CQ No. 25 SA Crime Quarterly no 31 • March 2010 3 Money down the drain The direct cost to government of alcohol abuse Debbie Budlender debbieb@mail.ngo.za Soul City has recently embarked on a multi-year campaign that aims to prevent and reduce violence through focusing on alcohol abuse. This article describes an exercise commissioned by Soul City to estimate the expenditure allocations and expected revenue of national and provincial government departments in South Africa that can be attributed to alcohol abuse. The article highlights, in particular, the estimates derived for the security-related departments, namely Safety and Security, Correctional Services, Justice and Constitutional Development, and the national and provincial departments responsible for policing. The article concludes with suggestions regarding areas that would benefit from further investigation. Soul City, the well-known South African multi- media health promotion and social change project, has recently embarked on a multi-year campaign that aims to prevent and reduce violence through focusing on alcohol abuse. As part of the preparation for this campaign, Soul City com- missioned a study to estimate the spending and revenue of South Africa's national and provincial governments that could be related to alcohol abuse. The estimate was to include budget allocations that aim to deal with the consequences of alcohol abuse (such as expenditure of departments of health), as well as allocations that aim to reduce the extent of alcohol abuse and its negative consequences (such as Drive Alive campaigns of provincial traffic authorities). This article highlights some of the key findings of the budget study. GENERAL APPROACH The estimation exercise did not attempt to calculate the full cost of alcohol abuse to society. Instead it focused only on the costs reflected in government budget allocations. It did not look, for example, at the costs incurred by private companies, or by households. Single et al note that estimates in other countries suggest that the money that government spends on policy-related costs such as treatment, enforcement, prevention and research in relation to drug abuse amounts to only about five per cent of even a narrow definition of the full costs incurred by society.1 The estimates given in this article thus represent only a fraction of the true cost that alcohol abuse imposes on society. The Soul City exercise also did not look at what government should spend, but instead on what government actually allocated in the 2009-10 financial year. Estimating the amounts allocated by government on account of alcohol abuse is not a simple exercise. There may be a few allocations that focus directly on alcohol abuse, such as money for anti-alcohol awareness-raising initiatives or research. However, even here there are challenges, as publicly available CQ No. 31 - Mar 2010 3/17/10 3:19 PM Page 7 budget documents may not provide the detailed breakdowns that allow the projects or initiatives directly related to alcohol to be separated out. Beyond the focused allocations, government also incurs expenses within general allocations such as those for enforcing the law, punishing offenders, providing treatment for health problems, and welfare and other services for those negatively affected by alcohol abuse. The challenge here is that the government agencies responsible for these functions do not allocate separate amounts for the police, magistrates, nurses, prison warders and other officials who deal with alcohol abuse. Instead, 'general' police, magistrates, nurses and prison warders deal with problems resulting from alcohol abuse alongside dealing with problems arising from a host of other reasons. For these areas we therefore have to estimate the proportion of time and other resources that is attributable to alcohol abuse. The Soul City study details the research or other information on which each such estimate is based. Some of the assumptions may well be challenged. By providing full details of the source and assumptions, the study laid the basis for future refinement of the overall estimate. After discussion with experts, the following departments were included in the study: • Health (national and provincial) • Social Development (national and provincial) • Safety and Security (national and provincial) • Justice and Constitutional Development (national) • Correctional Services (national) • Transport (provincial) • Economic Development (provincial) This article focuses on the security-related departments. It is, however, not only these departments that have allocations related to crime. For example, each of the nine provincial departments of Social Development has sub- programmes related to crime prevention and support as well as to victim empowerment. To the extent that we acknowledge traffic infringements as crimes, the sub-programmes of the provincial departments responsible for traffic enforcement also deal with crime. In addition to expenditure, there are some revenues that accrue to government on account of alcohol abuse (or use). Firstly, there are the fees paid to provincial governments for liquor licences. Secondly, there are the excise duties paid in respect of alcohol. Thirdly, there is the value- added tax (VAT) generated through sale of alcohol. The Soul City exercise included estimates for all these revenue sources. It did not include estimates in respect of income tax paid by those employed in the liquor industry. SOURCES The primary source for the estimates was the government's publicly available budget-related documentation. In particular, the paper draws on the budget numbers published in the annual national estimates of expenditure and provincial budget statements that are tabled in the legislatures on budget days. Each of the chapters of these provincial and national documents details the budget allocated for a particular 'vote', which is for the most part equivalent to a government department. Within each vote, the chapter disaggregates the budget estimates first into programmes, of which there are typically about five for each vote. Each programme budget is then further disaggregated into sub-programmes, which can number as few as one per programme, or, in a few instances, more than ten. South Africa's budget documents include fairly extensive narratives alongside the budget numbers. This narrative assists in understanding what particular sub-programmes cover. The narrative also provides an indication of issues that the department concerned considers most important. In addition to the budget documents, the estimation drew on reports of diverse research initiatives. These research documents helped in determining the proportion of each sub- programme budget that might be considered attributable to alcohol abuse. In addition, the initiative drew on information offered through interviews, e-mails and short discussions with many people. 4 Institute for Security Studies CQ No. 31 - Mar 2010 3/17/10 3:19 PM Page 8 THE OVERALL PICTURE The estimation exercise suggested that provincial governments allocated a total of close on R7 billion in 2009/10 on account of alcohol abuse, while national government allocated more than R10 billion. Against these amounts, we can offset the revenue gained through value-added tax on alcohol sales and liquor licenses to arrive at a net expenditure of more than R1 billion. The national and provincial departments of health accounted for R6 804,8m (35 per cent of total expenditure), with Safety and Security following close behind at R5 807,5m (34 per cent of the total). Correctional Services was the third largest contributor, at R3 355m (20 per cent of the total). SECTORAL ESTIMATES Safety and security For Safety and Security we included sub- programmes that involve active policing and investigation. We included the forensic science laboratory, despite it being a support service, because of the role it might play in linking particular criminal acts with alcohol. We drew on crime statistics patterns as the basis for the percentage attribution. These are different from patterns in respect of prisoners, as some crimes are more likely than others to result in imprisonment. Official police crime statistics for the period April 2008 to March 2009 show 36 per cent of reported crime falling in the category of contact crimes (crimes against the person), 7 per cent in the category contact-related (arson and malicious damage to property), 25 per cent property-related, 9 per cent crime 'heavily dependent on police action for detection', 22 per cent 'other serious crime' and 2 per cent 'other crime categories'. Within the category of crimes heavily dependent on police action for detection, are driving under the influence of alcohol or drugs (4 per cent of the overall total), and drug-related crime (4 per cent of the total). The category 'other crime categories' includes culpable homicide and neglect and ill treatment of children, while 'other serious crime' includes theft, commercial crime and shoplifting.2 Anecdotally, it seems that police attribute exceptionally high proportions of crime to alcohol abuse. Thus in a workshop with 20 priority precinct station commanders in the Western Cape during early 2009, participants agreed that at least 90 per cent of violent crimes were linked to alcohol.3 We assumed that contact crimes, crimes heavily dependent on police action for detection, and 'other crime categories' are most likely to have a link to alcohol abuse. For contact crimes, which include murder and sexual offences, we assumed 22,5 per cent are related to alcohol abuse. We based this proportion on the 55 per cent of cases in which murder accused are found to test positive for alcohol in the blood, but halved this percentage to acknowledge that alcohol would not have been the cause of the crime in all these instances. For crimes heavily dependent on police action for detection we assumed 50 per cent were related to alcohol abuse. For 'other crime categories' we assumed the same percentage as for contact crimes. Assuming further that each reported crime takes up the same amount of police resources, we then allocated 24 per cent of the police allocations for alcohol. This final assumption is unrealistic as one assumes that more resources will be used on a murder case than on one of shoplifting. However, we do not have any information on which to base a differential weighting of the different types of cases. SA Crime Quarterly no 31 • March 2010 5 Prog 2: Visible policing 2.1 Crime prevention 17 269,4 Prog 3: Detective services 3.1 Crime investigation 6 430,4 3.3 Forensic science laboratory 498,0 Table 1: Allocations by safety and security (Rm) Sub-programme Allocation After applying the relevant percentages discussed above, the figure attributable to alcohol abuse in the national safety and security budget for 2009/10 would be R5 807,4m. CQ No. 31 - Mar 2010 3/17/10 3:25 PM Page 9 6 Institute for Security Studies Community safety and social crime prevention At provincial level the departments responsible for safety and security have different names, and the programmes and sub-programmes are also diverse. The sub-programmes selected for the exercise are those that focus on social crimes, which we assumed to have a particular focus on violence against women and children. Community liaison and related sub-programmes also cover community police forums (CPFs) and the bulk of the funds allocated for these sub-programmes might well go towards this purpose. We included the sub-programmes because some of the CPF work will focus on social crime prevention, but the proportion is probably relatively small. Donson & Marais, in their study of injuries reported to three rural hospitals in the Western Cape, find that 70 per cent of injuries sustained through intimate partner violence were reported to be alcohol-related, as were two thirds of injuries from male-on-female violence more generally.4 (Male-on-male violence accounts for the overwhelming majority of violence discussed elsewhere in this report in respect of murder and assault). We assumed a proportion of 35 per cent for these sub-programmes in acknowledgement of the fact that the focus is not solely on violence against women and children, and that other crimes might well have some link to alcohol. For the community liaison sub-programmes we allocated a much smaller two per cent to acknowledge that there is likely to be some focus, but not a significant one, within CPF work on alcohol abuse. The fact that few of the narratives in the provincial budget statements explicitly mention alcohol supports the assumption that there is not a significant focus on this issue. However, the Gauteng budget statement does note that the Ikhaya Lethembu sub-programme provides services for survivors of violence against women and children, and that activities include prevention programmes on violence against women and children, and men acting as a network of safety protectors. The Gauteng sub-programme also provides for schools-based initiatives on substance abuse prevention and promotion of substance abuse prevention through youth desks. The KwaZulu- Natal statement refers to awareness campaigns on substance abuse and violence against women. Where a sub-programme is likely to deal with both social crime prevention and community liaison we allocated 18,5 per cent, half the sum of two per cent and 35 per cent. However, the names of the sub- programmes and the accompanying narratives often do not indicate at all clearly what a sub-programme covers. The proportions are thus 'best guesses'. The total figure attributable to alcohol abuse in the provincial community safety budgets was estimated at R43,9m. Justice and constitutional development For justice and constitutional development we included the lower courts' sub-programme, as general crimes are tried in these courts. Leggett reports that among arrestees in three metros of South Africa, nearly a quarter were arrested for property crimes, 20 per cent for violent crimes, 16 per cent for substance-related crimes, and 40 per cent for 'other' crimes.5 Illegal immigrants accounted for a large proportion of the last-named category. Half of the arrestees who were surveyed reported that they used alcohol or tobacco. Close on 20 per cent reported that they were under the influence of alcohol at the time of the alleged offence, but Leggett suggests that this might be an exaggeration as some might have claimed this to explain their behaviour. We used a proportion of ten per cent in acknowledgement of possible exaggeration as well as the fact that being under the influence of alcohol was not necessarily the cause of their committing the crime. The public prosecutions sub-programme under the National Prosecuting Authority is included because the focus is on sexual offences and community affairs alongside priority crimes and specialised commercial crime. This sub-programme is, however, also responsible for other issues. Thus the 2009/10 budget document states that in 2007/08 the service finalised 1 043 857 cases, including CQ No. 31 - Mar 2010 3/17/10 3:19 PM Page 10 SA Crime Quarterly no 31 • March 2010 7 diversions, while the regional courts dedicated to sexual offences finalised 'only' 4 365 cases, the Specialised Commercial Crime Unit finalised 3 031 cases, the Directorate of Special Operations resolved 178 investigations and the Asset Forfeiture Unit finalised 223 cases. We assumed, somewhat arbitrarily, that 20 per cent of the resources are spent on sexual offences and community affairs, and applied the 25 per cent used previously for victim empowerment to this 20 per cent, getting a resultant proportion of five per cent for this sub-programme. The witness protection programme is included because those covered include witnesses for domestic violence cases. We used a lower percentage of two per cent here on the assumption that only a small proportion of those in this programme are involved in domestic violence cases. probation and parole. We also include the allocation for the public-private partnership whereby private companies provide and manage prisons for the department on the basis that the service provided is similar to that in a programme such as corrections. All offenders at the two private prisons reportedly have risk assessment forms that indicate the role that alcohol abuse played in the crimes they committed. Unfortunately, however, these forms have not to date been analysed to give an overall profile.6 According to the Estimates of National Expenditure, virtually all sub-programme allocations are based on staff and inmate population. If we base the proportion attributed to alcohol on the proportion of inmates thought to be incarcerated on account of crimes related to alcohol abuse, then the same proportion could arguably be used for all the programmes listed below, as well as for the sub-programme related to prisons run by private companies through public- private partnerships. We use the same logic and approach described above in respect of crime prevention services under social development, but apply the calculations to all prisoners, whether already sentenced or not. The calculation yields an attributable proportion of 38,5 per cent. One might want to include some amount reflecting expenditure incurred in initiatives to help inmates overcome their addiction. Knowledgeable informants felt that such initiatives were rare. When respondents in Muntingh's small qualitative study of ex-prisoners of Western Cape prisons were asked if they participated in any programmes while they were imprisoned, the pre-release programme run by the Department of Correctional Services was the most frequently identified programme.7 Eight ex-prisoners said that they attended drug abuse programmes and three reported attending Alcoholics Anonymous. It is not clear if this occurred while they were incarcerated. According to the department, all substance abuse programmes for imprisoned offenders form part of other programmes and do not have a separate budget allocation.8 The Judicial Prog 2: Court services 2.5 Lower courts 2 559,3 Prog 4: National Prosecuting Authority 4.1 Public prosecutions 1 532,8 4.2 Witness protection programme 130,1 Table 2: Allocations by justice and constitutional development (Rm) Sub-programme Allocation After applying the relevant percentages discussed above, the figure attributable to alcohol abuse in the national justice and constitutional development budget for 2009/10 would be R335,2m. Correctional services In the correctional services budget, unlike that of other departments, most programmes consist of only one sub-programme. The corrections programme involves the accommodation of inmates. Care provides for inmates' welfare, including anti-retrovirals and moral regeneration programmes for staff and inmates. Development provides for the education and training of inmates. Social integration aims to prepare inmates for reintegration into ordinary society through CQ No. 31 - Mar 2010 3/17/10 3:19 PM Page 11 8 Institute for Security Studies Inspectorate plans to conduct a full audit of all programmes on offer in correction centres over the next five months, and this audit should provide better information for future estimates.9 programmes addressing alcohol abuse within prisons, and the number of prisoners attending such programmes • Traffic management: The proportions of offences and fines related to alcohol abuse, and initiatives undertaken to combat alcohol abuse • Community safety: Initiatives undertaken by community policing forums and similar initiatives to combat alcohol abuse • Disability: The extent to which disability, as reflected in disability grant recipients and among those benefiting from social services, is linked to alcohol abuse. Finally, we note that the estimations were based as much as possible on credible research. The aim was not to arrive at a figure that was as large and 'shocking' as possible. Shocking figures attract media attention, but they are too easily discredited, bringing other related research and advocacy into question. The problem of alcohol abuse is serious enough in South Africa – we do not need to exaggerate to show the importance of addressing the problem. To comment on this article visit http://www.issafrica.org/sacq.php NOTES 1 E Single, D Collins, B Easton, H Harwood, H Lapsley, P Kopp & E Wilson, International Guidelines for Estimating the Costs of Substance Abuse-2001 Edition, Geneva: World Health Organisation, 2003. 2 Available at http://www.issafrica.org/dynamic/ administration/file_manager/file_links/0909CRIMETOTA LS.PDF?link_id=24&slink_id=8302&link_type=12&slink_ type=13&tmpl_id=3 (accessed 2 November 2009). 3 Barbara Holtmann, personal communication, 25 October 2009. 4 H Donson (ed), A profile of fatal injuries in South Africa 2007. Pretoria: South African Medical Research Council & University of South Africa, 2008. 5 T Leggett (ed), Drugs and Crime in South Africa: A study in three cities. Monograph No 69. Pretoria: Institute for Security Studies, March 2002. 6 Louina le Roux: personal communication, 9 November 2009. 7 L Muntingh, Prisoner Re-entry in Cape Town – an exploratory study. Civil Society Prison Reform Initiative Research Paper No. 14. Cape Town: Community Law Centre, 2008. 8 Hennie Human: personal communication, 11 November 2009. 9 Gideon Morris: personal communication, 2 November 2009. Prog 2: Security 4 425,3 Prog 3: Corrections 1 112,1 Prog 4: Care 1 591,8 Prog 5: Development 448,7 Prog 6: Social reintegration 425,9 Prog 7: Facilities 7.1 Public private partnership prisons 710,4 Table 3: Allocations by correctional services (Rm) Sub-programme Allocation After applying the relevant percentages discussed above, the figure attributable to alcohol abuse in the national correctional services budget for 2009/10 would be R3 355m. GAPS AND RECOMMENDATIONS This article highlights some of the findings of the first known attempt in South Africa to arrive at an overall estimate of the costs incurred by government on account of alcohol abuse. The estimate is made up of a large number of sub- estimates, some of which are, unfortunately, based on assumptions that could have a wide margin of error. It will never be possible to arrive at an exact estimate, given that most of the expenditure occurs within general allocations rather than being specifically targeted at alcohol abuse. There are, however, several areas where there is weak information and which seem to merit further investigation, given their potential importance in terms of numbers involved as well as opportunities for expanding on what is currently being done. The areas include: • Employee assistance programmes for civil servants: The amounts allocated for these programmes, and the extent to which they address alcohol abuse • Alcohol and the criminal justice system: The extent to which crime can be attributed to alcohol abuse, the amounts allocated to CQ No. 31 - Mar 2010 3/17/10 3:19 PM Page 12