From Brexit to the election of Donald Trump, could 2016 really have turned out any differently?

What would have had to happen for 2016 to turn out differently? For many people, the Brexit vote and Donald Trump’s election were like experiments in counterfactual history gone wrong. Both seemed to be arbitrary, knife-edge moments when events could have gone the other way.

Arithmetically, that is more true of Trump’s victory than it was of Brexit. The President-elect would have lost if Wisconsin, which he won by a margin of 0.8 per cent, had gone the other way. If that had happened, he would presumably have also lost Pennsylvania and Michigan, which he won by smaller margins, and we’d be wondering how Hillary Clinton would escalate the cyber-war with Russia.

Clinton won more votes nationwide, but she needed to close that 0.8 per cent gap in those three states to change the outcome in the electoral college. The Brexit vote, on the other hand, was decided by a 4 per cent margin. Yet there has been much more commentary about how David Cameron ran the wrong campaign than about what Clinton could have done to win.

In America, a lot of attention has been devoted to the apparent injustice of the electoral college system, which has twice recently denied the presidency to the candidate with the most votes. This is mostly a waste of pixels, because you cannot change the rules unless you can persuade a two-thirds majority in Congress and three quarters of state legislatures.

Beyond that, the counterfactuals that I have seen most discussed are: what would have happened if James Comey, the director of the FBI, had not made his statement 12 days before polling day re-opening the investigation of Clinton’s emails; and, what would have happened if the Russians or their proxies had not hacked the emails of John Podesta, Clinton’s campaign chairman, and the Democratic National Committee.

Those are important questions, but the one that interests me more is how Clinton herself could have fought the campaign differently to deny Trump victory. I haven’t seen a convincing answer to that, beyond, “be a better candidate”. That includes Barack Obama, who modestly told David Axelrod, his former aide: “If I had run again and articulated [the vision of a tolerant, diverse and open America], I think I could've mobilised a majority of the American people to rally behind it.” (Not that a mere majority is enough, but we know what he meant.)

He probably could have done – which raises the question of why those Rust Belt non-voters who turned out for Trump could have been deflected or outnumbered if Trump had been up against a black liberal Democrat whose name was on Obamacare. Even if Clinton had gone on a bus tour of the Rust Belt for the last campaign fortnight and wept in cheap diners over the loss of American jobs I don’t know that she could have closed that 0.8 per cent gap.

On Brexit, on the other hand, every sofa pundit and her dog has an opinion on how Cameron should have done it. Starting with not promising the referendum in the first place. I’ve explained before why I think that criticism is misplaced: Cameron had to make the promise to win the 2015 election, and a referendum was coming come what may. Saying we shouldn’t have had a referendum seems like the equivalent of complaining about the US electoral college.

Cameron failed to grasp the importance of one variable: the election of Jeremy Corbyn, a Bennite Eurosceptic, as Labour leader meant that the Remain campaign would be seriously weakened.

But again, the question that I think matters is how Cameron could have fought the campaign differently. Fortunately, for those interested in the answer, we have Tim Shipman’s superb account of the referendum campaign and its fallout, All Out War. One of the great strengths of the book is that it discusses the counterfactuals so clearly, weaving them into the narrative.

Should Cameron have given 16- and 17-year-olds the vote, as the Scottish Parliament did in the Scottish referendum in 2014? Whether or not you agree with it, it wouldn’t have been enough. My rough estimate, assuming they turned out at the same rate as 18- to 24-year-olds and voted 80 per cent to Remain (as against 73 per cent for 18- to 24-year-olds), is that it would have reduced the Leave majority to close to 2 per cent.

Shipman is thoughtful in judging whether Cameron could have done more to keep Michael Gove and Boris Johnson on board, as Johnson in particular could have made the difference. Without him the Leave campaign would have been a collection of minor oddities. “Without Gove, Johnson may well not have had the courage of his wavering convictions and made the jump for Brexit,” Shipman concludes. “Between 19 February and 30 June, it was Gove’s decisions which shaped history’s path.”

And yet such counterfactuals are hard to know. If Johnson had jumped for Remain, Theresa May might have jumped the other way, as she would then have been the leader of the Brexit campaign and well-placed to win a leadership election among Eurosceptic Tory members.

The other big unknowable is whether Cameron could have got a better deal from other European leaders. Shipman writes: “He never forced Angela Merkel or the other European leaders to choose between compromise [on freedom of movement] and the UK leaving the EU.”

One problem is that neither he nor they really thought that the UK would leave. And if Cameron had asked for an emergency brake on immigration from the rest of the EU – which he thought about in February, and in the closing days of the referendum campaign – he wouldn’t have got it, which would have looked much worse than not asking for it.

The conclusion of this short tour through recent counterfactual history is that maybe events such as US elections and UK referendums are not so arbitrary and changeable as they seem. And that the “great men and women” theory of history is still important. What decided both those democratic choices were the characters of the players themselves: Donald Trump, Hillary Clinton, David Cameron, Jeremy Corbyn and, most unexpectedly of all, Michael Gove.​