In our school system, a lot depends on whether a 10-year-old can punctuate correctly.

The results of writing tests taken by primary school pupils for their key stage 2 Sats are scrutinised by parents and government officials, feeding into a high-stakes accountability system that carries the risk of school closures and job losses.

The writing tests are assessed by teachers, whose assessments are then moderated to ensure consistency across the country. But while maintained schools are told that their assessments must be moderated by their own local authority, academies are able to pick and choose which authority carries out the moderation. And Tes has unearthed evidence that shows many academies are using this power to “shop around” by opting for local authorities outside their area.

This year, a quarter of authorities have moderation arrangements with at least one school from another area, according to statistics obtained from the Standards and Testing Agency (STA) through a Tes freedom of information request.

There may be many reasons why academies choose to switch local authorities, including a perception that some are more “lenient” or provide a better service. Whatever the rationale, it appears that a hidden market in Sats moderation is in operation - one that offers choice to academies, but not to maintained schools.

The situation has been described as deeply unfair by critics. “There should not be one rule for some schools and another rule for others,” says Mary Bousted, joint general secretary of the NEU teaching union.

‘Cherry-picking’ LAs

Tes’ investigation reveals that half of Swindon’s academies have decided not to have writing tests moderated by their local authority. Perhaps they have chosen neighbouring Wiltshire, where 16 academies from other areas have “opted in”.

Meanwhile, 14 of Bromley’s academies have selected a different local authority for their moderation; 22 academies have chosen to be moderated in Devon - even though it is not their local authority; and Kent has 11 academies “opting in” to its moderation system, of which nine have KS2 pupils.

Why does this matter? Aside from issues of fairness, Bousted says that allowing some schools to “cherry-pick” their moderating authority fails to address inconsistencies in the process. “The focus should be on making sure moderation is consistent and gives valid and reliable judgements,” she says.

James Bowen, director of middle leaders’ union NAHT Edge, agrees. “Rather than schools shopping around for moderators, it would be better to have a system we can trust that removes the incentives to do so,” he says.

In this market, what are schools looking for and what are authorities offering? James Pembroke, Tes’ “Data Doctor”, says: “I reckon the key reasons for switching moderation provider are a poor relationship with the old local authority since conversion, a better package of support offered by the other LA and a moderation process perceived as too harsh.” He adds that it is the latter that may often prove to be the tipping point.

Bousted says that some local authorities might “provide a better service [or] there may be a perception that some are more lenient”.

Unsurprisingly, Kent’s improvement adviser, Margo Barraclough, says it is the former: schools are opting into the local authority’s moderation system, she says, because they consider its process to be robust and supportive. “Kent is a large local authority, so moderators and schools see a wide range of evidence and are thus able to arrive at secure judgements,” she adds.

Swindon, meanwhile, accepts that larger local authorities with more staff can provide a broader range of moderation “packages”, including training for teachers, than it is able to offer.

But surely these more varied packages and enhanced services would also be attractive to maintained schools? Pembroke thinks that if maintained schools had the choice, there would be even more movement between authorities. He says: “Some local authorities have very few advisers left in terms of school improvement and this has left some schools thinking they don’t get the support from their local authority, so they’ll get their moderation done by another local authority.

“And disputes over the harshness or leniency of moderation is possibly playing into [decisions to move to another local authority].”

Bowen thinks that most maintained schools would choose to stay put, although he adds “it’s also fair to assume that some would [move]”.

He says: “It seems unfair that local authority schools don’t have that option. That being said, it would become a pretty chaotic system if every school in the country could choose who moderates them; it is hard to see how that would work logistically.”

‘There is a lot of anger’

Assessing children’s writing is not easy, which is why moderation exists.

Teachers have to check their pupils’ writing against a national set of criteria, and moderators visit schools to check that teachers’ judgements are correct. But in 2016, these criteria were made tougher to match the new national curriculum - and the amount of flexibility teachers had in making judgements was drastically reduced.

The changes created chaos and confusion, as it was reported that authorities were taking different approaches.

This seemed borne out by local authority results. In 2015, between 83 per cent and 94 per cent of pupils reached the expected level 4 in writing across different local authorities. But in 2016, the gap widened, with between 59 per cent and 85 per cent reaching the new expected standard.

These huge variations may explain why maintained schools feel frustrated by the lack of freedom to choose a moderator, says Pembroke. “I think, post-2016, there is a lot of anger. Local authority schools didn’t have the choice to move, but MATs [multi-academy trusts] and other academies did,” he says.

Growing doubts about the system led to a government U-turn on the use of writing data and a consultation that ended with the Department for Education pledging to look into new ways of judging standards of writing.

Two years on, and the chaos surrounding the new Sats may have lessened, but it has not disappeared. Last year, an Ofqual investigation revealed that, in 2017, differences in the way local authorities approached moderation led to judgements that were “more inconsistent than they could have been” (see box, below). It found that visits to look at the work of the same number of children could take twice as long in some authorities as others, which “suggested variations in thoroughness between LAs”.

These differences can create problems for MATs that run schools across a range of authorities and want to ensure that they are comparing like with like.

The Cabot Learning Federation has two primary academies in South Gloucestershire, four in Bristol, and a KS1 academy in North Somerset (see box, below). The academy trust has chosen South Gloucestershire as its moderating local authority for its Bristol academies. Susie Weaver, interim executive principal at the Cabot Learning Federation, says: “We thought it would be helpful, given our schools are all in one trust, to all be moderated using the same process.”

But if MATs walk away from certain areas, could that compound the inconsistencies? Bousted believes so. She says: “The focus should be on making sure there is consistent, valid and reliable moderation of KS2 writing and that should require every school within an area to go through the same moderation process because, if there were inconsistencies, they would show up much more clearly.”

A competitive atmosphere

The STA has already attempted to improve writing assessment and moderation - but there is still more to be done. In this year’s round of moderator training, almost one-in-eight moderators still failed to agree with the STA when tested.

However, the STA and the DfE seem to want to say little about the issue of making the system more transparent. The STA refuses to release an evaluation of last year’s writing-moderation process, claiming it is “not in the public interest”.

The DfE, meanwhile, was asked to comment on the potential for a lopsided market in a system in which some schools have the choice to choose their moderation team, while others do not. A spokesperson commented: “Local authorities have a duty to externally moderate the KS2 teacher assessments of academies and, in 2018, 96 per cent of academies chose their geographical local authority to conduct their moderation.

“We are not aware of any local authorities that make a profit from moderation and in most cases they charge the minimum cost to cover their resources.”

Profit or not, some local authorities clearly perceive a competitive atmosphere. Several refused to reveal the cost of moderation in their area on the grounds that it was “commercial” information.

Overall, the charges to academies for moderation range from £60 to more than £1,000, according to figures provided to Tes by local authorities (see box, below).

However, even without a profit motive, and with completely consistent judgments, the added value that some authorities are offering, such as professional development, could tempt those schools with the ability to switch to do so. And the perception that moderation is harsher in some areas than others could accelerate the process.

Meanwhile, authorities with smaller budgets have fewer options - and maintained schools have none at all.