Most of us who attended low-performing, poorly-disciplined state schools can relate to the frustration of trying to pay attention to a lesson in a classroom full of truculent children who are running amok. At the time, it seems your chances of getting on are scuppered when the only way to grasp a quiet moment to question the teacher and properly do the work is in your lunch break.
It surely seems obvious that we would all do far better without a disruptive minority ruining for the rest of us entire crucial chunks of learning time. And this is before we even make the not-too-bold assumption that many pupils are tempted to abandon their work altogether and join in the detour from the lesson.
This leads one to wonder whether bad schools, due to a number of factors like tepid teachers and shoddy discipline, are allowing anarchy to ensue and therefore making pupils perform relatively badly. Or is the converse true? Are pupils are making schools bad by having upbringings that are against the grain and subsequently cementing that unruly culture in the classroom.
Without detailed evidence, the direction of causation is unclear. Some are skeptical of studies that show disruptive children impact their classmates. It could simply be, they say, that less smart parents who send their children to bad schools are likely to have less smart offspring and so those found in problematic classrooms were going to perform worse anyway.
But a fascinating new paper dispels this argument, convincingly showing that disruptive peers independently hurt classmates’ performance. It is not just that brighter children happen, luckily, to turn up in classrooms with less disruption. And it is not just that teachers who are less skilled at teaching generally are also less good at maintaining discipline. Disruption itself holds other children back.
How does the research tell us this? Well, it exploits the population variation in the proportion of children from families linked to domestic violence. This is useful because children with such backgrounds have been shown in previous studies to disrupt behaviour and learning in the classroom.
Domestic violence can only affect school outcomes by making the children subjected to it more disruptive in school. This makes it a very thorough way of measuring disruption as if we looked directly at the act of disruption, we could not be certain that that disruption was not just being caused by lax teachers.
In order to be ensure that the only factor actually influencing the results is how many disruptive pupils are in the year, the study also controls for any fixed effects. In order to do this, Carrell and Hoekstra ask whether schools with disruptive peers already have children who disproportionately do worse in education and the labour market later on in life – and then they factor this into the equation.
From this, we achieve some impressively far-reaching results. While there is increasing literature showing that peer effects are very important in education and impact short-term outcomes like grades and behaviour in school, there was previously little evidence that they affect outcomes later on in life. This called into question educational policies that were based on short-term effects of peers where the impact could not be shown to last until later life.
Astonishingly, the paper found that 5-6% of the rich-poor earnings gap was attributable to being in the classroom with children linked to domestic violence. Having just one additional disruptive peer throughout primary school was found to actually reduce one’s earnings between the ages of 24 and 28 by as much as 3-4%.
With free schools and academies having a reputation for succeeding at being highly disciplinarian, this adept investigation into the situation in our classrooms should be a strong case for opening more free schools and academies that experiment with and institute greater discipline, where state-run schools are failing.
We should sit uneasy with the prospect that another generation of young people are being held back by carry-on in the classroom.
Reddit this article ↓