When I heard about the proposal to change over to, some would say back to, a more rigorous O level style examination for students of mathematics, I was reminded of many conversations I have had over the past years on the subject of grade inflation. First I must declare an interest as a tutor of many students who have taken or are preparing for examinations at GCSE, A level and beyond.
Grade inflation has taken place. To convince yourself of this you only need to take a look at the exam papers taken by (mathematics) students over the last decades. The O level papers taken by students in the 1960s are certainly more demanding. You can view an example with my answers here. I took my O level mathematics in 1966, A level in 1969 and final degree exams in 1969. My children took GCSE and A level mathematics in the 1980s and 1990s. I have been tutoring mathematics since 2004.
Thinking about this, experience of this type may not be sufficient to make scientifically sound conclusions on grade inflation. However, my view remains that the mathematics exams are less demanding now that they were. If the results achieved have continued to rise and examinations are easier then, at the very least, we are not identifying the most able mathematicians. There will be punching at the top where the percentage of the student population achieving high grades has risen.
The percentage achieving the highest grade at A level used to be of the order of 5% or less of students taking the examination and similarly the percentage achieving a first class degree at university was very small - in my year just two in the class of 1969 at the University of Manchester. (I was third in the pecking order which is why I know the details so well!)
I do not believe that the distribution of mathematical ability will have changed over the years. I expect and hope that the general level is on the up - but I am convinced it is not so great as the higher percentage of top A level grades would today imply. I take nothing away from the work of all students but I do feel it is more difficult now to identify the most able which is a disservice to them and to mathematics in the UK.