Data shows dramatic decline in Key’s popularity

It’s become received political wisdom that John Key was a superhumanly popular politician, someone without parallel in modern history, an unmatchable asset for the National Party while he was its leader.

Not so, it turns out. Data from polling firm UMR, which has been referred to previously but not as far as I know published, shows that Key was no more popular than his predecessor, Helen Clark, and that his popularity took a massive dive over the last year of his reign – which may help explain his resignation in December.

The claims about Key’s success rest on his ‘preferred prime minister’ rating, which did indeed remain stratospheric. But the preferred prime minister ranking was relative: it asked people how much they liked Key compared to the alternatives. Because no one was ever very excited about the alternatives (Phil Goff, David Shearer, David Cunliffe and Andrew Little), Key’s ranking looked good.

But UMR has for years now asked a different, more straightforward question: do you have a favourable or unfavourable view of the prime minister?  This is equivalent to the much-discussed favourability ratings for American presidents. And it’s hard to argue with these numbers, which are derived from a simple, relatively objective question (people are asked if their opinion of a politician is very favourable, somewhat favourable, somewhat unfavourable or very unfavourable, with the results condensed into two categories) that has been asked consistently over a couple of decades.

The results, in the graph above, are striking. The dark red line at the top is Clark’s favourability rating; the dark blue line is Key’s. (The lighter coloured lines at the bottom are the natural inverses, their unfavourability ratings.)

What it shows is that, throughout their reigns, Clark often had a better favourability rating than Key – even in her last term when, according to received wisdom, everyone had fallen out of love with her. So the idea that Key was a politician without precedent simply doesn’t hold. What’s also striking is how much Key’s rating declined, not just from his first-term high but even after the 2014 election.

So what does this tell us? Probably that, for better or worse, personal attacks do work: Labour’s criticisms of Key’s behaviour eroded his popularity, just as National’s did for Clark.

Also that no one is untouchable or immune from that apparently basic law that the public always gets bored of you. The declines in Key’s popularity look to be linked to specific, well-known issues, such as the ponytail-pulling, the flag debate and the TPPA, so the things that some commentators thought should have dented his popularity in fact probably did, even if it wasn’t obvious from the preferred prime minister ranking.

These data may also help explain Key’s sudden resignation. He is on record as saying that, compared to successful traders, most people are too slow to sell failing stock. Given how much National invest in polling, I imagine Key knew something similar to these results, and may have concluded that he should get out now before things got worse. (The final jump in his rating is a post-resignation one; people like it when you go on your own terms and with dignity.) It’s hard to be certain, of course – but that seems like a more convincing explanation than some of the theories that have been circulating ever since he stepped down.