Should England make setting and streaming compulsory?
The Guardian has reported that Education Secretary Nicky Morgan is pondering a new policy for Ofsted, the school inspectorate.
Patrick Wintour says schools in England could now face a presumption in favour of setting and streaming.
To cut through the jargon, that means either sorting children by ability into different classes in individual subjects ("setting") or separating children into different groups by ability in which they stay for all classes ("streaming").
The former is commoner than the latter.
Ms Morgan told Parliament on Wednesday afternoon that there was "no truth" to the report.
Still there might be action in this area - even something close to this idea.
There is some enthusiasm for it at the top of the Conservative Party.
Lots of tinkering
As I wrote on Monday, ministers do really love micro-interventions.
Despite all the talk of autonomy for teachers and school leaders, there's a lot of tinkering.
But that is because this sort of thing is what voters understand.
And talk of setting and streaming is a way to express a preference for traditional education.
The construction of the idea also illustrates how, with so many schools now converted to academy status, the only lever the government has left is Ofsted.
Academies, after all, may opt out of the national curriculum. The inspector and the league table are, for most of them, their only real oversight.
Still, we can say some things about setting and streaming thanks to an international study called Pisa, run by the Organisation for Economic Co-operation and Development (OECD).
It is, in effect, a survey of schools combined with tests.
It asks participating schools about their habits, then tests their 15 year olds to see what lessons we can draw.
The survey fixates on maths. It also runs state and private schools together, and England is included with the rest of the UK.
But it gets some striking results; in the UK, only 6% of pupils are not "grouped by ability within their math classes".
The question leading to that answer might take in some practices that fall short of full setting and streaming.
But, even if that is the case, that's an exceptional number for that question, and tells us something about practices in this area.
Compare it with the OECD average for that measure of 49%. In South Korea, it is 28%. In Japan, it is 54%.
Those two countries are much better than us at maths, and are usually not considered to be at the woolly, progressive end of the pedagogical spectrum.
Pisa also finds that more than three-quarters of maths teachers in the UK say they would never use materials designed to be followed by pupils of different ability.
So, even when they have a range of ability in a class, UK teachers tailor what they're doing by ability.
The OECD average is 22%. For Japan and South Korea, it is 17% and 32% respectively.
I would expect that, in the UK, maths gets ability-grouped more than other subjects.
But I wouldn't expect that a country which is so unusually comfortable with setting for maths isn't doing more of it in other subjects than other countries.
So is it a good idea?
Well, whatever Japan and South Korea are doing to drive their maths scores up, it isn't setting.
Indeed, looking at crude Pisa results, maths ability-grouping tends to be linked to lower maths scores (even within the UK).
Satisfies an urge
Furthermore, as the Guardian reports, there is no strong academic evidence supporting more setting and streaming.
In fact there is some evidence it is harmful.
That might surprise a lot of British people.
Intuitively, it feels logical. It satisfies an instinctive urge to allow the top to prosper, in particular.
But in setting-free Finland, which does famously well in testing, the consensus is that the existence of lower sets may encourage teachers to lower expectations for lots of children.
So rather than hassling a struggling child all year to keep up, they subconsciously think "Oh well, they'll just go down a set next year".
This issue is still being reviewed in a British context via the government-backed Education Endowment Foundation.
We will know more about the wisdom of a very common British educational practice - and whether we do enough of it or too much - soon enough.