When a university lecturer, I was once asked to evaluate an application for graduate study from an Italian student. I was amazed to discover that her average grade for undergraduate courses was 99.7 per cent. She must be a genius.

Then I read the essay she had submitted in support of her application. It suggested that she was not a genius. A colleague explained the discrepancy to me. Grade inflation at Italian universities was so extreme that every student scored above 99 per cent. I should look at the decimal place and consider it a grade out of ten.

This example shows why one popular suggestion for solving A-level grade inflation is a short-term fix at best. Subdividing top grades – for example, by creating A-plus and A-plus-plus grades – may allow us to discriminate between good pupils this year. But if the causes of grade inflation persist, soon too many pupils will be scoring A-plus-plus and we will need to subdivide again. Changing a grading metric can no more eliminate grade inflation than crossing zeros off Zimbabwe’s dollar notes can eliminate its inflation.

The cause of grade inflation at many universities around the world is the crazy but common practice of lecturers grading their own students. This gives them two reasons to award high grades. It makes their courses popular and it helps people they have come to know and, often, to like into good jobs. But this cannot explain the inflation of A-level grades because these are awarded by examination boards. With independent examiners, why should A-level grades inflate?

The answer can be seen by analogy with the consumer credit business. A levels are to universities and students what credit scores are to banks and borrowers. Universities admit students on the basis of A levels and banks make loans on the basis of credit scores. Universities (banks) want A levels (credit scores) to provide accurate information about academic ability (credit worthiness) because admitting weak students (lending to risky customers) and rejecting clever students (declining safe borrowers) are bad for business. But only the best students (safest borrowers) want grades to be accurate. Most benefit by grades being skewed upwards.

Businesses tend to do what benefits their customers, since otherwise they will lose out to competitors. The credit ratings provided by companies such as Experian are purchased by banks, not borrowers. So we should expect them to be accurate rather than favourable. A-grades are provided by examination boards who charge students for their service and compete for business from schools. So we should expect A levels to be favourable rather than accurate.

The incentives for grade inflation are worse yet. Imagine that the Government involved itself in consumer credit as it does in education. Imagine that there were a department dedicated to improving the creditworthiness of the population, and that it had a stated target for increasing the number of people granted loans. And imagine, finally, that the Government also regulated credit agencies and the way they assessed credit quality. It would be amazing if credit scores did not inflate like A levels.

Here, then, is a simple suggestion for reform. Examination boards should be legally obliged to sell their products to universities, not students or schools. In all other respects, the provision of grades should be deregulated. Grade inflation would stop immediately. No company trying to sell its grades to Cambridge University would lump half its students into its top grade, as occurs with the further maths A level.

But this is only the reform’s most obvious advantage. Liberated from the guidance of Whitehall, examination boards would be free to develop new methods of assessing pupils’ academic ability – or, more precisely, of predicting students’ likely success at university. Competition would give rise to methods that are not only more accurate than the current A level but also more cost-effective, since universities would prefer the cheaper of two equally accurate grades.

As exam boards developed the predictive sophistication of credit rating agencies, we could expect to eliminate another serious defect. Well-designed credit scoring methods are difficult to “game”. Gaming occurs when a borrower does something that improves his score without improving his credit quality. An example could be borrowing from a lender whom the scoring method cannot observe, such as your mother, and depositing the money in a bank account that it can observe. This makes the scoring method overestimate your wealth and hence your credit quality.

Academic evaluation can also be gamed. Cramming is the obvious example. A period of intense, coached study just before exams improves results but it makes the student neither more intelligent nor more knowledgeable, as this rapidly digested knowledge is equally rapidly egested after the exam. A well-designed academic evaluation would be made immune to cramming, perhaps by randomising exam dates.

Truth would not be the only winner. Cramming for exams is a classic “arms race”. Once someone starts doing it, everyone has to join in to keep up. Relative positions and prospects of acceptance by universities end up unchanged, but you have all squandered hours of effort on the enterprise. Students, like universities, would benefit if grades could not be gamed by cramming.

The Government should pounce on this proposed reform. Not only do I offer it free, it is also cheap to do. It requires no public spending or expensive new agencies. They just need to sell their products to people with an interest in the truth.

Times Recommends

• Strip away the figleaf and reveal naysayers
• Dave’n’George: there may be trouble ahead
• I do not accept that £64,000 a year for politicians is peanuts