Race,attitudes and behaviour in racially-mixed,low-income neighbourhoods in Cape Town,South Africa

Introduction: From hope to scepticism

‘Apartheid’ in South Africa entailed systematic racial segregation between racial groups and discrimination against non-white groups. The Group Areas Act defined residential areas for the use of one or other racial group, and large numbers of people were forcibly removed into designated neighbourhoods (Field, 2001; Western, 1981). Schools, public transport and social spaces were strictly segregated. Immigration into the cities by African people was tightly regulated, especially in Cape Town. After the end of apartheid, racial discrimination other than affirmative action became illegal. The discourse of the ‘rainbow nation’ held out the possibility and hope that South Africans would overcome historic divisions and build a common identity and solidarity while acknowledging cultural diversity. In practice, however, hopes for desegregation have given way to disillusionment and scepticism. Residential and educational segregation persist, in practice, for most South Africans: most South Africans live in mono-racial neighbourhoods and their children attend mono-racial schools. Moreover, few South Africans have friends from other racial groups and inter-racial marriages are rare.

There is little inter-racial interaction even in social spaces that appear to be integrated, including public beaches (Dixon and Durrheim, 2003), the dining halls of integrated universities (Schrieff et al., 2005, 2010) and nightclubs (Tredoux and Dixon, 2009). In supposedly integrated schools, inter-racial relationships are often difficult, linguistic differences impede interaction and taste is racialized (Dolby, 2001; Schenk, 2009; Soudien, 2004; Vandayar and Jansen, 2008). Settlement by low-income ‘African’ people in otherwise ‘white’ middle-class suburbs has given rise to severe tensions (for examples, see Dixon et al., 1994; Saff, 1998). Even when new immigrants into a middle-class suburb are not themselves poor, some richer white residents move out (Morris, 1999) or at least take their children to schools elsewhere (Lemanski, 2006b). Indeed, white South Africans’ embrace of gated residential neighbourhoods is often interpreted as a way of limiting inter-racial interaction. For the overwhelming majority of children growing up in mono-racial neighbourhoods and attending mono-racial schools, the shopping mall might be the only racially-mixed space they encounter (Bray et al., 2010).

The persistence of informal segregation has led social psychologists working on South Africa to question the basic tenets of the ‘contact hypothesis’. Formulated by Gordon Allport and others in the 1950s (see Allport, 1954), the contact hypothesis held that, under certain conditions, contact with members of other racial groups would lead to reduced prejudice and conflict. The same logic applied also to interactions between other kinds of groups besides ‘racial’ ones. While the hypothesis has been ‘one of the most successful ideas in the history of social psychology’, the search for the full specification of the ‘optimal’ conditions for contact has led to it becoming more and more ‘detached from (and sometimes irrelevant to) everyday life in divided societies’, rendering it of little practical use in ‘understanding or promoting social change’; ‘there is little point in enumerating lengthy lists of boundary conditions for ideal contact without explaining how such conditions might be made relevant to the lived experiences of ordinary people in ordinary situations or without explaining how they might be implemented in particular contexts of inequality’ (Dixon et al., 2005: 697–8, 701). Close observation of interactions on beaches, in nightclubs and in university dining halls suggests that the ‘lived experience of contact often led to the reestablishment of racial boundaries’ (Dixon et al., 2005: 704; see also Dixon et al., 2008).

Accounts of interaction in racially-‘integrated’ neighbourhoods seem to corroborate this pessimistic conclusion. Desegregation has generally entailed upward mobility by non-white South Africans into richer neighbourhoods otherwise occupied by rich, white people – a process driven by the very rapid growth of the non-white middle classes and elite (Seekings and Nattrass, 2005). Racial ‘integration’ in such suburbs rarely entails much interaction, however, as residents live behind high walls. The best accounts of such suburbs are by young novelists grappling with the dilemmas of identity. In Coconut, Kopano Matlwa contrasts the gated residential estate where her African character Ofilwe now lives, with the African township where Ofilwe lived before her family became rich:

No toddlers with snotty noses and grubby hands play in the streets in Little Valley Country Estate. Groups of teenage girls in bright T-shirts, old torn jeans and peak caps do not sit on the front lawn pointing and gossiping about the guys that walk past the gates of their homes. Older sisters do not play the wailese loud, so that those who know the tune can sing along as each mops, dusts and sweeps their homes clean. In Little Valley Country Estate the neighbours are the cars you see parked in their driveways and the children are the tennis balls that fly over the wall and into your pool. Here at home, Tshepo [her brother] was my only company and I his. (Matlwa, 2007: 89–90)

Not much actual inter-racial ‘contact’ occurs in middle-class and elite neighbourhoods. In Coconut, when Ofilwe does have contact with white people, in the neighbourhood and at school, the interactions are infused with persistent racism.

There is more ‘community’, and more interaction, in less rich neighbourhoods, where walls are lower, gardens smaller and people walk on rather than drive down the street. In two neighbourhoods in Cape Town where upwardly-mobile lower middle-class coloured people moved in alongside struggling white families, inter-racial interactions were common and generally positive (Broadbridge, 2001; Teppo, 2004). In these neighbourhoods (and even in elite suburbs), however, patterns of inter-racial interaction might be affected by selection bias in that people might choose to live in these neighbourhoods precisely because they are unusually open to inter-racial interactions. Similar selection bias might occur in the small proportion of schools that are racially-mixed. Students who are more hostile to inter-racial interaction are presumably more likely to attend mono-racial schools.

The paucity of residential desegregation and the possibility of selection bias in the few neighbourhoods that are mixed means that it is difficult to assess the actual or potential quality of inter-racial interactions. There are, however, a few low-income neighbourhoods where more-or-less randomly selected poor coloured and African people became neighbours through no choice of their own. One such neighbourhood is Westlake Village, in Cape Town’s southern suburbs. In this case, relationships between coloured and African residents are reported to be generally good (Lemanski, 2006a). In this article, we extend the analysis to two other mixed, low-income neighbourhoods, Delft South (including the more recent neighbourhood of Delft Leiden) and Tambo Square. These neighbourhoods have almost equal numbers of poor African and poor coloured residents.

These neighbourhoods are interesting precisely because they are atypical not only within Cape Town but in South Africa as a whole. It is only in the Western Cape Province, and in the city of Cape Town within it, that the poor are racially diverse. Throughout the colonial period and into the apartheid era most poor people in Cape Town were coloured, i.e. they were the descendants of the Khoi farmers and San hunter-gatherers (who were the indigenous population of the Western Cape), or the slaves brought from the Dutch East Indies (‘Malays’) or of mixed (white–African) parentage. Only from the 1970s did large numbers of poor isiXhosa-speaking African people circumvent state restrictions and migrate into Cape Town from the rural Eastern Cape. With respect to religion, beliefs about ancestors, lifestyles and tastes, as well as language, poor African and poor coloured people in Cape Town are quite different to each other. In the early 2000s, coloured people comprised about one-half and African people about one-third of the total population of Cape Town, while white people comprised a small but rich minority.

Neighbourhoods that are mixed, but not by residents’ choice

Since 1994, the South African state has assisted a very substantial number of poor people to access formal housing, primarily in towns and cities. While the government’s own claims (e.g. South Africa, 2008: 28) are exaggerated, huge new neighbourhoods, comprising rows of very small, generally identical and often low-quality houses, have been built on the periphery of almost every town or city. About one in five households in Cape Town lives in a state-subsidized house in a post-1994 low-income formal neighbourhood.

These new houses are subsidized by the state in that the state subsidizes their construction. Access to a subsidized house is means-tested, supposedly limiting access to the poor (although, it is clear, some houses are allocated corruptly and others are obtained fraudulently). In most new neighbourhoods, some people have extended their houses themselves. In general, however, the poor lack the resources to extend their housing, and the ‘subsidy’ from the state covers the entire construction cost, which is why houses are very small and simple. The state expenditure is a subsidy primarily in that it is paid to private developers who install infrastructure and build the houses.

Houses in these new neighbourhoods are not allocated through the market, but instead through political and bureaucratic procedures. Most state-subsidized housing is earmarked for people living in selected informal settlements through an essentially political process. ‘Community’ leaders or organizations select the beneficiaries from within the settlement, supposedly using fair procedures. Subject to satisfying the means-test and other criteria, the beneficiaries are then allocated newly built houses by the project manager. In practice, in Cape Town, these procedures have resulted in new neighbourhoods which are as segregated racially as apartheid-era neighbourhoods. Because most informal settlements in Cape Town are almost entirely populated by African people, the new housing projects have almost entirely African populations. There is a large housing shortage for poor coloured people in Cape Town, but this is manifested in dire overcrowding in apartheid-era working-class neighbourhoods. Because they have not moved into informal settlements, very few coloured people have benefited from state-subsidized new housing.

Delft South and Tambo Square are exceptions to the mono-racial character of new state-funded housing projects. Delft is an area immediately to the east of Cape Town’s international airport. Because the land in Delft was previously owned by the government department responsible for ‘coloured’ housing, it was agreed that 50 percent of new houses built there had to be allocated to the mostly coloured families on the existing municipal waiting-list. The other 50 percent were allocated to the mostly African residents of selected informal settlements. New housing in Tambo Square, to the southwest of the airport, was similarly allocated equally to people from the neighbouring ‘coloured’ township of Manenberg and ‘African’ township of Guguletu.

Previous research in Delft South, very soon after the first new houses were occupied, found that the circumstances of obtaining houses had forged some trans-racial solidarity, but that residents otherwise retained their pre-existing, racially-segregated networks (Oldfield, 2000, 2004). A more recent study also found enduring racial mistrust and polarity in Delft South (Millstein, 2007: 30–1). The absence of substantial integration in Delft South has been contrasted with the case of Westlake Village, in southern Cape Town. Lemanski (2006a) attributed the deeper inter-racial interactions in Westlake Village in part to the fact that residents there were drawn from the surrounding area, and thus had both a shared place attachment and some prior relationships. Residents also shared the same spaces of employment and their children shared spaces of schooling. In Delft South, in contrast, people’s identities and primary relationships remained bound up with their previous and distant neighbourhoods.

During 2009 – 10 years after Oldfield conducted her research, and about five years after Millstein conducted hers – we interviewed 21 coloured and African adult residents in Delft South (including Delft Leiden). We interviewed another 10 residents in Tambo Square. Samples of interviewees were selected that were representative of these neighbourhoods in terms of geographical coverage. They included men and women, of various ages, employed and unemployed. The samples cannot be assumed, however, to be representative. Our purpose was to understand how changes have been seen and understood, not to measure precisely the extent of change. Interviews were conducted by Xhosa-, Afrikaans- and English-speaking interviewers, as appropriate, using semi-structured interview guidelines. Most interviews lasted about 45 minutes.

Our interviews revealed two aspects of inter-racial attitudes and interactions. We found, first, that racial stereotypes or generalizations reminiscent of the apartheid period endure. At the same time, however, we found evidence for many (but not all) residents that inter-racial relationships have deepened through proximity. While some racialized differences endure, we found that these are now, in part but not entirely, understood in terms of cultural diversity rather than social division. We found many positive signs of increased inter-racial toleration and neighbourliness, and little evidence of racial tension.

Racial generalization in the ‘new’ South Africa

Under apartheid, racial prejudice and stereotyping were probably widespread, even among low-income people. Indeed, there is evidence that poor coloured people are, in general, more prejudiced against African people than are richer coloured (or white) people. In the 1990s, most poor coloured neighbourhoods were strongholds of the National Party, while the ‘liberal’ opposition and the African National Congress drew more support in middle-class coloured neighbourhoods (Eldridge and Seekings, 1996). It would hardly be surprising if we found evidence of persistent prejudice and stereotyping in mixed neighbourhoods such as Delft South and Tambo Square, in 2009. We certainly did find some evidence of this sort. In general, however, we found much more evidence that reflected a discourse of multicultural toleration, of inclusion in the ‘new’, democratic South Africa.

Enduring stereotypes and generalizations often revolve around crime. In our interviews, blunt generalizations were voiced only by African interviewees. According to one African woman in Delft South, ‘I can’t trust coloureds, they steal, they steal everything’, whereas ‘black people, I trust them, they can’t steal’ (DS#25) (in fact, theft is common in African neighbourhoods). Another African woman in the same neighbourhoods told us that ‘I don’t feel good, because . . . coloureds are messing up a lot, I don’t feel good about them’ (DS#29). Coloured interviewees expressed racialized views in more subtle ways. Racialized views might be linked to differences within neighbourhoods. Tambo Square turns out to be heterogeneous, with a small southern section populated almost entirely by coloured people. Interviewees there describe it as ‘safer here than on the other side’, i.e. than the more racially-mixed area (TS#34). ‘We have this bond’, said another resident, ‘because we are more together on this side’ (TS#38).

Several residents told us that they thought that the other racial group disliked their racial group. This seems to reveal a group-based fear or sense of vulnerability. According to a coloured woman in Delft South:

Black people don’t want coloured people. Most of the black people don’t want coloured people. About two years ago we were reading [in] the paper [about] that guy [in] Khayelitsha, he buy a house that side but people don’t want him that side. (DS#18)

This seems to be a reference to an incident in which a coloured man acquired a house in the African township of Khayelitsha, but was prevented from moving in by local activists who insisted that housing there was for African people only. This woman, however, immediately qualified her generalization, by saying ‘that is why I say most of the black people is alright and other people is not alright’ (DS#18).

Against these generalizations and stereotypes, we found lots of evidence of residents embracing a discourse of multicultural toleration. The coloured woman in Delft South quoted above elaborated:

You know that time, in apartheid, the coloured people stay that side and the black people stay that side but it’s not like that anymore, we must love black people and the black people must love us also. But we cannot stay there in Khayelitsha. (DS#18)

Similarly, according to another coloured woman in Delft South:

You see the other people like to say ‘No, [I] don’t like blacks’. I say no, it’s a new generation now and you must understand. . . . In those days you didn’t used to live like this. I say, no, you are wrong, in this day we must learn to live together, you see. (DS#19w)

This was echoed by an African woman in Delft South:

We need to be people. To become one. If we can be separated and say this one must go and live there alone . . . that can create hatred and a war; because when it comes that your child go and mess there . . . everybody there will come and fight here. I like mixing a lot. (DS#31)

And an African man in the newer, Leiden section of Delft:

. . . years ago, the coloureds were live alone, the blacks were live alone, in the apartheid time, you see. Now it’s a new South Africa now. We are living together as one, you see. Nobody can say now I am a kaffir, if he say [that I] am a kaffir, I have a right to lay a charge against him there, because [I] am not a kaffir anymore. . . . So it’s possible to live together, so that they must understand that we are all in one here in South Africa, nothing higher than somebody. Nobody is higher than somebody, yeah we are equally now, yeah. (DL#2)

In these mixed neighbourhoods we found little evidence of racial animosity, or of pronounced and explicit racial prejudice. Rather, we found a positive attitude towards co-residence. In Tambo Square, for example, informants insisted that neighbours get on well with each other. According to one coloured woman: ‘There are Africans and coloureds and . . . Malawian people and so on, actually it is a mix now, . . . but we get along well, and they get along well with us’ (TS#38). This seems to be rooted in a commitment to non-racialism that is both strong, at least rhetorically, and widespread. According to an African woman:

I don’t see any difference. Because colour makes no difference. So if a person is black or white is the same. (TS#39)

Crucially, many residents emphasize that there are good and bad people in any racial group. According to one African man:

Not all the coloureds . . . use the drugs. There’s a right coloureds, there is a wrong coloureds, you see. Some of the coloureds use the drugs and they are silly. Some of coloureds, they don’t use the drugs. They are human being like me, you see. . . . I don’t trust coloureds, you see, because coloureds, some of them are corrupt, some of them are good, that’s why am saying to you. (DL#2)

Both coloured and African interviewees criticized their own racial group. According to one coloured woman in Delft South:

I can say I have a problem with coloured people because they steal our things. My toilet seat was stolen. I have put [in] a bath, a new bath, they took out that pipe. . . . While we are sleeping at night they are not sleeping. (DS#28)

An African woman in Delft South described coloured people as ‘people with respect’, and told us that, when she returned at night from church in Khayelitsha, coloured men would often enquire why she was walking at night; when she explained she was coming from church, they would escort her back to her house, whereas ‘black [i.e. African] people, they only want money’ (she had earlier been mugged by African adolescents, who took her bag). ‘I like coloureds because they are also people. I like them because they are not dangerous . . . and more over they are not people who make troubles. They know how to communicate with people’ (DS#31).

Overall, a recognition of intra-racial diversity seems to us to be decisive in rooting a ‘new’ and inclusive South African discourse in racially-mixed neighbourhoods. Earlier research found strong notions of ‘us’ and ‘them’, particularly in relation to differences in customs, culture and traditions and that cultural diversity was viewed as an obstacle (Millstein, 2007: 32). In the present study we found an acceptance of cultural diversity, and even a celebration of it (linked, in at least one interviewee’s mind, to the new South Africa being democratic – DL#2).

Learning multicultural coexistence

Several residents – especially coloured residents – spoke about how they had learnt to be more accommodating or tolerant. One coloured woman told us:

I can tell you in the beginning I was just always thinking, hey, kaffirs, how am I gonna stay with kaffirs? . . . And when I move in, they moved in, I say, oh, no, [it] is not what I was thinking, they are people like everybody, and they know that to respect for each other. (DL#6)

One coloured woman related how, unusually, she had lived in the large ‘African’ township of Khayelitsha – when ‘it was small, not [like] now’ – where she got used to living with African people, ‘that’s why it is not a problem for me, but for the other people it is a problem’ because they are ‘not used to the other people’ (DS#19w). Another coloured woman with no such background recalled that:

At the beginning it was not nice. Everyone thinks ‘this is my territory’, they just want to do anything. Okay, then you get people from Polokwane [in the far northern part of South Africa], fresh from the bush, now they want to come and take their bush manner out here! Broil their sheep’s head in their back yards! That is not right! (DL#3)

In some cases, the experience of living in more diverse neighbourhoods and of inter-racial interactions challenged people’s preconceptions. ‘Yes’, one coloured man told us, ‘I must say that some of the Xhosas living here are very decent people; we greet each other, we ask one another how things are, we have no problems with them’ (TS#35). Another coloured woman in Delft described her African neighbours as ‘very respectable’ (DL#3).

Children are, perhaps, better placed to learn attitudes. One coloured man told us ‘I think the most togetherness of races is all about kids, they don’t see it as we see it as big people. . . . It’s kids who can make a difference.’ He told us of a time when he gave his children some biscuits, and they happily shared them with other African boys from the street, without any hesitation (DL#1).

These statements about learning occur more often in our interviews with coloured residents, but there are cases of African interviewees voicing the same sentiments. An African woman highlighted the fact that all her neighbours were coloured and seemed to have no choice but to get to know people of a different population group. ‘Now that we are a mix I see it’s nice. . . . We get along, we borrow [from] each other’ (DS#26).

This commitment to non-racialism sits alongside multiculturalism. Most residents recognize that there are cultural differences between coloured and African people, but do not associate these with any normative hierarchy. In Delft South, one resident told us about her neighbours’ diverse customs:

We sometimes brew beer like in Eastern Cape. But we are always together, we are united. In such a way that . . . we don’t even worry about living with coloured people, we take them as our people. If they have a problem . . . or things like funerals . . . we go. . . . Ever since I came here, we as neighbours are living in peace. Sometimes you cannot live the way we do with your own family – I must put it that way. Because if you live with your family there are times where you don’t get along. . . . But as we are staying mixing with different people from different places, we are like a family. If you find us staying as neighbours, chatting . . . you would be sure that we are brothers and sisters. From that corner to that one [pointing] you won’t really differentiate where a person came from . . . we are like we came from one village and one family. (DS#23)

Language can be an impediment to interaction. A coloured couple, who have several African neighbours, said that ‘sometime the one says “no man, [I] am sick with Afrikaans”, then I say “no, you can speak English ” ’. The wife says that her husband ‘understands a little bit of Xhosa, . . . not so good but he can understand when the people is talking to him you see, and he’s got Xhosa friends’ (DS#19w). Asked whether she would prefer to live in an all-African neighbourhood, one African woman said ‘there would be more communication; it happens sometimes that as we are living with coloureds . . . perhaps you want to say something but you can’t because you can’t speak Afrikaans, their language; if we were all black people I think it would be very nice’ (DS#29).

In interview after interview, however, residents emphasize that relationships are good with neighbours from different racial or cultural groups. According to a coloured man:

My neighbours like you know they have got all kinds of races, . . . all kind of people: black, staying around me is muslims, my next door neighbour is muslim, my other next door neighbour is coloured and the opposite is black and the one next to them is white. So I have got four kind of race people here. But the communication with us is very good. I don’t got a problem with them at all. Yah, I think [relationships] are good, I think they are good, there’s no complaints about that. . . . That house the other house the other house [to the far left, across the road] there staying, not this one there where the tyres is, there’s a white lady and a coloured staying together there. Then there after the tyres in the zone there there is a black man and a white woman staying together. And the togetherness, they are staying together so but they are going maybe to the shebeen [bar] and sitting together and such things, walking past here. And there’s a muslim there staying with a white woman, yah. There’s plenty of white people from outside coming here to see the family here. (DL#1)

As in this interview, Muslims are often described as a different ‘race’; it is unclear whether the Muslims in question are coloured South Africans or immigrants from Somalia or elsewhere, but it seems to underscore the cultural understanding of race.

An African woman in Delft South expressed the predominant view:

I don’t have a problem with them. . . . I even live with them here in my yard [at the back of the house], I don’t have a problem. (DS#30)

A coloured woman went so far as to say:

We stay amongst Africans. There are Africans living opposite us. The majority here are Africans too. It is actually much better living amongst them. (DS#21)

Some people, she said, still held onto past attitudes and thought they were better than other people – meaning that some coloured people thought they were better than African people. Her retort to such racial arrogance was:

We stay in an under-privileged neighbourhood. Do not think that you are better then the rest, because you are not! What are you doing in Delft? . . . If you were better, then you do not belong in Delft. (DS#21)

When people encounter some unpleasant people in the other group, this can affect their view of the group as a whole. Above we quoted one African man who said that he did not trust coloured people because some, but not all, of them were untrustworthy.

Some of the coloureds . . . think they are richer than us. . . . They don’t take us as a human being like them. You see, they took us as just nonsenses. . . . The coloureds are rude. . . . Some of them are rude, some of them are not rude, but the most of them in the area here are very, very rude. They are using drugs, they are drinking liquor, so they are rude. (DL#2)

One coloured man drew a distinction between different African people in his immediate neighbourhood:

. . . these black people over there, my communication is more tight with them. The black people opposite me it’s more tight. . . . What I am trying to tell you is I am more intimate with the black people over there in the street and around here because they all know me. There’s plenty of black people . . . here, the people who are here is black people, and they know me, they know me very well, and we are very intimate. (DL#1)

The experience of living together is not uniformly positive. Other people say ‘I cannot live with these people’ (DL#3f), or ‘I cannot, we have been trying for years’ (DL#3). A coloured man complained that other people are resentful of inter-racial or inter-cultural interactions:

Jealousy, it’s very much here. Not jealousy of races, jealousy of cultures together. [I] am a coloured, you are a coloured, and I don’t like you, you as a coloured to go and talk too much to that white girl or white lady, white men or Muslim or whatever. Jealousy is the one thing I don’t like. (DL#1)

Many people, however, seem to have learnt to live together.

How close are inter-racial relationships between neighbours?

Most of our interviewees say they have ‘no problems’ with neighbours of other races, or at least that they have no more problems with them than with same-race neighbours, and almost everyone greets their neighbours Many have relationships that extend beyond these civilities. One coloured man tells us:

Yes, I must say some of the Xhosas living here are very decent people. We greet each other; we ask one another how things are. We have no problem with them. (TS#35)

This man’s wife then adds that she often visits their neighbours:

Yes! Many times I go and visit one, the other one will maybe come to me. So, we visit as if we are not different races. We feel fine asking each other things. (TS#35w)

They exchange gifts of food, and she also gives outgrown children’s clothes to her African neighbours:

Whatever is too small for my little ones I give to them, and they would come and say thank you very much here is a bottle of coke for you, something like that. (TS#35w)

Many residents have a close relationship with at least one of their other-race neighbours. Asked if she had any African friends, a coloured woman replies that she often visits and talks with one of her neighbours, an elderly African woman (DS#18). Indeed, some interviewees insist that an other-race neighbour is the neighbour to whom they feel closest, or on whom they rely in times of need.

Neighbours keep an eye on each other’s houses, which is especially important for those African people who return to their ‘home’ villages in the Eastern Cape over Christmas or at other times. Some entrust their children to neighbours if they are going out. Neighbours borrow from each other small items such as salt, sugar, and even small sums of money to pay for (pre-paid) electricity. Such interactions seem to take place between African and coloured neighbours as often, or at least almost as often, as between same-race neighbours:

I can leave them [his children] with anyone, either white or black or muslim, around me, I can leave them there, . . . I can be comfortable to leave them there. No that’s one thing I can tell you about my neighbours. I can leave [my children] with [them] comfortable [if] my wife [and] I want to go . . . (DL#1)

An African person says that a coloured boy stays in her house when she is away (DL#5). When there are African traditional ceremonies, coloured neighbours are also invited to attend, while African people say that they will accompany coloured neighbours to events like funerals:

We are always together – we are united. . . . we don’t even worry about living with coloured people, we take them as our people. . . . things like funerals, we go with them. (DS#23)

Some coloured people tell us that African people are more friendly and neighbourly than ‘our own race’.

At the same time, neighbours are rarely anyone’s closest friends. Regardless of race, and of the racial composition of the immediate neighbourhood, residents in these mixed-race neighbourhoods tended to have close friends living further away, sometimes in the same overall neighbourhood (for example, Delft South), often further afield. This was the case also in mono-racial neighbourhoods in Cape Town where we conducted further research. Indeed, our other research into the quality of ‘community’ in different parts of Cape Town suggested that, in most neighbourhoods, the quality of community is uneven or low. Our point is not to say that relationships between neighbours of different races are exemplary, but rather that race seems much less important than we expected in shaping everyday interactions and attitudes.

Relations between neighbours in both areas are constrained by the ever-present threat of crime. Perceptions of crime in both neighbourhoods were racialized in that most of our interviewees, whether African or coloured, associated crime with young coloured men. Most people seem to think that criminality is ubiquitous, but African people are more effective at disciplining their delinquent youth. Moreover, even African interviewees generally distinguished between good and bad coloured people. As one African woman put it, there are ‘wrong coloureds’, who use drugs, and ‘right coloureds’, who don’t, and who ‘are human beings like me, you see’ (DL#2). Indeed, the recognition that coloured and African neighbours alike included respectable people who were the victims rather than the perpetrators of crime helped people to reach across former racial divides.

Conclusion

The study of segregation and desegregation is dominated by studies of the persistence of segregation, albeit in changing forms (for example, for the USA, Charles, 2000; Logan and Zhang, 2010; Massey et al., 2009). The study of demographically integrated neighbourhoods is a relatively small (but growing) part of the broader literature (see, for example, on the USA, Ellen, 2000; Maly, 2005). In South Africa, residential desegregation has proceeded slowly, in part due to public policy on the allocation of state- subsidized housing. Furthermore, even those public spaces which appear to have desegregated turn out, on closer examination, to be characterized by continuing segregation at a micro level. Previous studies of unusual neighbourhoods suggest that demographic desegregation was not matched by any fundamental deracialization of everyday interactions. Studies of Delft South (Millstein, 2007; Oldfield, 2000, 2004) tend to see episodes of inter-racial cooperation as being contingent on specific, discrete stimuli. The exceptional case of Westlake Village, where there do appear to be close inter-racial interactions, is explained in terms of the strong links between and shared place attachment of coloured and African people prior to their becoming neighbours in the new public housing project. Overall, therefore, a sceptical tone has informed the literature on desegregation, especially but not only in South Africa.

Our findings are more encouraging. Coloured and African people in Delft South and Tambo Square do not have shared histories or place attachment or other ties prior to becoming neighbours in these new neighbourhoods. Nonetheless, our interviews provide plenty of evidence of increased racial toleration, reduced prejudice and race-neutral interactions. At the same time, racialized identities still matter: people see themselves and others as coloured or African (or ‘Muslim’, white and so on). The persistence of racialized identities does not seem to be associated with enduring racial division. Neighbours interact positively, and describe their neighbourhoods in terms of multicultural accommodation and toleration. Many of our interviewees appreciated their racial diversity. It is not clear to us, as it was to Oldfield (2004: 194), that coloured and African people’s lives are patterned so very differently. Perhaps the passage of time has meant that people’s lives are more oriented towards neighbours (although we did find that many residents retain extra- neighbourhood links, including to kin, old friends and churches). We did not find evidence to support Millstein’s finding that there is deep inter-racial mistrust. (Perhaps her finding is linked to her focus on formal politics, which might remain more racialized, rather than everyday interactions.) ¹ Although Millstein argued that little had been done for residents of different races to learn about each other in Delft, our study found that residents have been learning from each other during the course of their time as neighbours. The contact hypothesis should not be thrown out too quickly.

It is likely that inter-racial contacts are positive in these racially-mixed neighbourhoods precisely because the conditions are close to the ‘optimal’ ones identified in the contact hypothesis literature. Relationships are egalitarian rather than hierarchical, in that people are similarly income-poor neighbours in similarly small houses. They are not competitive, in that people are neighbours because they have already been allocated houses and receive the same, non-scarce municipal services. Contact is regular and frequent, and free from anxiety. Contact is also personalized in that relationships between close neighbours have real potential for change, in that strangers become acquaintances and acquaintances become friends.

Several specific features of these neighbourhoods enhance those factors that make interactions more positive. These are neighbourhoods in which almost everyone is poor. To be eligible for a state-subsidized house in any of our research sites, applicants had to have an income of less than R3500 per month. A small number of residents had improved their standard of living since moving into the area, but most remained dependent on a mix of government grants or pensions and meagre earnings from typically irregular or casual or informal employment. Unemployment is widespread. Most interviewees insisted that they had enough to feed themselves, but most also complained about their inability to afford much more. Shared experiences of poverty seem to be a social leveller, as residents see that their neighbours share their own economic difficulties. As one resident in Delft South told us (quoted earlier), ‘do not think that you are better than the rest, because you are not’, the proof being that someone who was ‘better’ would not be living in Delft (DS#21).

Residents not only see that economic hardship respects no racial boundaries, but neighbours’ responses to economic and social challenges also transcend racial lines. Among large numbers of both African and coloured people in our study sites, respectability is an important quality: respectable people do not get drunk or abusive, they attend church and strive to keep their children under control. In mixed neighbourhoods, most residents can see that neighbours with different ‘racial’ or geographical origins share their economic circumstances and strive for respectability. Conversely, jealousy occurred within as much as, and perhaps more than, across racial boundaries.

The threat and reality of violent crime is another important leveller. For most people, certain public spaces are dangerous at all times, and after dark even the immediate neighbourhood might cease to be safe. Most residents are aware of their vulnerabilities – and of the similar vulnerabilities of their neighbours, regardless of race.

The picture emerging from our research in Cape Town sits at odds with the conventional sociological wisdom, derived from the US experience, that homophily – i.e. that birds of a feather, flock together – limits inter-racial social interactions even among people living in racially-‘integrated’ neighbourhoods. In racially-mixed, low-income neighbourhoods in Cape Town, we find evidence that inter-racial interactions have become more commonplace and inter-racial attitudes have become more accommodating and tolerant, with the passage of time. A high degree of inter-cultural interaction in South African cities should perhaps not be surprising, given the long history in South Africa of harmonious inter-ethnic and inter-religious interactions in poor neighbourhoods. Indeed, while the apartheid state succeeded in imposing racial segregation, it largely failed to segregate African townships by ethnicity (for example, separating ‘Zulu’ from ‘Sotho’ families). It seems that the optimal conditions for the contact hypothesis are more commonplace in poor South African neighbourhoods than they are in either other public spaces in South Africa or in mixed neighbourhoods in the USA.

Recent research in the USA suggests that workplaces and voluntary organizations provide important fora for inter-racial friendships among African-Americans, while inter-racial friendships are more common among white people living in mixed neighbourhoods than among white people in mono-racial neighbourhoods (Briggs, 2007). South African research suggests that, in South Africa as in the USA, it is important to examine both the contexts within which inter-racial interactions do occur and those in which they do not. Comparative, cross-national studies might also help to clarify the conditions under which positive inter-racial interactions occur. Besides South Africa and the USA, such comparative work might usefully consider Brazil, where a high degree of ‘horizontal’ inter-racial social interaction coexists with a high degree of ‘vertical’ inter-racial inequality (Telles, 2005).