Nowhere have the fault lines in this arrangement been more harshly exposed than the attempt to implement the SNP’s manifesto pledge to reduce class sizes in the first three years of primary to a maximum of 18.

Different councils have pursued this goal with differing levels of zeal, depending partly on the circumstances in their particular area, partly on their existing policy objectives and partly as a reflection of their own political complexion.

So SNP-led West Lothian made great strides on class sizes, while West Dunbartonshire preferred their pioneering literacy programme involving

targeted support for slower learners, and Labour-dominated Glasgow declared the policy unsustainable within current budgets.

While falling rolls in some areas have enabled schools to meet the targets effortlessly, in other places an influx of population or pressure on popular schools have resulted in primary one classes getting bigger. Earlier this year it was calculated that at the current rate of progress, it would take until 2096 to reach the SNP target. The current situation is a shambles with some schools half empty and others bursting at the seams.

As revealed in The Herald today, in an attempt to hold the situation roughly where it is at the moment, the Scottish Government is to change the law to enforce a class maximum of 25, though only in primary one.

Essentially, this closes a legal loophole. The statutory maximum is 30 and parents making placement requests have been able to exploit it to gain places in classes of less than that number which the authorities had claimed were “full”.

As a result of such pressure East Renfrewshire decided it could not refuse placing requests until lower primary classes reached 30.

As only 6% of Scots pupils are in primary one classes of more than 25, the new legal maximum certainly looks attainable, even in the current climate.

Of course, the SNP’s political opponents are quite entitled to claim that this is another manifesto promise it has failed to deliver on.

Having increased numbers of teachers in training, partly to meet the challenge of smaller class sizes, hundreds of them now find themselves jobless after their probationary year, as schools clear them out to make way for a new generation of newly qualified teachers.

Putting the legal maximum at 25, with P2 and P3 staying at 30, will not help many of them into work, especially when hard-pressed local authorities may try to meet their new obligations by creating composite classes.

Smaller class sizes may look like a vote-winner but there was always a conflict between the mantra of choice in education that has been sold to parents for years and diminishing class sizes. Many would choose a class of 27 in a good school over a class of less than 25 in a

mediocre one. Bringing down the legal maximum to 25 will deny some parents their first choice of school.

However, education policy must be judged by what constitutes the greater good. Attempting to rationalise a situation described by the education secretary herself as “a dog’s breakfast” cannot be bad. Whether a more rational system is deliverable within the concordat is doubtful.

Meanwhile, we need more solid evidence that reducing class sizes further represents best value. There is some evidence that children thrive better in smaller classes when they start school.

Whether the enormous cost of building new classrooms and employing hundreds of extra teachers required to achieve 18 can be justified, especially during a recession, is questionable.