Why do exam grade boundaries change every year?

Because the boundaries are the mechanism that keeps the standard constant when the paper is not.

The underlying problem. It is impossible to write two exams of exactly equal difficulty. Question wording, topic selection, and how a mark scheme behaves in practice all vary. If boundaries were fixed at, say, 70% for a grade A, then a marginally harder paper would produce a collapse in grades and a marginally easier one an inflation — and neither would reflect any change in how much students had learned.

Moving the boundaries corrects for this. A harder paper gets lower boundaries; an easier paper gets higher ones. A student's grade should reflect their attainment, not the luck of which paper they sat.

How boundaries are actually set (England's system as the example):

Marking finishes first. Boundaries are set afterwards, once the distribution of marks is known — they cannot be published in advance.

Statistical evidence. Awarding bodies use prior attainment data for the cohort — how this year's candidates performed at an earlier stage — to predict roughly what the grade distribution should look like if standards are stable. This is comparable outcomes, and it prevents both grade inflation and unfair deflation.

Examiner judgement. Senior examiners review actual scripts at candidate boundaries to check that work at the proposed mark genuinely represents the standard for that grade.

Ofqual regulates the process, and boundaries are published on results day.

Two common misconceptions:

They are not set to fail a fixed proportion. The system is not a pure curve; it combines statistics with judgement about work quality.

Boundaries differ between papers and tiers within the same subject, because each component has its own difficulty.

Wide variation between years usually signals a paper that behaved unexpectedly, not a change in standards.