SA CRIME QUARTERLY No 18 DECEMBER 2006 17 S ince its transition to democracy, South Africa has gained a reputation as a dangerous country with one of the highest rates of violent crime in the world. But, as Antony Altbeker has observed, attempts to test the veracity of this common wisdom break down due to the fact that few other developing countries record and report reliable crime statistics.1 And among those that do, different legal definitions of crime, and differences in reporting and recording rates limit the usefulness of cross-national comparisons. Victimisation surveys offer one alternative, but of the 70 countries in which the International Crime Victims Survey (ICVS) is run, many – especially developing countries – only conduct surveys in their major cities. The most recent series of nationally representative data consists of 22 countries surveyed as far apart as 1991 and 2003, of which South Africa is the only non-western country.2 The Afrobarometer, a regular survey of public attitudes toward governance, democracy and economics in 18 African countries, offers a useful alternative with which to begin to place South Africa in a more meaningful African context (see www.afrobarometer.org). Snapshot of the Afrobarometer results The Afrobarometer contains a set of questions that tell us not only about people’s experiences of crime, but also about their experiences with the police and their evaluations of government performance on crime. To be sure, the Afrobarometer did not design its questions in order to measure the actual crime rate. Rather, its purpose has been to assess the extent to which people have been either directly or indirectly affected by crime and then to measure the extent to which these experiences affect other attitudes toward government and democracy. But because it asks the same question in each country and generally across time, useful inferences can be made about the relative magnitude of crime across these countries, if not the absolute level of crime, and associated trends. Robert Mattes, Centre for Social Science Research University of Cape Town Rmattes@commerce.uct.ac.za HOW DOES SA COMPARE? Experiences of crime and policing in an African context Afrobarometer’s regular surveys of public attitudes toward governance, democracy and economics in 18 African countries shows that experiences of crime and concerns over safety in South Africa are indeed quite prevalent, but are by no means exceptional. People in some African countries are as, or even more, fearful than South Africans, and there are several countries in which people confront crime more frequently than do South Africans. The results also show that the South African Police Service, despite having higher levels of physical and human capital than its counterparts to the north, often lags well behind in terms of transparency and community relations. Figure 1: Feared crime in own home in past year, 2005/06 Always Many times Several times Just once or twice And when these cross-African comparisons are made, it becomes apparent that levels of reported experience with crime and concerns over safety in South Africa are indeed quite high, but are by no means exceptional. People in some African countries are as, or even more, fearful, and more likely to see crime as a major problem than South Africans. Indeed, there are several countries in Africa in which people confront crime as, or even more frequently, than do South Africans. Common wisdom about South Africa creates expectations that South Africans would be highly critical of government efforts to fight crime. Equally, common wisdom about ‘Africa’ leads to expectations of uniformly higher levels of corruption and bureaucratic intransigence among the police forces elsewhere on the continent. However, the Afrobarometer results demonstrate that, while the South African Police Service is embedded within a relatively wealthy and modern state infrastructure and may have far higher levels of physical and human capital than its counterparts to the north, it often lags well behind many of them in terms of transparency and community relations, as well as other key but less tangible assets of popular trust and legitimacy. Finally, the trends over time revealed by responses in those countries that have undergone three 18 MATTES Afrobarometer surveys, suggest that any positive trends in South Africa may be less a result of improved policing by the SAPS, than a function of the much broader dynamics of a society moving beyond the massive social dislocations associated with political democratisation and economic liberalisation. Public insecurity Compared to other Africans, South Africans exhibit relatively high levels of fear, but by no means the highest. Just fewer than one half of all South Africans (49%) told interviewers in January and February 2006 that they had feared crime in their own homes at least once in the previous 12 months (Figure 1). Yet this number is statistically indistinguishable from the 46% of Basotho who said the same, and both numbers lagged far behind the almost six-in-ten Zambians and Kenyans who feared the same thing (59% each). And while just 40% of South Africans said that people were safer from crime and violence than “a few years ago,” respondents in at least four other countries were even more pessimistic: Zambia (33%), Cabo Verde (30%), Nigeria (30%) and Zimbabwe (22%) (Figure 2). South Africans are, however, the most likely respondents to list crime as one of the most important problems facing the country that government ought to address (22% in the last survey), and have been so since the first Afrobarometer surveys in 1999-2001. SA CRIME QUARTERLY No 18 DECEMBER 2006 0 10 20 30 40 50 60 Se n eg al M al i M ad ag as ca r M o za m b iq u e G h an a Ta n za n ia B o ts w an a So u th A fr ic a K en ya M al aw i C ap e V er d e B en in N ig er ia N am ib ia U ga n d a Le so th o Z am b ia % Figure 2: Improved safety from crime and violence compared to “a few years ago”, 2005/06 SA CRIME QUARTERLY No 18 DECEMBER 2006 19MATTES Exposure to crime The Afrobarometer data also suggest that South Africans’ experiences with crime in the form of theft and assault, while high, are lower than those endured by respondents in several other African countries. While almost one third of all South Africans (31%) say that they or someone in their family had “something stolen from [your] house” at least once in the previous 12 months, this only fell near the middle of the distribution of the 18 Afrobarometer countries and well below the 46% of Zambians who reported this (Figure 3). Sixteen percent of South Africans told Afrobarometer that they or someone in their family had been “physically attacked” in the previous year (Figure 4). But again, this was essentially the same rate as that reported in a number of African countries like Zambia, Uganda, Kenya and Nigeria, and well below the 27% recorded in Namibia. These reported levels of crime are far, far higher than those indicated by official crime statistics, and generally somewhat higher than those uncovered by victimisation surveys. One reason for this is that many forms of theft and personal attack are never reported to the police, especially if there is no compelling reason (such as an insurance claim in the case of theft) to do so. Many victims are pressured not to report attacks (especially when perpetrators are spouses or family members), or the incident is ‘sorted out’ by police on the spot and no complaint is filed. Another reason, especially for the smaller differences with victimisation surveys, is the somewhat more ambiguous phrasing of the Afrobarometer questions, both in terms of the definition of the crime and the inclusion of “anyone in your family” in the wording. As noted above, the original purpose was not to measure crime rates with any precision, but to investigate the attitudinal and behavioural consequences of being directly or indirectly affected by crime. Government officials have recently questioned survey results that show continuing high levels of insecurity in South Africa even as officially recorded crime trends rates are coming down. But as Altbeker has pointed out, there is no reason to expect a one- to-one relationship, since current levels of insecurity reflect the cumulative effects of personal experiences with crime over the past several years.3 Even so, the Afrobarometer survey found that those who reported being fearful were much more likely to have personally experienced theft or assault in the previous year.4 But levels of insecurity and concern over crime go well beyond direct or indirect personal experiences, and include a larger ‘socio- tropic’ concern with the national crime situation as Much better Better 0 10 20 30 40 50 70 60 U ga n d a Se n eg al Ta n za n ia M ad ag as ca r G h an a M o za m b iq u e M al aw i Z am b ia M al i Le so th o B en in B o ts w an a N am ib ia K en ya So u th A fr ic a C ap e V er d e N ig er ia Z im b ab w e % Figure 3: Something stolen, at least once, from house in past year, 2005/06 reported through the news media. Indeed, we find that national levels of fear and prioritisation of crime are both strongly related to the most visible form of crime – murder (at least for the African countries for which murder rates are publicly available).5 Public impressions of law enforcement What kind of policing do South Africans receive, compared to other Africans? Given the large amount of public attention focused on crime over the past decade, as noted above, public evaluations of government’s efforts to fight crime may well be quite negative. Yet, given the country’s relatively high level of development and the resources that have been poured into police infrastructure and training over the past decade, balanced against commonly held perceptions of levels of incompetence and corruption in other African states, one might as easily expect South Africans’ evaluations of the police to be the best in Africa. SA CRIME QUARTERLY No 18 DECEMBER 200620 MATTES The truth appears to be somewhere in between. South Africans do indeed have some of the worst views on the continent of their government’s efforts to fight crime. While popular majorities in ten Afrobarometer countries said their government was handling crime “well” or “very well” in the 2005- 2006 surveys, just 35% of South Africans were of that opinion, roughly the same number as the Nigerians (33%) and Zimbabweans (30%). Only Cabo Verdeans (23%) were more negative (Figure 5). On some aspects of police performance, South Africans are quite positive. For instance, respondents were asked “How likely do you think it would be that the authorities could enforce the law if a top government official committed a serious crime?” Undoubtedly reflecting the arrests and prosecutions of people like Jacob Zuma, Schabir Schaik and Tony Yengeni, 64% of South Africans think it is “likely” or “very likely” that the police (and prosecutors) would 0 10 20 30 40 50 M ad ag as ca r Ta n za n ia Le so th o B en in M al aw i So u th A fr ic a K en ya Se n eg al Z am b ia C ap e V er d e M al i G h an a N ig er ia B o ts w an a M o za m b iq u e N am ib ia U ga n d a % s ay in g “a t le as t o n ce ” 14 17 22 24 26 26 27 27 28 31 31 33 35 36 38 38 46 Figure 4: Physically assaulted, at least once, in the past year, 2005/06 0 10 20 30 M al i Ta n za n ia M al aw i Le so th o M o za m b iq u e Se n eg al So u th A fr ic a K en ya N am ib ia M ad ag as ca r B en in C ap e V er d e G h an a B o ts w an a Z am b ia U ga n d a N ig er ia % s ay in g “a t le as t o n ce ” 3 5 6 6 6 9 9 12 12 14 14 15 16 17 17 19 27 SA CRIME QUARTERLY No 18 DECEMBER 2006 21 do their jobs and act against such a person (Figure 6). In contrast, just one quarter of Zimbabweans (25%), Beninois (26%) or Kenyans (27%) believe this would happen. On the other hand, respondents in Lesotho (72%) and Ghana (70%) were even more confident than South Africans that their police and prosecutors would act against top officials. Experiences with police However, South Africans’ evaluations of the police turn far more negative than those of other Africans once asked about personal experiences. Respondents were asked: “Based on your experience, how easy or difficult is it to obtain help from the police when you need it?” While it would not be unreasonable to expect that this is basic stuff for the police, just 40% of South Africans say that it is “easy” or “very easy”. Thirty percent say it is “difficult” and 19% say it is “very difficult” to get assistance from the police (Figure 7). These negative experiences are statistically indistinguishable from those of Cabo Verdeans, Zimbabweans, Mozambicans, Basotho and Malawians. But they lag far behind those of the Batswana, of whom 58% say dealing with the Botswana police is easy. One reason why South Africans say it is difficult to get help from the police is that they are often asked for bribes or favours. One in ten South Africans (10%) interviewed had to “pay a bribe, give a gift, or do a favour” to “avoid a problem with the police.” While a far higher percentage of Kenyans (29%), Nigerians and Zimbabweans (22%) had to pay bribes, South Africans’ rates of victimisation at the hands of the police were significantly higher than in places like Cabo Verde (1%), Botswana (2%), Malawi (2%) and Lesotho (3%) (Figure 8). Not surprisingly, then, 47% of adult South Africans think that “all” or “most” police officials are “involved in corruption,” a figure that is statistically the same as in Mali (49%) or Namibia (44%) (Figure 9). Again, while this figure is far lower than the massive 75% of Nigerians or 70% of Zambians who say their police are corrupt, the South African result is substantially higher than in eight other African countries. Finally, the data reveal that in 12 Afrobarometer countries, popular majorities (as many as 84% of Tanzanians) say they trust their police force “a lot” or “somewhat”. By comparison, just 48% of South Africans say they trust the police (Figure 10). A legitimate police force? While the police can compel citizens to obey the law and cooperate with them to help fight crime, effective police forces (as with all effective political institutions) ultimately rely on a bedrock of popular legitimacy – that is, the belief that one ought to obey the law and MATTES Very well Fairly well Figure 5: Approval of government’s crime fighting performance, 2005/06 U ga n d a G h an a Se n eg al M al i K en ya M al aw i Z am b ia So u th A fr ic a B o ts w an a Ta n za n ia M ad ag as ca r M o za m b iq u e Le so th o N am ib ia B en in N ig er ia Z im b ab w e C ap e V er d e 0 10 20 30 40 50 70 60 % Very likely Likely Figure 6: How likely are authorities to enforce the law against top officials? 2005/06 0 10 20 30 40 50 70 60 Le so th o C ap e V er d e So u th A fr ic a M ad ag as ca r Ta n za n ia U ga n d a N ig er ia K en ya G h an a M al aw i M o za m b iq u e Se n eg al N am ib ia M al i Z am b ia B en in Z im b ab w e B o ts w an a Very easy Easy Figure 7: How easy is it to obtain help from the police? 2005/06 0 10 20 30 40 50 60 B o ts w an a Le so th o Z im b ab w e N am ib ia Se n eg al U ga n d a Z am b ia K en ya M al aw i M o za m b iq u e C ap e V er d e Ta n za n ia M ad ag as ca r M al i G h an a N ig er ia B en in So u th A fr ic a SA CRIME QUARTERLY No 18 DECEMBER 200622 MATTES % % work with the police regardless of whether or not one actually approves of the law or likes the police officer in question. While many South Africans do see the police as legitimate, these numbers are relatively low compared to many other African countries. Sixty-six percent of respondents agree or strongly agree that “the police always have the right to make people obey the law” – the fourth lowest result across all Afrobarometer countries. By contrast, 90% of Ghanaians and over 80% of Basotho, Ugandans, Malians and Senegalese also accord this form of legitimacy to their police. Comparing trends: the pace of improvement One might take comfort in the fact that, while South Africans’ evaluations of the fight against crime and the performance of the police are lukewarm at best, virtually all these measures have been improving since 2000, albeit from quite low bases. It may even be tempting to see these improving perceptions – as well as the concomitant improvements in the official police crime statistics – as a reflection of increased resources for the police, improved training, and better policing strategies. However, a comparison of trends in 12 countries that have had at least three Afrobarometer surveys SA CRIME QUARTERLY No 18 DECEMBER 2006 23 since 1999 provides little evidence for such an interpretation. What it reveals is that improving attitudes about crime are not unique to South Africa. In fact, public attitudes have been steadily increasing across a range of dimensions related to crime and law enforcement (perceptions of safety, evaluations of government performance on crime, trust in the police and courts) in nine of the 12 countries surveyed. Namibia and Zimbabwe show no consistent positive or negative movements across each of these dimensions, and only Nigeria shows consistent – and severe – declines. While further analysis is required to isolate the causes of these broadly positive cross-national and secular trends, one underlying reason might be that these countries are all in the process of moving beyond their interludes of transition from authoritarian states and command economies, to new political and economic regimes. These interregnums between old and new regimes brought massive social dislocation and attendant rises in crime. To be sure, only some of these countries have achieved fully-fledged democracies and market economies. But this does not mean that the others are locked in a prolonged, or permanent state of transition. Rather, they may be in the process of consolidating new hybrid regimes with real rules that serve to normalise political, economic and social life and thus reduce crime. Conclusion Lacking good, comparable official police statistics and victimisation surveys, it has been difficult to assess whether crime in South Africa is really more prevalent than anywhere else, and if it is, to what extent. Based on an alternative form of survey data provided by the Afrobarometer, it appears that South Africans’ levels of insecurity and exposure to crime are indeed quite high in comparison to other African countries. But they are by no means the highest. Moreover, such a comparison demonstrates that even with all the advantages provided by a relatively developed state and wealthy tax base, the South African police are often outperformed in the public eye by their counterparts elsewhere around the continent, most of whom have far less access to resources and training. What the SAPS may possess in terms of physical and human capital, it lacks in social capital. This strongly suggests that simply spending more money or throwing more officers into the fray will not bring down South Africa’s high levels of crime. Rather, the SAPS could do well to pay more attention Often A few times Once or twice Figure 8: How often people paid bribes or did favours to avoid a problem with the police, 2005/06 0 5 10 15 20 30 25 C ap e V er d e M al aw i Se n eg al M ad ag as ca r So u th A fr ic a Z am b ia M o za m b iq u e Z im b ab w e B o ts w an a Le so th o B en in Ta n za n ia N am ib ia G h an a U ga n d a N ig er ia K en ya M al i MATTES % SA CRIME QUARTERLY No 18 DECEMBER 200624 MATTES All of them Most of them 0 10 20 30 40 50 60 70 80 C ap e V er d e M al aw i M o za m b iq u e Le so th o So u th A fr ic a G h an a Z im b ab w e U ga n d a Se n eg al B o ts w an a M ad ag as ca r N am ib ia M al i B en in K en ya Z am b ia N ig er ia Ta n za n ia A lot Somewhat Figure 10: Trust in the police, 2005/06 0 10 20 30 40 50 70 60 80 Ta n za n ia Se n eg al M o za m b iq u e U ga n d a G h an a M ad ag as ca r B en in K en ya M al aw i M al i B o ts w an a N am ib ia C ap e V er d e So u th A fr ic a Z im b ab w e Z am b ia N ig er ia Le so th o Figure 9: Views about extent of police corruption, 2005/06 to the basics of creating a community-oriented police service that helps citizens rather than victimises them, and thus becomes a focal point for public trust and cooperation in the fight against crime. Endnotes 1 Antony Altbeker, Puzzling statistics: Is South Africa really the world’s crime capital?, SA Crime Quarterly No 11, March 2005. 2 Ibid. 3 A Altbeker, Change is fear, Mail & Guardian, 4 The relationship between fearing crime in your home and personal experience with crime is relatively strong (with home theft, Pearson’s r = .374, p. = .000, and with having been or knowing a family member who was assaulted is r = .268, p. =.000).. 5 African murder rates are obtained from A Altbeker, Puzzling statistics, op cit. % %