Mumbling about Beazley
Peter Brent from Mumble claimed in Saturday’s Fin Review that the conventional wisdom is wrong about Beazley’s re-election chances.
Opposition Leader Kim Beazley can reshuffle to his heart’s content, he’s still unlikely to win the next election, whenever it is. The government is too entrenched, the required swing (about 5 per cent) is all but insurmountable, Prime Minister John Howard’s grip is too firm and Labor stands for nothing.
That’s the accepted wisdom. But it’s wrong. It’s the coalition that will struggle at the next election.
Elections rarely turn on all those things commentators talk about.
One overriding rule that is rarely acknowledged is that young governments find re-election easy, while old ones find it increasingly difficult.
His thesis, put simply, is that age wearies. The electorate tires of the government of the day. Brent used the analogy of the seven year itch, noting that the average term of government since 1970 was around seven years.
Of the 21 governments elected since 1970 (ignoring incumbents), 18 made it to a second term. Of those, 10 won a third term. Of that group, only three made it into a fourth term. Just one of those made it five in a row and none made it to six.
That represents success rates, respectively, of 86 per cent, 56 per cent, 30 per cent, 33 per cent and zero. It’s called electoral gravity: re-election just gets harder.
Brent concluded, “on the question of likelihood, the evidence points in one direction. Labor will probably win the next federal election”.
I am less convinced.
Brent chose 1970 as his starting point because it “can be seen as the beginning of modern politics, with reasonably fair boundaries and bipartisan professionalism”. But as every statistician knows, the selection of a starting point is critical to any trend analysis. If Brent went back just one Liberal government his central thesis would have failed, as the following graph shows.
While the lines in the above graph for Whitlam and Fraser are consistent with Brent’s ageing-government come fed-up-electorate thesis, the lines for Menzies and Howard are not. Even the Hawke-Keating years are problematic for Brent’s thesis.
Brent’s analysis fails to factor in the impact of the economy on electoral cycles. There is a large literature on the impact of the economy on electoral cycles, and econometric models are often used to predict election outcomes. Election outcomes are not just about probability, they are also about performance.
Furthermore, Brent undersold the conventional wisdom. While the size of Howard’s margin is one element, there are others. A fuller rendition of the conventional wisdom would have noted that Prime Ministers who are recession-free and debacle-free usually win elections.
For my money, you cannot go past the betting market for an accurate assessment of Beazley’s chances of winning the next Federal election. The punters at Centrebet rate his chances at 36 per cent and the punters at IASbet have it at 40 per cent. It is competitive, but it is a long way from “Labor will probably win the next federal election”.