SA CRIME QUARTERLY No 19 MARCH 2007 7 The only other review of the impact of the minimum sentencing legislation was conducted in 2000 by the South African Law Reform Commission (SALRC). The SALRC study was arguably conducted too soon, as there is usually a delay of one to two years between a crime being committed and sentence being passed in our courts; and the legislation’s provisions apply only to crimes committed after minimum sentencing became law. One finding from the 2000 SALRC review that was quickly contradicted by the Hlakanaphila study was that magistrates and judges now do not feel their discretion is unduly compromised by the legislation. This was explained by a judgement from the Supreme Court of Appeal handed down in March 2001, which changed the way in which the courts had been interpreting the legislation. Malgas v S (Case Number 117/2000) essentially decided that a court could deviate from the prescribed minimum sentence if all the circumstances considered together were “substantial and compelling”. Earlier cases opined that a circumstance had to be extraordinary to justify a deviation from the I n 2006 Hlakanaphila Analytics was commissioned by the Open Society Foundation South Africa (OSF-SA) to investigate the impact of the minimum sentencing legislation. ‘Impact’ was considered broadly and included the impact on crime, on court procedure, on consistency of sentencing, on the rights of victims, on judicial independence, on public confidence, as well as issues such as the legislation’s constitutionality. A separate study of the impact on the size of the prison population was also commissioned by OSF- SA (and selected results will be published in the next issue of the SA Crime Quarterly). Methods used to assess impact The methodology for the Hlakanaphila study involved analysis of data including crime data, court data held by the National Prosecuting Authority, and Correctional Services data, supplemented by a survey of a sample of closed cases (328 records) drawn from three regional courts. Interviews with 50 relevant role players were conducted, including judges, magistrates and prosecutors. A review of relevant case law and international literature was also done. REACHING A VERDICT The impact of minimum sentencing The so-called ‘temporary’ minimum sentencing legislation introduced into South African law in 1998 is still in place. The legislation was passed largely in response to high crime rates at the time and the perceived leniency of the courts, and prescribes minimum sentences ranging from five years’ to life imprisonment for a variety of offences (including murder and rape and a range of other crimes, some of which are non-violent). Given the current furore over crime, it is highly likely that in April this year the legislation will be renewed for another year. But what has the impact of the legislation been and what legislative changes should be considered? Jean Redpath and Michael O’Donovan Hlakanaphila Analytics redpath@iafrica.com SA CRIME QUARTERLY No 19 MARCH 20078 REDPATH AND O’DONOVAN minimum, which was a much more restrictive interpretation. Presiding officers, particularly magistrates, largely feel this judgement gives them enough discretion to deviate when appropriate. Although it was not part of the brief to investigate overcrowding in prisons, the researchers had to consider correctional services data for the purposes of investigating changes in sentence length. It soon became apparent that current levels of overcrowding have little to do with minimum sentencing. Given that most accused sentenced for offences covered by minimum sentencing would in any event have received some sort of custodial sentence, the effect of minimum sentencing only becomes apparent at the point when prisoners remain in prison beyond the sentence and parole date they would otherwise have received – and in the majority of cases we have not yet reached that point. But it is also clear that in future years, the crisis in prisons will be far worse than what we have seen so far, as a direct result of minimum sentencing – and will only be ameliorated by mass early releases. Have minimum sentences helped reduce crime? According to the official line, crime has gone down since the legislation was passed. Can this be attributed to minimum sentencing? Murder, the most reliable indicator of crime levels and also one of the few crimes for which a minimum is prescribed under all circumstances, has indeed dropped since the legislation was passed. But this trend cannot necessarily be attributed to minimum sentencing. This is because the downward trend was established long before the legislation was implemented, may well have been overstated, and furthermore, may be in the process of reversing. The survey of closed cases highlighted the fact that whereas many countries consider ‘homicide’ rates when analysing crime trends, in South Africa we look at the ‘murder’ rate. There is a tendency in this country to think of culpable homicide only in relation to fatal traffic accidents. In the closed case sample there were no traffic accidents but instead cases of shootings in the chest and stabbings that were charged and convicted as culpable homicide. Surely we should be looking at the number of homicides – murder plus culpable homicide – when analysing crime levels? Plotting both the number of murder and culpable homicide cases over time showed that the two trends tended to broadly mirror each other until 2001/2 (Figure 1). This was the year of the uproar over ‘incorrect’ categorisations of crime and the Figure 1: Trends for murder and culpable homicide in SA, 1994/95–2005/06 0 5,000 10,000 15,000 1994/95 1995/96 1996/97 1997/98 1998/99 1999/ 2000 2000/01 2001/02 2002/03 2003/04 2004/05 2005/06 25,000 20,000 30,000 N u m b er Murder Culpable homicideSource: SAPS SA CRIME QUARTERLY No 19 MARCH 2007 9 retraining of police in how to classify crimes, as well as the explicit linking of crime rates to police performance. Thereafter murder continues to go down quite sharply, but culpable homicide goes up. The trend for the two together – referred to as ‘homicide’ – flattens out over the last three years, possibly indicating the end of the downward trend in homicides. It makes sense that minimum sentences have not necessarily impacted on murder levels because the international literature shows that the severity of sentence tends not to operate as a deterrent to crime. Instead, only high rates of detection of crime act as a deterrent. The best illustration of the point is that pick- pocketing was a capital offence in Victorian England, yet those who pick-pocketed regularly plied their trade during public hangings of people convicted of pick-pocketing, because they did not believe they would be caught. Sentences – even the death penalty – are irrelevant to criminals if they do not believe they will be caught. The literature did however find a small impact on crime via incapacitation – the impact of the removal of the offender from society. Internationally, this effect has been found to be small (because comparatively few offenders are caught, and the impact is only relevant in the case of serial offenders). Compared to other measures such as investment in social development and education, imprisonment is found to be expensive for the small amount of crime it prevents. Impact on court efficiency In the case of South Africa, for the same reason that the impact on overcrowding is yet to come, any small incapacitation effects of the legislation are probably still to come. Worse, the collateral impact of the legislation on court efficiency may be reducing the number of serious offenders being sent to prison, thus lessening any anticipated incapacitation effect. Since January 2002 the total number of admissions to prison has dropped by 25% for both long- and short- term sentences. There are several reasons for this drop, including an increased reluctance to send ‘petty’ offenders to prison. Another reason lies in an increase in court congestion. The increase in court congestion is a result of cases taking longer to conclude. The research, using NPA data, showed that both the regional court and high court – the courts in which these serious offences are heard – are taking longer to conclude cases and this may be partly attributable to minimum sentencing. By 2006 the average time between conviction and sentencing had risen to seven months, from four months in 2005. The time between the offence being committed and sentencing in the high court was in 2006 at an average of 25 months for cases heard by the high court and 27 months for cases heard in the regional courts and sentenced in the high courts (life imprisonment cases). Why are cases taking longer? Evidence suggests that accused persons facing stiff minimum sentences – of which they must be forewarned on first appearance – seldom plead guilty, tend to use Legal Aid more, and tend to appeal their sentences more often – lengthening case cycles, and adding to the overall burden on the courts. This view is supported by interview evidence, the sample data and relevant data on use of Legal Aid REDPATH AND O’DONOVAN 0 50 100 150 2002 2003 2004 2005 250 200 300 N u m b er Source: NPA Figure 2: Number of cases finalised with a verdict per regional court 260 240 201 179 and appeals. Thus there are fewer serious cases being convicted but they are taking up more of the courts’ time. The number of cases finalised per regional court is dropping (Figure 2), indicating the reducing efficiency of these courts and the increasing time taken to resolve cases. Interviewees furthermore consistently expressed the view that the split procedure for life imprisonment cases – which requires referral of the case to the high court for sentencing if it was heard in the regional court – is onerous and causing havoc. The referred cases are referred to as ‘section 52s’. Three copies of the regional court record must be prepared and sent to the high court for a s52. The high court must then satisfy itself of the conviction in the regional court before passing sentence. To do so the high court can and does re-hear evidence and often – more frequently than is comfortable, considering that these are life imprisonment cases – finds that the decision of the regional court to convict was incorrect. Available NPA data for 2005 shows that 12% of regional court convictions were set aside by the high court at sentencing stage. The high courts are not in favour of the minimum sentencing legislation, largely because of this split sentencing procedure, and have twice found it to be unconstitutional. The Witwatersrand High Court said the split procedure violated the right to a fair trial and caused inordinate delays. The Eastern Cape High Court held that the requirement to impose life imprisonment violated the principle of separation of powers, while the split procedure violated the right of an accused to have a case heard before the ordinary courts. In both cases the Constitutional Court declined to agree with the High Court (see S v Dzukuda & others, S v Tshilo 2000 (2) SACR 443 (CC), and S v Dodo 2001 (1) SACR 594 (CC).) Why is the number of short sentences also dropping? The data suggest that serious crime is being prioritised, in terms of the scarce resources of both prison time and court time. The number of persons serving terms of five years or less has dropped and continues to do so, which could suggest either that crimes which used to attract short sentences are getting heavier sentences, or that they are being dealt with outside the court and prison system. However, the number of people sentenced to long- term imprisonment has also been dropping, while there has not been a vast increase in capacity for alternative sentencing and diversion. This suggests that less serious crimes are being crowded out, and are either not entering the prison system, or are not entering the court system at all. Indeed, NPA data show that since January 2002 the total number of cases in the district courts has dropped by approximately 10%. Prioritising serious crime is laudable. But what message will an offender receive from the criminal justice system after committing a less serious offence, and what impact will that have on potential future behaviour? More consistent sentences? So far the research has established that the legislation has probably not had any desirable impact on crime and has probably reduced court efficiency. Have the prescribed minimums at least improved consistency of sentencing? Unfortunately, this is not generally the case either. The minimums have indeed increased the average sentence length per offence type. Analysis of Correctional Services data showed that between 1995 and 2005 the average term served by current prisoners for sexual offences rose from approximately seven months to just over ten years (126 months). But the minimum sentencing legislation appears at the same time to have increased the range of sentences passed per offence type, thus reducing overall consistency. This is because the minimums were generally set so far in excess of the earlier norms that in practice the ‘minimums’ are operating as maximums – an obvious result when the minimum sentence for some crimes is also the maximum penalty in our law (life imprisonment). The increase in the range of sentences is a result of the ‘maximum minimum’ being applied in some SA CRIME QUARTERLY No 19 MARCH 200710 REDPATH AND O’DONOVAN SA CRIME QUARTERLY No 19 MARCH 2007 11 instances, while sentences in line with previous practice are applied in others, leading to a greater range of sentences. For example, in the research sample of closed cases, murder charges resulted in an average sentence of eight and a half years, with sentences ranging widely from non-custodial sentences on the one hand, to 15 years’ imprisonment on the other. The exception to this finding is rape (but not life imprisonment rape). The minimum penalty prescribed for rape is ten years on first offence. (When rape is aggravated by specified violent circumstances or the vulnerability of the victim, life in prison is prescribed). The sample of rape cases drawn for 2005 showed the sentences for rape cases to be clustered around the ten year minimum, with a narrower range from seven to 15 years. The average sentence was ten years. Why should rape result in less of a range of sentences than other offences? Rape is different because it is the only offence apart from murder (and murder charges often result in culpable homicide convictions) for which the minimum of ten years applies to all rape offences. Most other offences require particular aggravating circumstances, such as use of a firearm or a threshold amount of money involved, for a minimum to apply. The minimum is also well known by prosecutors and magistrates and is unlikely to be overlooked. A ten- year sentence falls well within the regional court jurisdiction of 15 years, allowing for leeway on either side of the minimum. From the point of view of those calling for more appropriate sentences for rape, the legislation may well be viewed as a success. But that may only be from the perspective of those victims whose cases reach conviction stage. The harsher penalties have raised the stakes and reduced the likelihood of a guilty plea, increasing the proportion of rape victims who have to endure a full trial process. Rape cases still show a far higher withdrawal rate than other offences. Of the sample of rape cases opened in 2005 that had been closed by June/July REDPATH AND O’DONOVAN 2006, some 67% of rape cases were resolved by way of withdrawal, compared to 39% of all other offences. Victims of the worst kinds of rape (life imprisonment rapes) also face testifying in both the regional court for trial and the high court for sentencing, if the high court finds it necessary. Given that these cases include rapes of children, and the high court often feels compelled to be sure of the evidence of child witnesses, it is indeed the most vulnerable who are bearing the brunt of this two-step process. Public confidence in the justice system Has public confidence been buoyed by the lengthening of sentences? Magistrates and prosecutors agree that on the whole, victims of serious crime welcome heavy sentences – if they are in court for sentencing or are informed of the sentences. Broader public confidence, however, is linked to more than just sentencing. Earlier work conducted by Hlakanaphila Analytics for the United Nations Office for Drugs and Crime (UNODC) found that negative perceptions of the justice system are closely linked to perceptions about the general slowness of the justice system. So anything that slows down the operation of the courts and results in more postponements and referrals – like minimum sentencing – is likely to impact negatively on public confidence. It is unclear whether the positive impact of the heavier sentences offsets the negative impact that the legislation has on the length of cases. Public confidence may also be affected by raised expectations. Told that a prescribed minimum applies, victims may feel aggrieved if there is a deviation toward leniency in their case – even if the sentence ultimately handed down is in excess of what applied in the past. This is particularly likely in the most serious life imprisonment cases, when victims are often confused by the referral to the high court for sentencing. NPA data show that in 79% of cases in 2005 when life imprisonment was prescribed and a conviction obtained, an outcome other than life imprisonment was handed down in the high court. Even for less serious offences, the closed case sample found that the minimums still do not appear to be widely applied. The court records of the sampled 2005 cases were examined to see whether the offence noted in the charge sheet was one for which a minimum is prescribed. The majority of the cases classified as ‘minimum sentence cases’ were committed with a firearm and should, if the minimums were strictly applied, at the least have received a five-year sentence (the shortest of all the prescribed minimums). Yet in only 55% of all ‘minimum sentence’ cases was a sentence of five or more years handed down. Part of the problem is that the offences and circumstances in which minimum sentencing applies are quite complicated, and open to interpretation (except for ‘ordinary’ rape). In the 2005 sample, where there was a deviation from the minimum, ‘substantial and compelling circumstances’ justifying the deviation were only noted in writing, as required by the legislation, in one fifth of cases. This suggests that the courts frequently do not realise that, or consider whether, a minimum sentence applies in less well-known instances where they are indeed applicable. Interviewees suggested that the problem of the split procedure could be dealt with either by allowing the regional court to sentence to life imprisonment, or by allowing it to decide whether there are substantial and compelling circumstances and only referring to the high court those cases where there are not. The problem with the latter is that it would still be the most serious cases that would result in a split procedure. The problem with the former would be those 12% of cases where the high court has had to overturn the conviction of the regional court. No one is suggesting that all potential life imprisonment cases should be set down, heard and sentenced in the high court – because there are simply too many of them for our less than 30 judges to handle. The NPA data show that between May 2002 and March 2006, almost 10,000 cases in which a life sentence was a consideration were SA CRIME QUARTERLY No 19 MARCH 200712 REDPATH AND O’DONOVAN heard by the high courts (either from trial stage or for sentencing only). This means that in less than five years the high courts had the potential of filling almost 9% of South Africa’s prison beds through the life imprisonment minimum sentences alone. The impact of not renewing the minimum sentence Given the negative impacts discussed above, the inevitable question is, what if the legislation were simply not renewed? Interviewees think that the tariff for serious violent crime has been raised and that judges and magistrates would still hand down heavier sentences, even without the minimums. If that was the intention of the legislature, it certainly seems to have succeeded. Prosecutors would also probably be judicious in deciding in which cases they seek a life penalty, and would set them down in the high court. In all probability, fewer life terms would be handed down. The number of appeals would probably decrease. But there would still be a crisis in the prisons. The broader public would probably deplore the scrapping of minimum sentencing. Given the need for government to appear ‘tough on crime’, it is unlikely that the legislation will change, despite the damage it is causing. Tinkering with the legislation may result in paradoxical and unintended consequences, much as the legislation itself has done. Ultimately, a comprehensive overhaul of the justice system in general and the sentencing regime in particular is required. The courts do not appear to be coping, the prisons are increasingly overcrowded, and the public views mass releases from prison even less positively than it might view a change in sentencing law. Acknowledgement This article is based on research commissioned by the Open Society Foundation for South Africa (OSF- SA) to investigate the impact of the Criminal Law Amendment Act 105 of 1997. The relevant OSF-SA publication was launched on 28 February 2007 and is available on the OSF-SA website at . The opinions expressed in this article are those of the authors and not those of OSF-SA.