by Martin Butler
Imagine a world where the prison population was a rough mirror of wider society. In such a world there is a similar spread of rich and poor, highly educated and less educated, as well as a roughly equal proportion of men and women and those from deprived areas and well-off areas. The proportions of different ethnic groups reflect those in the surrounding society, as does the age profile, and having a mental health problem bears no relation to the likelihood of being in prison, neither does being in care in any systematic way increase the chances of ending up as a young offender. In addition, there seems to be no pattern from year to year. Some years there are low levels of crime and in other years the crime rate jumps for no discernible reason. The random nature of the prison population is recognised as providing good evidence for the belief that criminality is simply a result of individuals using their free will to make bad decisions, since we are all equally capable of this. After all, it could be argued, everyone is equal in possessing free will, and crime is a conscious and fully autonomous act in which social and psychological conditions play little part. Anyone, the argument goes, can be selfish or greedy and so succumb to criminality. In such a world, the general view is that prison exists to teach these individuals the error of their ways by providing them with extra motivation to retain their self-control next time temptation beckons.
It is instructive to ask how this imaginary world differs from, and is similar to, the actual world in which we live. What implications can we draw from the contrast? The obvious difference is that our prison population is nothing like that of the imaginary world, and I hardly need to go through the statistics that show how it’s very much not a cross-section of wider society with regards to gender, mental health, education level, family background, ethnicity, social deprivation, and so on.
One way we could go here would be to claim that the levels of crime committed are actually fairly evenly spread across the whole population (as in the imaginary world) but that the criminal justice system in our societies unjustly targets certain demographic groups. No doubt this happens. What is called ‘white collar crime’ is under-reported and often more difficult to prosecute; such crimes tend to be committed by those who fall outside the main prison demographic. But while this is a distorting influence, it’s difficult to believe that the distortion entirely explains the highly skewed nature of the prison population. For the sake of argument, let’s admit that this cannot be the whole explanation, and that those with particular backgrounds or from particular environments or with particular psychological profiles are more likely to commit crime. If this is the case, can we still hang on to what we might describe as ‘the conscious autonomous decision model’ outlined in our imaginary world, the idea that criminality is the result of a conscious and completely autonomous decision which is not the result of environmental or psychological factors, and for which the criminal is completely responsible.
In the real world, the picture is far more confused, and to some extent we want to have it both ways – for good reasons as we shall see. On the one hand most people would probably not go as far as claiming that social and psychological conditions play little part in crime – given the nature of the prison population such a view would obviously be implausible. On the other hand, there is an assumption that criminals have free will and are responsible for their actions.
There’s a clear tension here. The standard line to overcome this is to argue that having a particular background or being brought up in a particularly tough environment will mean the temptations and negative influences are so much stronger, therefore more individuals will succumb to crime. If you’re bought up in an environment where criminality is taken as fairly normal, you are likely to join in, and if those around you appear to gain status through their criminality, crime will be attractive. Many would argue however that none of this means that any given individual cannot still resist, or that crime is not a conscious autonomous decision to do wrong. After all, many from the most appalling backgrounds have ended up succeeding and have even become pillars of society. We can frame this in terms of equal opportunities: those who suffer from negative influences clearly do not have an equal opportunity to live a law-abiding life, compared with those who don’t. This doesn’t make criminality inevitable, nor does it mean that those who are not subjected to them will avoid criminality. We can see it in terms of risk factors – statistically, those who face more risk factors for crime are more likely to commit crimes than those who don’t.
This approach is still problematic. If human beings, possessing free-will, really are fully autonomous, surely they are able to float above these influences and risk factors? It must always be possible to say no to malign influences and overcome the risk factors. And if it is always possible to say no, why can’t everyone say no? How exactly does influence affect action? How do risk factors operate if everyone subject to them could overcome them? Here, we might consider the Aristotelian notion of akrasia, or weakness of will.[1] Some people are just weak-willed, and so readily succumb to malign influences. High social risk-factors for crime plus a weak will mean criminal behaviour.[2] So does this squeeze out free-will? The only way to preserve it here is to claim that weakness of will is in itself a choice. No one has to have a weak will. But this starts to look a little empty as an explanation, and seems to be no more than the simple observation that some people commit more crime than others under the same circumstances. Does this really takes us any further forward?
It’s useful here to compare crime with health. It is well established that there are a number of risk factors that will increase your likelihood of developing type 2 diabetes, including obesity, inactivity, a family history of diabetes and being elderly. These influence whether individuals develop diabetes or not, but it is of course perfectly possible to have all these risk factors and not develop diabetes. Can we see an analogy here with how risk factors increase the likelihood of crime but don’t make it inevitable?
With regard to both crime and diabetes, individual cases mean nothing. The idea of a risk factor only becomes apparent when we look at many individuals in a social context. If we take a hundred people all with the major risk factors for diabetes, we can say with a good deal of confidence that a certain percentage will go on to develop the illness but it would be impossible to identify which particular individuals. It’s also of course perfectly true to claim that of the hundred potential diabetes sufferers, it could be the case that none of them develop the disease, though this would be a case of such extraordinary good luck it would be like saying I could theoretically toss a coin a hundred times and always get heads. It’s an empty, academic ‘could’.
How does this differ from the case of crime and punishment? If we take a hundred children brought up in conditions associated with a high likelihood of criminality and imprisonment, we could predict with a good deal of confidence that a given percentage would end up in prison. Could we argue the possibility that not one would go on to commit a crime? Would this be just a matter of luck? If we want to retain the notion of free will and agency here, we would have to claim the possibility that every one of those one hundred individuals make the right decision and desist from crime, something that is completely within their control as conscious autonomous beings. But in practical terms is this really any different from the diabetes example? This ‘possible’ seems just as empty as the possibility of throwing heads a hundred times.
I think we have to accept that human beings acting in their normal social setting are subject to environmental and psychological influences, and it seems empty to insist that they could have acted differently if they had so chosen. In this context we are discussing statistical probability, and the language of free will and responsibility is out of place. Just as with health, the right thing to do is to work to reduce these risk factors which are as clearly identifiable with regard to crime as they are with type 2 diabetes. With both crime and health, prevention is better than cure. And in this context it seems ethical to actually drop the language of autonomy and free will.[3] This does not mean I’m advocating a wholly deterministic view of crime, the idea that criminals are purely the victim of circumstance. This would be simplistic. If you are assaulted and robbed, it would be absurd not to hold your assailant responsible for their actions and to expect consequences to follow for them from such criminal behaviour.[4] A criminologist or criminal psychologist wouldn’t, and certainly shouldn’t, respond any differently. In our individual interactions with other human beings, not only is it inevitable but also ethically right to treat them as responsible autonomous agents, and to do otherwise would be to regard them as less than human. In fact, with those who commit the most horrendous crimes that place them somewhere beyond simply being bad, we do actually struggle to see them as mentally healthy individuals who’ve just made bad decisions. We call them ‘monsters’ to dehumanise them. In any case, human beings can act badly and must be held responsible for their actions. Deterministic machines on the other hand, can merely cause harm. So how does this square with my earlier discussion of risk factors? It doesn’t. But this is precisely my point. I think we have to be openly inconsistent. In fact, there is a need to be inconsistent.
We struggle with the reality of this inconsistency because we don’t like to admit that in different contexts quite opposing perspectives are required. Those who want to focus on the role of the environment and psychology with regards to crime struggle with notions of blame and responsibility; those who want to emphasise moral responsibility balk at the idea that social and psychological factors play a role in crime. What I have called the conscious autonomous decision model of crime seems right when we are dealing with individual human interactions, but quite out of place with regard to public policy on crime and strategies to reduce crime.[5]
My attempt to describe an imaginary world in which crime really was disconnected from sociological or psychological causes is not, I think, entirely successful, if only because we would never be satisfied with the idea that crime is completely free-floating and no more than a matter of free will. The risk factors might be difficult to identify, but that’s not the same as denying that there are any. If such a world did exist it would be possible to consistently advocate the conscious autonomous decision model of crime in all circumstances. In our world however, inconsistency is both necessary and desirable.
***
[1] Other psychological or biological explanations could be used. For example, the view that high testosterone levels are associated with criminal behaviour. The issue remains the same with all these explanations even if they have more of a ‘scientific’ basis.
[2] This is what might be called a diathesis-stress model. A predisposition(psychological/biological) is triggered by external social factors(upbringing, social context, deprivation, etc.)
[3] Another perhaps more obvious context where blame and responsibility are used quite inappropriately is with regard to whole generations. For example, ‘Baby boomers’ blamed for such and such and ‘Millennials’ accused of so and so.
[4] This is a key claim of P.F. Strawson’s classic paper ‘Freedom and Resentment’
[5] The inconsistency I discuss is essentially Kant’s Third Antinomy.