When Lord Justice Leveson launched his inquiry into the press last year, he said that its prime purpose would be to decide “who guards the guardians”. It is not so different for us in education. Surely, the biggest issue we face now is assessing the assessors.

It is open season on exam boards. They have no friends. Nobody loves them. Before you reach for your hanky, you should pause. Frankly, they do not deserve your love. We all know that errors in examination papers are unforgivable. We all know that assessment at GCSE, AS and A2 is neither reliable nor robust. We all know that the responses we get from the boards when we query decisions vary from the helpful, through the tetchily defensive to the downright obstructive. We all know that the commercial activities of exam boards might compromise their professionalism. Above all, we all know that some of our candidates finish their courses with results that are not right - and not all of them are too low.

This is important, for it means that the quality of terminal assessment does not do justice to the teaching and learning going on in schools and colleges. That is really serious.

We - or our political masters, anyway - have chosen, over the past 30 years, to develop a national educational provision that is defined by its assessment regime. We can moan about it, but that is what has happened. The exams determine the teaching and learning that precede them. A weak assessment process, therefore, does not simply produce dodgy results; it undermines the very education we provide. It is as important as that.

So what do we do about it? Well, of course, we bash and blame the exam boards first of all. That always goes down well, but I do not think that it is enough. (Some of this is their fault, but by no means all of it.) Then we look at the nature of assessment itself. There have been welcome and intelligent calls for a move towards more assessment by multiple-choice questioning, which would create less scope for examiner error and a good test of fundamental understanding. I am sure that some extension along these lines is a good idea, but you cannot replace all assessment with multiple-choice tests. Think what it would do to teaching and learning in subjects such as English and history.

We need to go further. In the end, a regime of assessment must rely upon people making informed judgements about the quality of the work of students. We cannot avoid that completely. That is what assessment is. And, crucially, it relies on the quality of the people making the judgements. It is also what goes on in our schools all the time. It is a fundamental part of the teaching process. Experienced teachers use their critical judgement all day, every day.

So why do the exam boards find it so hard? Because they do not recruit the best people. There are lots of reasons why the best and most experienced teachers are not involved in the assessment business. They are very good reasons, too: money, time pressure, career structure and so on. Until we put this right, the assessment regime will continue to undermine education.

To their credit, the exam boards have been working hard over recent years to reduce the degree to which poor examiners can damage the results of students. They have developed increasingly precise and inflexible mark schemes. They have pioneered online marking that allows individual examiners to focus on specific questions rather than on whole scripts. I appreciate why they are doing this; they are trying to produce an assessment regime that can be successfully operated by inexperienced and unreliable assessors. But that cannot be right. Even if they are successful, they will simply have produced a second-rate assessment system. Assessment is only as good as the people who do it.

The best people are there. They are working every day in schools and colleges. They are teachers. So if we are going to salvage the situation, we have to bring the businesses of assessment and education closer together. Exam boards should be working far more closely with schools and universities - after all, this is where the expertise, the experience and the enthusiasm for teaching and learning are found. I suspect that will require a restructuring of the boards, but we shall see.

Assessment should be properly recognised as an integral element of teaching and learning, not just an extra bit at the end, conducted by outsiders. More attention should be given to the science of assessment within the training of new teachers so that we develop a reliable cohort of national assessors. If they do not come from the teaching sector, I do not see where else they are going to come from.

Every school should, therefore, have expert assessors on its staff, to guide and support colleagues. Of course, not all teachers will be interested but, if we do it properly, enough will. There must be a career structure, too. It should be possible to become a professional teacher-assessor, working within your school and with an exam board, just as it is possible to earn management allowances for so many other areas of professional work. This should not be something you do as an extra, squeezed in at night after a day’s work. It should be a recognised and rewarded part of certain teachers’ professional responsibilities.

Of course, it will take time to get this established. It need not be too long. The Chartered Institute of Educational Assessors already exists as a professional driving force. It will cost a certain amount of money, but we cannot afford not to do it.

Ken Durham is headmaster of University College School and chair of the Headmasters’ and Headmistresses’ Conference.