It's that time of the year once more: we have made it through the GCSE and A-level results season again. All over the country, school corridors have witnessed tears and hugs as teenagers got their results and journalists turned to their word-processers to bemoan falling standards or to bemoan the critics. But one feature of the exams system in this country rarely gets a mention: the ways the examinations are actually run.
In England, those teenagers will receive their results from one of three large "awarding bodies". All three are private enterprises: AQA has charitable status, OCR is a company limited by guarantee, and Edexcel is part of the Pearson publishing empire.
What this means is that the management of examinations in England is, perhaps uniquely, in the hands of three companies who depend for their business model on maximising market share and "selling" their syllabi to schools.
Schools, in turn, depend on a good performance in examination results for their reputation, standing with the Government and ability to recruit pupils.
Secondary exams also have a huge influence over the school curriculum, and this is something for Education Secretary Michael Gove to bear in mind if he seeks curriculum reform.
How did we get to a position in which pupil assessment was virtually a tradeable commodity?
School examinations were developed by universities first as an admissions tool, and later as an income-generating activity. The title of the once powerful "Joint Matriculation Board of the Northern Universities" gives us enough clues to the origin and purposes of external examination: they were ultimately about access to university.
This was the purpose of A-levels when they were introduced in the 1950s. As in so many other parts of the English education system, status, hierarchy and class dominated.
At the top of the system were the examination boards of Oxford and Cambridge; beneath them came those of London and the Joint Matriculation Board (JMB). Most children were not entered for external examination.
The introduction of the CSE in 1965 began to change this, but largely preserved the hierarchy of the examination system - CSEs were organised and provided by a multitude of often very local examining groups, and never gained the status of the O-levels and A-levels provided by the university-backed examination boards. Vocational qualifications, run through employer organisations, had the lowest status of all.
The development of GCSEs in the mid-1980s simplified the system. For the first time, there were national criteria as to what a school leaving examination in each subject should offer.
Smaller CSE boards became unviable and a process of merger produced large "examination groups". GCSE represented a centralisation of control over assessment just as the national curriculum, three years later, centralised control over curriculum. And it had much the same results - some innovative provision disappeared, but in return there was greater standardisation. As a by-product, universities' interest in examination boards declined substantially, though it had been waning for some time.
The 1990s saw further centralisation; raised expectations of exam groups and more aggressive market management produced the three awarding bodies we have now. Schools still had some choice. In a highly constrained and controlled market, they could decide which examinations to enter their pupils for.
Awarding bodies, then, are effectively commercial organisations running school examinations, highly regulated but concerned with market share and profitability.
If anything, the current complexity is likely to get worse, which will make the system less sensitive to learners. The haphazard way that the system has developed has led to poor coherence and lack of clarity about their role among awarding bodies.
Ministers have now said state schools may teach the IGCSE, an option preferred by some public schools because it is more traditional in its assessment structure - dependent on a content-based final exam.
Its introduction and the spread of the International Baccalaureate (IB) are likely not only to make the system more complex, but reinforce the lingering class issues in assessment. As largely unregulated exams, marketed on their status and distinctiveness, IGCSE and the IB will exaggerate hierarchy in the exam system: the exam you have done will matter as much as the grade you get.
The teenagers who received their results this August deserve congratulations and praise. They have worked hard. Exams are tough and demanding. Success is hard-won. Standards of assessment and marking are carefully policed by examiners who are diligent, professional and committed. But like so much in English education, external examining has been riven with issues of status and hierarchy, complicated in the past 20 years by a generous mix of marketisation.
The governance of the exam system - an amalgam of its historical roots and its contemporary market-based management - raises tough questions, especially in light of the power exams have over the way the curriculum is taught.
Should exams be run by private companies, charitable trusts or - as they once were - universities? How can we develop an exam system which is rigorous and fair, and puts the needs of students at the centre? Are the exam boards as we currently have them the best bodies to shape the post-14 curriculum?
If Michael Gove wishes to rewrite the school curriculum, he may have to think about such issues first. If he doesn't, the accelerating complexity of the exam system may undermine attempts to reform the secondary curriculum.
Chris Husbands, Dean of the Faculty of Culture and Pedagogy at the Institute of Education, University of London. He will become the IOE's director on January 1.