more traditional explanation of the function of the criminal law is that
the threat of punishment for violating that law operates as a deterrent.
Unlike the idea of utilising the criminal law to reaffirm social values
(often called 'educative deterrence'), the threat of punishment operates
here at a conscious level, inducing people to refrain from crime.
Deterrent punishment can operate in one of
Against the individual
With individual deterrence the hope is that the offender being punished
will find the experience so unpleasant that he will not re-offend.
It is extremely difficult to test how true this is. Statistics as to
the number of offenders not re-convicted (and who might therefore have
been deterred as a result of their previous punishment) are of limited
utility unless one knows why they did not re-offend (and what percentage
did commit further crimes but were not apprehended).
With general deterrence the punishment of the offender is aimed at the
public at large in the hope that the example and threat of punishment will
deter them from crime.
This operates in two ways.
First, regular and normal levels of punishment keep the constant threat of
similar punishment alive.
Secondly, in the past courts have occasionally passed 'exemplary
sentences'. These are disproportionately severe penalties usually imposed
when a particular type of activity is on the increase.
For instance, concern over the rise of football hooliganism led to the
exemplary sentence of life imprisonment for riotous assembly outside a
football ground in Whitton (1985) - (a sentence reduced to
three years' imprisonment on appeal, Whitton (1986).
Are people deterred from committing crime by
The theory of general deterrence rests entirely upon one assumption - that
people are in fact deterred from committing crime by the threat of
It is crucial therefore that the public know about punishments being
imposed. Such publicity is most easily obtained when exemplary sentences
are imposed (but is probably counterbalanced by the extensive coverage
given to the occasional 'lenient' sentence - as when a rapist is put on
probation (now a Community Rehabilitation Order)).
News media coverage of 'normal' punishments is far more selective and
mainly confined to local papers - and even there it is the circumstances
of the offence and the identity of the offender that attract most
attention (Walker, 1985).
Evidence shows deterrence works.
There are several celebrated examples purporting to demonstrate the
effectiveness of deterrent sentencing.
For instance, in 1919 the police went on strike in Liverpool. With the
chances of apprehension significantly reduced, the crime rate (especially
looting of shops) escalated sharply.
Similar results occurred when the Nazis occupied Denmark in 1944 and
arrested the entire police force; the general crime rate rose sharply and
Economic model - deterrence
Much of the research on deterrence has been based on an economic model.
The potential criminal is seen as a rational calculator who balances the
costs and benefits of his possible actions.
The 'costs' include the probability of apprehension, the severity of
punishment, the ease with which the crime can be committed, chances of
success, etc., while 'benefits' refer to the satisfaction, whether
monetary or otherwise, to be derived from the crime.
Feeding in other variables such as income, earning potential, environment,
taste, employment status and education level, the thesis is that a person
will only commit a crime if, according to his evaluation, the benefits
will outweigh the costs (i.e. there is a net utility to be derived), and
if this perceived utility exceeds that he could derive from alternative,
Becker economic calculation
Becker (1968), in a pioneering article, hypothesised that an
offence, O. could be expressed as the function
O = O (a, f, a)
p is the probability of conviction,
f is the expected punishment and
u is a composite variable representing all the other influences.
In order for deterrence to be effective and crime to be reduced, the costs
must be made to outweigh the benefits. Increasing the probability of
conviction or the severity of the punishment would make the prospect of
crime less attractive and induce the individual not to commit crimes.
As p, f and u are all functions of 0, albeit without precise values, it
follows that crimes with low detection rates would need correspondingly
The importance of the other variables must not be forgotten. For instance,
if education could be increased and unemployment reduced, the utility to
be gained from lawful employment might outweigh that to be derived from
Ehrlich, increase in likelihood of detection
and punishment reduces crime.
Several empirical studies based on this economic model of crime have been
carried out and indicate some support for the thesis.
For instance, Ehrlich (1973) found that a 1 per cent increase in p
(probability of conviction) was associated with a crime reduction of 0.99
per cent, and a 1 per cent increase in (expected punishment) was
associated with a crime reduction of 1.12 per cent.
Do criminals 'think' before acting?
Such an economic analysis of criminal motivation and the entire theory of
general deterrence is based upon an assumption that people are always
rational and think before they act.
They are 'rational utility maximises', which means that prospective gains
and losses are weighed against each other with decisions and choices being
based on such a calculus. While this might be true of some individuals and
some crimes (for instance, one might rationally calculate the gains and
losses in parking on a double-yellow line), it is clearly not true of all
individuals and crimes.
Many crimes (particularly those involving violence) are often committed by
persons in highly emotional states who are acting in an exceedingly
irrational manner. The man who, in a fury, beats his wife to death is
hardly a 'rational utility maximiser'; his motivation is anger, jealousy,
love, hate or whatever and he is simply not amenable to deterrence at that
Similarly, many other persons are not deterrable: the mentally disordered,
the intoxicated, the addict, etc. The criminal law might have some effect
with some of such persons in inducing some care in the commission of the
crime - for instance, a drug addict might not purchase drugs right in
front of a policeman - but it will never deter him from committing the
Prospect of detection affects actions.
Deterrence can only be effective if people think that there is a
reasonable prospect that they will be caught. A couple wishing to inflict
minor sadomasochistic injuries on each other in their own home are hardly
going to be deterred by the threat of imprisonment. They know that they
will not be apprehended and so will not be deterred. (See
The Spanner Trust for
information about promoting sado-masochistic rights.)
large numbers of crimes are never reported to the police or recorded by
them and that overall only about 2 percent of offences committed result in
a criminal prosecution.
The old adage appears true: crime pays; the criminal's chances of not
being caught are far higher than those of being apprehended.
Further, if one examines the clear-up rate per offence (again, of those
recorded by the police) one discovers that generally the offences for
which one is most likely to be apprehended (homicide - 95 per cent;
violence against the person - 76 per cent; sexual offences - 75 per cent)
are those that are often committed impulsively or by persons not amenable
to deterrence while the clear-up rate for crimes more likely to be
committed 'rationally' is even lower than the general average (burglary -
20 per cent).
The value of clear-up statistics.
Of course, in assessing the efficacy of deterrent punishment, the actual
statistics on clear-up rates are not helpful unless the prospective
criminal is aware of them.
What matters is the individual's assessment of his prospects of
apprehension. If he thinks there is a 95 per cent chance of being caught
if he burgles a house, then he will probably be deterred and it is
irrelevant that his assessment of the risk is hopelessly awry.
There are no reliable statistics on this. Willcock and Stokes (1968)
discovered that the young persons they surveyed tended to overrate their
chances of detection.
However, while many persons might be overcautious in their estimate of the
risks, it would not be unreasonable to assume that many others will think
they are 'too smart' to get caught - and once they start committing crimes
they will become part of the statistical pattern; their assessment of the
risks will become more accurate and, knowing the truth, they will be less
likely to be deterred.
Walker "Some Theory"
Apart from the anecdotal, it is difficult to obtain clear evidence as to
the deterrent effect of sentencing (Beyleveld 1980).
Judges, perhaps in desperation, continue to have faith in the deterrent
impact of their sentences.
Parliament has however become more sceptical.
The Criminal Justice Act 1991 and, particularly, the Government
White Paper (1990) preceding it has largely eschewed deterrence as a major
goal in sentencing.
Perhaps the most realistic view is that expressed by one of the leading
penologists in Britain, Walker (1985), who concludes that
'some people can be deterred in some
situations from some types of conduct by some degree of likelihood that
they will be penalised in some ways.'