some fleeting national attention. The journalists are pleased because they got to get out of Canberra for a bit, if only across the road to Queanbeyan. Only the Greens, whose vote fell 3.2 points to just 5.6%, seem to be without much consolation.
If you just focused on the fact that the Morrison government has an epic record of bumbling ineptitude (particularly for the people of Eden-Monaro, badly hit by last summer’s bushfire crisis), then you’d be surprised that Labor didn’t do much better. If, on the other hand, you focused on Scott Morrison’s poll ratings – and especially if you ignored Kevin Bonham’s warnings about them – you’d be surprised that the government didn’t take the seat.
But if you remembered that the crisis-induced bounce for governments is a fickle thing, and that the next federal election is the best part of two years away in any case – well, in that case not only wouldn’t you be surprised, but you probably wouldn’t care much about Eden-Monaro in the first place.
And that would be very sensible.
* Contested, that is, between the government and opposition. Once in a blue moon one of the other sort will produce an interesting result, but generally they can be ignored.