Analysis of crime statistics
6.22 Part of the basis for the conclusion reached by the United States Commission, which we have already mentioned, was a preliminary assessment of the trend in reported crime in Denmark following the liberalisation of the Danish laws on pornography in the late nineteen-sixties. The controversy surrounding the Commission prompted some people to question the thesis that the legalisation of pornography reduced the incidence of sex crimes in Denmark; and in time a contrary thesis emerged, that there is in fact statistical evidence indicating that the free availability of pornography can be linked to an increase in sexual offences. This is an argument which rests heavily on work undertaken by Dr John Court, Reader in Psychology at the Flinders University of South Australia. and Dr Court was good enough to submit to us several papers on the subject. The Nationwide Festival of Light and the National Viewers’ and Listeners’ Association urged us to pay special attention to Dr Court’s work; and others with a more specialised interest in the field of human behaviour, such as Professor Sir Martin Roth and Professor H J Eysenck, made reference to Dr Court’s work in the course of giving evidence to us.
6.23 It was clearly right that we should examine this evidence in some detail and we also think it right that we should discuss it in some detail here (1). It has not so far been subjected to close scrutiny and we think it important that the public should have the opportunity to assess what it shows. The issue of whether people are more likely to be sexually assaulted as a result of the circulation of certain kinds of publication or the showing of certain kinds of film is obviously a very important one. But we would not want to give the impression, by dealing with the issue at considerable length, that we think that it is the only or the decisive issue. We shall come in due course to other kinds of consideration. including other concepts of harm which have been put to us, and these are also important. A special feature of the argument about porngraphy and sex crimes is that the effects that have been suggested here are claimed to be statistically measurable.
6.24 But are they measurable? Here there are several sets of problems. The first set concerns the number of offences. Certainly we have detailed information about the number of sexual offences reported to the police over a long period. However. this is not the same as knowing how many offences occurred, since there is no means of telling how many victims decided not to go to the police to report what happened. It is often suggested that rape and sexual offences are peculiarly prone to non-reporting because victims often prefer to avoid the prolongation and reinforcement of the distress they have suffered which may result from questioning by the police and the giving of evidence and cross-examination at a trial. We cannot be sure, therefore, how many offences have been committed or what proportion never come to light. The particular difficulty in examining trends in offences is that there is no way of knowing whether the unreported proportion of offences is remaining constant – in which case the trend can accurately be assessed even though the incidence of the offences is not fully known – or whether fluctuations are occurring in the proportion that are not reported; victims may at different times be more or less ready to report offences, because of changing social attitudes. or because of changes in the law or legal procedure or police practice.
6.25 It follows that even apparently detailed information about the number of offences reported to the police gives an imperfect picture of the number of such offences; it may not even reveal whether, or by how much, the number is increasing or decreasing from year to year. The problem gets worse if one tries to make international comparisons. Legal systems vary considerably between countries and differences in what constitutes a criminal offence – and what name is given to it – often mean that comparisons which seem straight- forward are not in fact comparing like with like. Moreover, the form and reliability of criminal statistics varies widely from country to country. Finally, it is too easy to look at statistics in isolation from the procedures and social climate in the country concerned which may affect, for example, the readiness of victims to report offences.
6.26 The second set of difficulties concerns the other side of the correlation; the availability of pornography. The imperfections of the information about the incidence of sexual offences are insignificant compared with the lack of information about the availability of pornography. There are two kinds of problems here. One relates simply to the quantity of pornography available, about which there is very little reliable information. The second raises the question of what kind of material we should consider for these purposes as pornography, and whether the differences between different kinds or degrees of pornography are likely to produce different effects, and so should be taken into account in trying to make correlations with the incidence of sexual offences.
6.27 Much of the evidence we received assumed that pornography had worse effects, the more extreme it was. Thus, mild pornography consisting of little more than “pin-up” pictures is often regarded as harmless, while the strongest and most explicit material is commonly regarded as the most corrupting. Yet in considering a hypothesis that sexual offences are linked to the widespread availability of pornography, it is not obvious, on reflection, that hard-core pornography should lead to the commission of offences in a way that soft pornography does not. Indeed, many people might think that if a potential sex offender were going to be triggered off by something he saw, that effect would be produced at least as well if not better by the titillating and erotically arousing than by the clinical close-up. So how should one set about quantifying pornography? How should an index of availability reflect changes in the type of material over a period of time? How should one compare one country where soft pornography is widely available, but hard-core material is banned, with another where hard-core pornography is available for anyone who wants it but where pornography generally appears to be less pervasive?
6.28 These difficulties seem to us very real, and it is a severe shortcoming of much of the work which has been put to us that the sometimes detailed study of criminal statistics is accompanied by only the most generalised and unsupported statements about the availability of pornography. If no measures exist for what is supposed to be the causal factor, it is impossible to make any progress in attempting to relate cause to effect.
6.29 The relation of cause to effect brings in the third, central, and best- known, problem. Even if it is possible to provide an accurate measure of two variables, the existence of a correlation between them is certainly no proof that one is influenced by the other. The causes of crime are undoubtedly complex and elude isolation. Many factors are at work and much stronger indications than a mere correlation between variables are necessary before the existence of a causal link can even be suggested, let alone proved. The difficulties of proving or even plausibly arguing for, causal relationships between general or disuse social phenomena on a purely statistical basis are sufficiently well-recognised for us not to need to dwell on them further.
England and Wales
6.30 Having drawn attention to the general considerations which limit both the available information and the inferences which can be derived from it, we should nevertheless look at some of the information which exists and examine what our witnesses said about it. It is natural that we should start at home.
6.31 First, in relation to the availability of pornography in England and Wales, it needs to be said that no information exists to provide any kind of index. In the papers submitted to us, Dr Court did not attempt to provide such information. He does however treat Britain as an example of a “liberal” country in which the detrimental effects of pornography are to be seen, and he identifies two times at which, he suggests, pornography became increasingly available – first with a change in the law introduced by the Obscene Publications Act of 1964, and subsequently following the impetus of the American Commission Report on Pornography in 1970. Dr Court offers nothing to substantiate his statement and we find his explanation of the significance of these two dates less than convincing. As we have explained earlier, the Obscene Publications Act 1964 was a minor measure designed to strengthen the existing law by plugging two loopholes which had been found in the Act of five years previously. Dr Court cites as the only authority for his suggestion that pornography became more available after the enactment of the 1964 Act as an article by Mr Ronald Butt in The Times on 5 February 1976 which attacked what Mr Butt then found to be the ineffectiveness of the law in controlling pornography. Mr Butt argued that the intention of the 1959 and 1964 Acts had been systematically destroyed over the years by the exploitation of their letter, but nothing in his article supports the idea that the 1964 Act opened the way to the greater availability of pornography. Nor do we know of any authority for the suggestion that pornography became more freely available here after 1970 as a result of the influence of the Report of the US Commission. It seems to us that the choice of the years 1964 and 1970 as crucial in the increasing availability of pornography is purely arbitrary.
6.32 Our own enquiries have not placed us in a very much better position to assess the availability of pornography over a period of time. It is the experience of all of us that for very many years there has been a steady trend towards greater frankness in sexual matters and greater tolerance of the portrayal of nudity and sexual activity. This has been reflected in a wide variety of publications and films, including those which many people would regard as pornographic, but it does not help us to construct any kind of index with which to compare changes in the numbers of sexual offences. The Head of the Obscene Publications Squad of the Metropolitan Police told us that police activity in the last six years or so had resulted in a considerable reduction in the amount of hard-core pornography from overseas in circulation, at least in London, though the police were unable to quantify the amounts concerned. The same period, however, has seen a considerable increase in the publication of British magazines with an explicit sexual content which, particularly in the years 1975 to 1977, moved very quickly to adopt a character much less restrained than had previously been acceptable. These magazines were distributed throughout the country and clearly reached a much wider audience than the undercover trade in overseas pornography ever had; it was suggested to us by some witnesses that if sexual offences were indeed linked to pornography, a noticeable effect should have occurred over the last five years. However, as we have described in Chapter 4, there are indications that the circulation of these magazines reached a peak in 1976 and has subsequently suffered a decline, aided by increasing police action against publishers from early in 1977 and, in London, against retailers from early in 1978. The managing director of the largest publisher in this field stated in the trade press in June 1978, and confirmed in correspondence with us, that the effect of this action had been to reduce the sales of his magazines by 40 per cent. in the first quarter of 1978.
6.33 With that patently inadequate basis for assessing the availability of pornography, let us nevertheless turn to the information available about the incidence of sexual offences, Figure 1 shows the numbers of indictable sexual offences reported to the police in every year since 1946. This gives a longer perspective than Dr Court has offered and enables the point to be made that sexual offences (or reports of such offences) were significantly increasing long before pornography is alleged to have become more widely available and well before the “ineffective” Acts of 1959 and 1964 had a chance to exercise their influence. Indeed, it would appear that the number of reported offences levelled off between 1959 and 1966; though if the figures for 1964 to 1970 are taken in isolation, an increase becomes noticeable, even if it is less marked than that in the period 1946 to 1959. But what is perhaps even more significant is the trend since 1973; until the increase last year there had been a steady fall in the number of sexual offences known to the police. In the light of the suggestion we reported in the preceding paragraph, this should be interesting to those who are impressed by such correlations; a period of apparently rapid increase in the availability of pornography seems to have been accompanied by a reduction in the number of sexual offences reported, which was reversed in the year in which increased police activity reportedly reduced the availability of sex magazines.
6.34 We have also used Figure 1 to make another point of some substance, which is that it is futile to seek to examine one class of offence in isolation from the trend in crime generally. In our society, as in others throughout the world, rising crime has been a matter of much concern for many years and it is pointless to seek a special explanation for a rising trend in sexual offences if that trend merely reflects an increase in the general level of offences. But is that the case? In fact it is not, because sexual offences seem to have been significantly out of step with other offences, in that they have not shown anywhere near the same increase in numbers over the years. This is shown dramatically in Figure 1 by comparison with offences against the person, which have continued to rise steeply when sexual offences have stabilised and actually fallen. The papers submitted to us by Dr Court do not face up to this point and we think it a weakness of his work that he has concentrated on sexual offences to the extent of divorcing them from their context of crime generally. However, we note that in an article published in The Times on 7 August 1976, Dr Court put forward the proposition that the increase in serious sexual offences is greater than the general rise in crime in countries such as the USA and Australia, but that in Britain this has been “masked in recent years by the upsurge of violent offences associated with the Irish problem”. Given the fact that the trend in violent crime, as Figure 1 shows, goes back at least a quarter of a century, and given the scale of that increase, we doubt whether Dr Court would wish to stand by that statement.
6.35 It can fairly be said that offences against the person have not been representative of crime generally because they have increased proportionately more than other offences. However, indictable offences as a whole have, in comparison with just after the war, increased in number over five times, whereas sexual offences have rather more than doubled in number. In 1946 sexual offences accounted for 2 per cent of all indictable offences recorded by the police. This proportion rose to almost 4 per cent in 1955 but has fallen steadily since, until in 1977 and 1978 sexual offences accounted for less than 1 per cent of all indictable offences recorded by the police. Whether compared with thirty years ago or twenty years ago or ten years ago, sexual offences are now relatively less common among crimes known to the police.
6.36 It can be argued, of course, that to take sexual offences as a whole produces a misleading result, since the figures reflect changes in the law or in sexual morality with regard to offences such as those involving homosexual activity or sexual intercourse with under-age girls. Dr Court pays special attention to rape as an offence which may be potentially pornography- induced, and it is reasonable to regard indecent assault, among the other sexual offences, as appropriate to the same category. Figure 2 shows the numbers of these offences reported in each year since 1946. Offences of rape have increased faster than sexual offences as a whole, there having been five times as many reported cases in 1978 (1,243) as in 1946 (251), but this is the same rate of increase as for all indictable offences and a much smaller increase than for offences against the person, which increased over twenty- fold. Offences of indecent assault on a female, although more common, increased less dramatically than rape, very much in line with the increase already referred to for sexual offences as a whole. A marked fall since 1973 has been reversed in the last two years. As with Figure 1, Figure 2 also shows how the trend for another offence of the same order of frequency, robbery, is in contrast to that for the sexual offences examined.
6.37 Dr Court did not devote much attention to the situation in England and Wales. Yet he confidently observed in a published paper that “highly significant increases are apparent for rape and attempted rape for the period in question”. The period in question, in Dr Court’s thesis, is presumably that between 1964 and 1974, then the latest year for which he quoted figures, when the number of reported offences of rape and attempted rape rose by 103 per cent. This period of ten years was clearly a bad one for offences of rape, but the assertion ignores the fact that offences of violence against the person rose over the same period by 172 per cent, and it is not helped by the stabilisation of the number of rapes, despite the unrelenting rise in other violent crime, in the three years after 1974. The “significance”, moreover, in Dr Court’s view, is presumably linked to his own totally unsubstantiated surmise that pornography began to be widely available after 1964.
6.38 In his papers, Dr Court has paid rather more attention to the figures for the Metropolitan Police District, on the ground that a greater concentration of pornographic publications is associated with some parts of London than with the provinces. It is certainly not true that pornography has been confined to London but neither we nor Dr Court can offer any evidence of the relative circulation of pornography in London and in provincial centres. With that deficiency acknowledged, let us examine the statistics of offences known to the police in London. Dr Court has concentrated on the figures for rape, and Figure 3 accordingly does the same but also shows (on a different scale) the numbers of other sexual assaults in the years since 1947. The figures for offences of rape recorded by the police since 1947 indicate a slightly different picture from that for the country as a whole. Rapes have increased twice as fast in London as in the rest of the country and have continued to increase when, in the years 1975 to 1977, they came nearer to remaining constant elsewhere. This is not because crime in London has increased faster than elsewhere – the contrary is the case – and the effect is that rape has increased much faster than indictable offences as a whole, though still much more slowly than offences of violence. The same is not true, however, in respect of other sexual assaults, which have increased in London only by about the same extent as in the country as a whole, and they showed a significant decline between 1970 and 1977. Moreover, what Figure 3 also shows in a longer perspective than Dr Court gives is that the rising trend in rape and sexual assaults started well before what Dr Court alleges was the first date significant for the availability of pornography.
6.39 While we were hearing oral evidence it was reported that sexual offences known to the police in London had risen substantially in the first six months of 1978 compared with the same period of 1977 (rapes from 75 to 122 and sexual assaults from 550 to 741). These reports prompted some of our witnesses to draw the conclusion that the rise was related to the onset of a more active enforcement policy in the Metropolitan area, in which, from the beginning of February 1978, scores of raids on retailers by divisional police officers had significantly reduced the amount of pornography available in London (we have already reported one publisher’s statement of the effect on his sales). A reduction in sexual assaults during a period when pornography is widely available, followed by an increase when curbs are placed on the trade, hardly bears out the hypothesis that pornography stimulates sex crimes; but we do not think that the figures for one period of six months or for a year (now that the statistics for 1978 as a whole confirm the increase apparent in the half-year figures) can be used as the basis for any conclusions. Arguments of a similar kind were put to us in relation to the rate of sex crime in Manchester, where there has also been a policy of rigorous enforcement of the law on obscene publications, but we came to the conclusion from a more detailed examination of the statistics that the arguments were insufficiently supported.
6.40 We must at this point refer again to the problem of knowing how large the “dark area” of sexual crime is. The figures for rapes known to the police certainly underestimate the incidence of such offences. However, there may be grounds for thinking that there has recently been an increase in the level of reporting of cases of rape, and this could be reflected in the latest figures. There is a greater consciousness of the problem of rape, partly associated with the development of organisations such as Rape Crisis Centres, and greater protection is now afforded to complainants in rape cases under the Sexual Offences (Amendment) Act 1976 (which provides for their anonymity and imposing certain restrictions on their cross-examination); because of these factors, rape victims may be more willing to go to the police, and this may have increased the official statistics.
6.41 We now attempt to summarise what has emerged from this study of the arguments linking pornography and sexual offences in England and Wales. First, there is no firm information about the availability of pornography over the years, and there is no foundation to suggestions that have been made about particular times at which pornography became increasingly available. Second, the rising trend in sexual offences generally, including rape and sexual assault started long before it is alleged pornography began to be widely available. Third, the increase in sexual offences generally, and in rape and sexual assaults, has been significantly slower (though in the case of rape alone the difference is less significant) in the last twenty years than that in crime generally. Fourth, the contrast between the upward trend in crime generally and the greater stability in the numbers of rapes and sexual assaults has been most marked in the years from 1973 to 1977 (except in the case of rapes alone reported to the police in London where the increase is consistent with that in other forms of crime), when this period appears to have been the one when pornography was most available.
6.42 So, what conclusions can be drawn? We have already referred briefly to the extreme difficulty of identifying causal factors in the incidence of crime but, in the light of the findings we have set out above, we see no need to develop that argument. The case put to us by Dr Court, and by other witnesses who have looked to Dr Court’s work for support, cannot on the basis of the situation in England and Wales, we believe, even survive as a plausible hypothesis. Each of the findings in the preceding paragraph stands opposed to it. Consider Figure 4, which shows the relative importance of the numbers of rapes and sexual assaults in the totality of crime. We think it unlikely that any fairminded person would, on the basis of the information depicted in that graph, feel the need to construct a special hypothesis to explain the increase in sex crimes, still less that the hypothesis constructed would be that an alleged increase in the availability of pornography first after 1964 and secondly since 1970 (increases for which there is no independent evidence) provided a crucial influence in the growth in such offences.
6.43 We are not denying the possibility that pornography could be linked to the commission of sexual offences. Nor are we suggesting, on the other hand, as the data in Figure 4 encourage some to suggest, that the availability of pornography is associated actually with a decline (or at any rate a reduction in the rate of rise) in sexual offences. It might be, for example, that if only information were available about the amount of pornography in circulation some kind of correlation might be possible; it might be that whatever the factors are which have given rise to the substantial increases in most other forms of crime, they do not operate in respect of sexual offences and therefore some other separate explanation of the more modest growth in sexual crime has to be sought; it may be that there has been a substantial reduction in the willingness of victims of rape and sexual assault to complain to the police, so that the true number of offences, and the trend, is quite different from what appears in the criminal statistics. These are, we say, possible; but we do not think they are probable. It is not possible, in our view, to reach well-based conclusions about what in this country has been the influence of pornography on sexual crime. But we unhesitatingly reject the suggestion that the available statistical information for England and Wales lends any support at all to the argument that pornography acts as a stimulus to the commission of sexual violence.
1 We received assistance in analysing the statistics and the arguments based on them from the Home Office Research Unit, in particular in a paper prepared by Mr Stephen Brody.
Â© Crown Copyright 1979. This extract republished in the public interest.