I enjoyed chatting to Professor Pam Snow last night for the latest Filling The Pail podcast. At one point, Pam spoke about the youth justice system in Victoria – an issue I am not overly familiar with – and she prompted some tangential thinking on my part about aims and measures.
Imagine, for instance, that a politician promised to reduce crime. Once elected, the politician changed the law or guidance to judges in such a way as to significantly decrease the number of convicted criminals given custodial sentences. The politician then pointed to the reduced number of people sent to prison as evidence that they had fulfilled their campaign promise and reduced crime.
Would you be convinced? Probably not.
It may strike you as transparently circular, but even it it did not, you may have other reasons for scepticism. For a start, you would have developed your own impression of the level of crime from your membership of the community. This impression may be right, wrong or exaggerated and it may be influenced by media reports. However, you would still have a perspective from within the issue and would be likely to have an opinion on it. Perhaps more critically, you would, if so inclined, be able to draw on independent statistics to confirm or refute the politician’s claims. This could be raw reported crime, or it could be something more stable and less affected by changes in reporting procedures. In the UK, for instance, you could look to the Crime Survey for England and Wales to see how the prevalence of various crimes had changed since the politician took office.
Now consider the position of school exclusions. I think everyone would like to see fewer school exclusions for similar reasons to why we would like to see fewer people sent to prison. However, there are differences about how this could be achieved. Some people argue that we need top-down pressure to reduce the number of school exclusions whereas others argue we need to work harder to improve students’ behaviour at an earlier stage in order for fewer school exclusions to be necessary. Perhaps there are those who would argue you need both. There is a lot of room for debate and few people would sit on either extreme.
Whatever your view, an important question arises as to exactly how you improve school behaviour and whether it is even possible. I would argue highly structured models that actively teach pro-social behaviour, similar to the Michaela Community School approach, coupled with interventions to address deficits in e.g. literacy, are likely to be more effective than more laissez faire approaches, but this is open to debate.
Regardless, we repeatedly see initiatives that are schematically similar to the politician thought experiment. School systems go through repeated cycles where top-down pressure is imposed to reduce exclusions and then the resulting reduction in exclusions is heralded as evidence of improved behaviour. It is a bureaucratic self-delusion that maintains until some scandal occurs that draws people’s attention to the fact that school behaviour is as bad, if not worse, than ever.
So how do the bureaucrats get away with it? I think there are two reasons. Firstly, people are not all situated in schools in the same way that we are in the wider community. If you are not a teacher and don’t have children, or you have children at a school where behaviour is not a significant issue, then it is relatively easy to be in the dark. Teachers do sometimes make the public aware of issues around behaviour, but there is pressure not to do so. In many schools, the culture is such that poor behaviour reflects badly on the teacher – the teacher is not engaging enough or they are not able to form good enough relationships. And school management are likely to take a dim view of a teacher going on-the-record to complain about behaviour in the school where they work.
What we really need is an independent check, similar to the Crime Survey of England and Wales. If teachers, students and parents were systematically surveyed on a yearly basis about a range of school behaviours then we would have an independent measure of the effects of any policy changes. This could be reported at a system level or, if we want to be pretty radical, it could be reported at a school level, alongside statistics on exclusions and academic achievement. That way, we may be able to identify schools that perform better than the rest and we could test my hypothesis about schools like Michaela.