GCSE wailing and gnashing of teeth

Much sound and fury resulting from today's GCSE results:

Up to 10,000 pupils are believed to have missed out on C grades in English — considered a good pass — as results registered their only annual decline since 1988.

Ok, let's reset expectations here. A C grade in 1987 English 'O' Level was a good pass, demonstrating a reasonable understanding of spelling and grammar. The pass rate for 'O' Levels generally at grade C or above in 1987 was just north of 30% (Figure 3.7) Note that boys and girls had virtually identical scoring averages at 'O' Level; only when GCSE was introduced did girls start to rise significantly above boys.

To see the trend, look at the 2001 GCSE results:

The National Association of Head Teachers, looking forward to the day when every subject is awarded at least a grade C, said such a "quantum leap" would require "urgent solutions to the teacher recruitment crisis".
Apparently agreeing, the CBI said it was alarming for competitiveness that more than 40 per cent of the entries in English, maths and science failed to achieve a grade C or above.

So just under 60% of students taking GCSE English (and that's pretty much all students) got a grade C or better in 2001. What about in 2012?

The drop in the number of pupils awarded good results in English proved controversial. Nationally, 669,534 sat GCSEs in English language or a joint language and literature paper, but the proportion of C grades dropped from 65.4 last year to 63.9 per cent. It equates to a fall of just over 10,000 on the number of pupils expected to gain good marks.

Oh noes! Only 63.9%? How unfair! So, how unfair was it to the 2001 pupils of whom less than 60% scored C or above? Was the class of 2001 noticeably more thick than the classes of 2011 or 2012? If not, why the discrepancy in grades? What about the class of 1987?

You do have to feel sorry for some of the pupils in this year, though. The goalposts moved, but not in the direction that their teachers expected; therefore their grade expectations (on which university applications were based) were overly ambitious. Against that, however, every student this year endured the same penalty; I expect a frantic scrambling by 6th form colleges to reassess acceptable entry levels.

Grading on a curve is (IMHO) the least worst solution to GCSE. Nearly every child in the age bracket takes GCSEs, so by grading on a curve for widely taken subjects you provide a relatively objective assessment of the ability of that child against their peers; since one year is unlikely to be much smarter or dumber than another year, there's reasonable read-across. Note that this doesn't work for more selective subjects where the academic profile of students is very different from the main, e.g. Latin and Greek, but let's solve one problem at a time.