“Walk with the gods is over” proclaimed Fairfax yesterday. “It seems more likely that the last year and a half [of polling results] has been the aberration, that we have now entered normalcy”, declared Peter Hartcher in the Sydney Morning Herald, with the sort of biblical authority usually reserved for Paul Kelly on a gravitas bender.
The phenomenon previously known as the Rudd Honeymoon had come to a sudden end. Sky News and the ABC dutifully followed, the free to air broadcasters and radio jumped on the bandwagon and by about 9.00am, we had a national narrative — The Budget had hit Rudd hard! It was, or course, helped along every step of the way by ALP types who spun the Nielsen results for their own political benefit, again, like they always do (memo to media: a Labor apparatchik bearing bad news is attempting to play you like a violin — they are better at this game than you are, wake up to yourselves, it’s embarrassing to watch).
The problem of course was that Nielsen never actually demonstrated the political change that Fairfax had projected onto it — it was just poor analysis. While the two party preferred estimates had come in at 53/47, the context Fairfax was using here was the Nielsen poll taken nearly two months previously — at a time when Labor was having one of its mini vote booms and where that particular Nielsen came in at the high end of the phone poll spectrum with a big 58/42. It wasn’t a bad poll — it’s called sample variance. Any change in public opinion over that two month period was slated home by Fairfax columnists as being a consequence of the budget.
The silliness of this position became crystal clear by yesterday afternoon with the release of Essential Report, which showed Labor increasing its two party vote by a point — coming in at 62. Then last night Newspoll came in, showing a 1 point increase to the ALP, up from 55 to 56.
Fairfax got itself into this mess because they’ve cut back on the number of times they send Nielsen into the field for them. If they’d commissioned a poll last month, it would have been expected to come in somewhere around the 55/45 mark (for the simple reason that all pollsters move together most of the time). It would have meant that the context for yesterday’s Nielsen poll was one of small to no change across the budget period — the reality of what actually occurred when we look at all the polls together — rather than the 47th declaration that Rudd’s honeymoon is over.
As a result of this silliness, we had to endure 24 hours of the political news cycle talking nothing but complete horsefluff, while Labor went about exploiting the dodgy Fairfax polling analysis as evidence for how tough they’ve been with the budget. The other problem, or course, is that for the next 24 hours in the cycle we’ll have to put up with journalists talking about how confusing it all is.
The sad part, the truly sad part, is that it’s not actually confusing at all — Nielsen last polled toward the end of March and since that period every pollster has shown that the broad approval levels of Rudd (approval ratings for Nielsen and Essential, satisfaction ratings for Newspoll) have declined by between 9 and 10 points while his broad disapproval levels have increased between 8 and 10 points.
Over the same period, Turnbull’s broad approval and disapproval levels have remained unchanged (between a 0 and 2 point increase in both approval and disapproval from all pollsters).
Meanwhile, over the exact same period, the ALP two party preferred average of the three pollsters that have polled since the budget has reduced from 59.7 at the end of March down to 57 as of last weekend.
Most importantly, the change in all of these metrics was a gradual process spanning nearly two months — the budget didn’t have a great deal to do with it, for most budgets don’t.