The Promise We Made About Piano Lessons
Somewhere in the pitch for your child's music lessons there is usually a claim about maths. When researchers pooled 54 studies and nearly 7,000 children, the claim fell apart - and the way we defended music may be exactly why it is disappearing from state schools.
Somewhere in the pitch for your child's music lessons, there is usually a claim about maths.
It comes from the visiting teacher, or the school newsletter, or another parent by the gate. Learning an instrument builds the brain. It helps with numbers, with patterns, with concentration, with holding things in your head. Nobody is being dishonest. It is just that when researchers went looking for the effect, they had real trouble finding it.
The idea has a birthday. In 1993, a short paper reported that 36 university students did slightly better on a paper-folding task after listening to Mozart. The lift lasted about a quarter of an hour. Within five years, an American state was posting a classical CD to every newborn baby.
There was no consistent advantage for the music group. Where a difference showed up in one trial, it failed to show up in the second.
Then came the bigger check. Giovanni Sala and Fernand Gobet pooled 54 studies covering almost 7,000 children aged three to sixteen, none of whom had learnt music before. Across all the studies there was a small apparent benefit. But when they separated out the trials that compared music against another activity, rather than against doing nothing at all, the benefit collapsed to roughly zero.
Their wording is unusually blunt for academics. Music training, they concluded, does not improve general thinking skills or school results. Learn music, and you get better at music.
This is not really a fact about music. Psychologists call it the far transfer problem, and it turns up everywhere. Chess clubs, memory apps, brain-training games and sudoku all tell the same story: practising a thing makes you better at that thing, and almost never at anything else.
You will still meet the musical child who is good at everything, and the pattern is real enough in the data. But lessons cost money and time, and they need a parent willing to nag about practice. Those conditions rarely turn up on their own. Music is doing some of the sorting, not much of the causing.
The instinct was understandable. Once every subject has to prove itself against a measure, the arts reach for the nearest number available. Music grabbed maths and reading, because those were the numbers that counted.
Something like that has now happened. GCSE music entries in England have fallen by 35 per cent since 2010, and A-level music by 47 per cent. One recent report found that 42 per cent of state secondaries enter no pupils at all for GCSE music.
What is left has drifted towards families who can pay. About half of pupils at private schools get sustained instrumental tuition; in state schools it is closer to 15 per cent. The average yearly budget for a music department is around £1,800 in a maintained school and £10,000 in an independent one. Four in ten primary schools offer no chance to pick up an instrument in a group during the school day.
The government's own music plan, published in July, notes that music GCSE has the largest gap between disadvantaged pupils and their classmates of any subject. A subject sold as brain training for everybody has quietly become an extra for the well-off.
The social claims deserve more care. A well-known 2010 study found that four-year-olds who drummed and sang together were more likely to help each other minutes afterwards. A careful field experiment in 2021 tried much the same thing and found nothing at all. Anyone who tells you the evidence here is settled has not read much of it.
Harder to measure is what an ensemble asks of a child. In an orchestra you cannot be the loudest. You have to listen while you are playing. And when you lose your place you carry on anyway, because thirty other people are still going.
There is also the plain business of being bad at something in public and continuing. A ten-year-old scraping through a piece at the Christmas concert is doing something most adults now arrange their lives to avoid. None of that appears in a league table, which is exactly the problem.
Money and timetable minutes will decide most of this. But the argument decides what survives the next squeeze, and our argument has been wrong for a generation.
If the reason your child learns the cello is that it might nudge a maths grade, that reason has an expiry date, because sooner or later somebody runs the trial. The other reason is smaller and much harder to write into a report. They will be able to make music happen in a room, with other people, for the rest of their life. Nobody needs a study to sign that off.
It comes from the visiting teacher, or the school newsletter, or another parent by the gate. Learning an instrument builds the brain. It helps with numbers, with patterns, with concentration, with holding things in your head. Nobody is being dishonest. It is just that when researchers went looking for the effect, they had real trouble finding it.
The idea has a birthday. In 1993, a short paper reported that 36 university students did slightly better on a paper-folding task after listening to Mozart. The lift lasted about a quarter of an hour. Within five years, an American state was posting a classical CD to every newborn baby.
What happened when somebody checked
In 2013 a team at Harvard ran the test properly. Preschoolers were randomly put into six weeks of music classes, six weeks of visual arts classes, or nothing at all. Then they were assessed on four things that music-trained children are supposed to be better at: spatial reasoning, shapes, number, and vocabulary.There was no consistent advantage for the music group. Where a difference showed up in one trial, it failed to show up in the second.
Then came the bigger check. Giovanni Sala and Fernand Gobet pooled 54 studies covering almost 7,000 children aged three to sixteen, none of whom had learnt music before. Across all the studies there was a small apparent benefit. But when they separated out the trials that compared music against another activity, rather than against doing nothing at all, the benefit collapsed to roughly zero.
Their wording is unusually blunt for academics. Music training, they concluded, does not improve general thinking skills or school results. Learn music, and you get better at music.
This is not really a fact about music. Psychologists call it the far transfer problem, and it turns up everywhere. Chess clubs, memory apps, brain-training games and sudoku all tell the same story: practising a thing makes you better at that thing, and almost never at anything else.
You will still meet the musical child who is good at everything, and the pattern is real enough in the data. But lessons cost money and time, and they need a parent willing to nag about practice. Those conditions rarely turn up on their own. Music is doing some of the sorting, not much of the causing.
The promise that cost music its place
Here is why the bad argument matters. For thirty years, music was defended in schools on the grounds of what it did for other subjects. That is a fragile place to stand. If the violin is there to lift maths results, then a headteacher under pressure on maths results can quite reasonably move the money to maths.The instinct was understandable. Once every subject has to prove itself against a measure, the arts reach for the nearest number available. Music grabbed maths and reading, because those were the numbers that counted.
Something like that has now happened. GCSE music entries in England have fallen by 35 per cent since 2010, and A-level music by 47 per cent. One recent report found that 42 per cent of state secondaries enter no pupils at all for GCSE music.
What is left has drifted towards families who can pay. About half of pupils at private schools get sustained instrumental tuition; in state schools it is closer to 15 per cent. The average yearly budget for a music department is around £1,800 in a maintained school and £10,000 in an independent one. Four in ten primary schools offer no chance to pick up an instrument in a group during the school day.
The government's own music plan, published in July, notes that music GCSE has the largest gap between disadvantaged pupils and their classmates of any subject. A subject sold as brain training for everybody has quietly become an extra for the well-off.
The case that doesn't need a study
So what is the honest argument? Start with the obvious one we somehow stopped making. Music lessons make children able to play music. That is a real skill, it lasts for decades, and it does not need a footnote.The social claims deserve more care. A well-known 2010 study found that four-year-olds who drummed and sang together were more likely to help each other minutes afterwards. A careful field experiment in 2021 tried much the same thing and found nothing at all. Anyone who tells you the evidence here is settled has not read much of it.
Harder to measure is what an ensemble asks of a child. In an orchestra you cannot be the loudest. You have to listen while you are playing. And when you lose your place you carry on anyway, because thirty other people are still going.
There is also the plain business of being bad at something in public and continuing. A ten-year-old scraping through a piece at the Christmas concert is doing something most adults now arrange their lives to avoid. None of that appears in a league table, which is exactly the problem.
What happens next
England is in the middle of a reset. The National Centre for Arts and Music Education opens this month, backed by up to £13m over three years. The EBacc measures that squeezed music off so many timetables were dropped last year. A new national curriculum reaches classrooms in September 2028, and £25m is going into putting more than 130,000 instruments into schools.Money and timetable minutes will decide most of this. But the argument decides what survives the next squeeze, and our argument has been wrong for a generation.
If the reason your child learns the cello is that it might nudge a maths grade, that reason has an expiry date, because sooner or later somebody runs the trial. The other reason is smaller and much harder to write into a report. They will be able to make music happen in a room, with other people, for the rest of their life. Nobody needs a study to sign that off.
Sources & Further Reading