One of the good things about elections is they are a good opportunity to stress-test your assumptions. Here are five of mine, written down so if they come unspooled I won’t be able to go back on them.
Campaigns don’t matter
I’m a bit of a sceptic about the effect that general election campaigns have in terms of deciding the outcome. Mostly the rival campaigns cancel each other out, and what really matters are the opinions set about politicians over the long haul.
That’s not to say that events in the campaign period don’t matter. The Tories only hit on the “vote against us, get Ed Miliband bossed about by the SNP” message late on, which was particularly effective against the Liberal Democrats in the South West, giving them their majority.
But it only worked because of perceptions about Ed Miliband set years before – that he was weak and indecisive – and reinforced by events in his leadership.
My strong sense is that Theresa May will not have a good campaign by any objective standard – her interviews will be halting, if she does come into contact with any actual voters they will go badly – but it won’t matter, because people fit the facts to fit their opinions of the leaders, rather than the other way around.
The Liberal Democrats will outperform their polling
One of the underreported trends of the last two years has been the Liberal Democrats’ remarkable recovery in local council by-elections, well in excess of what we’d expect from their modest revival in the polls.
Their success in the local elections in 2016 and their good results in the Witney and Richmond Park by-elections also outperformed their standing in the polls. It should be noted, however, that their revival seems to be confined to England.
I expect that the Liberal Democrats will outperform their poll share in the local elections and in the general elections as far as their vote share is concerned, even if they return the same or fewer seats.
The local elections will be a good guide to the general election
When the academics give us the PNS from today’s contests – that’s the projected national share, how the country would have voted if there had been elections everywhere and not just in some parts of the country – I think we will know two things. The first will be what Labour’s ceiling is. They will not get a better result on 8 June than they do today. The second will be what the Tory floor is. They will not get a worse result on 8 June than they do today. That pattern has held for every local election contested in the same year but not the same month as the general election – in 1983, 1987 and 1992.
Ukip will collapse, the Tories will be the major beneficiaries
Ukip are a busted flush, with a sub-par leader. They can’t even make headlines anymore. They will fall back pretty much everywhere and the biggest net winners from that will be the Conservatives.
You can’t win an election if you are behind on leadership and the economy
If fewer people think your candidate is a better Prime Minister than the opponent and if fewer people want your Chancellor than the other guy’s Chancellor, you won’t win an election, no matter what the headline poll figures say.