Mental disorder is a cause of crime: what is meant by "cause"?

In psychiatric terminology, behaviour is considered an aspect of the mental. The easiest way to deal with the relationship between mental disorder and crime would therefore be just to consider criminal acts to be a form of mental disorder. This stance has never been met with much enthusiasm, however, neither in clinical psychiatry, nor in science or jurisprudence.

Thought of as two distinct phenomena, the connection has been postulated as leading from mental disorder to crime and to be, at least in some respect, causal. At the same time, it is evident that causation in this context cannot mean that mental disorder is a necessary or sufficient cause of crimes.

Modern medicine has increasingly come to work with probabilistic models. Probabilistic theory defines the relation between “risk” factors and effects as an increased probability of the effect in the presence of the risk factor (cf. Reichenbach, 1956; Cartwright, 1979). In our context, probabilism would mean that particular forms of mental disorders are likely to be associated with particular forms of criminal acts. The risk factor may then be assumed to be a (full or partial, see below) cause of the event (meaning that causation is “attributed” to the factor) if there is a temporal relation so that the risk factor can be shown to generally precede the effect, if covariation with other factors (referred to as “confounders”) can be accounted for by logistic or other multivariate statistical models, and if reasonable models are at hand for understanding how the causation operates. 

In other cases, risk factors can be judged to be coincidental to or reflections of common causes. By using probabilism in this way, scientific exploration has been made possible beyond experimental models testing causation. The terms “risk” and “risk factors” are assigned to the cardiologist Dawber (Kannel, Dawber, Kagan, Revotskie & Stokes, 1961) as a model to identify background factors, such as elevated blood pressure, cholesterol, and smoking, behind coronary heart disease. They have become central to medical research and have even come to represent a paradigmatic feature of today's society (Beck, 1992). The concept of risk is therefore a means of avoiding statements of causation, and “explanatory value” in this context will mean “proportion of the variation statistically related to the variation in the risk factor”, which does not necessarily “explain” it in the common, causal meaning of the word.

In the vast majority of those afflicted, mental disorder does not lead to crime. A possible definition of causation in this context would therefore be that a mental factor is a cause of a crime if the mental factor is an insufficient but necessary part of a set of conditions that together are unnecessary but sufficient for the crime (a so-called INUS condition, Mackie, 1965, 1974). Suppose, for example, that a lit match causes a forest fire. The lighting of the match is not by itself sufficient; many matches are lit without bringing about forest fires. But the lit match is in this case a part of a constellation of conditions that together are sufficient for the fire. The match was dropped on a pile of dry leaves, and a gust of wind contributed to the lighting of the fire. Each of the components, the match, the pile of leaves, and the wind, is an INUS condition, each was insufficient, each was necessary, and all together were sufficient for the forest fire, even if other sets of conditions also could have led up to the same effect.

Counter-factuality is thus a prerequisite for a factor to be an INUS condition under the given set of conditions (it should be possible to conclude that “if the mental factor had not occurred or been present, then the crime would not have occurred”, cf. Lewis, 1973; Mackie, 1965, 1974). From this follows manipulability, that it is possible to change the effect or the probability of the effect by changing the cause.

Mackie’s model provides a useful framework to deal with causation behind complex human behaviours such as violent crime. The way we attribute causation even in the sense of INUS conditions in complex chains of events has to be considered. Singling out one of the INUS conditions as the cause of a certain event is often a matter of choice and not based on rigorous scientific investigations. Since each factor, by definition, forms a necessary part of the overall condition, we do not really have any grounds for pinpointing one of them as contributing to the effect to a higher degree than the others. 

Human minds, however, strive to attribute causes in order to be able to predict what will happen in the future. Only in very rare instances are such attributions of causation based on experiments or strict, logical deductions. As the factors that may be shown to cause human actions in the INUS sense are invariably numerous and interact in complex constellations, the way we identify causes and assign importance to them is in itself the object of psychological research (Cheng, 1997).

As for crime and punishment, there is every reason to believe that mental disorders attract undue attention among possible explanatory factors. Generally, we have a strong tendency to assign causation of undesired events to factors that are strange or exotic in relation to ourselves, classically to other ethnic groups or to people with features that in one way or the other make them different from us. This powerful force directs our attention towards mental disorders among all the possible INUS conditions that may be discerned in the background to a crime. 

In forensic psychiatric research and expert opinion, the attribution of causation has no doubt been influenced by ideas developed within the professional psychiatric paradigm. And for the causation that is to be judged by the lawyer, counter-faction will be non-informative. How could any mental condition (i.e. inner experiences, cognitions, and/or behaviour patterns) be ruled out as a contributing factor in the very complex sets of factors influencing human action?

This blog post is to a large extent excerpted from the paper "Mental disorder is a cause of crime" I co-authored with Susanna Radovic, Christer Svennerlind, Pontus Höglund and Filip Radovic in 2009.