Figuring Crime

Quantifacts and the Production of the Unreal

Publication Date: 01/04/2006

Journal: Public Culture

Reissue Date:

Page start: 209

Page End: 246

Volume: 1

Edition: 18

Publisher City: Raleigh Durham

Nothing rings with more authority to South African ears than a crime statistic. It is the music of our spheres: what the sound of an accordion is to a Marseilles sailor, or the jaunty plinkety-plunk of the banjo to a Louisiana woodsman, so is the rhythmic, measured refrain of a crime statistic to a South African.

Darrel Bristow-Bovey, 20001 In an address to the national parliament in 1999, Nelson Mandela voiced an old cliche, a deep truth from both the heart of modernism and the fraught history of numbers in South Africa. “Figures,” he said, “are meaningless in the context of people’s concrete ex- periences.”2 But, in the contemporary world, the opposite may be true: that figures render large abstractions concretely meaningful to personal experience, speaking with authority about the connection of human beings to otherwise incomprehensible phenomena. Being assertions of the real, they fill the space between the unknowable and the axiomatic, imagi- nation and anxiety. Viewed thus, the statistic is a medium of communication and a species of commodified knowledge, one whose value and veracity accumulates as it circulates.

Part fetish, it has also become a term in the ordinary language of being.
The rise of contemporary Western perceptions of society, Ian Hacking (1990:1-5) has famously argued, was closely tied to the “avalanche of numbers” produced, publicized, and deployed for purposes of governance by nineteenth-century states (cf. Canguilhem 1989). The obsession with counting, and with calculating probability, he suggests, had pro- found epistemic effects. For one thing, “society” itself “became statistical.” For another, the appeal to law-like regularities began to replace other kinds of causal explanation, such as “human nature,” in making sense of, and acting upon, the world. Which, in turn, made rates of “deviancy” – of criminality, suicide, madness, illness – especially salient. To wit, post-en- lightenment ideas of the social, the moral, the normal, and the rational owe a lot to the crime statistic, a fact made evident in the actuarial underpinnings of much early detective fiction. Take, for example, Edgar Allen Poe, a fine organic anthropologist with an explicit interest in the “public mind.” In common with many writers of good mysteries, Mark Seltzer (2004:561) observes, Poe often invoked numbers in the interests of sociological realism; for him, “the death of God” left us with mathematics, “the death of Satan,” with forensics. The Victorian impetus to quantify deviance, in both realist fiction and social science, pre- supposed a rule-governed social order whose positive outlines were most visible in the negative: in lawlessness and vice. It was these “social pathologies” that would become the urgent object of sociology and social engineering (cf. Hacking 1990:118).

So much for Euromodernity, in which, from the first, the very idea of governance de- pended on statistics, the “science of [a] state” whose operations were enabled “by the ac- cumulation and tabulation of facts about the domain to be governed” (Rose and Miller 1992:185). Wherein lies the significance of crime statistics at the dawn of the twenty-first century? How do they figure, so to speak, in an age in which foundational assumptions ab- out society, citizenship, and order are called into question, in which social engineering is ever more suspect, in which the discourse of deviance is deeply discredited, in which go- vernment and public alike appear more concerned with personal security, suffering, rights, and risk than with social pathology – and in which the ontological status of “the” state is it- self is a matter of argument?

The Neo South Africa, like most nation-states today, produces its own avalanche of numbers. The tide of statistics made publically available by the police service is swelled, in this Age of Neoliberalism, by the ever more state-like exertions of non-governmental or- ganizations and the private sector. Those figures feed a thoroughly modernist “lust for precision” (Hacking 1990:5), a fervid faith in the panacea of probability, and a populist sen- se that countering disorder begins with counting it properly. It is hardly surprising, then, that the crime statistic has taken on unprecedented sovereignty in this postcolony. Not only has it become diagnostic of the national health. It is also a discursive currency by means of which government speaks to its subjects, citizens speak among themselves, experts speak to every-persons, everyone speaks back to government – and the media mediate all the incessant talk, adding their own inventions, inflections, inflations.

Three things are especially noteworthy about the rising sovereignty of the crime sta- tistic in this context. And about the pivotal place of quantifacts – statistical representations that make the world “factual” – in its public discourses. All of them are critical to our broad- er theoretical concerns here.

The first is the paradox of dis/trust: while crime statistics constitute a widely-cited measure of the condition of the postcolony, they tend also to be distrusted, due largely to their susceptibility to abuse. They are, in short, at once a fetish and the object of a lively hermeneutic of suspicion. The second has to do with alienation and intimacy: counter to the commonplace that numbers displace visceral experience into the realm of pattern and probability – vide Mandela’s observation – it is arguable that they do just the opposite here. As they circulate and are mediated, these statistics reduce a mass of faceless felonies, awful things that occurred elsewhere, into the objects of first-person affect: revulsion, revenge, pain. The third thing arises out of the phenomenology of figures: for all the ambivalence with which they are regarded, crime stats are treated here not as a representation of reality, but as a reality in themselves, a reality in which is congealed material facts, moral order, collective identities, even the quality of life (cf. Urla 1993).

In this excursion into the criminal anthropology of South Africa, then, we explore what, exactly, it is that crime statistics make real, how they take on a public life, by what means they convert the abstract into the intimate, tertiary knowledge into primary experien- ce, quantity into quality. And why it is that they have become so much more than the tools of criminologists and reformers, so pervasive a public passion, so deeply inscribed in nar- ratives of personal being, so vital to the construction of moral publics, so integral to deba- tes about the meaning of democracy, freedom, security, human rights. Conventionally fra- med as value-free information, these numbers appear to be taking on ever more political heft as the modernist state deregulates the functions of governance, as sovereignty is par- sed and privatized, as control over the means of violence is rendered ambiguous, as a cul- ture of “popular punitiveness” gains credence (Bottoms 1995; Haggerty 2001:197). As they do, and as citizens and communities claim more responsibility for their own welfare, modes of producing and deploying crime statistics themselves proliferate. Which – as we intimated a moment ago and shall see in exquisite detail – sets in train processes whose effects, of- ten unremarked, are deeply implicated in remaking the postcolonial nation-state, the nature of its governance, and citizenship within it. But we are running ahead of ourselves.

SOVEREIGN STATISTICS: the alchemy of numbers

Critical readings of the history of the industrial revolution and the rise of democracy return repeatedly to the dangerous alchemy of numbers. At work in processes of commodification and bureaucratization, they suggest, was the power of arithmetic to abstract value, to turn people into ciphers, to enable the alienation of humans from their essence and experience. Lefort (1988:18-19), for one, argues that the paradox of the modern idea of “society” is that it can never be made real: far from materializing “the people,” institutions like universal suffrage turn citizens into statistics as “[n]umber replaces substance.” Simmel (1978:297f., 444) thought differently. For him, money was the currency of counting. And while it served to reduce distinction and weaken personal ties, it also enabled the translation and commensuration of difference, giving rise to a society of morally interdepen- dent, self-sufficient persons in which individuals become real in relation to an impersonal mass. Novel modes of accounting, in other words, enable the production of new qualities, subjects, sensations. Even under the most rationalizing of conditions, the traffic between quality and quantity occurs constantly in the production of meaning and value.

The contentious life of the crime statistic in South Africa makes it plain that enume- ration is never a mere flight from substance: measurement is always mediated by historical conditions as numbers are made to signify in various ways. Some of those ways – the idi- om of the actuarial state, the principle of majority will – have long been integral to the scaf- folding of democracy, sui generis; shades here of de Tocqueville, Weber, and much of the history of liberalism. Others are more specific. Thus apartheid South Africa tried to legiti- mate itself by dis-counting, literally, large sections of the population; its opponents used statistics, from land distribution through poverty datum profiles to mortality rates, to argue against “unrepresentative” rule. More recently, activists striving to persuade the state of its responsibility toward the homeless and HIV/AIDS sufferers have advocated “emancipatory” enumeration (Robins 2003:259-60).3 Indeed, in the “new modernity,” Beck’s (1992) “risk society,” where personal destiny is read in figures, improving one’s chances requires imp- roving the odds (Crawford 2004: 522). Even in colloquial terms, the quality of life is calcula- ted in relation to the law of large numbers: “I want government to provide more security for people like me,” an elderly Mafikeng man told us in 1999, “before I too become a statistic.”

While the truth-value of numbers has been a focus of bitter debate between the “hard” and “soft” sciences, there have also long been efforts to grasp the statistic as a so- cial construction (Kitsuse and Cicourel 1963; Poovey 1998). One influential development of this concern has been inspired by Foucault’s analysis of enumeration as a means of go- vernance (Cohn 1987:224f.; Anderson 1991:163f.; Appadurai 1996); another emerges from science studies (Haggerty 2001:53: cf. Rose 1999), which have examined the produc- tion of various modes of calculation as “artifacts” yielded by the interplay of actors, institu- tions, and technologies (Latour and Woolgar 1979; Deleuze and Gauttari 1987). Each of these approaches offers insight into the nature of crime counts in post-apartheid South Af- rica; each has limitations. Thus we might ask, a la Foucault, why measures of lawlessness and victimhood have become so salient in defining populations here, overlaying class, race, and political disposition. Yet, pace Foucault, it is equally necessary to take note of a critical politics that has arisen in response to criminal accounting on the part of the state, a politics of the kind that challenges the Idea of “governmentality” as all-purpose social control. Similarly, in charting the tortuous life of numbers, there is much to be gained from tracing the manufacture and circulation of statistical artifacts in the science studies mode: in this instance, by following the argument that rages about official crime tallies, about their implications for the rights of citizens and the responsibilities of rule – itself a fractal drama in which government, the media, citizen-subjects, and other parties conjure the meaning of normality and emergency, state and nation. At the same time, as we shall see, this argument bears the imprint of larger forces than can be grasped from the perspective of scien- ce studies, with its stress on the contingent outcomes of actor networks.

But let us begin our excursion into the criminal anthropology of the “new” South Afri- ca with the numbers themselves. In so doing, we enter the concrete world of quantifacts inhabited by its citizens, into the paradox of dis/trust with which statistics of state are nego- tiated, into the intricately scaffolded reality that crime counts make of life here.

EXCURSIONS INTO THE UNREAL: counting crime in the postcolony

From within the country, crime rates are seen as anything from dire to catastrophic – and, by many, as deteriorating.4 An American study conducted in 2004 found that 81% of the population take lawlessness to be a “serious threat to democracy.”5 Well, how bad is it? How awful a picture is painted by the numbers? In order to answer the question, we suspend disbelief for a moment and draw a synoptic, quick-and-dirty portrait, courtesy of the Criminal Information Analysis Centre (CIAC) of the South African Police Service – with commentary and complementary counts from critical criminologists and non-governmental research organizations. Our figures are for 2002.6

Violent crime in South Africa is, by any measure, very high indeed. Almost 22,000 people were murdered in 2002, about 48 per 100,000 of the population. Even in the best of years since 1994, some 58 intentional homicides occurred on an “average” day. Add 35,012 attempted murders and 11,087 culpable (i.e. non-intentional) homicides, and the tally of dead and wounded bodies is considerable. Save, of course, next to those laid waste by AIDS, which, at last count,7 claimed 39% of all mortality in the country. By our calcula- tion, homicide, attempted homicide and culpable homicide have, together, yielded 590,098 victims since the transition – more or less 180 per day. Nor is this all the bad news. Three other serious crimes of violence are also noteworthy: in 2002, there were 119,185 recorded armed robberies, 264,399 grievous assaults, and 52,107 rapes. This last figure, important because sexual crimes are the object of particular national anguish, is a lot lower than some of the more dire ones that circulate as mythostats, according to which an incident oc- curs every 4, 11, 17, 35 or 36 seconds.8 The police aggregate is 16.81 times less than one every 36 seconds, the lowest of those faux figures; which means that 17 out of 18 would have to go unreported. The SAPS believes that it is one in three;9 independent victim sur- veys have it that “approximately half of all respondents who [suffer] rape report” it (Hirscho- witz, et al 2000:2). Still, the figure is frighteningly high. So is carjacking, at about 15,000 cases per annum, although they are counted under other offences by the police.

To what does all this add up? Over half a million (533,870) serious violent offences in 2002 – murders, attempted murders, culpable homicides, rapes, armed robberies, agg- ravated and indecent assaults, kidnapings, and abductions – in a land of 44.8 million souls. And this excludes the 315,623 residential burglaries and 256,593 common assaults. No wonder South Africa has a very anxious population. Vernacular realism here counts its truths, unsurprisingly, in the dramaturgy of damaged persons and seized property.

Looked at over space and time, though, these gross numbers take on a slightly dif- ferent cast. First space. Perhaps the most striking thing about crime in South Africa is how unevenly distributed it is. For example, almost two-thirds of all murders occur in just three provinces: KwaZulu-Natal, which has the highest incidence, Western Cape, which has the highest rate, and Gauteng, which is reputed to be the homicide capital of the nation, but actually is not. Hijacking is also heavily concentrated: Gauteng (115.4 per 100k) is a hund- red times worse than the least afflicted province and almost four times worse than KwaZulu-Natal, which comes second.10 These patterns of distribution are not always predictable. Take a figure that most South Africans would treat with incredulity: the most violent province in the nation, measured by rates of attempted murder, rape, and grievous assault. KwaZulu-Natal? No, it is the least. Gauteng or the Western Cape, those scenes of televisual mayhem? Wrong again. Limpopo or Mpumalanga, site of frequent witch purgings and vigilante deaths? No. It is the Northern Cape, usually seen as a sleepy, featureless flatland in the arid heart of the country.11

Patterns of variance become even yet marked as we move to micro-topographical levels. As the Commissioner of Police put it in his report for 2002/3 (see n.6), a huge dis- proportion of serious “contact” crimes occur in a few “township precincts”; specifically, those with high levels of “urbanization” and poverty, large informal settlements, and prolon- ged unemployment.12 This resonates with the received wisdom that, in a still segregated country, whites and Indians suffer attacks on their property, blacks and coloureds, attacks on their person. In fact, many of the latter also fall prey to robbery. And bloodshed does spill over into white and Indian neighborhoods. But, taken across the statistical terrain of South Africa, patterns of crime and victimhood are typically taken to follow lines of race and class; “extreme poverty and extreme risk,” notes Ulrich Beck (1992:41) tend everywhere to “attract” each other. Thus it is that, in December 2003, the Minister of Community Safety for the Western Cape – a portfolio that tells its own story about the contemporary imperati- ves of governance – assured the public that the “flashpoints for murder and rape” in the province “are Khayelitsha and Nyanga,”13 just the kinds of “township precinct” that the Commissioner had in mind. As if the containment of disorder to two of the poorest, most segregated parts of Cape Town should reassure everyone who does not live in them. Giv- en the proclivity to pin lawlessness on young black men – to situate it, that is, at the conflu- ence of race, gender, and generation – the “fact” that violence is highly enclaved plays into a mass-mediated view of “the” townships as breeding grounds of brutality. It also has the effect of erasing the implications of white collar crime, crimes for which whites tend to be collared, in producing the conditions for lawlessness in post-apartheid South Africa.

As with space, so with time. “Crime, and the fear of crime,” notes Jonny Steinberg (2001:2), “is as old as South Africa itself.” But shifts in the manner of its counting, them- selves deeply political, make the assessment of long term trends difficult (see below). No- netheless, according to the SAPS, most violent offences have either leveled off recently or are on the decrease, some quite steeply; murder, for instance, has dropped by 19% or so since 1994. Nor are the police alone in asserting this. A 2003 national victim survey, done by the independent Institute for Security Studies (ISS 2004), makes even more opti- mistic claims. Its figures – which, like all victim surveys, elicit a high degree of trust among those who traffic in crime stats – show that the only felony to have increased between 1998 and 2003 was housebreaking, and then only slightly. The study also turned up other telling quantifacts. Of the decreasing percentage of people (24.5% to 22.9%) who fell victim to crime, by far the majority were hit by property offences. None of the five most common fe- lonies involved force; only two of the top ten did, assault and robbery. A victimization rate of less than 1% was recorded for murder, sexual assault, and car jacking (ISS 2004:102f.).. Despite these trends, most South Africans – among whom there is an inverse rela- tionship between fear of crime and the “actual risk of violence” (Leggett 2004b:6)14 – belie- ve lawlessness to be on the rise; although profiles of terror are inflected by race and other social considerations (ISS 2004: 40f.). Perhaps because it measures a wide range of inse- curities in a world of changing power relations, few accept at face value information that calls into question the national trauma. For example, in 1998, The Star, a major daily, publi- shed a feature entitled, “The Surprising Truth About Crime.”15 It argued, on strong empirical grounds, that the country had become safer than it had been under apartheid; that, with some exceptions, lawlessness had not “increased drastically in former white suburbs”; that most categories of offence were on a long-term downward spiral. But “people get angry when you say it,” noted Benty Naude, a professor of criminology. “The public perception is wrong,” she added. “Crime is not really as high as people think.” The inflation of mass fear, the article went on to say, is owed to the fact that media reporting is incident, not inci- dence, driven; its stress on sensational acts of violence has the effect of generalizing the singular, thus to feed the impression that criminality has come to saturate the country. The Star predicted the manner in which all this would be read. A host of angry letters to the paper asserted that South Africa remained the most most crime-afflicted place on earth. Is it? How, criminally-speaking, does South Africa measure up to the world? A parenthetic comment first, though. A “common way of ratifying shared belief,” says Seltzer (2004:560), “is to share in [its] suspension”; Americans, he observes, affirm some- thing by saying the opposite, like “its really unbelievable.” This is no less true in South Afri- ca, where the terrifying reality of violence is materialized in words like “unreal.” But a collec- tive sense of selfhood in the face of disorder is also conjured up, here, by evoking the argot of exceptionalism. This, it is said, is the “murder capital” of the world. It has the highest rate of rape. And so on. Nor is this altogether fanciful. While international comparisons are no- toriously difficult, among countries that report their crime figures in more or less commen- surable ways – less than half of those on the planet – South Africa does seem to have among the worst rates of most violent felonies;16 although there are some surprises, like the fact that a recent UN report showed Estonia and Canada to have over double its per capita incidence of rape.17

On the other hand, South Africa ranks much lower for property offences. According to the Seventh United Nations Survey on Crime Trends (see n.16), it came only 18th, in 2000 for all categories of theft – behind many “developed” countries, including those to which scared South Africans have chosen to emigrate. Against its rate of 1,287.21, Austra- lia (3,514.65), New Zealand (3,313.25), England and Wales (3,257.52), and Canada (2,220.77) look far worse; so do the Netherlands (4,580.26), Norway (4,281.34), France (3,963.83), Germany (3,701.56), Denmark (3,633.73), Iceland (2,662.99), and Finland (2,207.71), all of them commonly seen as havens of peace, order, and civility. Similar pat- terns and rankings obtain for other property offences. In fact, on the United Nations “grand total of recorded crimes” for 2000, which includes violent offences, South Africa was 10th, after New Zealand, England and Wales, the USA, the Netherlands, and Canada. This rank is inflated, moreover: the “grand total” omits several nations, many of them thought to be crime-ridden, that did not submit figures to the UN. Over the long run, too, South Africa’s relative position seems to be improving, not least because felony rates have been going up fast across the world; note, in this connection, the argument, made by John Gray (1998) and others, that the downside of “the neoliberal project” is a destructive escalation in global incidences of lawlessness and imprisonment. Thus, between 1997 and 2001, robbery inc- reased on average by 24% in Europe and 128% in Japan; violent criminality, by 22% in the EU (50% in France, 35% in the Netherlands, and 26% in England and Wales) and 79% in Japan (Barclay and Tavares 2003:2). Against this, if we believe the SAPS and the 2003 victim survey, the South African picture looks less dire. But could this more positive, less exceptionalist gloss be a product of the fact that less crimes are brought to the attention of the police here than elsewhere? Not really. All the comparative evidence suggests that South Africans report crime as often as citizens in other countries (Louw 1998:12f.).

There is, however, a qualification to be made at this juncture. A decade ago, Lorrai- ne Glanz (1994:10) declared, peremptorily, that it is impossible to “know precisely how much crime takes place in South Africa, or anywhere else.” Glanz should know. Or, rather, not know. She was to become the Director of Crime Statistics in the post-apartheid Department of Justice. What she was pointing to, implicitly, is the fact that police statistics every- where are erected on an edifice of indeterminacies and impossibilities.

The most pervasive of those indeterminacies are pragmatic. Official figures are composited from cases either brought to the police by the public or turned up by their own investigations. As a result, they inevitably under-count the incidence of crime; in South Afri- ca, this is made worse by a historic mistrust of the law, by the belief that conviction rates are low, and by the uneven bureaucratic capacities of the SAPS. But felonies may also beover-reported. This is less remarked in South Africa, with its strong public predilection to stress the pathological. According to both the police commissioner and the press,18 there has been a palpable increase in false claims of late, largely for financial motives; bearing false witness to having suffered crime is itself becoming a major category of criminality. Nor is it confined to minor misdemeanors. It extends to robbery, hijacking, housebreaking, ass- ault, sexual violation, arson, even homicide.19 It is impossible, of course, to establish the precise extent of either under- or over-reporting. Together, they make the quantifactual world of crime impossible to plumb.

But indeterminacies inhere in other things as well. Some are purely actuarial, like the fact that crime rates are necessarily calculated against size of population – which, in South Africa, is affected by the presence of an unknown, if often wildly-guessed-at, number of “aliens.” Others lie in record-making itself. For example, the eventual disposition of felony figures often depends on whether a reported episode is coded as, say, attempted rape, sexual abuse, or aggravated assault, as murder or culpable homicide, kidnaping or abduction. A high level of arbitrariness is unavoidable. What is more in this respect, official statistics do not re-classify dockets once they have gone through the courts. Thus a death listed as murder that turns out to have been suicide, a car jacking or burglary or rape that morphs into a fraud case, an abduction that becomes a homicide; each of these will remain where they were originally tallied, there to disfigure official numbers in ways that are well-nigh impossible to ascertain – numbers that can, at best, paint a highly schematic, distorted picture of law and disorder in South Africa.

Yet another species of indeterminacy derives from a quite different source. Because crime statistics are the dual products of police work and public reporting, it follows that they will fluctuate in proportion to the efficacy of the SAPS and the trust placed in it by the popu- lation. Hence, paradoxically, rising crime rates may be less an indicator of rising crime than it is of increased confidence in, and the success of, policing; less, in other words, a function of lawbreaking than of law enforcement. Or, which is most likely, of a complicated dialectic of the two. But the disfiguration of the real lies not only in the production of crime statistics. It also inheres in their circulation. There is, it appears, a proportional relationship between the generality of a quantifact and its capacity to travel as knowledge. The more inflected or refined any number, the less likely it is to survive the rigors of movement or mediation. It is only the most gross, least qualified – those that erase the differences which make a difference – that become national truths. Thus, to return to our example for all seasons, South African murder figures are cited constantly as a symptom of the state of disorder in the land. Rarely is reference made to the highly inflected geography that restricts most kill- ing to relatively confined spaces, spaces with a specifiable sociology. In this respect, the social geography of violence resembles those of countries like the USA or Brazil, which have lower homicide rates, but whose death zones in inner cities and favelas are not all that dissimilar from the sites of greatest danger here; indeed, relative enclaving is, arguab- ly, a better key to the calculus of risk than are national aggregates.

When fully parsed, then, crime patterns in South Africa – even read through a haze of indeterminate numbers – are less exceptional than public discourse allows. But such ni- ceties are erased in the circulation of quantifacts. Among other things, this reasserts the contemporary mythos of a society at risk, its populace equally endangered. Which, as a large body of research shows, the population here is not. It also leaves intact the racial ar- chetype according to which the source of disorder lies overwhelmingly at the hands of young black men who are – in a metaphor that will prove both apt and unfortunate – hold- ing the country to ransom. Lost, too, is the fact South Africa has a social history, and a poli- tical sociology, in which statistics have come to mean different things to different people.

To the extent that the crime stat is a mode of objectification, alike a means by which realities are realized and a lens through which a nation may see itself, they take on special salience in South Africa. Why? Most immediately, because, under the ancien regime, they did not circulate as a free currency of social knowledge; matters of law and order were ma- naged by the state as part of its racial politics of security. Besides which, black-on-black crime was under-counted because blacks did not count. Indeed, it has been said that, be- fore 1994, “the ‘real’ state of crime” was never reported (Marsh 1999:76; Shaw 1997; Em- mett 2000:290). But something more profound has happened than the mere provision of new, improved information. It is only with the opening up of a democratic public sphere – or, rather, of variously articulated spheres – that South Africa, as an undivided polity, has been able to reflect on itself, on its social order, on the spirit of a polity reborn. And in that process, crime statistics have become a standardized measure of disorder, marking out a new emergency to replace those of old. They are, as it were, a medium of populist dis- course that appears, also, to be a message.

To say that the “real” extent of lawlessness in South Africa is unknown, that official statistics are a fact-making fiction, is not to deny the reality of crime. Or the brutal truths to which the numbers speak. Nor is it to deny that South Africans encounter criminality in visceral, concrete terms – whether it be at first hand, by anecdotal narration, or by a barra- ge of mediated incidents that, in their repeated re-telling, take on a perceptual mass of their own. As they do, these encounters morph from incidents to incidence, read in “the public mind” as the state of the nation, as the law in disorder, as a world under threat. Thus it is that numbers become a vehicle for the experience of the unreal, an experience that transforms the abstract into the sensate, general into the particular, the unknowable into the known. All of which serves to conjure up, once more, our reality-producing aphorism: Crime in South Africa is unreal. Therein, if we may be forgiven the pun, lies its truth value.


“They are now treating [crime statistics] like they are national security secrets.” Ted Leggett, BBC News, 7 October 2003.20

In July 2000, Steve Tshwete, then Minister for Safety and Security, announced a moratorium on all official statistics. Due to grave doubts about their accuracy, he said, no more would be published for the while: the Police Commissioner had ordered a revision of prevailing procedures of data gathering and calculation. During the hiatus, government would continue to collect figures for its own purposes. But not for public consumption. Then, almost a year later, and as suddenly, the silence came to an end. On 31 May 2001, Tshwete declared that a re-tooled technology of number production was in place: the SAPS had a new, computerised Crime Administration System, new manuals, new staff, new training arrangements and equipment, more precise definitions of felonies, even fresh crime statistics.21 A New Era of Enumeration had dawned. The flow of figures began again.

Once more, South Africa was awash in a stream of stats.
When he announced the moratorium, the Minister seems to have sensed what was to come. The ban, he insisted, was “not an attempt at secrecy.” But then, as if to confirm suspicions to the contrary, he berated those who “bl[e]w South Africa out of existence by blowing up the levels of crime.”22 As he anticipated, the announcement was greeted with outrage. Opposition politicians accused the African National Congress of “an absolute abuse of power…outrageous in…a democratic country.”23 Independent researchers, noting that South African figures were the best on the continent, said that the withholding of crime information would “foster mistrust between the rulers and the ruled.”24 Could it be, asked one, “that the police…do not like what the statistics say?”25

Perhaps the loudest protest came from the print media, for whom crime figures are a daily staple, their febrile fluctuations marking the pulse of national life. In the void left by the blackout, reporters set about doing their own counts “in the public interest.” The Cape Times called on readers for accounts of violent incidents on the notoriously crime-ridden N2 highway, a vital artery that has long been a corridor of conflict between the dispossess- ed and the prosperous who drive by to and from Cape Town. By compiling “an accurate picture” of risk, the paper hoped to goad the SAPS into action.26 Soon after, Independent Newspapers, arguing that the moratorium violated the Constitution, took legal action ag- ainst the Minister of Safety and Security and the Commissioner of Police for the Western Cape Province.27 Joining the chorus, the opposition Democratic Alliance alleged that, in the absence of proper statistics, it was impossible to “evaluate the performance of government” in discharging one of its most fundamental duties: ensuring the security of citizens.28 Public opinion was as vociferous: “Surely one of the best ways to counter the negative impact of crime…is the greatest possible degree of transparency from our public representatives?” wrote a typically angry correspondent to the Cape Times, indignantly offering a tally of his own recent encounters with crime.29

The argument continued as long as the ban itself. Nor did it abate fully when the moratorium ended. This furore makes clear how crime statistics have become objects of struggle, how they crystalize debate among disparate parties in an expansive political process. As gauges of disorder, they are now the diagnostic index of the effectiveness of governance; which is why their publication is a vital sign of the transparency now synony- mous with democratic rule. As lawlessness eclipses other threats in the popular ima- gination, citizens regard this information as an inalienable right. And government, itself an increasingly complex amalgam of public and private interests whose functions are defined in terms of service provision, is seen above all as a guarantor of law and order to the citizen-consumer – even in South Africa, where there is still some commitment on the part of the state to public welfare amidst neoliberal devolution.

Small wonder, then, that ruling regimes should feel compelled to produce credible crime counts. And to put a favorable spin on them for public consumption. Small wonder, too, that scholars, commentators, and the population at large should be wary of these counts, suspecting them of various sorts of misinformation. Nor is this a specifically South African problem. In September 2002, criminal justice experts in the U.S. expressed con- cern that the Attorney General was “exerting political control” over previously independent agencies responsible for collecting crime statistics.30

To count crime, in short, is to produce the stuff of politics (cf. Dixon 2002). This has three dimensions. The first is epistemic. It lies in the nature of measurement itself. As we saw above, “true” rates of lawlessness are unknowable, “accurate” comparisons in space and time are all but impossible, and felony figures are an indeterminate product of the in- terplay of law-breaking and law enforcement. All of which lays the ground for a more calcu- lating play on quantifacts, one that seeks make them signify in different ways for different ends. This is the second dimension of the politics of crime statistics, the strategic, one much in evidence in the controversy over the moratorium. Here all parties “argued with numbers,” using them to confirm their own assertions and to discount those touted by others. Thus, where critics censured the withholding of data that reflected poorly on the state, and victims published counts in order to highlight statistical subterfuges, government accused the media of an unpatriotic inflation of levels of crime.

The play of political strategy has been most palpable in relation to the national homi- cide rate, not least because it remains the standardized nightmare of most South Africans, a measure of the obscenity of a people wallowing in its own blood. Predictably, the state has reacted strongly to the assertion that the country is the “murder capital of the world.” One official response has been to try to counter South Africa’s alleged exceptionalism by placing its statistics in favorable light. Thus, in its first quarterly report after the moratorium (CIAC 2001), the SAPS presented a set of bold global contrasts; among them, the fact that, where Washington, D.C. had a homicide rate of 69.34 per 100,000 for January-June 1998, Pretoria’s was 41.12. While the logic at work in this comparison might not be totally arbitrary, it drew a sharp riposte: D.C. was the city with the steepest rate in the US, chided criminologist Ted Legget (2003); Pretoria was far from that here. Local figures for murder, rape, and aggravated robbery, went the counter-argument, are still “terrifyingly high” (Louw and Schönteich 2001:45). But the efforts of government to “spin” South African statistics continue unabated: housebreaking and assault are worse in Australia, we are told, car theft more prevalent in Canada, personal safety levels lower in the UK, and so on. At issue, in this counterfactual conjuring with comparison, is the production of a state of normality – and, hence, a normal state. To argue, both for domestic and foreign consumption, that rates of violent disorder are declining, at worst “stabilizing,” or that the police are “winning the war against crime,” is to assert that state, to assert the state.

As this suggests, crime statistics have become inflated political tender. Among other things, they are an index of national worth in a global marketplace in which southern poli- ties must meet northern scrutiny in respect of democratization, stability, and credit-worthi- ness, and the like. In contention are such prizes as investment, foreign aid, tourism, world sports events. Spiking the public argument about numbers is the sense that what is “really” at issue are unspoken and unspeakable matters of race. African politicians and cultural commentators often suggest that the exaggeration of lawlessness in white media is an insi- dious form of racism, disparaging the ability of a black government to maintain order, of black police to enforce the law, of black youth to behave with civility. Conversely, many whites feel themselves to be the victims of a system in which violent crime is a form of racial revenge (van Rooyen 2000).

Tangible in the controversy over crime figures has been the paradox of dis/trust of which we spoke earlier. Allegations about the abuses of enumeration seem only to affirm the faith in their revelatory potential, raising their value and intensifying the quest for ever more rigorous measurement. Thus it is that, despite suspicions about the distorting effects of statistics, especially in the service of a calculating government, hope remains in the re- demptive power of “accurate” – i.e. unmediated – numbers. Herein lies their quality as fe- tish. Note, in this respect, the enthusiasm for two internationally-acclaimed techniques of enumeration recently introduced in South Africa, victim surveys and geographical informa- tion systems (GIS). Far from being subjected to critique, they are widely held to offer a more exact purchase on levels of disorder. But these are not purely technical instruments: they have social and political implications, marking notable shifts in the ways that crime is conceived and counted. And how it is made to signify in public discourse.31

It is precisely at the level of discourse that we must seek the third, the constitutive dimension of crime statistics. Because of the way they circulate, these numbers translate large vectors of danger into personal and collective markers of risk, subjectivity, identity. In so doing, they render numinous forces of disorder into concrete, communicable “facts,” conjuring up citizens, moral communities, the nation. Herein lies their “politics” in the lower case: their capacity to contrive or reproduce meaningful social categories. Nor is this merely a matter of official figures, victim surveys, or geographical information systems. Vernacular statistics, everyday forms of counting crime, are also tropes of popular interaction in South Africa; they feature centrally in narratives of lawlessness, themselves a vital idiom of public culture here. The semiotics of popular accounting turn out to be integral to the figuration of self, community, and society in the postcolony. It is to this, the constitutive dimension of statistics, that we devote the rest of this essay.

“I want to reclaim my right to be wounded without my pain having to turn me into an example of woman as victim.” Njabulo Ndebele, The Cry of Winnie Mandela, 2003

Victim surveys have gained currency across the world over the past thirty years; evi- dence, this, of a shift of attention in criminology from perpetrators to those whom they harm (Fattah 1986; Maguire and Pointing 1988). In 1987, an International Crime Victim Survey was proposed by the Council of Europe; a standardized version has now been carried out in over 50 countries. In the US and UK, national surveys are done on a regular basis. The method, which involves questioning a sample of people about their experience of a range of offences over a stipulated period of time, is said by its advocates to enhance official sta- tistics, casting light on the “dark figure of crime”: offences, like sexual abuse and assault, routinely under-reported, most of all among the poor and powerless (Banks 1997:11). These claims have been given wide credence, though some critics insist that underclasses are less likely to report “minor crimes” (Cramer 1995; Kleck 1991:175-6), that subjective recollection is always unreliable, and that the method is “not ideally suited” for collecting incidents of an intimate sort, registering a lower rate than do police statistics in respect of offences such as rape (ISS 2004:126).

For those who champion them, these surveys are more than just an improved measure of crime. The technique has been propelled by an international movement, part of a world-wide politics of suffering, that has lobbied against a global “establishment” allegedly prejudiced against victims, victims held to be excluded from criminal justice processes for being “emotional and vindictive” (van Dijk 1996:1; Camerer 1997:1).32 In South Africa – where the rights of the injured were dramatically endorsed by the creation of the Truth and Reconciliation Commission soon after the end of apartheid – the first national survey was conducted by the Department of Safety and Security in 1998.33 A second, the one mention- ed above, was done by the Institute for Security Studies in 2003 (ISS 2004).

The methods employed in victim surveys resemble those of market research. As neoliberal political culture comes ever more to define citizens as consumers, casualties of crime seem most aptly depicted as customers ill-served by government. For advocates of “victimology,” the consumer perspective is crucial to rethinking received, state-centric app- roaches to law and disorder. “Get tough” policies have had little success in curbing crime, they note; evidence from across the world suggests that police resources are better spent protecting those at risk (Camerer 1997:1; Farrell and Pease 2001). But more than address- ing state policy, these surveys aim to be vehicles of popular empowerment. “By shifting the focus of the inquiry from the offender” say Camerer et al (1998:1), they “provide information which enables victims themselves to take preventive action against further victimization.” Victims themselves turns out to be the key phrase.

Displaying a generically neoliberal suspicion of government, the victim movement is wary of having its cause hijacked as an alibi for harsher “law and order.” It accuses the state of “stealing” crime from those who suffer it – and of redefining it as an offence against itself. “Victims,” proponents maintain, “are victims in their own right,…an end in themselves” (Camerer et al 1998:2). They seek to mobilize “civil society” behind “victims’ charters,” de- manding compensation from ruling regimes for their failure to maintain order.34 This stress on self-advocacy resonates with a more general tendency to displace politics into the do- main of the law and to reduce it to redress and punishment (Brown 1995:27; J. Comaroff n.d.; Comaroff and Comaroff 2003). Victim surveys, then, are part of an encompassing set of forces reshaping governance, public culture, and politics at the point at which the risk- bearing subject meets the right-bearing subject.35

From the perspective of these surveys, “each member of society is an indirect or vi- carious victim of crime” (Glanz 1989:1), either a victim or a potential victim. While their data inform prevention strategies, they are directed largely at the public, being widely circulated by the media, often in dramatic form. In this manner, they become one many mechanisms – from NGO’s promoting civil protections to ubiquitous ads for home security – that encourage people to view themselves as prey; in this manner, too, they may create what they measure. Thus one study of Durban (Robertshaw et al 2001), having asked respondents about their perceptions of lawlessness, concluded that “many South Africans feel helpless and trapped by the fear of crime.” A wealth of evidence confirms that citizens identify them- selves ever more in terms of vulnerability to violation. And they organize on this basis too: support groups mobilize around rape victims, neighborhood watches form in the wake of attacks, petitions circulate in the name of the injured. Even felons see themselves as casu- alties: “Criminality, you have made me your victim!” declared an inmate at a spirited hiphop and poetry performance inside Cape Town’s Pollsmoor prison.36 Similarly an article in The Big Issue, organ of the “marginalized and unemployed,” which described youths in custody as “helpless animals, dumb and resigned behind their bars” (Kretzmann 2004:17).

The victim surveys published and publicized in South Africa to date offer insight into the unintended ways in which enumeration implies evaluation. The results of the national survey of 2003, for example, were released so as to have maximum impact on public dis- cussion in the run up to the general election of 2004, in which crime was a emotive issue. They received prominent coverage. The media stressed that, despite the stabilization of lawlessness since 1998, the public felt less safe than it had five years before.37 Whatever else they conveyed, their reports gave graphic, pigmented accounts of people in fear. “Race Group’s Views of Safety Differ,” said a typical headline.38 In surveys of this sort, by contrast to official statistics, victimization and the anxieties to which it gives rise are heavily profiled in terms of hard-edged classifications of race, gender, age, and location. While they reveal persisting inequities in risk and security, these measures also reify old catego- ries at a time when identities and relations among them are under reconstruction. A widely reported study by Ted Leggett, for example, asserts that “Coloureds [are] twice as likely to be murdered,” and are much more likely to be jailed, than are blacks, whites, and Indians; his evidence is based on numbers from the Cape Flats, a terrain on which identities are often rendered highly ambiguous. The Director-General of Labor in the Western Cape, who found the study “disturbing,” remarked that it lent itself to the “demoniz[ation] and iso- lat[ion] of a whole group of South Africans.”39

Undoubtedly, victim surveys expand received understandings of crime. Undoubtedly, too, they provide a counterpoint to official data, albeit sometimes as much for their similarities as for their differences. But, like all social measures, they are also complex discur- sive processes. As Wendy Brown (1995:21) argues, efforts to protect those with injury-for- ming identities may entrench the very thing they denounce. In seeking to counter insecurity by showing who is most susceptible to attack, profiling reinforces risk by reproducing cate- gories of victim in need of protection. And by instilling a culture of fear. This sense of vulne- rability is the ironic underside of liberalization, of a climate in which subjects are held to be “active participants in their own governance” – and government is, in theory at least, a me- chanism for the adjudication of decision-making (Rose 1993). Which is why politics, increa- singly, is centered on recognition, often through the law, for the entitlements of the relatively disadvantaged; why the disempowered are no longer a class but claimants in class actions; why social movements, formed out of shared disabilities, pursue rights to reverse wrongs. Citizenship here is defined as personal fulfilment: as a process of ongoing risk avoidance that rests on “incessant calculations” (Rose and Miller 1992: 201). But devolution has its limits, which are made manifest where freedom and choice run up against the real threat of victimhood.

It is at this juncture that the paradoxical role of the state becomes visible. In calling for more policing, ordinary South Africans insist that an ordered world is impossible without regulation on the part of an interventionist state; hence the mantra that democracy is en- dangered by criminality. While the victim movement keeps its distance from government, a heightened awareness of the entitlements of citizens translates into a demand that the administration get tough on crime. Not only is there broad skepticism about “human rights policing,” believed to favor criminals over their prey, but there are continuing demands to bring back the death penalty, a topic we revisit below. This mood of popular punitiveness is the under-side of “minimal governance”: it is the animating force that allows the latter to transpose itself into its opposite when, in the name of emergency, executive authority over- rides civil liberties. The line between neoliberal government and fascism-lite, as the history of the present shows, is gossamer thin.

In all this, the victim has a double-valence. S/he is claimed both by the state as it declares war on crime and by civil activists seeking recompense for the injured (van Dijk 1996:123). “The public,” meanwhile, becomes an aggregate of the imperilled, in which identity and difference are defined in terms of hurt and trauma. And entities like “community” and “neighborhood,” constructs with little other sociological salience, are given fresh life in the search for security, a search that begets new public-private institutions, new forms of surveillance, new modes of control.


If victim surveys reduce the sociology of disorder to probabilities of injury over time, geographical information systems plot insecurity in space. In essence, GIS overlays one slice of information on another produce a single layer of previously unseen associations. Focusing primarily on criminal incidents, offender movements, and profiles of “targets, it has increasingly incorporated the “victim” as a category of analysis – for example, in dis- cerning the role of location in repeat victimization. As this implies, the two approaches may complement one another: one charts patterns from on high, the other from subjective ex- perience. Rose and Miller (1992:202) see this kind of complementarity as typical of the re- gulated autonomy of neoliberal polities, in which a stress on self-assertive citizenship exists alongside mechanisms for “governing at a distance.” But the two modes of measurement can also come into conflict. Some critics contrast the grounded truth of victim experience with the reified data generated by GIS, questioning the fetishism of its technology and the faith placed in its explanatory powers.

In the era of “intelligence-led” policing (Read 2001:xxi), the expanded use of GIS un- derlines the ever more prominent role of statistics as vehicles of communication across the dispersed agencies – state, public, corporate, academic – that comprise the “partnerships” of neoliberal rule. Crime statistics circulate not merely as information, but as mediated discourse about the nature and efficacy of governance itself. Unlike victim surveys, which have been used as much to challenge state institutions as to inform them, GIS is a synoptic instrument (Obermeyer 1995:81) most useful in the centralized management of disorder. It has also been deployed in the development of methods of surveillance-from-a- distance, like CCTV, the advent of which has raised questions across the world about violations of privacy (Cho 1998:162). The recent announcement that new British crime-fighting initiatives would include the satellite tracking of serial offenders prompted at least one journalist to remind the Prime Minister of the Human Rights Act.40 But for the most part, such scruples succumb to the popular desire for security. In South Africa, GIS has largely been deployed, more modestly, to map incidents reported to the police. Yet enthusiasm for it has been marked: in 2001, the Minister for Safety and Security announced that it had been introduced at 340 police stations and would “greatly enhance…operational planning.”41

A virtue of GIS, it is said, is its power to “influence people” by translating statistical analysis into accessible visual displays (Hirschfield and Bowers 2001:6). It is also held to reveal otherwise invisible risk factors for various sorts of offence by situating hotspots of reported attacks in their physical locale. But how exactly does an elaborate mapping of the distribution of lawlessness advance our understanding of its incidence? Ken Pease (2000: 228f), reflecting on the British scene, worries that the visual seductiveness of GIS – and a measure of technophilia – may inflate its importance at the expense of other ways of ap- proaching crime. Mapping belies the fact that victimization is not always structured along spatial lines, he argues. Nor does it capture the experience of those who share anything other than location.42 In short, the claim that GIS enables information to be “seen at a glance” is misleading. Note though, that “victimization,” the actual experience of the injured, re- mains the naked truth that GIS ought to be able to explain.

Pease has hit on a critical problem here. The invocation of space as source of deter- mination in GIS derives not from a principled criminological theory. It is owed to techniques capable of making patterns of disorder seem so tangible as to be self-explanatory. But space is always, in the first instance, a product of human practice, both material and mean- ingful. As Merleau-Ponty (1962) and others (e.g. Soja 1989) have insisted, it does not exist in unmediated form, as a thing with determinations of its own. If this is true, it cannot, in it- self, reveal hidden causes – however it may be mapped.43 Yet GIS and victim surveys pre- sume that they can, casting their data into categories – “inner city,” “suburb,” “township,” “informal settlement” – that turn race and class into space. And attribute to it an inordinate explanatory weight (see e.g. Robertshaw et al 2001:68; Louw et al 1998:20f.).

The deployment of crime maps in official reports and public media says much about their role as vehicles of inscription and communication – and about their effects on popular perceptions of space, danger, identity. If maps of state mark out the terrain of the nation, GIS “hotspots” show, vividly, the limits of governance. They delineate where order ends, where disorder prevails. Here the panoptic eye is ambiguous. It is not that of the criminal justice system. It is the dispassionate gaze of a technology that may be deployed, alike by private interests or by ruling regimes, for policing, scientific research, commerce. Even organized crime. If these qualify as instruments of “government at a distance,” to return to a more general issue to which we alluded above, what exactly is government? Foucauld- ians, eager to avoid the reification of the state, deploy the term to gloss the “shifting al- liances between diverse authorities in projects to govern a multitude of facts of economic activity, social life and individual conduct” (Rose and Miller 1992:174). But surely this reifies “governance” instead, fudging critical questions about the effects of privatization and technicization on organized power; questions about the collaboration, in our Brave Neo World, of different sorts of regulation – from law enforcement, through the regulation of intellectual property, to the management of markets. The term seems even less adequate in disclosing how technologies like GIS serve as modes of social control.

Take an instance of the deployment of crime maps in the South African media: a full-page spread, “Pinpointing Our Crime Hotspots,” in the Saturday Star in 1999.44 It cen- ters on a map of Gauteng, embellished with graphics of a gun, prone bodies, a relict infant, a pair of cuffed hands. Box inserts tabulate figures for the “top-ten hit parade”: the policing areas with the highest concentration of serious crimes. The text situates the reader on this chart of terror: “Tourists and locals alike are familiar with Gauteng’s reputation as South Africa’s crime capital,” it begins, “but there are certain areas where you are more likely to be the victim of a specific type of crime than others.”

Here it becomes clear how statistical renderings intersect, in “real life” situations, with other genres of representation, giving rise to the total crime fact. The communicative powers of GIS are extended for maximum impact, being overlaid by a montage of melodramatic images drawn from a standard repertoire. This graphic is framed by headlines that sport the grit-and-flash of classic crime reporting. The prose translates the panoptical data on the map into the implications for the reader-as-potential-victim. Together, they illuminate how so-called risk-reducing data information triggers trauma by tying victimhood to identity. “Women living in Johannesburg, Soweto, and Vaalrand should exercise extra caution be- cause in these areas between 43 and 216 women are raped for every 100,000,” says the text. “If you live in Pretoria or Johannesburg, ensure you have a good security system or a fierce dog, because the two cities are equally susceptible to housebreaking.” Finally, “[i]f you are able to dodge hijackers, be careful where you park your car: Avoid Pretoria and Johannesburg’s city center….” and so on. The report garnered considerable attention. In Mmabatho, where the Saturday Star has a large readership, it prompted one Tswana family to consider bringing their daughter back from college in Gauteng. Such “authoritative” interventions, like terror alerts issued by the US Department of Homeland Security since 9/11, are a common feature of official communication in the risk society. They announce serious threats that – while highlighting vulnerable people and places – are agonizingly vague. Risk, as the word itself suggests, is terminally uncertain, at once “both real and unreal” (cf. Beck 1992:29). Which is why it is always open to exploitation.

The technical aura of the map in the Saturday Star coexists with the mis-en-scene of sensational media crime. Risk avoidance is written in a register calculated to induce fear: the hapless addressee is asked to pondering odds of surviving in Johannesburg, where simply being female is a mortal threat, where owning property requires a fierce dog, where driving a car invites a joust with jackers. No wonder that more than one survey has shown that South Africans, especially bourgeois city-dwellers, are becoming ever more fearful of crime. This imploding terror is the outcome of various intersecting processes, among them, the desire to perfect the measurement of disorder, an active private market in security, and an eagerness to sell newspapers.

But the value that accrues to these maps as they circulate, as they invoke appre- hensive subjects and publics, is conditioned also by a wider social context: one in which violence has become a currency of politics and of economic enterprise; in which govern- ment both abhors the threat of disorder, yet uses it for its own ends; in which citizens are at once repelled and fascinated by lawlessness; in which other urgent threats, like poverty and ill-health, are eclipsed, making it more likely that crime serves as a routine mode of production and representation. The publication of evocative maps owes much to a deregu- lation of the means of enforcement, to an effort, in the words of a community policing spokesman, to get individuals and neighborhoods to “take responsibility…”45 Yet as we have said, these attempts seem to achieve precisely the opposite, fueling ever more insis- tent calls on government to treat crime “like the national emergency it is” – calls that make repeated appeal to statistics to substantiate their cause. 46

TRAUMA TESTIMONIES AND MYTHOSTATS: figuring popular experience

Crime rates, we have argued, are a psychic barometer. They measure the fall of old lines of separation, old securities – and the rise of new modes of integration, expectation, redistribution. But how exactly do they feature in popular discourse? What prompted us to contradict Nelson Mandela in asserting that, far from rendering it alien, numbers give substance to experience?

A telling statistic of the recent South African past is that 58% of its metropolitan population (49% overall) had discussed crime in the prior two weeks (ISS 2004:49f.). Narratives of lawlessness rival the weather and sport as a topic of common talk, cutting across existing lines of difference: testimonies to transcendent truths, they root their cre- dibility in the fidelity of ordinary experience. These narratives are mundane melodramas whose endless retailing appears to wrest order from chaos. Tokens of a type, they evince several common features. One is to distill circumstantial evidence into fact by enumerating incidences. Not only does this signal veracity. It imparts primary value to second or third- hand happenings, translating a multiplicity of events into a shared subjective reality. The play between quantity and quality involves a humdrum hermeneutic that seeks to spell out the social proportions of risk and uncertainty. Vide this story, told us in June 2000 by a Tswana woman in her mid-70’s:

Taxis have become very dangerous here in Mafikeng. My daughter had a terrifying experience last November: she hailed a taxi that was traveling to town. By the time it got to the main road, she was the only passenger in the van. The driver, who was drunk, turned sharply in the wrong direction and picked up speed. When L. called out to him, he pointed a gun over his shoulder and kept going. Luckily, he was for- ced to slow down behind a donkey cart, and she leapt out, rolling dow n a bank at the side of the road. She was badly bruised, and lost her purse. This is now the third case I have heard of recently in which women have had trouble in taxis. We really can’t travel around the way we used to in this town, even to wakes or funerals.

This testimonial evokes the complicity of the listener by voicing several indices of verisimi- litude. It begins with a fact for which the tale is evidence: the risks of taxi-riding, itself iconic of the mobility brought by liberation and liberalization. The legitimacy of the speaker as witness, and the status of the story as evidence, derives from the relationship of narrator to victim; in this case, mother to daughter. Detailing personal connections assimilates the events to the terms of intimate experience. “We” – black, petit-bourgeois, peri-urban women – are meant to feel terror as an ordinary ride becomes a nightmare journey, death just the flick of a drunken trigger-finger away. And so crime produces its own subjects and sensations of the real, its culture of vicarious victimhood. Her story is our experience. It is all too real: we are all victims of taxi assault, we have all lost our freedom to violence.

Tales of this kind are less about uniquely horrifying events than about the “fact” that such things have become commonplace. The incident in the taxi invoked other instances, however imprecise their comparability. “This is the third case I have heard of” turns random occurrence into a condition of public being and suffering. A letter to the Cape Times from a recently-returned emigre, written during the moratorium on statistics, deploys a similar semiotics of (ac)counting:

In the 12 weeks I have been back…I have personally “encountered” 10 criminal acts against friends and acquaintances…At nearly one incident per week these statistics are mind-bending. If this is happening in my relatively affluent and sheltered life, I can only imagine what life must be like in less protected areas.47

Again we see how the idiom of calculation makes visible the limits of order and, thereby, gains purchase on a people united in its vulnerability against the state. Street statistics are a means of embracing otherwise unimaginable sectors of society (“the less protected”) into a nation in search of security. The writer computes a home-made victim count, multiplying his “personal encounters” tenfold, thus assimilating third-hand incidents to first-hand experience. The numbers here make evident a hidden truth to which “the public” has a right – and, at the same time, inflate the menace.

Here, as in the taxi tale, quantity turns into a qualifact: that the nation is in trouble. In this world, shared truths, about crime or anything else, come from the victims. “How many more innocent people have to die before the government will admit that its so-called law and order is a sick joke,” wrote Patsy Tyler to the Cape Argus in July, 2003.48 Tyler challenges the public to “stand up and be counted.” Anyone who has suffered crime should submit an account, and the media should declare a day on which to publish them, so that “every story has been told.” A “comprehensive list” of offences inflicted on its citizens would “shame our government in the eyes of the world.”

But the popular calculus of crime also works by magnifying epic events into mytho- stats: events that, in their singularity, come to signify collective being and trauma – and, in turn, give rise to the most terrifying statistic of all, namely, that everyone “knows” one or more people who have suffered brutal attack. Spectacular violence tends to be a ne- cessary condition for an incident to have this effect. Other social considerations also add uncommon sign value to particular episodes; their occurrence, for example, in relatively en- closed populations hitherto largely free from violence. Like the slaying of 52-year-old Mot- lhabane Makolomakwa in the quiet North West village of Matlonyane. Known as “Ten-Ten” after a famed soccer star, the victim was the most prominent resident of his impoverished community; a successful farmer, former mid-level government employee, and chair of the local “tribal council,” he had also sponsored a football team. In 1994, five youths hacked and burned him to death in his own truck. They insisted that he had killed their fathers and turned them into zombies – this being a common explanation nowadays for why some become rich while others, especially young males, remain jobless. Found guilty, the “boys” were each sentenced to twenty years.49 The murder and High Court trial, which were avidly covered by local media, riveted the Tswana-speaking population of the province. They made plain just how much the division between young and old had been exacerbated by fears of escalating crime. Older citizens returned repeatedly to the incident as evidence of a loss of community and respect for authority among “the children.” For them, it underlined the virtues of “traditional” discipline. Young men, on the other hand, took the case to illust- rate the negative effects of the monopolization, by their seniors, of the means of producing wealth. Among both, it hardened lines of generation and intergenerational conflict.

Another set of collective identities materialized when, in January 2003, nine men were murdered in a male massage parlor in Sea Point, on Cape Town’s Atlantic seaboard, their throats cut before they were shot in the head.50 The site, scale, and bloodiness of the killing pointed to a hate crime. The Gay and Lesbian Equality Project of South Africa quick- ly became involved, its representatives meeting with the assistant police commissioner. Police reported a huge response to the request for leads, and, unusually, sex workers and gay organizations “joined hands” with the ANC in a memorial service.51 While the case re- mained unique in its proportions, its meaning grew murky as the SAPS became convinced that the motives for the crime were more routine: that it was drug-related, perhaps, or the aggravated consequence of a robbery.52 Yet the atrocity retained its mythic proportions, entering public discourse as an index of heightened risk to all gays. Just how such episo- des generate terrors of their own became clear in June 2004, when the Gay and Lesbian Alliance lodged a complaint against a website that, among other offensive statements, said – with obvious reference to the massacre – that homosexuals should be “Sizzled.”53

This case makes plain how epic crimes may evoke and intensify a sense of citizen- ship in communities of identity. Mass mediation is integral to the process, converting extra- ordinary happenings into a generic intimacy, a shared sensation of fear. Their victims are at once unusual and horrifically commonplace: you, I, could be next. Derrida (2002:248) observes that each murder is “singular, thus infinite and incommensurable” (original em- phasis); yet, ironically, its capacity to bespeak the general, to instill a sense of common predicament, derives from that very singularity. Mythic crimes signify by means that go bey- ond identification or transference. In their suffering and annihilation, the socially-marked casualty substitutes for the self and the collectivity for which s/he comes to stand: youth, the elderly, women, gays, whites, property-owners, the nation. Hence the tendency to claim those who, like sacrificial victims, redeem their kind with their innocent blood. By their dramatic erasure they spell out the minimal conditions of humanity and moral community.

Particularly galvanizing in this regard was the kidnaping of Leigh Matthews, 21-year- old daughter of a Johannesburg businessman in July, 2004.54 In South Africa as we have seen thousands of persons, young and old, go missing each year. Yet this case mobilized an “overwhelming” response from people across the land.55 Why?

What made this event so compelling was not just the media melodrama or the voy- euristic curiosity aroused by the private suffering of the attractive and the affluent. Neither was it the appeal of its blond victim, pictured again and again in a satin ball-gown, on the poignant brink of womanhood. It was the way in which these things, occurring at the inter- section of race and class, worked together to make it seem at once exceptional – “Touched By an Angel,” ran one headline56 – yet capable of typifying the state of the nation-as-trau- ma. Ms. Matthews was singled out by social circumstance: reports noted that her kid- nappers had probably done research into her father’s income.57 Yet, as You magazine not- ed on its cover, “not only [the] rich” are targets. “You could be a victim.”58 The dialectic between singularity and representativeness at work in the image of Leigh Matthews is evident, too, in a report in a national Sunday paper:59 “Hundreds of South African families are tormented by the anguish of not knowing when their missing loved one might be found,” it read. “More than 1800 adults and nearly 1300 children were reported missing…last year” Missing, but not kidnaped. There is a wealth of difference between abductions, runaways, and kidnapings. A wealth of difference and a difference of wealth. Child abductions have long been prevalent in poor communities all over the country.60 These ca- ses seldom merit media attention, a point forcibly made by several black callers to talk shows in Cape Town at the time; one community network, Bush Radio, refused to cover the event for this reason. The discrepancy is underscored in Michael William’s perspicacious detective thriller, The Eighth Man (2002), which follows the parallel deaths, in the Western Cape, of two young people separated by race and class. By contrast to those of middle-class white children, Williams’s story tells us, the bodies of impoverished young blacks often still do not count.

Leigh Matthews is set apart from these faceless victims. Her case had proved a challenge, said the head of the Missing Person’s Bureau, “because kidnaping for ransom [i]s not…common.”61 This is disputable. SAPS figures indicate that it became much more frequent during the 1990’s; though just how frequent is a subject of disagreement.62 A spokesmen for a corporation that markets “risk assessment” and “crisis management” across the world suggests that it has become a brazen industry in search of quick turn- over.63 This corporation, which trains wealthy, nervous South Africans in “pre-incident” and anti-kidnap techniques – its courses are covered by ransom insurance – insists that the vast majority of those kidnaped are returned alive.

Not in this case. A cash ransom was delivered by Mr. Matthews.64 But his daughter was not returned. As the days went by, the “public” coalesced ever more visibly around the incident. Police set up a 24-hour line to deal with the “flood of information” coming in; hund- reds of South Africans, including many abroad, posted messages of support on a special website.65 Strangers in the street, in elevators, in taxis exchanged information, particularly in Johannesburg. The tragic discovery of her body eleven days later drew forth an outpour- ing of emotion: “Country mourns for Leigh,” intoned a typical headline, above a photograph of a diverse group of fellow students signing a book of sympathy on the elite private univer- sity campus from which she was snatched.66 A petition was launched on the internet calling on government to reinstate the death penalty:67 “Do it for Leigh Matthews,” it read. Nelson Mandela, the State President, and representatives of all major political parties offered public condolences. “Leigh Matthews was a flower, the lifeblood of the nation,” said an ANC spokesman. “We can’t unite…if these sick elements are in our midst.” But, ironically, it is precisely around events like these that the nation does unite, hyperreal events that become the measure of a traumatized citizenry transcending its differences, if only for an instant. “Something like this could happen to any of us,” wrote the Matthews family to the Sunday Times; “we need to come together as a community, as a nation, and take a stand against such wickedness.”68 Obscured here is the place of this tragedy in a larger pattern of lawlessness, one heavily inflected by race, class, and gender: one in which, as we have seen, rates of injury remain heavily concentrated among the black urban poor. The fact that the suspect subsequently arrested for the killing was a young white man, the son of a Baptist minister and a fellow student – not, as widely anticipated, a Nigerian gangster – merely adds to the irony.69

Such is the alchemy of numbers in everyday discourse – in which crime stories are a privileged register not just for calculating the quality of life and the state of the nation, but for mobilizing mass action. In making the singular into the plural and vice versa, we repeat, this arithmetic bypasses the sociology of crime. Narratives of lawlessness might console, enthrall, indict, particularly among the classes that take personal safety as a right and have property to protect; one effect of the Matthews kidnaping, after all, was to magnify levels of terror among paler, more prosperous citizens. A politics of indignation comes not far be- hind: “It is our country and our right to be able to live free of fear,” said that same internet petition in the name of Ms. Matthews. While such exchanges, mulled over in private, may do little to address the structural bases of insecurity, they do a lot to prime the culture of dread, and the sense of immanent victimhood, that they seek to redress.


“It is a terrible space to be living in; in what must be one of the best societies on earth and not allowing yourself to enjoy it.” Mondli Makhanya, Sunday Times, September 200470

In Lewis Nkosi’s play, The Rhythm of Violence, published in 1964, the main action takes place in a makeshift police station in the Johannesburg City Hall, overlooking a square in which a protest meeting is in progress. The crowd yells (p.3),

“Freedom in our lifetime!”
Jan ( [a young white policemen] nervously): ”How many of us are here? “
Piet (his senior) ”Two hundred men at the ready to shoot…”
Jan (more nervous still): “You think that number is enough?”
A voice from the square, across the dialogue: “Can they rule by the gun forever?”

In this brilliant cameo, Nkosi captures a telling moment in the tortured history of numbers in South Africa. Critical episodes in this history are figured in stark fractions, the proportions of which have not changed as much as we might like to think. The brutal percentages of apartheid rule – 13% of the people with 87% of the land – have morphed into new mea- sures of inequality: grossly uneven rates of victimization, a close correlation of poverty with violence and disease, and the like. Now, as before, politics crystalizes in ratios, and not just in respect of crime. Few issues loom as large in the battle over AIDS than statistics that purport to describe the pandemic. This, too, is evidence for one of our contentions: that, in their very abstraction, numbers make real phenomena otherwise ungraspable in human experience. But, in so doing, they produce new indeterminacies. And P/politics in both the upper and the lower cases.

These politics, we argue, are more complex than is implied by the Foucaultian link between numbers and biopower. This is not only because “governmentality” fails to grasp neoliberal collaborations between the private and public sectors. It also glosses over the degree to which quantifacts are the stuff of contestation. As we have seen, official felony counts are routinely debunked, and often appropriated, in the cause of new species of so- cial action. When governance is reduced to law, order, and service delivery, and when citi- zenship hinges on security, crime rates become a prime currency of public culture, a prime index of order, a prime gauge of effective rule – and a prime suspect in efforts to detect a politics of bad faith.

In South Africa, in sum, levels of lawlessness are made to bear witness to the suc- cess or failure of post-apartheid democracy and, by extension, to the legacy of colonialism. They are the benchmark against which reconciliation and enfranchisement are evaluated. And the specter against which a sociological imagination is shaped, moral reform is moot- ed, and grass-roots mobilization takes flight. In all this, whatever they may or may not actually measure, crime statistics fill the space between the unknowable and the axiomatic, defining subjects and populations as they do. As this suggests, they are uncommonly good to think with – which, in part, is why enumeration is not limited to state bureaucracies or even to those who count in the public interest. Tallying lawlessness is a mundane communicative practice, a mode of figuring the relationship between incidents and incidence, epic events and everyday existence, uncertainty and order.

But, as we have seen, vernacular accounting is no more capable of eliminating un- certainty than is any other form of calculation. The endless recycling of experience – its re- counting, in the double sense of the term – may seek to probe reality and reduce ambigui- ty, to tame horror by resort to repetitive narration. But, by making third-hand happenings into first-person sensations, it amplifies vulnerability, reiterating a state of apprehension, a frisson to be replayed, a trauma hard to transcend. Which may explain why such anger is evinced by evidence that South Africa is not as exceptional as it may seem. At the same time, numbers do not mean the same everywhere across the social spectrum. Recall one last time that those most at risk are also least likely to see crime as a transcendent danger – in part because they face other life-threatening challenges. It is also the case that the vic- tims who count most, who escape “becoming a statistic” by attaining mythic status, are sel- dom those for whom suffering is part of their generic condition. But it is such epic felonies, like the Makolomakwa burning, the Leigh Matthews kidnaping, or the Sizzlers massacre, that become iconic of collective identities, of publics, of a nation said to be sinking into a sea of criminality. Therein lies the standardized nightmare, the abomination, from whose jaws social order it to be wrested. Therein lie the perverse politics and the peculiar produc- tivity of crime. Therein lies the criminal anthropology of the Brave Neo South Africa.


Just after this essay was completed, in September 2004, the South African Police Service published its new statistics for the financial year 2003/4. They showed that the incidence of most violent felonies – the major exception being armed robbery, much of it street mugging – had decreased perceptibly. What is more, figures for the provinces indicated that patterns of variance had increased; for example, the homicide ratio of Khayelitsha, the township mentioned earlier, and Camps Bay, a famously rich, white seaside neighborhood, was 358:1. In sum, the patterns described above were, if anything, further emphasized. Also affirmed was our observation of the tendency of South Africans, especially South Africans of privilege, to disbelieve “good news” or to discredit official numbers that might salve the national trauma. The reaction to these new statistics was striking: they were greeted, wrote the Cape Times, with the same “outrage,…anger and displeasure” that annual police reports always seem to evoke.71 The flood of letters to the press is best captured, perhaps, in a single caption: “We, the People, Victims of Horrific Crime, Know the `Good News’ is a Lie.”72 Nor did such responses come only from the lay public. Women’s and children’s rights activists, offended at the under reporting of rape and abuse, expressed themselves volubly. So much so that President Thabo Mbeki counter-attacked at length in his weekly letter published on the African National Congress website, ANC Today.73 “When,” he asked, “is good news bad news?” When it is read, he said, answering his own question, through the kind of psychosis that produces a fear of freedom. Perhaps. The argument continues.

Other Essays