Crime Quarterly No. 4 SA CRIME QUARTERLY No 4 JUNE 2003 17 I n the last issue of the SA Crime Quarterly, Joan van Niekerk highlights the increased incidence of reported child sex abuse and the declining age of the average sexually abused child. The declining age of offenders is also of concern: she notes that Childline statistics show that “43% of all cases of sexual assault reported to Childline nationally are committed by children under 18”. Given the concerns raised by statistics such as these, the Children’s Rights Project of the Community Law Centre at the University of the Western Cape commissioned a research project to further understand the trends with regards to arrest, custody and reporting in respect of child sex offenders. This article is based on that research, which was also designed for the debate on the Child Justice Bill1 in Parliament in early 2003. Important to note is that rape, indecent assault involving grievous bodily harm, and indecent assault of a person younger than 16 years – all crime types considered in this research – fall under Schedule 3 of the Child Justice Bill. Schedule 3 offences are more serious offences that have stricter provisions applicable than for other offences, including those relating to imprisonment of children before trial.2 This article first considers custody data for children in prison for sexual offences, followed by data on the arrest of children for sexual offences in the Western Cape. Both these sets of data provide an indication of the extent to which, and the manner in which, the criminal justice system deals with child sex offenders at present. The third section looks at data with regard to arrests for sexual offences committed against children in one metropolitan police area of the Western Cape, as this subsection of sexual offences is a crime type that is causing concern and which has been linked to children. Lastly, we consider data in respect of abusers identified by persons across South Africa contacting Childline for help, as this provides some indication of trends in respect of sexual crimes against children that may not reach the criminal justice system. Children in custody for sexual offences Custody trends are important when considering the future management of such offences, particularly Jean Redpath, criminal justice research consultant redpath@iafrica.com SOUTH AFRICA’S HEART OF DARKNESS Sex crimes and child offenders: some trends It is a source of great concern that the number of sexual offences committed by children is apparently on the increase. Given that the Child Justice Bill provides for a separate criminal justice process for children, and stricter provisions for child sex offenders, this article explores some of the relevant trends in this regard. The management of child sex offenders appears inconsistent, and very often no appropriate intervention is made at all. Calculating the actual incidence of child sex offenders is difficult, but some data suggests that children might be responsible for a significant proportion of sexual crimes committed against other children. SA CRIME QUARTERLY No 4 JUNE 2003 18 REDPATH since the Child Justice Bill provides for imprisonment under certain conditions of children suspected of sexual offences. Comparison with adult figures highlights some problems with the management of these offences. What are the overall custody trends with respect to all sexual offences and all sex offenders?3 As of 31 March 2001, persons in prison for sexual offences comprised 13% (22,524) of all prisoners, 12% (13,724) of sentenced prisoners and 16% (8,797) of unsentenced prisoners.4 Children comprised only 2% (525) of persons in prison for sexual offences, 4% (314) of the unsentenced prisoners, and 2% (211) of sentenced prisoners. Plotted over time, the number of all sex offenders in prison shows an even increase, with the number of sentenced sexual offenders increasing steadily at a rate of about 72 prisoners a month. At all times there are more sentenced than unsentenced prisoners (Figure 1). This is a very different picture from the one that appears from the custody data with respect to children only (Figure 2). This shows significant fluctuations on a month-by-month basis, unlike the even trends for all sex offenders. Another notable difference is that since January 1999 there have been more unsentenced than sentenced child sex offenders. Furthermore, the number of sentenced children in prison for sexual offences has been decreasing over the period under consideration at a rate of about one child every three months. What do these differences suggest? The irregular data for children suggests that there is erratic management of children suspected of sexual Figure 1: Prisoners (all ages) in custody for sexual offences in South Africa, 1998-2001 Source: Department of Correctional Services Figure 2: Children in custody for sexual offences in South Africa, 1998-2001 Source: Department of Correctional Services 0 5,000 10,000 15,000 20,000 25,000 Ja n ‘ 9 8 M ar ch M ay Ju ly Se p t N o v Ja n ‘ 9 9 M ar ch M ay Ju ly Se p t N o v Ja n ‘ 0 0 M ar ch M ay Ju ly Se p t N o v Ja n ‘ 0 1 M ar ch M ay Ju ly Se p t N o v Ja n ‘ 9 8 M ar ch M ay Ju ly Se p t N o v Ja n ‘ 9 9 M ar ch M ay Ju ly Se p t N o v Ja n ‘ 0 0 M ar ch M ay Ju ly Se p t N o v Ja n ‘ 0 1 M ar ch M ay Ju ly Se p t N o v N u m b er p er m o n th Unsentenced Sentenced Total 0 100 200 300 400 500 600 700 N u m b er p er m o n th Unsentenced Sentenced Total SA CRIME QUARTERLY No 4 JUNE 2003 19 offences. Particularly worrying is that more unsentenced, rather than sentenced, children are in custody for sexual offences (the reverse of the norm for adults). This is problematic as it suggests that many children are in custody for sexual offences due to the vagaries of the criminal justice system, rather than because a court of law has determined that prison is the most appropriate place for the child in each case. The fact that the number of children sentenced for sexual offences in prison is decreasing, albeit at a low rate, is probably to some extent a reflection of the courts’ increasing unwillingness to sentence children to prison time. This decreasing number would not be cause for concern if it were clear that children accused of sexual offences were instead being dealt with in an appropriate manner (for example, via diversion to a youth programme), as prison time is seldom the most appropriate sentence for a child. However, as we shall see below, the data does not suggest that this is the case: in the Western Cape the proportion of children arrested for sexual offences who are subsequently diverted, remained at 5% or less between 1998 and 2001. Another point to note is that the peaks in numbers of unsentenced children around December/January and April/May each year correspond with Christmas and Easter holidays. It is unclear whether this implies that more children commit sexual offences over this time, or that staff (such as probation officers and social workers) who would ordinarily be available to keep children out of prison pending their appearance in court, are less available. Arrests of child sex offenders in the Western Cape While custody data provides some insight into the numbers of children ending up in prison, arrest data gives an idea of how many children reach the stage of arrest, but might not reach prison. It is important to note that arrest data does not give a picture of the incidence of a crime type: many other factors such as the preparedness and ability of police to arrest have an impact on arrest data. However, arrest data is important in order to give an indication of the extent to which police officials might in future be called upon to implement the provisions of the Child Justice Bill in respect of child sex offenders. It may even be the case that once the Bill becomes law, the existence of an appropriate framework for child offenders will lead to a larger number of arrests by police. This data is important to illustrate the position prior to enactment of the legislation. The office of the Provincial Commissioner of the South African Police Service (SAPS) in the Western Cape provided figures on persons 17 years and younger arrested for sexual offences during the period 1 January 1998 to 31 December 2001. Over this period, more children (498) were arrested in 2000 than in any other year (Figure 3). The high arrest figure in 2000 corresponds with the highest custody numbers noted in the section above. Although there was a drop in 2001 from the high in 2000, the overall trend is toward an increase in these arrests in the province, by about 29 arrests per year. Of the 498 children arrested in 2000, 470 were arrested for rape, attempted rape or indecent assault. To place this number in perspective, this represents 5% of the total number (8,971) of rapes, attempted rapes and indecent assault cases reported to the police in the Western Cape in 2000. REDPATH Figure 3: Children arrested for sexual offences in the Western Cape, 1998-2001 Source: SAPS Provincial Crime Information Analysis Centre 0 100 200 300 400 500 600 1998 1999 2000 2001 42 223 312 463 498 396 323 387 287 60 78 88 N u m b er o f ar re st s Total Rape Indecent Assault SA CRIME QUARTERLY No 4 JUNE 2003 20 REDPATH This would tend to indicate that a higher proportion of adults arrested for sexual offences are prosecuted than are children arrested for these offences. Diversion of children can explain only a very small proportion of this difference. And since the withdrawal rates are similar for adult and child sex offenders, the difference in prosecution rates boils down to the outcome for child arrestees classified as ‘unknown’ by the SAPS. This category is cause for concern: what is happening to these children? Arrests by Child Protection Units in the Western Cape In this section, the ages of perpetrators arrested by the Child Protection Units or Family Violence, Child Protection and Sexual Offences Units (CPUs and FCSs) in the Western Cape are considered. There are 45 such units in South Africa, two of which are in the Western Cape: one located in Mitchells Plain, which serves the West Metropole of the Cape Metropolitan region, and the other at Goodwood, which serves the East Metropole.6 The East Metropole CPU has, since April 2000, kept a handwritten record of the ages of all alleged perpetrators arrested in cases referred to the CPU where the victim is under the age of 18 years. The data from the East Metropole CPU, which covers the period from April 2000 to end January 2002, was reviewed. A total of 865 arrests were analysed according to monthly trends, age, and type of offence. The record covered 22 months, which implies an average of 39 cases per month. The majority of the 865 arrests recorded by the East Metropolitan CPU between April 2000 to end January 2002 were for indecent assault and rape. Forty-nine per cent (428) of perpetrators were charged with indecent assault, while rape (excluding attempts, but including statutory rapes and ‘sodomy’) accounted for 43% (373) of all arrests. Other arrests were for attempted murder, assault with intent to cause grievous bodily harm, assault, kidnapping, abduction, and cruelty towards children. Figure 4: Outcome of arrests of children for sexual offences in the Western Cape, 1998-2001 Source: SAPS Provincial Crime Information Analysis Centre 0 1998 51 10 20 30 40 50 60 % o f ar re st s 30 13 5 43 29 24 4 45 36 16 3 47 22 26 5 1999 2000 2001 Prosecution Withdrawal Unknown Diversion What is the outcome of arrests of children for sexual offences in the Western Cape? About 40%- 50% of all these arrests lead to prosecution, while 20%-35% of such cases are withdrawn (Figure 4). Only 5% or fewer of children arrested for sex crimes are diverted from the criminal justice system. It is worrying that between 10% and 25% of all arrest outcomes are classified as ‘unknown’. If the ‘withdrawn’ and ‘unknown’ cases are added together, this implies that in all the years considered, around 40% of child sex offenders, who are faced with enough evidence against them to warrant an arrest, are simply not being dealt with by the criminal justice system at all. How does the situation depicted in Figure 4 compare to the situation for all sexual offences in the Western Cape, and in South Africa? Unfortunately the data in Figure 5 regarding the Western Cape and South Africa is in respect of reported cases, as opposed to arrests. However, if we assume that the number of arrests is equal to the number of reported cases, less those reports that are categorised by the SAPS as ‘unfounded’ or ‘untraced’, we find that about 65% of these ‘arrests’ are prosecuted in the Western Cape and South Africa, and that 35% are withdrawn in the Western Cape while 30% are withdrawn in South Africa. Monthly arrest patterns The first noteworthy trend is that the monthly data does not show a consistent pattern over time, which tends to suggest that the number of arrests each month is not predominantly dependent on factors which change evenly over time, such as normal population growth. Instead, it suggests that other more erratic factors, such as staffing levels, have a major impact on the number of arrests each month. Despite this variation, a trend toward an increase in numbers of arrests can be discerned. Linear equations fitted to the data show that the trend is toward an increase in the total number of arrests of almost one extra arrest per month. The monthly trend with respect to arrests of children is toward an increase of one child arrested every three months, while the trend for adults is approximately one every second month. The fit to a straight-line graph is poor for all three (total, adults, and children), however, so these rates of increase must be seen only as showing the general trend toward an increase in arrests and are not predictive of values in a future month. In other words, there is a large variation recorded from month to month. Figure 6 shows the number of arrests in each month. The variation is from a minimum of 22 arrests in April 2001 to three times as many, that is, 66 arrests, in November 2001. With respect to child perpetrators, the minimum number of arrests per month is three (in April 2000) and the maximum is 21 – seven times as many (in September 2001). However, various other local peaks and troughs occur before and after these high and low points. Note that there does also appear to be a suggestion of a seasonal pattern, with higher numbers of arrestees in the latter part of the year (August to December). However, only just less than two years’ worth of data is available, which is not enough to see whether this trend is valid. What is clear is that a number of other factors must contribute to the highly fluctuating number of arrests per month, such as, for example, the number of police and social workers 0 Prosecuted 45 10 20 30 40 50 60 % 54 21 29 30 19 8 10 Withdrawn Guilty South Africa Western Cape SA CRIME QUARTERLY No 4 JUNE 2003 21REDPATH on duty, available resources (for instance vehicles), and police priorities. What is also clear is that these factors are the same for both children and adult arrestees, as the peaks and troughs are roughly in tandem for children and adults. Age of arrestees As is evident from Figure 7, the 18 - 30 age group made up the largest group of arrestees, accounting for 30% (261 arrests), followed by the 31 - 40 age group with 25% (212 arrests). Children (persons aged 17 or younger) were arrested in 201 cases (23%). Once the Child Justice Bill is enacted, it is therefore likely that the East Metro CPU will be required to follow its framework in around 23% of their cases – those that deal with child arrestees. Another question worth considering is whether the proportion of children arrested by the East Metro CPU (23%) represents a high or low figure for perpetrators of sexual crimes against children. One way of attempting to answer this question is to analyse arrest rates. The extent to which certain age groups might be expected to contribute to arrests for sexual crimes against children in the East Metro area can be assessed by analysing arrests as a proportion that each age group makes up of the Western Cape population.7 The various arrest rates (covering the 22 months considered) for each population age group are illustrated in Figure 8. Figure 5: Outcome of all sexual offence cases reported to the police in South Africa and the Western Cape in 20005 Source: SAPS Unfounded/ untraced SA CRIME QUARTERLY No 4 JUNE 2003 22 REDPATH It is clear that children aged 10 to 17 and those aged 31 to 40 both have the highest East Metro CPU arrest rate (32 arrests per 100,000). For children aged 10 to 17, this might be expected, as only crimes committed against other children are considered; and it is more likely that if a child is to commit a sexual crime, that it be committed against someone of comparable age or younger. Furthermore, given the power relations when an adult offender is involved, it may be more likely for a child to report a sexual crime when it is committed by a child, than when it is committed by an adult. It is also noteworthy that the rate for children aged 10 to 17 is higher for indecent assault (17 per 100,000) than for rape (13 per 100,000). This suggests that children of this age who commit a sexual crime are either more likely to commit indecent assault than rape, or more likely to be arrested for indecent assault than for rape. It may be postulated that the high overall rate (32 per 100,000) for the 31 - 40 age group could lie in the greater exposure of persons of that age to children – they are often parents, family relatives or other caregivers of children. As we shall see in the next section on Childline statistics, callers to Childline identified the abuser as a parent in 47% of calls and in 85% of calls the abuser was a person known to the child. Older people (41+) have the lowest rate (19 per 100,000) for all East Metro CPU crimes, yet they may also be expected to have exposure to children as parents or other caregivers. One explanation for this low rate could lie in the ratio of men to women in the older age groups – as men commit more sexual crimes than do women, so a smaller proportion of men in the population may lead to a lower arrest rate. In the Western Cape only 46.5% of persons in the 41+ age group are men (there are 15% more women than men in the 41+ age group), while 49.6% of the population under 41 is comprised of men (there are only 2% more women than men in the under 41 group). While persons aged 18 - 30 have a comparatively low indecent assault arrest rate (10 per 100,000), their rape arrest rate (13 per 100,000) is similar to the 10 - 17 and 31 - 40 age groups. This suggests that persons in the 18 - 30 age group who commit sexual crimes against children are either more likely to commit rape than indecent assault, or are more likely to be arrested for rape than for indecent assault. Why this should be so is unclear, but it may be that rape victims are more likely to report the crime than are victims of indecent assault. Childline statistics Statistics from Childline are useful in that they count incidents of sexual offences against children that may Figure 6: Arrests of adults and children for all offences committed against children, East Metro area of the Western Cape, April 2000 to January 2002 Source: SAPS East Metropole Child Protection Unit 0 10 20 30 40 50 60 70 2000 April 24 21 3 May 27 23 4 July 29 25 4 Aug 45 33 12 Sept 30 26 4 Oct 57 42 15 Dec 30 22 8 2001 Jan 37 32 5 Feb 32 28 4 Apr 22 16 6 May 43 31 12 June 31 27 4 Aug 46 37 9 Sept 53 32 21 Oct 43 29 14 Nov 66 51 15 2002 Jan 35 25 10 N u m b er o f ar re st s p er m o n th Children Adults Total fewer than 40 per day per call centre. Two thirds of calls, or 66%, were from children aged 11 - 15. Only one third (13,842) were calls with a specific need – the remainder were what Childline refers to as ‘test calls’. The analysis deals only with ‘non-test’ calls. The child victim knew 85% of perpetrators, and of these 47% were parents while 8% were stepparents. Calls to report sexual abuse amounted to 1,734, or 13% of the non-test calls. This included both intra- familial and extra-familial abuse. Figure 9 is based on non-test calls only, and shows the rate at which people of different age groups commit offences against children, based on callers’ reports to Childline between July and December 2000. Because not all callers identified the age of the perpetrator, this graph analyses only the 2,190 (16%) of callers reporting all types of abuse, and the 584 (34%) of calls to report sexual abuse in which the age of the perpetrator was identified. Those aged 15 - 20 years stand out as being most likely to abuse children, whether for all types of abuse (13 per 100,000) or sexual abuse in particular (4 per 100,000). Unfortunately, we do not know how many of these alleged abusers are under 18 years of age (and would therefore be treated as a child under the Child Justice Bill), and how many are in the 18 - 20 age group. The age group next mostly likely to abuse children is that between 10 and 15 years (11 per 100,000 for all abuse and 2 for sexual abuse). The rate for the under- 15 population is calculated excluding those younger than 10.9 Indeed, the Childline research shows that 54% of all child sex abusers, and 51% of all abusers of children, are themselves younger than 20 years of age. Unfortunately, these age categories are different from the East Metro CPU arrest data, so direct comparison with that data is not possible. However, similar to the East Metro CPU arrest data, the Childline figures show that people in the 40 - 50 age group were least likely to be involved in child abuse. Conclusion The relatively high and increasing number of unsentenced children in custody is worrying. If the 13 SA CRIME QUARTERLY No 4 JUNE 2003 REDPATH 23 or may not have been reported to the police.8 In other words, they provide an indication of the extent to which such instances may be under-reported. During the six-month period of July to December 2000, Childline’s six provincial call centres received 42,234 telephone calls, that is, approximately 230 calls each day, an average of just Figure 8: Arrest rates for sexual offences against children in the East Metro area of the Western Cape, by age Source: SAPS East Metropole Child Protection Unit Figure 7: Age of persons arrested for offences committed against children in the East Metro area of the Western Cape, April 2000 - Jan 2002 Source: SAPS East Metropole Child Protection Unit <18 years 23% 18-30 years 30% 31-40 years 25% 41+ years 22% 0 All CPU crimes 32 5 10 15 20 25 30 35 N o . o f ar re st s p er 1 0 0 ,0 0 0 p eo p le i n a ge g ro u p 32 19 25 13 13 7 17 16 10 11 Rape Indecent Assault 10-17 yrs 18-30 yrs 31-40 yrs 41+ yrs children that are reported; the Childline research indicates that the true extent of this problem may be somewhat hidden. The provisions of the Child Justice Bill appear to be necessary to provide for a clear policy framework and consistent management for child sex offenders. Acknowledgement The study that this article draws from was commissioned by the Community Law Centre at the University of the Western Cape, website www.communitylawcentre.org.za Endnotes 1 B49-2002. 2 For example, an inquiry magistrate may order that a child older than 14 accused of a Schedule 3 offence be detained in prison under certain circumstances, after the child’s first appearance at an inquiry proceeding, and before trial. Sections 15,33,36, 58 and 81 of the Child Justice Bill all refer to Schedule 3 offences. 3 The Department of Correctional Services provided all custody data. 4 The Department of Correctional Services provides ‘snapshot’ data of the number of persons in prison at a particular date (e.g. at the end of the month). It does not count the total number passing through its doors over, for example, a year. 5 Figure 5 does not add up to 100% because of rounding, but also because some of the matters which were withdrawn or prosecuted in 2000 may have been reported in 1999 or earlier, so there may be more of those all together than there were reports in a year. The justice department is also making some progress on the backlog of cases. 6 The following police station areas fall under the East Metropole: Atlantis, Bellville, Bellville South, Bishop Lavis, Bothasig, Brackenfell, Delft, Durbanville, Elsie’s River, Goodwood, Gordon’s Bay, Cape Town Airport, Khayelitsha, Kraaifontein, Kuilsrivier, Macassar, Melkbosstrand, Milnerton, Parow, Ravensmead, Somerset West, Strand and Table View. 7 All provincial population estimates were calculated on the basis of Census 96. The predictive table compiled by Calle Hedberg, HISP was used to calculate rates. The predictions take into account estimates of deaths due to HIV/AIDS. Population estimates for mid-2000 were used throughout. 8 All the statistics in this section are drawn from a Research Report compiled by Childline and kindly supplied by Lynne Cawood of Childline Gauteng. 9 The numbers of alleged abusers younger than ten is small, as is also clear from the East Metro data. The rates including the under ten population: for all abuse the rate is 3.5 per 100,000 and for sexual abuse 0.73 per 100,000. SA CRIME QUARTERLY No 4 JUNE 2003 24 REDPATH aim is to keep children out of prison except where it is appropriate, it makes no sense that fewer are being sentenced, while more are spending time in prison unsentenced. Inconsistent management of child offenders is also suggested by the fluctuating custody figures discussed above. There is a need to determine what is happening to children suspected of, and arrested for, sexual offences. A significant proportion appears not to go to prison, and it is not clear from the data to what extent diversion or another appropriate intervention is occurring. The data seems to imply that most child sex offenders are simply going through the system without any intervention that might affect their future behaviour. The arrest figures show that the number of arrests of children for sexual crimes committed against other children is increasing; however, the number of arrests for all age groups of such crimes is also increasing, and at a faster rate. Children do nevertheless appear to commit a significant portion of sexual crimes against other Figure 9: Rate at which people of different age groups commit offences against children in SA, based on calls to Childline, July-Dec 2000 Source: Childline 0 All Abuse 13 2 4 6 8 10 12 14 N o . o f p eo p le n am ed a s p er p et ra to rs p er 1 0 0 ,0 0 0 p eo p le i n e ac h a ge g ro u p i n S A 11 7 4 5 4 2 2 11 Sexual Abuse <15 yrs 15-20 yrs 20-30 yrs 30-40 yrs 40-50 yrs