This could become another Poll Tax moment for the Conservatives

In both cases, the failure to question assumptions allowed a sensible idea to become a total disaster

One of my favourite books about how leaders make mistakes is The March of Folly by American historian Barbara Tuchman, published in 1984. Her case studies began in ancient legend with the fatal decision by the Trojans to take the wooden horse within their walls, proceeded through the disastrous conduct of eighteenthcentury British leaders that led to the loss of America, and culminated in the long humiliation of the US in Vietnam.

In these and many other cases, Tuchman argued, intelligent leaders took seemingly rational decisions that nevertheless added up to catastrophic failure and persisted in their chosen course of action despite clear warnings about the consequences. Quite often the cause of failure was starting out with an incorrect assumption – the Trojans considered horses to be sacred animals and were easily persuaded that they were being given an offering to the gods. Sometimes it was overconfidence – George III assumed that a few colonists could never match his military might. And very often there was a feeling of being in so deep that there was no choice but to persist, in an atmosphere where dissent was not welcomed, characterised by Lyndon Johnson’s decisions to commit more and more forces against the Vietcong.

During my own political lifetime, the most striking example of this was the Conservative persistence in introducing the poll tax, or community charge, in 1990. All of the above factors were at work. There was a strong initial assumption that a domestic rating revaluation was unpalatable, although alternatives such as simply not doing it were not adequately considered. The party had won such crushing consecutive victories that it seemed a little thing like local taxation could not possibly bring the house down. Most of all, once the prestige of Margaret Thatcher – a prime minister famed for sticking to her guns – was attached to the policy, it was very hard for people around her to start saying it had all been a terrible mistake.

Once the initial decisions had been made, ever greater complexity was created to try to make it work, with transitional subsidies, means-tested rebates and a great mass of wellintentioned new rules. The intended effect of spreading the cost of local government more widely and improving accountability was briefly attained. But the unfairness for millions of individuals who saw their bills rise sharply swamped these advantages, producing a situation that MPS in their constituencies found impossible to defend. The result was that our greatest peacetime leader was overthrown for the sake of a policy not central to what she was trying to achieve, and it was abandoned anyway as soon as she was gone.

This pattern of a flawed starting assumption and perfectly rational decisions that flowed from it has been once again plain to see in the debacle of the A-level results. Officials no doubt came to Gavin Williamson months ago to say that if the grades predicted by teachers were used there would be massive grade inflation. Such inflation was exactly what the Government had been trying for years to prevent. Therefore, the predicted grades would have to be adjusted downwards. And that required a computer program, an algorithm, to do it. Everything then proceeded from that assumption.

From that point on, it is easy to see how each decision was made. Someone will have said: “surely we can rely on the predicted grades where a teacher only had a few pupils”. Oh fine, would be the response – exempt those in groups of 15 or fewer. But the result of that was to favour elite institutions. Then the Royal Statistical Society, concerned about the lack of outside expertise on the advisory body working on the algorithm, offered two distinguished fellows to help. The officials at Ofqual, however, would have worried that outsiders would open up a public debate on how to make the calculations, so demanded a non-disclosure agreement. The distinguished fellows refused and went away. Thus diversity of advice was lost.

By the time the Scottish results came out to utter condemnation, the whole business had reached the familiar point of being so far advanced as to make it difficult to turn back. Hence the move by the Education Secretary to try to save the entire exercise by allowing mock results to be used, but bringing a new layer of complexity that baffled Ofqual and led to them producing and then withdrawing their guidance for appeals in a single day.

In a striking parallel to the poll tax, the overall outcome was broadly defensible but the individual impact was not. A-level results on average improved a little. Sixty per cent of grades received by students were exactly the same as those submitted by schools, and 96 per cent were within one grade. A record number of 18-year-olds from the most disadvantaged backgrounds were accepted into university. The trouble is that qualifications are individual, not collective. This still left tens of thousands aggrieved, with results that were difficult to explain because they were adjusted by an algorithm developed behind the scenes.

There are some policies that are so vital that they have to be pursued no matter what, or that will be proved right in the end. There are others that are just a balance of judgment, where critics have a fair point, and which cause injustices that were never intended. The A-level algorithm was one of them, and the Government is right to abandon it and base grades on predictions, including for this week’s GCSES.

But there are also lessons for the future. Ministers must base policies on expert advice, as in this case, but expert officials, left to themselves, will keep adding to a policy that might have been an error. Someone has to ask: “Is this all based on the wrong assumption from square one?” Or say: “This doesn’t fit with our political objective of delivering fairness and opportunity for young people in difficult catchment areas”. That someone is a minister, listening to diverse advice.

All of us who have served in government have made mistakes. The important thing is to learn from them. The ministers at the Department for Education should now start wargaming what can go wrong with the reopening of schools in September and ensure there are no further unforeseen crises for pupils who have already lost crucial months of education.