Alex Crowley was Research Director and then Political Director of Boris Johnson’s London Mayoral election campaigns in 2008 and 2012. He is also the author of Victory in London, a book about how those campaigns were won.

The general post-election consensus amongst commentators and insiders is that, however you regard this result, one thing is certain: another election this year would be bad for the Conservatives. Most confidently predict Labour would do very well (and by implication even better than they did on 8th June).

But is this a considered response to the available evidence, or is it a product of what usually happens after an election – namely, the hindsight bias where we convince ourselves that we knew what was happening all along?

This mental shortcut tends to produce an ‘inevitable’ analysis – where the event that happened was always going to happen. This shapes our analysis of similar events to come, thanks to another mental shortcut known as the Availability Heuristic. That is, we tend to judge something as more likely to happen that springs more readily to mind — usually something that has just happened.

Are these psychological tendencies leading us to be more negative about Conservative prospects in another election than is actually the case?

Cast your mind back to 18th April, when Theresa May called the election. The overwhelming view – within the Westminster bubble at least – aas that she was right to do so because she was inevitably going to win a large majority. This was based on three main factors.

1) A double digit Conservative lead in most opinion polls for many months

2) Large personal approval ratings for May and,

3) Low personal ratings for Jeremy Corbyn, and a settled consensus that he was a bad leader who would perform badly in any election.

These factors led to a very high expectation of a big May win and a heavy defeat for Corbyn. Up until the Conservative manifesto launch, (and in some cases after it) this was the prevailing narrative.

The actual result showed May and the Tories underperforming expectations and Corbyn and Labour overperforming. This led to a prevailing analysis that May was always going to campaign poorly, and Corbyn was always going to campaign well.

Indeed, you could be forgiven for thinking that Corbyn ‘won’ the election and May ‘lost’ it. Much post-election analysis has focused on how the voters ‘rejected’ the Conservatives and ‘embraced’ Labour. This has led to the narrative that, if another election were held this year, Corbyn would be the most likely winner. But I’m not sure this is the certainty many are stating it to be.

Granted, Labour did much better than anyone thought they could, but if the voters had truly ‘rejected’ the Tories, then Labour would now be in a position to form a government. Voters actually rejected Labour for the third time in a row.

It is also true that Labour is now ahead in the polls, and Corbyn has a higher personal approval rating than May. But are these factors enough to conclude, with certainty, that Corbyn is the most likely winner of the next election in such terms that imply close to 100 per cent probability?

I think the answer is no, or rather it is more unlikely than people assert and here is why.

  • The hindsight bias afflicted analysis of the election campaign. The most accurate polling model from YouGov showed that up until the launch of the Conservative manifesto, the party was on course for a comfortable majority (if not the 100+ one many were expecting). After the manifesto, the lead fell considerably and never quite recovered. Whatever you think of the policies, the manifesto drastically altered a Conservative message and strategy that was working. It is reasonable to think that had the strategy been consistently adhered to, then the party would now have a majority. However, much of the post election analysis has asserted (implicitly or otherwise) that May was always likely to have underperformed and the outcome was always likely given voter dissatisfaction with ‘austerity’ and the mobilisation of young voters. Thus, a new election is now considered to be highly dangerous for the Tories for much the same reasons: they will have a poor campaign as last time, and voters will punish them further on economic grounds. I believe this is down to heuristics, not thorough analysis. These things may be true, but they are not inevitable.
  • We don’t know who the next Conservative leader will be.  This should be obvious, but it is a factor strangely absent from most analysis. If you believe that the perceived strengths and weaknesses of both leaders were significant factors in the result, then it surely follows that a new Conservative leader will have an influence, and could outperform Corbyn. Most analysis of a potential new election is based on the implicit assumption that there is a close to 100 per cent probability that no alternative Tory leader could do better. On random chance alone, this cannot be right. Similarly, we have to consider that Corbyn may underperform in another election because expectations about him now are dramatically higher than they were when the election was called.
  • The impact of the result on Labour’s internal politics. Pre election, the working assumption was Corbyn would go down to an historic defeat, and thus the strategy of Labour moderates was to let him hang by his own rope. Therefore, his internal critics stopped public criticism. This had the affect of, for the first time, allowing Corbyn a clear run from his own side, and making Labour seem more united than they are. Post election the moderates have either u-turned completely or remained quiet, aware their strategy has backfired. But in a new election, who knows what calculation they will make and thus what impact it will have on the Labour campaign?
  • The electoral arithmetic shows that Labour needs more than ‘one last push’. The party needs to gain 64 seats and achieve a swing of  six per cent to be sure of an overall majority (they could govern with less, but let’s assume this is the benchmark of outright victory). Thus, they need to significantly outperform the result they achieved on 8th June. Any new election would see raised expectations of their likelihood  of winning. There will be even more intense scrutiny of their campaign, and voters will know there is a real chance of Corbyn winning, which may influence their decision. When Corbyn was looking like a loser, voting Labour was risk free. Who knows what level of risk voters will attach to a Labour vote if they assume Corbyn might win? In these circumstances, is a six per cent swing achievable for Labour? Yes, it is possible, but I wouldn’t state it with absolute certainty.

Does all this matter? Well, if you are a Conservative MP pondering whether to trigger a leadership contest, your view on the wisdom of another election – which must increase in likelihood with a new Prime Minister – will have a big influence on your decision. In the weeks that have followed the result, a common theme in briefings to journalists is that May should not be challenged because another election risks allowing Corbyn into Number 10. This is possible, but it is completely certain? Can such an outcome be altered?

There are plenty of good reasons to avoid another election – cost, impact on Brexit negotiations, general political instability – but the absolute certainty that Corbyn will win should not be one of them.