The subject of primary education is, indeed, one of great and serious importance,” said Bradford MP William Forster in 1870, speaking in a gaslit House of Commons to introduce one of the greatest social movements of the age.

Since then, under the terms of the Elementary Education Act passed that year, schools in England have had to provide places for children from the age of 5 and, since 1880, it has been compulsory for them to attend.

But decades of studies on how children fare in this system - set up before widespread electric lighting, the telephone or aeroplane flight - have found, of course, that some do much better than others.

Part of the reason for this seemingly obvious fact is that children arrive in school on Day 1 with certain characteristics already in place. Being able to identify and act upon these traits is the skill of teaching.

But to a teacher in the UK faced with a class of 30 four- and five-year-olds, some of the factors that will have a bearing on their learning are more visible than others. It is obvious whether or not a child is a girl (girls tend to do better than boys in literacy) or has an Indian or Chinese background (and will therefore tend to score more highly than white students). But what about age? Without surreptitiously glancing at the pictures of students’ faces smiling down from the months-of-the-year wall display, how many teachers would know Ruby, Arun or Shawn’s birthdays?

Yet research has shown that being born at the end of the school year - whether it is August in England, March in Japan or December in Tasmania - has a significant effect not only on how well children do at reading, writing and mathematics but also on whether they are held back a year, assessed as having special educational needs or have confidence in their own abilities. Children who are young for their year, researchers say, are more likely to be bullied and less likely to take up leadership positions.

In the lottery of life, the location of your birthday in the school calendar is a big deal.

This is not a new argument. But it is a fascinating one. In the 1980s, for example, Canadian psychologist Roger Barnsley noticed that many ice hockey players had one characteristic in common: they were born in January, February or March. They were the oldest in the “hockey year” and, at age 9 or 10, when professional teams start showing an interest, that birthdate made a difference in terms of strength and stamina. Those children deemed the most talented received better training, and that small advantage snowballed into a wider gap by adulthood.

A similar birthdate effect has been found among European footballers and US politicians - and, of course, in education.

Another influential study on the subject was carried out in 2006 by Kelly Bedard of the University of California, Santa Barbara, and Elizabeth Dhuey, now of the University of Toronto. They looked at the effect of birthdate on mathematics and science scores across 18 countries. They found that at age 9, the size of the effect was between 4 and 12 percentiles. That means that if a child born in August was ranked in the 78th percentile of those tested, a child of the same intelligence but almost a year older might reach the 90th percentile.

Never, perhaps, has this issue been more widely recognised. In his celebrated best-seller Outliers, Malcolm Gladwell picks up on these studies and says: “We prematurely write people off as failures ... We overlook just how large a role we all play - and by ‘we’ I mean society - in determining who makes it and who doesn’t.” The chattering classes lapped it up.

Surely the solution, then, is to make sure that those labels are not prematurely applied? To tell children’s first teachers not to designate students who are old for their year “gifted” or those who are young for their year “struggling” too early. But it’s just not that simple.

The political becomes personal

A heatwave has rolled over England a few days before the school summer holidays start. At The Wroxham School in Potters Bar, just north of London, teachers have opened the doors and let their class of 30 five-year-olds out on to the field. They are scattered in small clumps, some sprawled out on the ground, others sitting half-hidden in the long grass, quietly chatting. One group of four boys sits in a circle, their legs apart and their feet touching, forming a star shape. They have a ball that is going to be rolled between them.

This Reception class is, like most such classes, made up of children who are confident writers and children who are shy at speaking, those who make friends within minutes and those who still miss their afternoon naps.

Take the little ring of boys, who are discussing the rules of their game and have not yet started playing. Which are the two most academically able children? Teacher Sam Revell points to the oldest of the group, Alex, born in November, and to the youngest, Aaryan, who has a June birthday. And the two most socially confident? The same pair.

Summer-born children may do worse on average but that by no means predicts the outcome for individuals. And for teachers it is the individual that counts. “In Reception (children aged 4-5), every child is an individual,” Revell says. “We don’t label by ability and all the children are mixed up in groups.”

As the months progress, children encounter phonics, counting and social expectations. At the end of the Reception year, teachers are expected to assess them on these, and other, measures. Teachers at Wroxham take birthdate into account, but in a class setting it is quite possible that summer-born children will be among the high achievers, so why would they need extra support when others are still unable to hold a pencil?

Similarly, if a summer-born child is struggling would you want to sit back and put their performance solely down to age? It is just one factor and others, particularly socio-economic background, have a bigger impact. This is one reason why the newest member of the Royal Family, Prince George, is not doomed to failure simply because he was born in July (nor was his father, who was born in June).

Like a pointillist picture, it is only when you step back from focusing on the individual that the entire image emerges.

Ellen Greaves, research economist at the Institute for Fiscal Studies, is one of the co-authors of a series of studies on birthdate effect. She says that any discussion on the subject inevitably begins with someone saying: “Well, my child was born in August and is incredibly successful.”

But she says it is worth emphasising that the relationship is an average. “It is not the case that every August-born child will be below every September-born child. But, on average, this is the relationship we see,” she continues. “When you look at a large sample the differences are clearly there. We had national test data on all children in England and survey data covering 10,000 to 14,000 children. The differences are stark.”

For individuals, then, the solution seems obvious. If a child born in August (or July or June) is going to spend their time at school in a futile race to catch up with older children, hold them back a year so they are the oldest in their class instead.

Journalist Pauline Hull runs the campaign website summerbornchildren.org, “for parents who want their child to begin Reception aged 5, and not aged 4”. She argues that parents wanting to defer entry are not looking for an advantage but trying to eliminate a disadvantage.

“Report after report is emerging saying that there is a problem for summer-born children,” she explains. “So why not try to do something quite painless and see if it helps the situation? Starting Reception at 4 is not for every child.”

Hull has caught the attention of those in power. Liberal Democrat MP Annette Brooke has secured a half-hour parliamentary debate on the matter in September, after the success of an early day motion noting that summer-born children have “long-term disadvantages as a result of England’s inflexible school starting age”. The motion was signed by the chairman of the Commons Education Select Committee, Conservative MP Graham Stuart, and its former chairman, Labour MP Barry Sheerman.

“There are going to be quite clear-cut cases where it is a good idea to defer entry and there is no reason not to do it,” Brooke says. “It’s about seeing children as individuals. I think it is important that parents should have that choice.”

How early is too early?

But, in fact, there is a whole other side to this discussion. Perhaps the issue is not one of “summer-born” or “not summer-born”. Perhaps the question is so emotive in England only because children start school so early.

In England, children must attend school from the start of the term after they have turned 5. In the past, most schools asked them to arrive instead at the start of the term in which they turned 5 to attend a “Reception” class. But over the past decade, two important changes have taken place: first, an increasing number of schools have moved towards a September-only intake for Reception; and second, the Reception year has been given its own play-based curriculum, protecting it from the formal teaching that begins in Year 1.

When an investigation into summer-born children found that being in school for longer gave children a slight academic advantage, provided the curriculum was suitable, the system was changed. As a result, since September 2011, all schools have had to offer four-year-olds a full-time September start.

These changes have left the statutory school starting age of 5 somewhat redundant: the actual starting age is now 4. Parents have the right to request that their child attends part-time or waits until they are at statutory school age, but the decision to delay entry in order to change year group is made on a school-by-school basis.

Scotland has a similar starting age but a different system. Deferral is a right for some parents and discretionary for others, and between 7 and 12.5 per cent of Scottish parents do choose to defer their child’s entry year. But recently, moves have been made to limit the practice.

In 2011, education officials at City of Edinburgh Council realised they had a problem. The number of parents asking to hold their children back was exploding: three times as many parents as in similar-sized authorities. The officials investigated and found that parents were concerned that the curriculum in primary schools was not flexible enough and that their children were not ready for it.

Parental guilt is a powerful force. Although there may be good medical or emotional reasons for some children to be held back, evidence from the US suggests that parents are not always the best judges of this.

A study by Daphna Bassok of the University of Virginia and Sean F Reardon of Stanford University, published this year, found no evidence that children with lower cognitive or social abilities were likely to be “redshirted” (as deferral is known in the US). In fact, students who deferred were more likely to be white, wealthy and male.

The researchers put forward the possibility that after Outliers became a best-seller, and the disadvantage of being young was widely discussed, parents may have felt that letting a child enter school as the youngest was akin to relegating them to a life spent on the “junior” team. They also warned: “If redshirting rates are on the rise, and if differences in this behaviour persist across groups, poor, black and Hispanic children may be negatively impacted by redshirting practices.”

But what if the factors behind age-skewed ice hockey teams are not the same ones as those causing educational bias? After all, in ice hockey the birthdate advantages seem to multiply as children age, whereas in mathematics tests they fall away. What if it is not purely the fact of their relative youth that is causing summer-born children to fall behind?

Another recent study - this one in the UK by the Institute for Fiscal Studies - considered how school starting age, length of schooling, relative age and age at test affected children’s performance. They found that all four factors mattered (and are obviously closely related): test scores at age 7 were higher, on average, for students who had received more schooling, had started school at an older age, were older when they took the test and were relatively older in their class.

National test data for seven-year-olds found that 45 per cent of the difference in scores between August- and September-born children was because they were different ages when they were tested, and 32 per cent of the difference was put down to relative age.

These factors may sound identical, but they are not. The impact of relative age is thought to be because of the Matthew effect (named after the passage in the Bible that states “to the one who has, more will be given, and he will have an abundance”). According to this theory, the oldest children get greater access to challenging work, while the youngest are demotivated by being behind. In contrast, age at test is just that.

But what happens if you test children outside school, on non-curriculum questions that have nothing to do with how they have previously been assessed by teachers? When the team did this - looking at British Ability Scales scores of children taken from the Millennium Cohort Study, which tests children at a certain age rather than at a certain time and uses questions that are not linked to a particular curriculum - they found that age at test explained 78 per cent of the difference in test scores and relative age accounted for just 10 per cent of the difference.

Put simply, “relative age matters more for tests taken inside than outside school”, the researchers concluded.

Making allowances

The evidence is mounting that for summer-born children, the key word is not “age” but “tests”. To put it another way, deferral will not work for the system as a whole but age-adjusted testing would.

The ultimate age-adjusted system is perhaps that of New Zealand. There, most children start school on or near to their fifth birthday (although they do not have to attend until they are 6) and are assessed after they have been at school for a set amount of time, such as 20 or 40 weeks, rather than at the end of the school year.

The system supports personalised learning, especially in the country’s many small primaries, says Brigit Stevens, a Year 01 teacher at Tawhai School, a primary in Stokes Valley, Wellington. “From a practical point of view, with all the children starting in one block you get all your class rules and routines firmly established with all the children present.

“Our staggered starts mean you are re-establishing those rules and routines regularly and you have to have very clear expectations for the older children to help the newer ones to settle. This is fantastic for helping to develop role models and allows the older five-year-olds an opportunity to have responsibility in the class.”

And linking school start to a student’s age can make it a more personal rite of passage. “It was very exciting,” says New Zealander Megan Pacey, 38, now chief executive of UK charity Early Education. “There were expectations of you: ‘Now you are 5 you are going to school, you are going to have to tie your own shoelaces.’ I remember on my birthday being packed off to school with my birthday loot: a new lunch box, drink bottle and school bag.”

This individualised approach lasts until the end of students’ second year of schooling. Then children progress at age 7 to Year 3, having been at school for between one and a half and two and a half years.

So does it work? The simple answer is that there is no simple answer. In New Zealand, a birthdate effect still exists.

But the Bedard and Dhuey study showed that maths scores at age 13 did not show an age effect in two countries: Denmark and Finland. What these countries have in common is that they do not ability group in primary schools. So, alongside individual and system-wide measures, changes to school structures could also help to eliminate birthdate effect.

Speaking to TESS, Bedard outlined three key points that should be considered by those with an influence on school systems. First, the younger the starting age, the more room there is for relative age effects. Second, ability grouping or streaming at young ages should be avoided. Third, school readiness across different socio-economic groups should be carefully considered to ensure that children are prepared when they enter school.

The fact is that when you are born matters, even though children are assessed against norms that pretend it does not. The effect diminishes as children age, but it remains. The simplest - and cheapest - quick win is age-adjusted testing, which is not perfect but appears vastly to improve the lot of summer-born children.

If age-adjustment made the system fairer, would we be happy with it? The fact is that the political will does not seem to be there. Could it be that the problem is not the system, but us? Perhaps more than wanting the system to be fair, we want our children to be first.