SA CRIME QUARTERLY No 20 • JUNE 2007 21 B etween 1995 and 2004 the South African prison population grew from 116 846 to 187 036, while available accommodation remained the same at approximately 113 000. But this is not simply the result of more people being imprisoned, or even, as is often argued, the large number of unsentenced prisoners. It is a widely accepted fact that South Africa’s prisons are overcrowded and that this is unacceptable from a human rights perspective. However, there is far less agreement among Correctional Services stakeholders on the reasons for prison overcrowding. This analysis focuses on the effect of sentencing on the size and profile of the sentenced prison population. This focus was motivated, at least in part, by important legislative amendments made in 1997 and 1998; namely the introduction of mandatory minimum sentences (Criminal Law Amendment Act 1997) and the expansion of the sentencing jurisdiction of the district and regional courts. (Magistrates’ Court Amendment Act 1998) It is accepted that the size of the total prison population is influenced by a number of factors or ‘prison population drivers’, for example macro- scale population trends, trends in crime, law enforcement, and sentencing. Sentencing emerges from this analysis as a critical driver of the prison population. Using quantitative data (made available by the Department of Correctional Services) on the size and profile of the prison population over an eleven- year period, 1995 to 2005, the relationship between sentencing and the size of the prison population is explored. On closer inspection it appears that fewer offenders are sentenced to imprisonment, but for longer. It was also found that, apart from the dramatic change in the profile of sentence lengths, the offence profile of the prison population has changed significantly, and this has had a material impact on services rendered to prisoners, security issues in prisons, and infrastructural requirements. It is concluded that these three factors, working in tandem, resulted in the rapid change in size and profile of the sentenced prison population, especially after 1998. THE NUMBERING OF DAYS Sentencing and prison population growth On 30 May 2007 the Criminal Law Amendment Bill (15 of 2007) was tabled in Parliament, proposing amendments to what has become known as the ‘minimum sentences’ legislation. The proposed amendments herald another chapter in the prison overcrowding debate in South Africa and will focus attention on the impact of sentencing on the size of the prison population. Chris Giffard, Independent Consultant chrisgiffard@gmail.com Lukas Muntingh, Project Coordinator Civil Society Prison Reform Initiative (CSPRI) Community Law Centre, University of the Western Cape lmuntingh@uwc.ac.za Average for January of each year % increase 22 GIFFARD AND MUNTINGH SA CRIME QUARTERLY No 20 • JUNE 2007 Using basic mathematical projections, a projection is made of the size of the prison population by 2015. This article is a summary of a more detailed research report commissioned by the Open Society Foundation (SA) and only the key findings are highlighted here. Overview of the prisoner population Figure 1 provides an overview of the total South African prison population for the period 1995 to 2005, and three trends are discernable. The first is the overall increase, as noted above, from less than 120 000 in 1995, to more than 180 000 by 2004. The second is the fairly rapid increase in the unsentenced prison population from 1995 to 2000, Figure 1: Total, sentenced and unsentenced prisons population, 1995 to 2005 Source: DCS Management Information System Table 1: Overview of the South African prison population Sentence category 1995 2000 2005 1995-2000 2000-2005 1995-2005 Unsentenced 24 265 61 563 52 313 154 -15 116 0-6 months 5 831 5 717 5 674 -2 -1 -3 >6-12 months 6 374 6 598 5 416 4 -18 -15 >12-<24 months 3 765 6 156 5 763 64 -6 53 2-3 years 12 854 13 846 17 816 8 29 39 >3-5 years 21 066 16 162 16 731 -23 4 -21 >5-7 years 15 068 13 882 12 137 -8 -13 -19 >7-10 years 12 193 18 418 21 233 51 15 74 >10-15 years 6 168 10 442 23 139 69 122 275 >15-20 years 2 660 4 603 10 586 73 130 298 >20 years 1 885 4 919 9 197 161 87 388 Life sentence 443 1 086 5 745 145 429 1 197 Other sentences1 4 274 3 031 1 706 -29 -44 -60 Total sentenced 92 581 104 860 135 143 13 29 46 Total prisoners 116 846 166 423 187 456 42 13 60 Source: DCS Management Information System 80 000 100 000 19 95 /0 1 19 95 /0 7 19 96 /0 1 19 96 /0 7 19 97 /0 1 19 97 /0 7 19 98 /0 1 19 98 /0 7 19 99 /0 1 19 99 /0 7 20 00 /0 1 20 00 /0 7 20 01 /0 1 20 01 /0 7 20 02 /0 1 20 02 /0 7 20 03 /0 1 20 03 /0 7 20 04 /0 1 20 04 /0 7 20 05 /0 1 20 05 /0 7 60 000 40 000 20 000 0 120 000 140 000 160 000 180 000 200 000 N o . o f p ri so n er s Unsentenced Sentenced Total that thereafter levelled and declined steadily until 2005. The third is the steady increase in the sentenced population that started in 1995, dropped as a result of two amnesties in 1995 and 1998, but then accelerated until 2005, when the Department of Correctional Services (DCS) implemented a remission of sentence programme. It is evident that post-2000, the sentenced prison population became the main driver of the increase in the total population, taking over from the unsentenced prison population. From 1995 to 2005 the profile of the sentenced prison population showed remarkable changes, as presented in Table 1. The most significant trend was the rapid increase in the number of prisoners serving sentences of longer than seven years. For example, the number of prisoners serving sentences of 10-15 years increased by 275%, the 15-20 year category by 298%, and life sentences by a phenomenal 1 197%. The general trend was that the longer the sentence, the greater the increase. On the other hand, prisoners serving sentences of less than 12 months showed a decline, as well as prisoners SA CRIME QUARTERLY No 20 • JUNE 2007 23 serving sentences of between three and seven years. Figure 2 shows very clearly how the proportion of prisoners serving sentences of less than seven years declined, while those serving sentences of longer than seven years increased from just less than 26% in 1995 to 52% at the end of 2005. Legislative changes and their effect Two key legislative changes were introduced in the late 1990s. The first was the Magistrates’ Court Amendment Act (1998) that extended the sentencing jurisdiction of the magistrates’ courts at both district and regional levels. In the case of the district courts, the maximum penalty that they may impose was increased from twelve months to three years’ imprisonment, while the jurisdiction of the regional courts was increased from 10 years to 15 years’ imprisonment. The second was the Criminal Law Amendment Act (1997) that provided for mandatory minimum sentences for specific offences, and came into force in May 1998. The minimum sentencing legislation has a number of features to ensure that the intended severity of the prescribed sentences is not undermined by sentencing officers or by the executive. No part of the sentence can be suspended (Criminal Law Amendment Act 1997:S 51 (5)). Nor can the time spent in prison awaiting trial be deducted from the prescribed sentence (Criminal Law Amendment Act 1997: S 51 (4)). An offender sentenced in terms of the minimum sentencing legislation must also serve four fifths of the sentence before s/he can be considered for parole, whereas the majority of other offenders can be considered after serving half of their sentence (Correctional Services Act 1998:S73(6)(b)(v)).2 Judicial officers may impose a lesser sentence than the prescribed minimum only if they find ‘substantial and compelling circumstances’ which justify a departure from the mandatory sentence. (Criminal Law Amendment Act 1997:S 51(3)(a). See Ehlers L & Sloth-Nielsen 2005:12-13) As will be shown below, this legislation made a substantial contribution to changing the profile of South Africa’s sentenced prison population. But while the impact of the legislation on prisoner numbers is likely to be felt for decades to come; its GIFFARD AND MUNTINGH Figure 2: Proportion of prisoners serving sentences of shorter and longer than 7 years Source: DCS Management Information System 10 20 30 1995 50 70 40 60 80 % o f se n te n ce d p ri so n er s 0-7 years >7 years 74 26 2000 61 39 2005 48 52 minimum sentencing legislation had a consolidation effect on trends that started pre- 1998. A closer inspection of the data shows that sentence lengths started increasing prior to 1998; a trend that was interpreted as a more punitive attitude from sentencing officers. This is demonstrated in Figures 3 and 4, that show the increase in the number of sentences longer than ten years prior to 1998, and the acceleration from 2000 onwards. SA CRIME QUARTERLY No 20 • JUNE 200724 effects were not immediate. For the first two years of its operation, there was no impact at all, due to the time lag from the date of implementation of the minimum sentencing legislation, the commission of offences after that date and detection, prosecution, conviction and sentencing. From 2000 onwards the combination of the Magistrates Court Amendment Act (1998) and the GIFFARD AND MUNTINGH Figure 3: Prisoners serving sentences of life and >20 years 1995-2005 Source: DCS Management Information System Figure 4: Prisoners serving sentences of >10-20 years 1995-2005 Source: DCS Management Information System 4 000 5 000 19 95 /0 1 19 95 /0 7 19 96 /0 1 19 96 /0 7 19 97 /0 1 19 97 /0 7 19 98 /0 1 19 98 /0 7 19 99 /0 1 19 99 /0 7 20 00 /0 1 20 00 /0 7 20 01 /0 1 20 01 /0 7 20 02 /0 1 20 02 /0 7 20 03 /0 1 20 03 /0 7 20 04 /0 1 20 04 /0 7 20 05 /0 1 20 05 /0 7 3 000 2 000 1 000 0 6 000 7 000 8 000 9 000 10 000 N o . o f p ri so n er s Life >20 years 19 95 /0 1 19 95 /0 7 19 96 /0 1 19 96 /0 7 19 97 /0 1 19 97 /0 7 19 98 /0 1 19 98 /0 7 19 99 /0 1 19 99 /0 7 20 00 /0 1 20 00 /0 7 20 01 /0 1 20 01 /0 7 20 02 /0 1 20 02 /0 7 20 03 /0 1 20 03 /0 7 20 04 /0 1 20 04 /0 7 20 05 /0 1 20 05 /0 7 0 5 000 10 000 15 000 25 000 20 000 N o . o f p ri so n er s >15-20 years >10-15 years 25SA CRIME QUARTERLY No 20 • JUNE 2007 An analysis of the data reveals that in early 2000 the number of prisoners serving life sentences began to increase dramatically. It is also apparent that the longer than 20-year category did not have as sharp an increase at that time. The increases look similar in the >10- 20 year sentences (see Figure 4). The >10-15 year category showed a far sharper increase, starting in May 1999. There are two probable reasons for the increase in this latter category. The first is the extension of the sentencing jurisdiction of the regional courts to a maximum of 15 years from the end of 1998. Secondly, a fifteen year sentence is mandatory for first offenders convicted of ‘less severe’ instances of murder, and robbery when aggravating circumstances are involved, or the taking of a motor vehicle (vehicle hijacking). Given the increase in these types of offences in the 1990s,3 it is probable that these offenders make up a significant proportion of the increasing numbers in this sentence category. Figure 5 shows that offenders serving prison sentences of three to seven years decreased in terms of real numbers and proportional share. From the late 1990s to 2005, it is more specifically the increase in the general sentencing tariff that played the major role in increasing the size of the prison population. In general, the number of prisoners serving long sentences increased, while the number of those serving shorter sentences decreased. The turnover of prisoners thus slowed down, and regardless of the fact that fewer offenders were being sentenced to imprisonment; they were staying there for longer. It is also concluded that, at this stage, the increasing size of the sentenced prison population was not caused by the minimum sentencing legislation. The increase in the number of prisoners serving longer sentences preceded the promulgation of the minimum sentencing legislation and thus also its delayed impact from 2000 onwards. It is possible that this increase was facilitated and consolidated by the minimum sentencing legislation and the increase in sentence jurisdiction, but the initial impetus came from elsewhere. A combination of public and political pressure on the courts to increase the severity of sentences, and the increase in the jurisdiction of the magistrates’ courts provided this impetus. The increase in the number of prisoners serving two- to three-year sentences due to the increased jurisdiction of the district courts is clearly visible in the data. It is however the >10-15 year sentence category that made a greater contribution than any other to the rise in the prison population, and this is attributed to the increase in jurisdiction of the regional courts. GIFFARD AND MUNTINGH Figure 5: Number of prisoners serving sentences of >3-10 years, 1995-2005 Source: DCS Management Information System 19 95 /0 1 19 95 /0 8 19 96 /0 3 19 96 /1 0 19 97 /0 5 19 97 /1 2 19 98 /0 7 19 99 /0 2 19 99 /0 9 20 00 /0 4 20 00 /1 1 20 01 /0 6 20 02 /0 1 20 02 /0 8 20 03 /0 3 20 03 /1 0 20 04 /0 5 20 04 /1 2 20 05 /0 7 0 5 000 10 000 15 000 25 000 20 000 N o . o f p ri so n er s >3-5 years >5-7 years >7-10 years 10 20 30 50 40 60 Aggressive Sexual 42 47 58 10 12 16 Changes in the offence profile A study of the offence categories for particular sentence groups also indicates changing trends. The DCS currently uses five offence categories: the three major categories being economic, sexual, and aggressive, while the remaining two categories – narcotics and ‘other’ – are relatively insignificant. The minimum sentencing legislation is aimed primarily, but not solely, at sexual and violent offences. SA CRIME QUARTERLY No 20 • JUNE 200726 GIFFARD AND MUNTINGH The first important trend is that while the number of offenders serving sentences for economic offences fell, the number of both aggressive and sexual offenders rapidly increased. In 1995, offenders serving sentences for economic offences made up 42% of the sentenced prison population, and sexual and aggressive offences 46%. By 2005, the figure for economic offences had fallen to 20%, while that for sexual and aggressive offences had increased to 75%. (see Figure 6) Figure 7: Sexual offenders serving sentences of >15 years, 1995-2005 Source: DCS Management Information System Figure 6: Percentage shares of crime categories Source: DCS Management Information System Economic % o f se n te n ce d p ri so n er s 1995 2000 2005 Narcotics Other 19 95 /0 1 19 95 /0 7 19 96 /0 1 19 96 /0 7 19 97 /0 1 19 97 /0 7 19 98 /0 1 19 98 /0 7 19 99 /0 1 19 99 /0 7 20 00 /0 1 20 00 /0 7 20 01 /0 1 20 01 /0 7 20 02 /0 1 20 02 /0 7 20 03 /0 1 20 03 /0 7 20 04 /0 1 20 04 /0 7 20 05 /0 1 20 05 /0 7 0 1 200 1 000 800 600 400 200 1 400 1 600 2 000 1 800 N o . o f p ri so n er s >15-20 years >20 years Life sentence 20 05 /1 2 40 32 20 4 4 2 3 6 3 since 1995. Only the special remissions of 2005 brought some respite. Figure 8 shows that even though the proportion of prisoners living in conditions of between 100% and 200% occupancy4 slowly decreased from 1996 to 2004 (a trend ended by the remission), this decrease has been at the expense of others. The number of prisoners in prisons with occupancy rates of more than 200%, increased from just 1% in 1995 to 36% in 2004. Of equal concern is the proportion of prisoners detained in institutions in which there are three times as many prisoners than capacity allows. There were no prisoners in this category until 1997, but by 2004 as many as 5% of all prisoners (a total of over 9 000) were held in such facilities. The special remissions reduced this number only slightly, to just less than 8 500. This is because it is largely long- term prisoners and unsentenced prisoners who experience these conditions, and neither of these categories benefited from the remissions. Staff and prison management are also adversely affected by the increase in numbers of long-term prisoners – but it goes further than that. An increasing proportion of offenders are classified as maximum security prisoners, due to the length of their sentences being based on the formula used by DCS. The result was a sharp increase in the number SA CRIME QUARTERLY No 20 • JUNE 2007 27GIFFARD AND MUNTINGH In particular, the number of offenders serving sentences for aggressive and sexual offences significantly impacted on the numbers in the longer sentence categories. The increase in both categories is directly related to the minimum sentencing legislation. Figure 7 shows that the increase in the total number of sexual offenders serving life sentences began in August 1999, 15 months after the introduction of the minimum sentencing legislation. This is as expected. But there is a significant difference in the timing of the increase in numbers of those serving life sentences for aggressive offences and those serving life sentences for sexual offences. The numbers serving life for aggressive offences began to increase slowly from 1995, due to political and public pressure, and then increased more rapidly from 1999, as a result of the minimum sentencing legislation. The number of sexual offenders serving life, on the other hand, did not increase until in late 1999, when the legislation forced the hands of sentencing officers. The impact on prisons Such increases in prisoner numbers impact on prisons in a variety of ways. Prisoners themselves are of course most directly affected. The total number and proportion of prisoners living in prisons that are overcrowded have increased substantially Figure 8: Prisoners living in different overcrowding conditions, 1995-2005 Source: DCS Management Information System 10 20 30 50 40 60 70 80 90 0 % o f to ta l p ri so n er s 0-100 101-200 >200 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 of maximum security prisoners, from 14 229 in 1995 to 38 663 in 2005 (see Figure 9). However, for the first five years, until 2000, the increase merely kept pace with the total number of sentenced offenders: in fact, until 1997, the total number of maximum security prisoners declined, as did this category as a proportion of total sentenced prisoners (11% for 1997). From 1998 maximum security prisoners as a category increased in proportion by 2-3% each year, and by 2005, 30% of all sentenced prisoners were in this category (see Figure 10). Projecting the prison population Projecting the size of the South African prison population is fraught with difficulties (Giffard and Muntingh 2006:39). The projection entails basic mathematical forecasting. Firstly, it assumes that all social, political, health and other variables are to Figure 9: Major security classifications 1995-2005 Source: DCS Management Information System Figure 10: Percentage share of major security classifications, 1995-2005 Source: DCS Management Information System SA CRIME QUARTERLY No 20 • JUNE 200728 GIFFARD AND MUNTINGH 19 95 /0 1 19 95 /0 7 19 97 /0 1 19 97 /0 7 19 98 /0 1 19 98 /0 7 19 99 /0 1 19 99 /0 7 20 00 /0 1 20 00 /0 7 20 01 /0 1 20 01 /0 7 20 02 /0 1 20 02 /0 7 20 03 /0 1 20 03 /0 7 20 04 /0 1 20 04 /0 7 20 05 /0 1 20 05 /0 7 0 50 000 40 000 30 000 20 000 10 000 60 000 70 000 90 000 80 000 N o . o f p ri so n er s Maximum Medium Non-board 19 96 /0 1 19 96 /0 7 20 05 /1 2 19 95 /0 1 19 95 /0 8 19 97 /0 5 19 97 /1 2 19 98 /0 7 19 99 /0 2 19 99 /0 9 20 00 /0 4 20 00 /1 1 20 01 /0 6 20 02 /0 1 20 02 /0 8 20 03 /0 3 20 03 /1 0 20 04 /0 5 20 04 /1 2 20 05 /0 7 0 40 30 20 10 50 60 80 70 % o f se n te n ce d p ri so n er s Maximum Medium 19 96 /0 3 19 96 /1 0 20 05 /1 2 Table 2: Projection of the total prison population 2005 to 2015, two scenarios Current Projected totals Current Projected totals Dec 2005 Dec 2010 Dec 2015 Jan 2005 Dec 2010 Dec 2015 Unsentenced 46 327 46 327 46 327 46 327 46 327 46 327 0-6 months 4 189 4 189 4 189 5 674 5 674 5 674 >6-12 months 3 812 3 812 3 812 5 416 5 416 5 416 >12-<24 months 3 089 3 089 3 089 5 763 5 763 5 763 2-3 years 9 654 9 654 9 654 17 816 17 816 17 816 >3-5 years 10 675 10 675 10 675 16 731 16 731 16 731 >5-7 years 9 089 9 089 9 089 12 137 12 137 12 137 >7-10 years 18 298 18 478 21 462 18 298 18 478 21 462 >10-15 years 23 740 33 743 43 489 23 740 33 743 43 489 >15-20 years 11 122 15 996 20 627 11 122 15 996 20 627 >20 years 9 486 13 103 16 884 9 486 13 103 16 884 Life sentence 6 615 10 441 14 050 6 615 10 441 14 050 Total 156 096 178 595 203 348 179 125 201 624 226 377 Source: DCS Management Information System Sentences of 7 years and less stable as from December 2005 Sentences of 7 years and less stable as from January 2005 (pre-remission figures) remain constant for the projection period. Secondly, only sentences longer than seven years have been forecast. Shorter sentences tend to be more variable over time, and thus projections are less reliable. Thirdly, the final projection assumes that sentences of seven years and less remain the same over the entire forecast period, as well as the unsentenced prison population. Table 2 shows the projections for the entire prison population. The three columns on the left show the projection, if it is assumed that these totals are frozen from December 2005, after the special remissions. The three columns on the right show the projection based on the assumption that these sentence categories return to their pre-remission levels and then hold stable. The difference is substantial, and reflects the total number of prisoners released during the special remissions. It would be extremely optimistic to expect that, in the absence of systemic solutions involving the entire criminal justice system, the total number of these short-term prisoners will not increase again, as they have done after previous executive releases. While the very long sentence categories seem to have experienced the most spectacular increases (with the total of those serving life more than doubling over a ten year period, for example), it is the >10-15 year sentence category that is providing the bulk of the increase. The projections suggest an increase of nearly 20 000 offenders serving >10-15 year sentences between 2005 and 2015. The increases in all the sentence categories of longer than seven years have serious implications for prison overcrowding. The projections suggest that, assuming 9 000 new prison places by 2010 and a further 9 000 by 2015, the proportion of prison places taken up by prisoners serving sentences of longer than seven years will increase from 61% currently to 75% in 2010 and 88% in 2015. In 1995, this sentence category took up only 26% of the available capacity, and in 2000, the corresponding figure was 45%, as shown in Figure 11. SA CRIME QUARTERLY No 20 • JUNE 2007 29GIFFARD AND MUNTINGH 30 SA CRIME QUARTERLY No 20 • JUNE 2007 Conclusion The Criminal Law Amendment Act (1997) has thus far had some impact on the nature of the prison population, but has not yet impacted on total prisoner numbers. The full impact of the longer sentences brought about by the minimum sentencing legislation is likely to be felt only in five to ten years from now, as these prisoners need to serve four-fifths of their sentence before being considered for parole. The Magistrates Amendment Act (1998), which extended the jurisdiction of the regional courts from 10 to 15 years, had a more immediate effect. Projections suggest that it is this category of prisoners, many of whom were sentenced in the regional courts, that is likely to provide an increasing proportion of the projected increase of over 40 000 sentenced prisoners over the next ten years. Acknowledgements This article is based on research that was commissioned by the Open Society Foundation of South Africa into the impact of the Criminal Law Amendment Act 105 of 1997. The relevant OSF-SA GIFFARD AND MUNTINGH publication was launched on 28 February 2007 and is available on the OSF-SA website at www.osf.org.za. The opinions expressed in this article are those of the authors and not those of OSF-SA. The authors would like to express their sincere gratitude to the Department of Correctional Services for their assistance in making the quantitative data available. References Correctional Services Act 1998 (Act 111 of 1998). Criminal Law Amendment Act 1997 (Act 105 of 1997). Ehlers, L and Sloth-Nielsen, J 2005. A pyrrhic victory? Mandatory and minimum sentences in South Africa. Occasional Paper 111. Pretoria: Institute for Security Studies. Giffard, C and Muntingh, L 2006. The Effect of Sentencing on the Size of the South African Prison Population. Open Society Foundation for South Africa. Available at http://www.osf.org.za/home Magistrates’ Court Amendment Act 1998 (Act 66 of 1998). Masuka, S 2002. Prevention is better than cure: Addressing violent crime in South Africa’. SA Crime Quarterly (2). Pretoria: Institute for Security Studies. Endnotes 1 Over 80% of the category ‘Other Sentences’ consists of indeterminate sentences for ‘habitual criminals’. Others include death sentences, day parole, periodic imprisonment, ‘Other mental instability’ and prevention of crime. 2 There are numerous provisions for different parole consideration dates, but for the majority of offenders not sentenced under Act 105 of 1997, the minimum period to be served is half of the sentence. 3 For detailed data see http://www.issafrica.org index.php?link_id=24&slink_id =2797&link_type=12&slink_type=12&tmpl_id=3. See also Masuka, S 2002. ‘Prevention is better than cure: Addressing violent crime in South Africa’. SA Crime Quarterly (2). Pretoria: Institute for Security Studies. 4 An occupancy rate of 100% refers to an institution that is operating at its intended capacity. An occupancy rate of 200% means that there is double the number of prisoners than originally intended (400 in a prison intended for 200 prisoners). In Figure 8, the category 0- 100 refers to the number of prisoners living in institutions in which there is no overcrowding. Figure 11: Projected percentage of total national capacity used by prisoners serving sentences of longer than 7 years Source: DCS Management Information System 10 20 30 1995 50 70 40 60 80 90 80 % o f to ta l ca p ac it y 26 2000 45 2005 61 2010 75 2015 88