SA CRIME QUARTERLY No 11 MARCH 2005 33 T he analysis of cost is of obvious interest during any public policy debate, but what benefit does it offer us in South Africa in relation to crime policy? Cost analysis studies of a broad range are popular internationally and have come to be considered an integral part of the policy development processes of most Western governments. National ‘costs of crime’ studies have been undertaken in countries such as Australia1 and cost is considered to be a useful component in the overall analysis of crime as well as deciding how the state should respond to it. This article aims to provide a brief overview of cost analysis and its value in relation to the issue of crime and the policy making process in South Africa. Why try to calculate the cost of crime? For Cohen, the ability of the state to take care of its citizens can be maximised through reducing the costs of crime to society.2 He assumes here that studies of this nature enable this process. Similarly, the assumption underlying many crime-costing studies is that cost analysis can inject greater efficiency into policy making in relation to crime. Essentially, the idea is that by generating information of a specific nature, policy makers will be enabled to make more ‘rational’ choices, because their decisions will be based on empirical research. Notwithstanding the fact that these assumptions may be naïve in terms of the true nature of political decision making, they have spawned literally hundreds of studies internationally. It is interesting, therefore, that there are relatively few studies that set out to test these assumptions, and to document the nature of policy change that may be attributed to such studies. Types of studies Three broad categories of cost related studies are presented below, followed by a discussion of the value of each, specifically with the South African situation in mind. It should be noted that while the literature refers broadly to costing exercises as ‘cost- benefit’ studies,3 this is inaccurate because many studies do not include a focus on the ‘benefit’ side of the equation. Accounting studies and cost analysis This first group of studies are essentially accounting exercises that seek to represent the total monetary costs of crime to a defined area or group. However, because no study is able to calculate the costs related to all categories of crime, choices must be made about which specific crimes to examine. Often these choices are influenced by what information is available and the extent to which this is reliable. Cheryl Frank Open Society Foundation – South Africa cdfrank@worldonline.co.za COSTING CRIME What value is there for policy making? Studies that purport to count the cost of crime have wide appeal in the private and public sectors, and in the media. Information on the cost of a particular problem and its solution can no doubt assist decision makers. But in the case of crime, assessing the ‘cost’ is so fraught with difficulty that the results hardly seem worth the effort. Some kinds of ‘cost of crime’ studies are more beneficial to the policy process than others, and the findings need to be used with great care. SA CRIME QUARTERLY No 11 MARCH 200534 FRANK The reliability of baseline data is another key problem. In attempting to measure the costs of particular crime categories, the ‘real’ rates of these crimes must first be established. The studies are therefore dependent on the existence of a reasonable reliable set of data e.g. official crime statistics or victimisation surveys. It has been acknowledged that “the empirical basis for assessing costs is weak in many cases”,4 due to the difficulty of estimating actual crime levels. This is rendered even more difficult given the many crime categories that are problematic in terms of underreporting (e.g. rape, domestic violence and child abuse). Another challenge is the range of costs associated with a specific crime, and that it is simply not possible to count all these costs. The costs associated with a criminal incident may be either direct or indirect. The direct costs of a car-hijacking could include those related to the replacement of a stolen vehicle, the medical costs related to direct injuries, and the costs incurred by the criminal justice system when responding. Indirect costs could include a heightened fear of crime, the costs related to using a different route, and the loss of work time (and income) from attending court proceedings. A further consideration is whether costs are tangible or intangible. This difference is critical methodologically, and leads to one of the more controversial problems faced by studies of this nature. In relation to a rape, for example, the tangible costs would include: the costs of the physical injury (most often measured in terms of medical costs), the costs of the criminal justice processes applied to the crime incident, etc. The tangibility of these costs makes it possible to identify and measure them with relative ease. Other costs are, however, significantly intangible. Examples include the immediate emotional trauma experienced by the victim, the ongoing trauma and fear that may result from the incident (possibly emerging some time after the event and lasting over an indefinable period), and the overall emotional and behavioural impact for the victim e.g. avoiding certain areas, avoiding certain kinds of work, etc. These are impossible to quantify in monetary terms, and both ideological and methodological problems arise in attempting to do so. Various strategies have been employed to represent these intangible costs, but they all remain fundamentally flawed and are a source of particular discomfort for social scientists and social service practitioners. Overall, this endeavour is loaded with problems, and Mayhew notes that “intangible costs are subject to great uncertainty, and vary considerably under competing costing methodologies”.5 The main question is what utility do such studies offer? There is no doubt that, given the high levels of crime in South Africa, an accounting study of this nature would produce a high and shocking bottom- line rand figure. However, such a number is likely to restate the problem rather than offer any solutions, and as such its utility is questionable. At best, such a figure may offer a useful media sound bite, and at worst, avail itself to misuse by opportunistic role players in the policy process.6 Overall, the problems of measurement are significant enough to warrant extreme caution in relation to studies of this nature. This is especially so in countries like South Africa where resources are limited and ought to employed in ways that serve the delivery agenda as directly as possible. It is probably more strategic to obtain cost related information in other ways. For example, victimisation and cost questions could be embedded into household surveys undertaken by organisations such as Statistics South Africa and the Human Sciences Research Council, and cost related questions could be included in the victimisation surveys undertaken by the Institute for Security Studies. Cost-benefit and cost-effectiveness studies These studies differ from the first category on the basis of their comparative orientation. More specifically, they aim to compare the costs of inputs with outputs or outcomes (in terms of benefits). Essentially, these studies attempt to measure the impact of interventions, and whether the financial, human resource and other inputs that were made amount to an effective employment of resources. These studies are also used to compare the relative impact of two different kinds of interventions. Such SA CRIME QUARTERLY No 11 MARCH 2005 35FRANK efforts can produce valuable information for policy making and operational management. This would enable us to understand, for example, that one pattern of police patrolling could prevent one rape, while another could prevent three burglaries. This is more useful data, but how this information is ultimately used cannot be controlled. For example, in a context that is dominated by the interests of business, it may be considered more important to prevent the three burglaries (into business premises) than the one rape in the neighbouring residential area. These studies consider the details of programme inputs, delivery and outcomes. The ability to clearly define the impact of the intervention is critical, and evaluations depend on the availability of accurate data on the programme’s impact. Such information, especially in relation to crime prevention, is weak and evaluation is still not considered to be an important aspect of programme implementation. This use of cost-benefit analyses can, theoretically, also enable more reasoned reflection on policy questions of a broader nature, such as whether a country should invest in social crime prevention or the criminal justice system. The choice to invest in diversion programmes for young offenders rather than prosecuting them is critical, and these studies present an opportunity to encourage this kind of reflection. Costing new policy and legislation This category of cost related activity warrants particular attention in South Africa. It involves projecting the costs of draft policies and legislation, and a good example is the costing of the Child Justice Bill.7 While this kind of costing is used in many countries as a standard part of the policy and law making process, overall, its value for policy making and analysis has received much less attention than the accounting studies discussed earlier. It seems that this approach is best applied as an integrated part of a structured policy development process, rather than an afterthought as is currently the case in South Africa. Studies of this nature can best serve the policy process when they allow policy makers to juxtapose a range of policy choices, their possible outcomes and their projected costs. Overall, this approach can play an important function by offering information on the viability of a particular policy choice. Cost projections may also serve poorer countries by enabling the progressive realisation of particular services over time. One of the dangers, however, is that cost may become the overwhelming concern for policy makers, and may override other more important considerations such human rights. It is critical, therefore, that a policy process is rooted in a strong human rights paradigm, and that procedures exist to prevent policy makers from choosing the cheapest option over that which may be the ‘best’. South African crime-costing efforts South Africa has had an interesting history with costing exercises, although there has been no attempt to undertake a comprehensive national costing of crime. The most significant work has been done in the public health sector, where attempts have been made to estimate the costs of injury, some as a result of crime. This work has ranged from national to very localised studies. Methodologies that have been used are modelled on international public health techniques. Overall, however, there has been no national assessment of the economic burden of injury. It should also be noted that these studies have predominantly measured direct costs, with almost no focus at all on indirect and intangible costs. These are discussed in some detail by Bowman and Stevens.8 Other challenges Costing is an ideologically loaded endeavour While the origins of some kinds of costing in World Bank structural adjustment programmes point to ideological questions,9 it is also true that the reduction of (often traumatic) human experience to a monetary value may be very uncomfortable for human rights or crime prevention activists. More specifically, engaging policy makers on issues of cost may detract from the fundamentally humanistic origins of many policy positions. While cost analysis remains a natural and necessary SA CRIME QUARTERLY No 11 MARCH 200536 FRANK component of management in the public sphere, it needs to be done with due consideration for a great many other factors, not least of which is the human rights imperative. What the studies do not tell us One gap in the information is in the area of the criminal economy and the ways in which this may relate to the issues of cost and investment. Other complicated questions that these kinds of studies are unable to answer relate to economic under-investment as a result of crime, the reduced productivity of individuals, businesses and government, lowering of labour force participation and costs associated with ‘victimless’ crimes such as gambling, drug crimes, and prostitution. Most importantly, ‘hidden crimes’ such as corruption, fraud and sexual offences cannot be easily quantified as there are no methodologies that produce reliable estimates.10 These studies also do not include information about the costs of the impact of the criminal justice system. Examples include the impact of imprisonment on the offender, his or her family and society, or the costs of long periods of time awaiting trial. How the results are used What value then do these assessments offer? Pure accounting studies offer relatively little value for policy analysis and should be used with great caution given some of the methodological problems. The question of how cost analyses are used raises the related issue of how empirical research findings are viewed and used within policy processes, both to develop new policy and evaluate existing initiatives. Perhaps the paucity of local research on crime and crime prevention has neither fuelled public debate on these issues nor assisted in the processes of policy development and review. This is linked to the weaknesses in policy processes as well: while policy making continues, the monitoring and evaluation of the impact of these policies is both ad hoc and superficial. From a civil society perspective, cost analysis helps to analyse the actions and impact of government, especially in terms of comparing policies to budgets, and budgeted spending to actual spending. There have been several developments in South Africa over the past few years, such as the introduction of the Public Finance Management Act,11 that indicate the development of structures of accountability in government service delivery, and these can promote internal monitoring as well as facilitate external oversight. Overall, cost analysis in relation to crime is one of a range of tools available for enhancing the policy development and monitoring process. It is best employed among a range of other tools. In relation to crime, cost-benefit and cost-effectiveness studies that have a comparative objective are more useful for policy making and review as well as operational decision making, and may offer a more valuable investment than merely counting the costs of crime. This article is based on a discussion paper written by the author for the Alliance for Crime Prevention, August 2004. Endnotes 1 P Mayhew, Counting the Costs of Crime in Australia, Australian Institute of Criminology, 2003, 2 M Cohen, The Cost of Crime: Methods and Findings from Past & Recent Research, European Seminar on Costs and the Distribution of Costs of Crime and Disorder and Crime Prevention, Helsinki, Finland, 1-2 April 2004. 3 For example, see MA Cohen, Measuring the Costs and Benefits of Crime and Justice, Criminal Justice 2000, Vol 4, 2000, pp 264 – 315. 4 P Mayhew, op cit, p 1. 5 Ibid, p 2. 6 MA Cohen, 2000, op cit. 7 C Barberton with J Stuart, Re-Costing of the Child Justice Bill: updating the original costing taking into consideration changes made to the Bill, Afrec, University of Cape Town, 2001. 8 B Bowman and G Stevens, in S Suffla and A van Niekerk (eds), Crime, Violence and Injury Prevention in South Africa: Developments and Challenges, Pretoria, UNISA and the Medical Research Council, 2004. 9 Ibid. 10 E Shapiro, Cost of Crime: A Review of the Research Studies, Information brief for Minnesota House of Representatives, 1999. 11 The Public Finance Management Act, Act No 1 of 1999.