I’ve been surprised to see how simplistic the journalistic discussion has been about the ALP’s choice of Latham in December 2003. The assumption has been that:
- Beazley would probably have gotten a similar result to 1998 and 2001 (ie a narrow loss).
- Latham oversaw a much bigger loss.
- Therefore the ALP were wrong to choose Latham, and the pundits and punters who thought otherwise are misguided.
Just talking about expected outcomes misses the the fact that candidates also differ in their variance. As my friend Justin Wolfers has pointed out, Latham was the ALP’s catch-up football candidate – who might have won big, but actually lost badly. But if you’re behind at half-time, choose to play catch-up football, and lose by an even bigger margin, it doesn’t necessarily mean you opted for the wrong strategy. (Or to put it another way, when you’re ahead, you want the safe guy. When you’re behind, you want the risky guy.)