In May this year in Crikey, I expressed scepticism about the ALP’s decision to run Maxine McKew in Bennelong.
It wasn’t that I didn’t think she could win. I always put it about 50-50, but that was less than the chances I gave Labor of taking government – about two in three. I just doubted the value of star recruits in marginal seats in general, and didn’t believe she stood a better chance of taking the seat from Howard than any old (hard-working) party hack would have.
So now the results are (nearly all) in, can we say whether her recruitment was a stroke of genius or not? She won the seat, of course, which is all that counts at one level. But how did her swing compare with others’?
McKew’s swing of 5.5 percent swing (the current figure according to the AEC) is just a smidgin above the national one of 5.45 to Labor and a similar smidgin below the New South Wales one of 5.6. Let’s call that neutral evidence.
Perhaps we can get a better idea by comparing Bennelong’s swing with those in other Sydney seats. That’s what this table does.
Sydney seat | Swing to ALP |
1. Macarthur | 10.4 |
2. Lindsay | 9.7 |
3. Mitchell | 9.1 |
4. Chifley | 8.7 |
5. Werriwa | 8.3 |
6. Banks | 7.9 |
7. Parramatta | 7.8 |
8. Prospect | 7.1 |
9. Greenway | 6.9 |
10. Cook | 6.8 |
11. Macquarie | 6.5 |
12. Hughes | 6.3 |
13. Watson | 5.8 |
14. Bennelong | 5.5 |
15. Berowra | 5.0 |
16. Fowle | 4.9 |
17. Kingsford Smith | 4.8 |
18. Reid | 4.8 |
19. North Sydney | 4.6 |
20. Barton | 4.5 |
21. Lowe | 4.4 |
22. Bradfield | 3.9 |
23. Grayndler | 3.7 |
24. Mackellar | 3.1 |
25. Blaxland | 2.9 |
27. Sydney | 2.2 |
28. Warringah | 1.8 |
29. Wentworth | -1.3 |
Bennelong, at number 14, is bang in the middle. Again, not much evidence either way.
But as in every election, different parts of Sydney moved by different amounts. In particular, Sydney’s outer fringes swung big time to Labor, while those further in tended to stand relatively still.
How did Bennelong compare with other electorates in its area – with neighbouring seats?
From this AEC map we can identify Bennelong’s neighbours, and the table below ranks them only.
Sydney seat | Swing to ALP |
1. Parramatta | 7.8 |
2. Bennelong | 5.6 |
3. Berowra | 5.0 |
4. Reid | 4.8 |
5. North Sydney | 4.6 |
6. Lowe | 4.4 |
7. Bradfield | 3.9 |
Now we’re getting somewhere. Out of all neighbouring seats, only Parramatta swung to the ALP by a larger amount than Bennelong. And Parramatta was the only electorate in the list which changed members in 2004; the new member Julie Owens probably accrued a personal vote in that time, which would account for part of her swing.
This exercise has been less than robust, and involves quite a few assumptions of the “everything else being equal” kind. Maybe the cretinous behaviour in Lindsay in the final week had a greater than average effect on Bennelong’s Asian voters. And all sorts of other qualifications.
But the evidence does seem to point to the conclusion that putting Maxine McKew in as Bennelong candidate was a very clever thing for the Labor Party to do.
Full marks to everyone involved.
Admitting you were wrong is good. Providing the evidence to prove it even better. Full marks to Mr Mumble.
Yeah, nah. Can’t compare even neighbouring seat swing. No other candidate had the PR reach of ‘the PM’ free media (esp Ruddock in Berowra – nil). And 2004 was a serious try (refer Oakes), but the last 2-3% was probably exponential hard to get. Max Axer.