This week, 319 secondary schools have been told that they are coasting and so will become eligible for intervention. This includes a staggeringly high 22.6% of schools in the east Midlands.

For those deemed to have insufficient capacity to bring about improvement this could be the start of a long road to forced academisation. And for the third or so who are already sponsored academies, re-brokering becomes a possibility.

There may well be schools where teachers and heads are not working as effectively as they possibly can to secure the best outcomes for students. Or where another leadership team could take over and run the school far better than the current one. If there are, we don’t think the government has managed to identify these schools.

We have written this before, but it needs to be repeated: secondary schools serving more affluent communities cannot be deemed to be coasting. This is because the current definition of coasting has a combination of metrics with strong social gradients (such as progress measures) and threshold hurdles that poorly performing affluent schools can always reach:

  • In 2014, fewer than 60% of pupils achieved 5 A*-C at GCSE (including English and maths), and the school has less than the national median percentage of pupils who achieved expected progress in English and in mathematics; and
  • In 2015, fewer than 60% of pupils achieved 5 A*-C at GCSE (including English and maths), and the school has less than the national median percentage of pupils who achieved expected progress in English and in mathematics; and
  • In 2016, the school has a Progress 8 score below – 0.25 and the upper band of the 95% confidence interval is below zero.

The dark pink dots on the chart below highlights that schools deemed to be coasting are overwhelmingly serving more disadvantaged communities (higher deciles correspond to greater deprivation).

Coasting schools by decile of disadvantaged pupils

This social gradient will improve in years as the threshold measures are gradually replaced by Progress 8, but they will not be removed entirely because Progress 8 itself is always likely to retain a social gradient.

We think there are fairer ways to identify coasting schools.

There are schools that are persistently achieving worse GCSE outcomes than other schools with exactly the same ability and social profile of intake.

We don’t know why their results are worse, but if we must choose schools that we think could improve somewhat, we would pick these.

We do this by measuring each school’s performance on a contextual value added (CVA) measure that evaluates GCSE outcomes (best eight GCSEs in 2014 and 2015; Progress 8 in 2016 using provisional data) given the prior attainment and social profile of the pupils.

We find that 240 schools were in the lower quartile on these measures for each of the last three years. Of these, 113 met the DfE coasting definition.

The CVA measure, by construct, ensures that the probability of having a negative score does not rise the more disadvantaged the schools is (see here). This is shown by the blue dots in the above chart. On the whole we think this measure is much fairer.