Tactical voting, here and there

On Friday in Crikey Bernard Keane had an interesting story on the role played by tactical voting in the victory of “teal” independents in last month’s Australian election. He refers to results from the Australian Election Study, presented by researchers at ANU (reported here in the Guardian), that showed “voters who supported teal independents were mainly former supporters of Labor and the Greens, not the Coalition.”

This is a good example of why I don’t like the AES; it’s not just that it produces unreliable data, but that its proponents persist in using it when they could be using proper scientific polling or, as in this case, actual election results.

But Keane’s discussion is well worth a read. One point that I think he misses is the difference between the underlying strength of the respective parties’ vote in Sydney and Melbourne. If we look at the 2016 election, the last one with no teals in the field, the Labor and Greens vote in Sydney was already very low in the Liberal heartland seats: 27.0% combined in Warringah, 31.4% in Mackellar, 31.8% in North Sydney, 32.6% in Wentworth.

In Melbourne, those figures were significantly higher: 37.8% in Goldstein, 38.7% in Kooyong, 40.3% in Higgins.

That meant that when the teals appeared, Labor and Green voters in Sydney didn’t have to worry much about tactical voting. They knew that if the independents were going to have any chance, they’d be getting ahead of Labor or Greens as a matter of course. In Melbourne they had no such assurance; it was possible that a teal would be knocked out before getting a chance to garner Labor and Green preferences.

So in the latter the Labor and Greens vote dropped sharply, as voters switched to the teal: down 23.5 percentage points in Goldstein and down 25.4 points in Kooyong. In Higgins, the teals didn’t even risk contesting and Labor won the seat instead. But in the Sydney seats the drop in the Labor and Green vote was much less pronounced.

If you want to see really serious tactical voting, though, you need to look at Britain. There it’s not a question of just staying in long enough to get preferences – with first-past-the-post voting, you have to plump for the candidate with the best chance of beating your less-favored option.

Last Thursday saw by-elections in the two Conservative-held seats of Wakefield and Tiverton & Honiton. I noted earlier this month that they would be a key indicator for the survival prospects of prime minister Boris Johnson. He is still hanging on, but more than ever it seems that his days are numbered.

Wakefield, historically a Labour seat until the Tories won it in 2019, was less of a surprise: Labour increased its vote by about eight points to 47.9%, almost five thousand votes ahead of the Conservatives’ 30.0%. That was towards the high end of expectations, but not an unprecedented sort of margin for a by-election. Tactical voting was no issue; no-one else reached double figures, either this time or in 2019. An independent (ex-Conservative) took third place with just 7.6%.

Tiverton & Honiton was quite different. The Conservatives had held it and its predecessors for nearly a century, and in 2019 had a margin of more than 24,000 votes. They also benefited from divided opposition: Labour came second last time with 19.5%, followed by the Liberal Democrats on 14.8%.

But as in North Shropshire six months earlier, Labour decided this was not the seat for it, and its candidate received only a token 3.7%. Both Labour voters and disaffected Tories flocked to the Lib Dems, who won the seat with 52.8%, a very comfortable 14.4 points or six thousand votes ahead of the Conservatives. It’s believed to be the largest majority ever overturned in a by-election.

The size of the defeat speaks for itself, but in its way the fall in the Labour vote is at the heart of the government’s problem. It shows that – in contrast to the 2019 election, when they were mostly at one another’s throats – Johnson’s opponents are capable of co-operating against him. A majority Labour government may be a slim prospect, but a reformist coalition looks very much a possibility.

As I keep saying, voting systems matter. By-elections can throw up odd results anywhere, but this kind of sudden seismic shift is very much a feature of first-past-the-post systems, where votes for a candidate who can’t reach the top two are simply wasted. Voters have to be ready to change their long-standing practice in order to have an effect.

If the teals are to become a permanent fixture in Australian politics, at least that’s one problem they won’t have to deal with.