John Key is always making calculations. He does this incredibly well, just as he did as a money trader before entering politics. Key knows to sell stocks when they are high, and get out before there's a good chance of them going down in value. It's clear that he saw those stocks as having a good chance of falling in 2017 or soon after.
It is likely John Key calculated that a fourth general election and even a fourth term in government had a very high chance of ending badly. Hence he has bailed out early. This is a decision that has been made on the basis of what is good for John Key, not what is good for National, let alone for New Zealand. The Prime Minister has essentially decided that it is better to go now, while on a high, rather than to go through a difficult fourth general election, which he may lose, or go into a fourth-term government that might require sharing power with, and managing, Winston Peters and New Zealand First.
A good chance of losing the 2017 general election
Every election under MMP is relatively close - and often much closer than the public realises. The reality is that in both 2011 and 2014 National was returned to power only having just been able to cobble together a majority of votes in Parliament. The National Party itself has never won enough seats itself to govern, and has only been able to get its majority through the support of the minor parties. It's governed on a knife's edge, having won about 47% of the vote, and then got over 50% of the votes in parliament via the Maori Party, Act, and United Future. There is a possibility that it might have been able to govern with support in some way from New Zealand First, but it has never come to this.
So although National has impressively retained the support between elections of 45-50% of the public, this has been the absolute minimum it has needed to then win the election. Without large coalition partners, this high support has been a necessity.
There have been a couple of developments lately that have looked to threaten that precarious coalition keeping National in power. First, the closer relationship between Labour and the Greens has increased speculation that those parties might work together in the Ohariu seat to ensure that Peter Dunne is finally pushed out of Parliament. His majority of 710 was vulnerable to a single candidate from the left parties. Dunne's departure would be a loss to National's post-election coalition building.
Second, the growing rapprochement between the Mana and Maori parties is also a concern for National, as it might mean that a resurgent Maori force would have as much as five seats to potentially offer a Labour-led coalition (and, in fact, a larger overhang in Parliament).
This all means that National's precarious hold on office is even more in danger. And therefore there is an increased chance - that Key would have been aware of - of National falling short of being able to put together a coalition government in 2017.
And of course there's another year of possible political and economic turmoil to get through as well. Various factors that Key is aware of might have made his governing situation more difficult. For example, are there further by-elections coming up? Could this necessitate an early election? What if Nikki Kaye doesn't return to politics? Would National win a by-election in Auckland Central? And what would a loss there mean for Key's ability to govern?
A good chance of a National-NZ First coalition government
There is an increased chance that in 2017 Winston Peters and New Zealand First will be the kingmakers after the election. They might well choose to keep National in power. But just how happy would John Key be with that arrangement? He's never had to deal with Peters in a government, and it's unlikely that Key would enjoy any such power sharing or management of Peters.
Key would have been highly cognisant of the likelihood of a fractious coalition government if it was supported by Peters and his party. In addition, there would be a significant chance of the coalition disintegrating, and that Key would have a hard job managing the fights and processes involved. He would not have wanted to repeat Jim Bolger's 1998 experiences with Peters.
Key's selfish calculation
Key has probably seen the writing on the wall. The chances of winning a fourth election, and forming a government without Peters has been growing slimmer. He has therefore decided that it is better to get out now rather than have his legacy eroded by involvement in such failure and fracture.
Hence Key has made a decision based on his own needs - certainly not those of National. There is no doubt that National would have a better chance of winning a fourth term under Key's leadership than anyone else. His leadership is probably worth 3-5 percentage points for National. We might expect that a National Party under Bill English or Paula Bennett would struggle to get even 45 per cent of the party vote. So if Key really believed in helping keep Labour or New Zealand First out of office, he would have stayed to fight on.
Instead he's thrown a "hospital pass" to the rest of the caucus. Bill English will get the job as Prime Minister. He's probably got the best chance of keeping National in with a chance. And in the way that Key has timed his announcement - leaving his caucus only seven days to make a decision - and in his very significant endorsement of English, has effectively acted to keep any other rivals from taking the job.
New Zealand will remember John Key mostly as a successful prime minister. Part of this has been down to his strong forte in making political calculations. There should be no doubt that he used these skills in weighing up whether to go into a fourth election. And he has probably made the right call in judging that the risks to his legacy, reputation, and enjoyment in office were seriously negative if he stayed in the game. But the National Party might have reason to question the consequences of his latest decision.