Rebecca Long Bailey inadvertently posed a good question yesterday, when she tweeted:

Rishi Sunak the new Chancellor is a former Goldman Sachs banker who has backed anti trade union regulations; tax cuts for big business; opposing tax avoidance clampdown; cutting tax for wealthiest. The mask is slipping already, this isn't a different kind of Tory government.

This raises the same question that is being begged by the Labour leadership contest: to what extent does a person's past behaviour and beliefs predict their future behaviour?

We have countless examples of it not doing so. All of you will have worked with somebody who arrived at your firm with a glittering CV but who turned out to be a dunderhead, and conversely with somebody of no great reputation who turned out to be excellent. Indeed, this is pretty much the history of top-flight football managers.

Similarly, in politics, people have defied their reputations. Martin McGuinness had a history of terrorism, but he became essential to the peace process. The racist Lyndon Johnson did more for civil rights than JFK who had a liberal reputation. Theresa May arrived at Number 10 carrying hopes she would fight "burning injustices", just as Gordon Brown did to hopes he would be a Viking warrior, whilst Thatcher was initially an under-estimated housewife. Francois Mitterrand became president amid socialist hopes, but was soon forced to make the austerity turn. And few people in 1997 predicted that Blair would become a reckless military adventurer.

And this is not to mention the many "future Prime Ministers" who shrank into obscurity - who today remembers John Moore? - or the ministers hoping to make reforms only to "go native" under pressure from the Sir Humphrys.

So, why might a man's past be an unreliable guide to his future?

Sometimes, it's because he can use his reputation to do things that others can't. McGuinness's links to the IRA allowed him to persuade men to put down their arms, just as Richard Nixon's reputation as a cold warrior gave him the political capital to seek détente with Russia and China.

A second reason lies in William Goldman's eternal truth: nobody knows anything. What we know of somebody - especially if only by CV or reputation - is only a partial guide to the reality and their future. As investors in Woodford's equity income fund know, even a good long track record is an imperfect indicator of the future.

Thirdly, and perhaps most importantly, we all face constraints and incentives that break the link between character and future conduct. As Marx said:

Men make their own history, but they do not make it as they please; they do not make it under self-selected circumstances, but under circumstances existing already, given and transmitted from the past.

But this is not an insight confined to Marxists. Quite the opposite. Warren Buffett made the same point when he said:

When a management with a reputation for brilliance tackles a business with a reputation for bad economics, it is the reputation of the business that remains.

And Charles Prince, former CEO of Citi, said the same thing when he famously said: "as long as the music is playing, you've got to get up and dance." If incentives are strong enough, one's past choices don't predict one's future ones.

Herein lies the reason why Ms Long Bailey might be wrong. Even if Mr Sunak is a head-banging inegalitarian, incentives and social structures constrain how far he can implement his desires. Acting upon his past beliefs is no way to retain the support of people who have only "lent" their votes to the Tories.

Now, I don't say this with great certainty: I prefer to leave futurology to people who are paid enough to compensate them for making fools of themselves. I do so, however to flag up a danger for Labour - that they might be mischaracterising the Tories by painting them as simple-minded Thatcherites.

Indeed, Ms Long Bailey might be making a very anti-Marxist error - a version of the fundamental attribution error, of attributing too much to character and underplaying the importance of social structure.