Back to table of contents

Chapter 4. Deterrence

4.1. The twin ideologies

As one of those taking an active part in slaying the ideas and practice of treatment within the framework of penal law, I look with considerable anguish and anxiety on how deterrence is thriving on the death of its competitor. For a considerable period I have been lecturing on the fallacies of treatment during morning sessions in Oslo, while Johs. Andenæs(1) (1974) has argued for general prevention in the afternoon in the same room - to a very attentive audience. Of course they are attentive. These students are to man the system of crime control. They are in need of rational substitutes for the ideology of treatment. Good, rational, scientific substitutes - such as they are used to. They get them. And they get them in increasing numbers. During the last years we have received important proposals for a change in the penal system both in Finland (Straffrättskommitteens betänkande 1976) (Anttila 1977) and in Sweden (Brottsförebyggande Rådet 1977). Both declare the treatment ideology to be dead. And both find a most welcome substitute in deterrence - or "general prevention" as we call it in Scandinavia - as the basic foundation for the penal system. Dichotomies rule the world. Treatment-ideas have faded out, so there is a need for general prevention. To me this seems to be the major weakness in the otherwise inspiring and inspired report on "New penal system" from Sweden (Brottsförebyggande Rådet 1977). As if treatment and deterrence were the only way of coping with conflicts.

It is too simple. But at the same time it is quite natural that ideas of general prevention are replacing ideas for treatment. The two sets of ideas are often presented as essentially different. But actually they are closely related on many points. They are both, in their recent stage, the outcome of an epoch of rational, useful thought. They have in common a manipulative element. Treatment is something intended to change the criminal; deterrence is an attempt to change other people's behaviour. It is in both cases pain with a purpose. In both cases some sort of behaviour-modification is supposed to take place.

Another common element is that they are both soundly embedded in science. But it is no longer any fun to measure the effects of treatment. They are all negative. Thus the researchers have moved to the new promised land. Is private killing deterred if state killing is introduced? Isaac Ehrlich (1975) claims 7 to 8 persons saved for each murderer executed, while other authors argue that Thorsten Sellin (1967) was right when he found that the death penalty was of no importance for the rate of murder in a state. The problems are more complex than with evaluation of the effect of treatment, but in principle the problems and potentialities for measurement are the same within general prevention and treatment. So also are the groups taking part in the new quarrels. Since there are questions of fact, of science and applied social engineering, we are again in the hands of measurement experts, as we shall later be in the hands of "social technicians" to get the results translated into action.

The similarities between treatment ideology and deterrence explain why they are so easily exchangeable. But there are also differences between the two. Particularly striking is the greater survival ability of the ideas within the field of deterrence or general prevention. The theory lays claim to empirical validity, just as treatment did. But it is a much more difficult topic to research. First, even its basic concepts are vaguely delimited. The unpreciseness within the area is amply illustrated by the simple fact that the key concepts "general prevention" and "deterrence" are used interchangeably throughout most of the literature, and therefore also here (cfr Andenæs 1974, Appendix 1). Furthermore, skimming the literature, one will see how everything from police activity to hanging can be classified as stimuli in a system of general prevention or deterrence. By and large, I think it is fair to say that everything that might be conceived of as elements in formal social control might also be characterized as elements in general prevention. And lastly: even in cases where the stimuli are pinned down to manageable proportions, the effects are often more troublesome to measure than when it comes to treatment. The reasons are again simple. Treatment effects at least had a formally clear target: those receiving treatment. With general prevention or deterrence these questions are more complicated. The general population, in whole or in part, is the target. That population might shift their activity from one type of crime to another, or move from one country to another, or might or might not have received the message of increased or decreased dosage of the stimuli.

Conceptually, as well as empirically, ideas of general prevention or deterrence are thus more cumbersome to handle than ideas of treatment. Looseness regarding definitions, stimuli, and target makes it next to impossible to disprove these ideas. The theory is strengthened by claims that it is founded on science, but survives empirical scrutiny. It is probably these aspects of the theory that make it possible for general prevention to fill the void after treatment has gone, and which makes the ideology suitable in an epoch where the infliction of pain would otherwise have been problematic.

4.2. Scientification of the obvious

It is obvious that punishment directs action. We know. We do not touch a red-hot oven. We do (often) change our behaviour if someone who means a lot to us blames us for misbehaving. More of us use seat-belts if it is expensive not to. What we know from our personal lives we tend to transfer to our public life. My experience in my family and circle of friends is given validity in the discussion on how to deter the thief, the drug addict, the violent offender. Why should not they become prevented by punishments, as I am by the hot oven?

There are actually rather good reasons why they are not. In the public arena there is no question of immediate control and punishment, but of formal sanctions following a long time after the possible crime. Here there is no question of punishments applied by one who has some relationship with the offender and therefore also greater possibilities than merely the creating of pain. Nor is it usually a question of punishments versus nothing, but of an increase versus a decrease of some sorts of punishments for specific types of crimes. To apply our daily experiences of pain avoidance to a general debate on deterrence, we would have to raise questions of the deterrence value of a red-hot oven of 200 C° versus one of 300 C°, or of a few minutes of scolding by a father against 15 minutes. And lastly and most important, a discussion of general prevention is not a discussion of the immediate effects of pain, but mostly of the effects on person A of the fact that person B is being punished. Some of us are not all that clever at learning from other people's experiences with their peculiar types of hot ovens.

Still, it is obvious that punishment does deter. Some sorts of punishment deter some actions in some situations. Without any punishments, chaos might emerge. When the police go on strike, trouble ensues. I agree completely. In its most elementary form, the basic premise of deterrence-theory is completely valid. If no action were taken against any of those who break the law, this would certainly affect the general level of crime in the country.

But that elementary form is not where ideas of general prevention or deterrence are applied in practice. In practice, these ideas are applied when politicians need arguments for increased penalties for certain crimes, or when judges want to be particularly harsh, e.g. increase a punishment from one to two years of imprisonment. Innumerable sentences in my country begin with the formula: "By reason of general prevention, it is here necessary to apply a severe sanction." It is a safe way out, one based both on intuition and on science.

And here we are at the core of the problem: theories on general prevention or deterrence are completely acceptable when it comes to the extreme examples, all or nothing. If no action were taken against those who break the law, this would certainly affect the general level of crime in the country. If tax-evasion were systematically met with capital punishment, it would likewise improve tax-behaviour. But these are not the cases where the ideas are used. Nearly all the concrete applications concern small additions or subtractions from usual standards of pain. Here theories and empirical studies within the field of general prevention give us no guidance. But by constantly using their extreme examples, the workers within the area give us an impression of having useful theories and results. They do in other words, give the process of pain delivery a false legitimacy. They could have said: We are of the opinion that criminals should suffer. This is a statement of values which is open to attack and we could have engaged in a moral debate on suffering. But they do not do that. They maintain, after complicated scientific discussions - and vivid demonstrations that ideas of treatment are without any scientific foundation -, that they base their own ideas on empirical science. And they produce all the classical cases of obvious effects of certain forms of pain to support their argument. By couching the obvious in scientific terms the impression is created that the choice of punishment is based on reason, and that the crime picture would have been different if the methods of punishment were different. Pain is given a scientific legitimization. We are not quite happy with what we are doing, but we persevere in the name of science!

4.3. Level of pain delivery

There are just now some 1 800 Norwegians in prison. This means approximately 44 per 100,000 of the population. But why just 44? Why not 115 per 100,000 as in Finland? Or why not take the large industrialized nations as a model? USA has half a million incarcerated just now, that means 230 per 100,000 of the population. USSR seems to have 1.7 million prisoners, that is 660 per 100,000 inhabitants, according to Neznansky (1979). But we could also make a complete turn-round and look at the case of a small, highly industrialized country with large minority problems, drug problems and crime problems, situated in the very heart of Europe: Holland with less than 20 prisoners per 100,000 of the population. That is in other words half the level of Norway, but exactly the same as Iceland, which for historical and geographical reasons ought to have resembled Norway. Internally in Eastern Europe we can find the same extreme variations with USSR and Poland at the top, while both Chechoslovakia and DDR seem to have very small numbers of prisoners.

Historically the picture shows even greater variations. At the time when Henrik Ibsen attempted to pass his high school exam - he failed - Norway had a prison population that was five times as high as in Denmark. Finland and Sweden had the same high figures. Our numbers of prison inmates subsequently fell dramatically until the turn of the century, since when - with the exception of Finland - they have kept at a fairly constant level, right up to the present day. They have kept constant, regardless of the fact that the figures in several crime indicators have doubled and re-doubled.

The point here is not to explain variations, or no-variations. The point is only to say that there is nothing new in an increase in the numbers of registered crimes without a corresponding increase in the number of prison inmates. Conversely, neither is it new for the level of registered cases to go down without a corresponding fall in the number of prisoners. There is no inevitable connection between the levels of crime and punishment. The two seem to exist in a complete and badly understood relationship to each other. There is little to indicate that the rate of crime in a country decides the rate of imprisonment. On the other hand, there is little to indicate that it is the rate of imprisonment, or the striking pow er of the police, which determines the level of crime. No doubt they do influence each other, but within a very wide area. For this and other reasons, to regard punishment simply as a means of control against undesirable activities is to adopt a far too restricted point of view.

4.4. Crime control as the goal for crime control?

Reading through the new wave of literature on deterrence, one is struck by the simplicity in the more basic reasons given for it all. It is the same simplicity that once pervaded the field of treatment. Then it was obvious that criminals had to be healed away from continuing their anti-social activities. Now it is equally obvious that the population, through examples of the sufferings of the sinners, have to be kept on the narrow track. It is as if the goal of the crime- control system is to control crime.

It cannot be. Or, let me restate it like this: if general prevention or deterrence had been the major goal of the operation, the crime- control system ought to have been constructed in quite a different way from what we find in our countries. If the goal of punishment was to create conformity, the crime-control system would have had to put hardly any effort into controlling those crimes which are perceived as serious crimes in our society. Most cases of murder might go unpunished; we know it is wrong, it would be more than sufficient to have some formal ceremony declaring who was to blame. Instead all the energy could be spent on the protection of weak norms. In Norway we have recently outlawed the selling of skate-boards. We have also passed a law making it compulsory to use seat-belts in the front seats of cars. Here are noble tasks for deterrence. Just a few five- year prison sentences, and we would have done it. Several hundred people would be spared severe bodily harm every year if the safety-belt law were obeyed; maybe thirty lives would be saved. That equals all the cases of murder in Norway in an average year.

It cannot be that simple.

A more sophisticated goal is suggested by Klaus Mäkelä (1975). He is one of the key persons behind the proposal for a new penal law for Finland, and his main argument is that the penal system, through general prevention, ought to create a priority list with regard to values in society. This is an interesting possibility. I shall soon come back to it. By and large, the idea has, however, been used simply to say that the more severe acts have to be met by the more severe punishments. Or, as stated in the major Swedish proposal quoted above: "The sanction ought to be dependent upon how dangerous or deplorable the crime is" (p. 200). So, most pain to those who have committed the most deplorable acts.

And who is to establish the priorities, that is, draw up the list? - Parliament is, in my part of the world. The pattern is simple: the penal law draws up the list of sins, Parliament categorizes and ranks them and decides in detail the amount of pain to be inflicted for each possible breaking of the law, and of course then allocates the greatest pain to the highest ranking sins.

With Klaus Mäkelä's formulation of a priority list with regard to values, the task has been broadened. The goal is not simply to control crime, but to give priorities to values reflected within penal law. This is also what the Swedish proposal for a "New Penal System" puts forward. It is not the need for control, but what the offence deserves which must determine the severity of the penalty - "The consequences must emphasise the danger or the atrocity of the crime" (p. 200).

But if this is the aim - and it can be a perfectly respectable aim - then new questions arise, in particular in relation to the understanding of what we actually are doing while talking like this. Is this really a discussion of general prevention? What are the scientific as well as social consequences of insisting that this is a discussion of general prevention? It is claimed that the murderer is executed not to prevent murder, but to denounce the evil character of murder. But why discuss this in the framework of an empirical analysis? Pain delivery is here used as a moral declaration. Why not say so?

By saying so, we would first of all weaken the position of general prevention. Furthermore, it would become even more obvious that pain delivery was intended to function as a sort of language. That would prepare the way for the idea that other, less painful languages, might be used. It would weaken the legitimization of inflicting pain when the pain was perceived as signals instead of concrete forms of controlling behaviour.

1. Johs. Andenæs is clearly the most important scholar in the development of ideas of general prevention. His first article on the topic is from 1950: AlmenprevensJonen - illusjon eller realitet? (General Prevention - Illusion or Reality?) Most of his writings on the topic are published in English in the book Punishment and Deterrence< (1974). For more critical Scandinavian views, see Aubert (1954) and Christie (1971).

Back to table of contents